Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Re: Blew another damn transformer on my Trane XB80

14 views
Skip to first unread message

robb

unread,
Apr 9, 2011, 7:59:31 PM4/9/11
to
"Steve Turner" wrote in message news:inoakf$qin$1...@dont-email.me...

> A continuation of the "Why does the 115V->24V transformer keep
> blowing on my Trane XB80?" discussion I started on 04/02/2011.
>
> Yep, My A/C unit blew another transformer. Pictures (and wiring
> diagram) here:
>
> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>
> Trane's manual for the unit is here (for perhaps better viewing of
> the wiring diagrams that I also copied to my above flickr site as
> jpg images):
>
>
> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-1666-07_04012009.pdf
>
> As you can probably see in the pictures, there is visible charring
> of the 115V leads going into the transformer, and of course the 115V
> circuit is open (again). If you didn't see my first thread, this is
> the third transformer the unit has blown. In the previous
> discussion, it was discussed that perhaps the first one just blew
> because of old age (6 years), and the second blew because it wasn't
> a proper replacement (poor quality, made in China, etc.). This
> third unit is most certainly a proper replacement, and it's most
> certainly indicative of a real problem I have somewhere else in the
> unit. I didn't see any such charring on the previous two units, at
> least not like this.
>
> I never got a real chance to test out the system after installing
> this third transformer. We had cool weather for several days, and I
> never tried to force the system to come on so I could monitor it;
> that was probably a mistake. Unfortunately, I was also absent from
> the premises during the extended times when the unit was most likely
> operational, so that didn't help either. However, my family tells me
> that it WAS working and cooling the house rather nicely, for at
> least a day, perhaps two. I'm getting 115V in all the right places,
> so it doesn't look like an over-voltage condition to me. Perhaps
> it's an overheating condition? It looks to me like the only real
> load on this circuit is the blower motor; could the motor be causing
> this? The blower spins freely when I turn it by hand. Start
> capacitor on the motor maybe? Relay on the control board perhaps?
>

Just a thought,

When you put the new transformer in does the "Diagnostic Light" LED
turn on ?
Does the LED blink or flash in one of the patterns indicated in the
"diagnostic codes" section that you posted a link to ?

The "diagnostic codes" imply that the control board can detect
several of the problems mentioned by others.
Maybe a good starting point as there is not enough time to test with
test equipment.

robb

Jim Yanik

unread,
Apr 9, 2011, 8:29:30 PM4/9/11
to
"robb" <so...@where.on.net> wrote in
news:WfKdnQOvu_d2bz3Q...@earthlink.com:

a bad start cap would not affect the 24v control transformer,that just
powers the controller board and relays.
either something is loading the transformer or a problem with input
voltage.


>>
>
> Just a thought,
>
> When you put the new transformer in does the "Diagnostic Light" LED
> turn on ?
> Does the LED blink or flash in one of the patterns indicated in the
> "diagnostic codes" section that you posted a link to ?
>
> The "diagnostic codes" imply that the control board can detect
> several of the problems mentioned by others.
> Maybe a good starting point as there is not enough time to test with
> test equipment.
>
> robb
>
>

a line voltage monitor may be needed to see if there's some short-term
overvoltage applied to the transformer,and a oscilloscope may determine if
the input waveform is sinusoidal or otherwise,because 60hz iron core
tranformers don't like extreme distortion on their input,it gets converted
to heat,not output voltage. a DMM will not show line distortion or short
term overvoltages.
are the wires burned close to the transformer,or over their entire length?

--
Jim Yanik
jyanik
at
localnet
dot com

cl...@snyder.on.ca

unread,
Apr 9, 2011, 9:45:34 PM4/9/11
to
On Sat, 09 Apr 2011 19:29:30 -0500, Jim Yanik <jya...@abuse.gov>
wrote:


Or just another crappy transformer.
You do not just ASSume it is the transformer


>>>
>>
>> Just a thought,
>>
>> When you put the new transformer in does the "Diagnostic Light" LED
>> turn on ?
>> Does the LED blink or flash in one of the patterns indicated in the
>> "diagnostic codes" section that you posted a link to ?
>>
>> The "diagnostic codes" imply that the control board can detect
>> several of the problems mentioned by others.
>> Maybe a good starting point as there is not enough time to test with
>> test equipment.
>>
>> robb
>>
>>
>
>a line voltage monitor may be needed to see if there's some short-term
>overvoltage applied to the transformer,and a oscilloscope may determine if
>the input waveform is sinusoidal or otherwise,because 60hz iron core
>tranformers don't like extreme distortion on their input,it gets converted
>to heat,not output voltage. a DMM will not show line distortion or short
>term overvoltages.
>are the wires burned close to the transformer,or over their entire length?

Transient overvoltages of less than 20% or there-abouts are unlikely
to blow the transformer primary of the furnace without manifesting
themselves elsewhere in the house - and over 20% would definitely
manifest themselves elsewhere.

Steve Turner

unread,
Apr 9, 2011, 9:55:37 PM4/9/11
to

I peeled the insulation back from the hot and neutral wires and the burning
only occurred right at the terminals, no more than 1/4" into the wiring.

John Robertson

unread,
Apr 9, 2011, 10:06:18 PM4/9/11
to

So the problem is the joint of the transformer wiring to the lugs. They
may have cold solder joints, or the lug may be slightly loose - enough
to crack the solder connection over time. Your transformer may still be
good!

When you are examining the transformer wire make sure it is stripped
enough to get good solder coverage. Burn off the insulation (enamel) -
do not sand, scrape, or try to erode it because if you scratch the wire
it will break at the scratch.

Re-solder the transformer connectors after first making sure the wire
has at least two complete wraps around the solder lug and the solder
makes a good meniscus joint to the wire and lug. Also use a good grade
of solder, not regular 60/40, find some 63/37 solder (Kester is best) -
this solder resists cold solder joints due to movement when cooling by
solidifying very fast.

John :-#)#

--
(Please post followups or tech enquiries to the newsgroup)
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
Call (604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

Phil Allison

unread,
Apr 9, 2011, 10:22:13 PM4/9/11
to

"robb"

>
> Yep, My A/C unit blew another transformer. Pictures (and wiring diagram)
> here:
>
> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>

> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-1666-07_04012009.pdf
>
> As you can probably see in the pictures, there is visible charring of the
> 115V leads going into the transformer, and of course the 115V circuit is
> open (again).


** From the damage to the tranny visible in the photos - the primary has
developed an internal short ( due to insulation failure) and since there is
no fuse link in series, the resulting current was high enough to make an
exposed wire leading to the terminals explode.

An internal short could develop due to heat alone because of an overload on
the tranny - but this requires a fault to exist on the secondary side which
seems not to be the case.

High voltage spikes on the primary could also cause insulation failure
leading to the damage seen in the pics - lightning does this sort of thing.
So also could back emfs from the blower fan if the is a bad connection in
the AC supply feed.

I suggest you provide the next replacement for that vulnerable tranny with
some "protection" - firstly an in-line fuse of say 1/4 amp AND a
capacitor wired across the primary of say 1uF rated for continuous use
across the AC supply.

If there is an overload on the tranny, the fuse will blow.

The 1uF capacitor should suppress spike voltages enough to save the tranny
from harm.

..... Phil

Grant

unread,
Apr 9, 2011, 10:54:12 PM4/9/11
to

Is that a dual primary 115V (230V) xformer? Did you wire both primaries
in parallel for 115V operation?

Grant.

Phil Allison

unread,
Apr 10, 2011, 2:00:35 AM4/10/11
to

"Phil Allison"

> firstly an in-line fuse of say 1/4 amp ....


** Must be a "slo-blo" type fuse.

.... Phil

Jeff Thies

unread,
Apr 10, 2011, 10:25:29 AM4/10/11
to

1 uF sounds a little high.

I would think .01 uF at 400V would be a better bet. Or an MOV (metal
oxide varistor) or three (across the line and then from each side to
ground).

I don't think this is a voltage spike problem though. Voltage spikes
take out other components first.

Jeff
>
>
>
> ..... Phil
>
>
>
>
>
>
>

tra...@optonline.net

unread,
Apr 10, 2011, 10:43:29 AM4/10/11
to
On Apr 9, 8:29 pm, Jim Yanik <jya...@abuse.gov> wrote:
> "robb" <s...@where.on.net> wrote innews:WfKdnQOvu_d2bz3Q...@earthlink.com:
>
>
>
>
>
> > "Steve Turner"  wrote in messagenews:inoakf$qin$1...@dont-email.me...

It also could be powering a thermostat. Some of them will take power
from the transformer circuit. For example, I have a Honeywell
VisionPro
that has an optional 24V connection to power the thermostat, with
batteries then being the backup. Doing it that way allows the display
to be backlit 24/7 too.

Also, it's not unusual to have humdifiers tied into the 24V circuit
too.

Again, he needs to start doing some basic current measurements,
starting at the transformer and find out how much current the whole
thing is drawing and then if it's high, work to isolate it. If the
current
is normal, then I'd put in a 1.5 amp fuse in the secondary
temporarily.
Contrary to the suggestions to use a fast blow, I'd probably just use
a regular one, as whatever is capable of destroying a transformer
should be most capable of opening any fuse.


>
>
>
>
>
> > Just a thought,
>
> > When you put the new transformer in does the  "Diagnostic Light" LED
> > turn on ?
> > Does the LED blink or flash in one of the patterns indicated in the
> > "diagnostic codes" section that you posted a link to ?
>
> > The "diagnostic codes"  imply that the control board can detect
> > several of the problems mentioned by others.
> > Maybe a good starting point as there is not enough time to test with
> > test equipment.

That's a good idea too.


>
> > robb
>
> a line voltage monitor may be needed to see if there's some short-term
> overvoltage applied to the transformer,and a oscilloscope may determine if
> the input waveform is sinusoidal or otherwise,because 60hz iron core
> tranformers don't like extreme distortion on their input,it gets converted
> to heat,not output voltage.

I don't see how he could have a waveform that is so out of shape that
it
burns up this one specific transformer, yet there are no apparent
complaints of any other problems in the house. Or how you'd get
such a badly distorted waveform in the house in the first place.


> a DMM will not show line distortion or short
> term overvoltages.
> are the wires burned close to the transformer,or over their entire length?
>
> --
> Jim Yanik
> jyanik
> at
> localnet

> dot com- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -

Mark

unread,
Apr 10, 2011, 11:21:07 AM4/10/11
to

To OP
this is a long shot but...
I noticed on the Trane wiring diagram that you posted that the
connections for the IGNITION circuit are shown to be near to the
connections for the primary of the transformer.

If there is some insulation problem on the ignition wires and the VERY
high voltage from the ignition circuit is sparking or jumping over to
the transformer primary circuit, this could cause the symptoms you are
seeing.

Make sure the wires that are part of the ignition circuit are not even
close to the anything else. and inspect them (with the power off) for
any sigh of cracks or other faults. These have very high voltage
like in a car spark plug and can jump several inches if there is an
insulation problem. And the problem would be intermittent. The
transformer primary voltage would look perfect except when the high
voltage spark jumps over to it and this could easily damage the
insulation on the transformer which is not designed for very high
voltages.

If you are an electrical novice, you may want to think about calling
in for some help at this point.

Mark
>

David

unread,
Apr 10, 2011, 11:55:01 AM4/10/11
to
>"Mark" wrote in message
>news:fe82a1bf-396e-4db1...@e9g2000vbk.googlegroups.com...

>To OP
>this is a long shot but...
>I noticed on the Trane wiring diagram that you posted that the
>connections for the IGNITION circuit are shown to be near to the
>connections for the primary of the transformer.
<snip>
>Mark

Everything I see on the diagrams says this is a HSI (Hot Surface
Igniter) type of furnace. There is no high voltage in that type
of igniter.
David

Smitty Two

unread,
Apr 10, 2011, 12:18:04 PM4/10/11
to
In article <fs62q6djakj0jdki5...@4ax.com>,
Grant <o...@grrr.id.au> wrote:

>
> Is that a dual primary 115V (230V) xformer? Did you wire both primaries
> in parallel for 115V operation?
>
> Grant.

This thread started on a.h.r (alt.home.repair) and I'm not sure the OP
monitors s.e.r. Someone cross-posted it here, but the bulk of the thread
(several hundred posts) is on a.h.r. only.

So if you want to be helpful, you might wander over there. ISTR reading
that the OP taped off two wires.

Jim Yanik

unread,
Apr 10, 2011, 1:08:15 PM4/10/11
to
Grant <o...@grrr.id.au> wrote in
news:fs62q6djakj0jdki5...@4ax.com:

> On Sat, 09 Apr 2011 20:55:37 -0500, Steve Turner
> <bbqb...@swtacobell.net.invalid> wrote:
>
>>On 4/9/2011 7:29 PM, Jim Yanik wrote:
>>> "robb"<so...@where.on.net> wrote in
>>> news:WfKdnQOvu_d2bz3Q...@earthlink.com:
>>>
>>>> "Steve Turner" wrote in message news:inoakf$qin$1...@dont-email.me...
>>>>
>>>>> A continuation of the "Why does the 115V->24V transformer keep
>>>>> blowing on my Trane XB80?" discussion I started on 04/02/2011.
>>>>>
>>>>> Yep, My A/C unit blew another transformer. Pictures (and wiring
>>>>> diagram) here:
>>>>>
>>>>> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>>>>>
>>>>> Trane's manual for the unit is here (for perhaps better viewing of
>>>>> the wiring diagrams that I also copied to my above flickr site as
>>>>> jpg images):
>>>>>
>>>>>
>>>>>
>>>> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-16

>>>> 66- 07_04012009.pdf

another poster on the alt.home.repair NG made that suggestion/comment.
A very good one,that I had overlooked. good catch!

Mark

unread,
Apr 10, 2011, 1:36:04 PM4/10/11
to

David,
I think you are right...

well it sounded good anyway....

regards
Mark

Phil Allison

unread,
Apr 10, 2011, 7:47:24 PM4/10/11
to

"Jeff Thies"


** No it ain't.


> I would think .01 uF at 400V would be a better bet.

** Think again.

> Or an MOV (metal oxide varistor) or three (across the line and then from
> each side to ground).

** Bad idea.

> I don't think this is a voltage spike problem though. Voltage spikes take
> out other components first.

** Not on the AC supply they don't.


..... Phil


Tony Miklos

unread,
Apr 10, 2011, 8:05:50 PM4/10/11
to

Hi John, fancy meeting you here! Wait, I didn't realize this is cross
posted to sci.electronics.repair Now I know why you are here. Glad you
liked the package.

Tony

Tony Miklos

unread,
Apr 10, 2011, 8:14:08 PM4/10/11
to
On 4/9/2011 10:22 PM, Phil Allison wrote:
> "robb"
>
>>
>> Yep, My A/C unit blew another transformer. Pictures (and wiring diagram)
>> here:
>>
>> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>>
>> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-1666-07_04012009.pdf
>>
>> As you can probably see in the pictures, there is visible charring of the
>> 115V leads going into the transformer, and of course the 115V circuit is
>> open (again).
>
>
> ** From the damage to the tranny visible in the photos - the primary has
> developed an internal short ( due to insulation failure) and since there is
> no fuse link in series, the resulting current was high enough to make an
> exposed wire leading to the terminals explode.
>
> An internal short could develop due to heat alone because of an overload on
> the tranny - but this requires a fault to exist on the secondary side which
> seems not to be the case.

Oh my gawd, someone agrees with me. Looking at the picture says a lot.
Burnt on the primary side, looks like new on the secondary. And this
is the third transformer with an open primary!


I don't know about a the 1uF cap, seems way too high so I'll delete what
was below and pretend I didn't see it.

Phil Allison

unread,
Apr 10, 2011, 8:19:27 PM4/10/11
to

"Tony Miklos don't know much"

>
>> ** From the damage to the tranny visible in the photos - the primary has
>> developed an internal short ( due to insulation failure) and since there
>> is
>> no fuse link in series, the resulting current was high enough to make an
>> exposed wire leading to the terminals explode.
>>
>> An internal short could develop due to heat alone because of an overload
>> on
>> the tranny - but this requires a fault to exist on the secondary side
>> which
>> seems not to be the case.
>
> Oh my gawd, someone agrees with me. Looking at the picture says a lot.
> Burnt on the primary side, looks like new on the secondary. And this is
> the third transformer with an open primary!
>
>
> I don't know about a the 1uF cap, seems way too high

** Why ?????

It ( likely ) needs to absorb a transient back emf generated by a powerful
blower fan.

Non puny size cap will ever do that and using 1uF cases no harm whatsoever.


> so I'll delete what was below and pretend I didn't see it.

** Nothing like doing the old Ostrich trick when the brain gets overloaded.

Wot a jerk off.


.... Phil


Ron D.

unread,
Apr 10, 2011, 8:39:44 PM4/10/11
to
Put 3A automotive fuse on secondary. Use light bulb trick to find
shorts. Two 12 V light bulbs in series that draw no more than 40 VA.

Jeff Thies

unread,
Apr 10, 2011, 9:47:52 PM4/10/11
to

I thought you had simply misspoke and that this was an honest error.

The reactance of a 1uF cap at 60Hz is: 2652 ohms (1/(2*pi*F*C)
Online calculator:
http://www.kusashi.com/reactance-c.php?f=60&c=1&stage=results

V^2/R = W

Assuming primary, as why would you put it on the secondary:

120^2 / 2652 = 5.43 W

Does that not seem wrong to you?

If not then go buy a 200V non polarized 1uF cap. It is no easy chore.

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=131081

Jeff

Tony Miklos

unread,
Apr 10, 2011, 10:37:17 PM4/10/11
to

I have some big ass mylars, I think they are 4uF @ 200v.

Phil Allison

unread,
Apr 10, 2011, 10:46:59 PM4/10/11
to

** Completely.

Such a cap dissipates no energy at all.

What planet do you come from ??


> If not then go buy a 200V non polarized 1uF cap. It is no easy chore.


** 1uF caps for use across the AC supply are cheap and plentiful.

Typical examples are metallised polypropylene " class X2 " types and sell
for $1 or $2 each.

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=399-5466-ND


What planet do you come from ??

.... Phil

Phil Allison

unread,
Apr 10, 2011, 10:48:36 PM4/10/11
to

"Tony Miklos"

> I have some big ass mylars, I think they are 4uF @ 200v.


** They won't last long if connected across the AC supply.

.... Phil


Tony Miklos

unread,
Apr 10, 2011, 10:55:00 PM4/10/11
to

I know, I was just about to reply to my post saying that the voltage
rating is VDC.

Phil Allison

unread,
Apr 10, 2011, 11:01:54 PM4/10/11
to

"Tony Miklos"
Phil Allison wrote:

>>
>>> I have some big ass mylars, I think they are 4uF @ 200v.
>>
>>
>> ** They won't last long if connected across the AC supply.
>>
>

> I know, I was just about to reply to my post saying that the voltage
> rating is VDC.


** Film caps rated at 600VDC will generally last a fair while wired across a
120 volt AC supply ( but not with a 240VAC supply ) - but is it far better
to use a purpose designed and agency approved "class X1" or "classX2"
capacitor.


... Phil


cl...@snyder.on.ca

unread,
Apr 10, 2011, 11:15:34 PM4/10/11
to
On Sun, 10 Apr 2011 12:22:13 +1000, "Phil Allison" <phi...@tpg.com.au>
wrote:

>
>"robb"
>
>>
>> Yep, My A/C unit blew another transformer. Pictures (and wiring diagram)
>> here:
>>
>> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>>
>> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-1666-07_04012009.pdf
>>
>> As you can probably see in the pictures, there is visible charring of the
>> 115V leads going into the transformer, and of course the 115V circuit is
>> open (again).
>
>
>** From the damage to the tranny visible in the photos - the primary has
>developed an internal short ( due to insulation failure) and since there is
>no fuse link in series, the resulting current was high enough to make an
>exposed wire leading to the terminals explode
>

>An internal short could develop due to heat alone because of an overload on
>the tranny - but this requires a fault to exist on the secondary side which
>seems not to be the case.
>
>High voltage spikes on the primary could also cause insulation failure
>leading to the damage seen in the pics - lightning does this sort of thing.
>So also could back emfs from the blower fan if the is a bad connection in
>the AC supply feed.


>
>I suggest you provide the next replacement for that vulnerable tranny with
>some "protection" - firstly an in-line fuse of say 1/4 amp AND a
>capacitor wired across the primary of say 1uF rated for continuous use
>across the AC supply.
>
>If there is an overload on the tranny, the fuse will blow.
>
>The 1uF capacitor should suppress spike voltages enough to save the tranny
>from harm.
>
>

A capacitor across an AC supply??????????????? As a surge
protector????
Have not heard of that before.

A capacitor across the AC line would appear as a load - and could
form a resonant l/ci tank circuit, which would also appear as a low
resistance -causing high current to flow

There are 2 other POSSIBLE issues here though - - -.

Both are perhaps long shots - but mabee worth investigating.

The transformer primary APPEARS to be saturating.
Primary current on an unloaded transformer CAN, in some cases, excede
full load current. Possibly the transformer requires MORE load on the
secondary than it is getting.

An example is a microwave oven transformer. With no load on either the
high voltage or low voltage secondary, the primary will generally
saturate and overheat. If you remove the high voltage secondary and
add your own windings to make a "custom" transformer, it is not
uncommon for the primary to saturate at no/low loads - overheating the
transformer.

Like I said - a long shot, but possibly worth investigating. adding a
small 24 volt pilot light across the secondary MAY solve that kind of
a problem.

The other POSSIBILITY is a DC bias on the primary, which WILL cause
saturation on 1/2 cycle of the AC. Need a scope to check that
effectively - or something like a 10uf nonpolarized capacitor and a
100K ohm 1/2 watt resistor in series across the primary, with a DC
voltmeter connected across the cap. Make all connections BEFORE
turning on the mains power. You should expect to see readings of +/-
approx 25-35mv across the cap in a normal residential situation.

Lets say you read 275mv DC on the line, and the transformer primary
resistance is 2 ohms.. That will put a DC current of 137.5ma through
the primary - which when added to the normal AC current on the one
half cycle will greatly excede the saturation current of the primary.

A half wave rectified load on the same circuit could put a DC
component across the line.

An AC (nonpolarized) capacitor IN SERIES with the primary would
remove the DC component from the primary winding, but finding a
capacitor that would ballast the primary properly (allow full rated
primary current) while not causing a series resonance (which would
appear as a short circuit across the mains) is not something I would
try to calculate.
>
>..... Phil
>
>
>
>
>
>

Phil Allison

unread,
Apr 10, 2011, 11:25:02 PM4/10/11
to

<cl...@snyder.on.ca>
"Phil Allison"

>
>
>>High voltage spikes on the primary could also cause insulation failure
>>leading to the damage seen in the pics - lightning does this sort of
>>thing.
>>So also could back emfs from the blower fan if the is a bad connection in
>>the AC supply feed.
>
>>I suggest you provide the next replacement for that vulnerable tranny with
>>some "protection" - firstly an in-line fuse of say 1/4 amp AND a
>>capacitor wired across the primary of say 1uF rated for continuous use
>>across the AC supply.
>>
>>If there is an overload on the tranny, the fuse will blow.
>>
>>The 1uF capacitor should suppress spike voltages enough to save the tranny
>>from harm.
>>
>>
>
> A capacitor across an AC supply??????????????? As a surge
> protector????
> Have not heard of that before.


** Then your ignorance is showing.

The stated reason for the capacitor was in relation to the "blower fan"
inside the same unit as the small tranny.

The event the cap has to deal with is a back emf surge generated by that
fan when the AC supply is suddenly disconnected - for whatever reason.


> A capacitor across the AC line would appear as a load

** Draws 45mA continuously.

Yawnnnnnnn....

> - and could form a resonant l/ci tank circuit,

** Yawnnnnnn....

(snip absurd drivel)


> There are 2 other POSSIBLE issues here though - - -.
>
> Both are perhaps long shots - but mabee worth investigating.
>
> The transformer primary APPEARS to be saturating.


** The primary appears to be EXPLODING !!

You ridiculous wanker.


.... Phil


Jeff Thies

unread,
Apr 10, 2011, 11:27:02 PM4/10/11
to
On 4/10/2011 10:46 PM, Phil Allison wrote:
> "Jeff Thies"
> Phil Allison wrote:
>
>>>>> I suggest you provide the next replacement for that vulnerable tranny
>>>>> with
>>>>> some "protection" - firstly an in-line fuse of say 1/4 amp AND a
>>>>> capacitor wired across the primary of say 1uF rated for continuous use
>>>>> across the AC supply.
>>>>>
>>>>> If there is an overload on the tranny, the fuse will blow.
>>>>>
>>>>> The 1uF capacitor should suppress spike voltages enough to save the
>>>>> tranny from harm.
>>>>
>>>> 1 uF sounds a little high.
>>>
>>>
>>> ** No it ain't.
>>
>> I thought you had simply misspoke and that this was an honest error.
>>
>> The reactance of a 1uF cap at 60Hz is: 2652 ohms (1/(2*pi*F*C)
>> Online calculator:
>> http://www.kusashi.com/reactance-c.php?f=60&c=1&stage=results
>>
>> V^2/R = W
>>
>> Assuming primary, as why would you put it on the secondary:
>>
>> 120^2 / 2652 = 5.43 W
>>
>> Does that not seem wrong to you?
>
> ** Completely.
>
> Such a cap dissipates no energy at all.
>
> What planet do you come from ??

What a jerk you are. You do realize that is 45mA running through that.

Jeff

Phil Allison

unread,
Apr 10, 2011, 11:29:46 PM4/10/11
to


** What sort of know nothing JERK thinks that capacitors dissipate energy
??

Then proceeds to calculate the reactive impedance and treat it the same as
resistance ??

Big bad.

Zero out of ten.


.... Phil


cl...@snyder.on.ca

unread,
Apr 11, 2011, 1:33:11 AM4/11/11
to
On Mon, 11 Apr 2011 13:25:02 +1000, "Phil Allison" <phi...@tpg.com.au>
wrote:

>

Phil - your mamma should wash your mouth out with soap.
I SAID the other two scenarios were long shots - but so is everything
else that has been suggested. The windings of the trasnformer do not
APPEAR to be overheated - looks like just blackened at the connections
between the winding and the connecting wires.

It is definitely a strange failure - and I don't think it has been
properly analyzed to determine exactly what/where the problem is.
As Arthur Conan Doyle said, "after you have eliminated all the
possibilities, whatever remains, however improbable, must be the
truth"

Phil Allison

unread,
Apr 11, 2011, 1:48:43 AM4/11/11
to

<cl...@snyder.on.ca>

>>** The primary appears to be EXPLODING !!
>>
>> You ridiculous wanker.
>>

> Phil - your mamma should wash your mouth out with soap.


** You need to get your hands off of it.


> I SAID the other two scenarios were long shots ..

** And I said they were ridiculous drivel.

> - but so is everything else that has been suggested.

** Your opinion is based on your ignorance only.

> The windings of the trasnformer do not
> APPEAR to be overheated - looks like just blackened at the connections
> between the winding and the connecting wires.


** See the vaporised metal coating deposited on the plastic cover next to
the tranny?

That is a damn EXPLOSION !!

It happened very suddenly and made a loud bang too.

I said:

" High voltage spikes on the primary could also cause insulation failure
leading to the damage seen in the pics - lightning does this sort of thing.
So also could back emfs from the blower fan if the is a bad connection in
the AC supply feed."

If the insulation on the enamel wire of the primary is punctured by a HIGH
VOLTAGE SPIKE, effectively shorting out most of the primary - then the
120 AC supply ( no fuse exists remember ) will easily turn the two exposed
wire ends into metal vapour !!!

Cos they just became the fuses.

It just so happens that many small transformers made in China, India & Sri
Lanka etc are very prone to this sort of failure - due to bad manufacturing
practices.

.... Phil


Stormin Mormon

unread,
Apr 11, 2011, 8:15:46 AM4/11/11
to
What is this "old light bulb trick"?

--
Christopher A. Young
Learn more about Jesus
www.lds.org
.


"Ron D." <ron.d...@gmail.com> wrote in message
news:874969f6-f77c-41f5...@k9g2000yqi.googlegroups.com...

Phil Allison

unread,
Apr 11, 2011, 9:14:48 AM4/11/11
to

"Stormin Mormon"

>
> What is this "old light bulb trick"?


** You need to ask Agent 86 that one....


... Phil


Jeff Thies

unread,
Apr 11, 2011, 11:12:17 AM4/11/11
to

It doesn't have to dissipate energy to explode. That is a lot of joules.


>
> Then proceeds to calculate the reactive impedance and treat it the same as
> resistance ??
>
> Big bad.

So I left out the PF, so what?
>
> Zero out of ten.

My first encounter with you, but no doubt others have had the same
reaction. Telling an inexperienced OP to put in a 1mF cap without any
specifics, and that for the dubious goal of "suppressing" the back EMF
of a motor running off the line has a disconnect from reality.
Transformers are generally much hardier devices than the solid state
components they feed.

Say what you want. You've already made your mark and I really don't
care what you add. I've taken away what I need to know about yourself,
and the OP is off somewhere else.

Happy now?

Jeff
>
>
> .... Phil
>
>
>
>

Smitty Two

unread,
Apr 11, 2011, 11:25:45 AM4/11/11
to
In article <inv5oj$nkt$1...@news.albasani.net>,
Jeff Thies <jeff_...@att.net> wrote:

> My first encounter with you

Jeff, since you don't hang out on s.e.r., I'll clue you in about Phil.
Regardless of what he knows or doesn't know, he calls everyone a
fuckwit, and worse, all the time. Virtually every sentence of every post
is chock full of derision. If you choose to dialogue with him, you'll
get great heaping portions of his demented abuse. So don't take it
personally.

Jeff Thies

unread,
Apr 11, 2011, 12:57:57 PM4/11/11
to

Thanks. That's about what I surmised. A North American version of Rod
Speed.

Some people I'll read consistently, I'll put him in the other group.

Jeff

David Nebenzahl

unread,
Apr 11, 2011, 2:30:33 PM4/11/11
to
On 4/11/2011 8:25 AM Smitty Two spake thus:

That's true, and it's also true that he often acts--seriously, no
joke--like he's off his meds.

But it's also true that he knows more about electronics than probably
all the "experts" here (a.h.r.) put together. And it is possible to
engage him in serious dialog on electronic topics, as shown by recent
threads in s.e.r. So while he can be abusive, you ought to listen to his
technical advice, which is golden.


--
The current state of literacy in our advanced civilization:

yo
wassup
nuttin
wan2 hang
k
where
here
k
l8tr
by

- from Usenet (what's *that*?)

Meat Plow

unread,
Apr 11, 2011, 2:46:49 PM4/11/11
to
On Mon, 11 Apr 2011 11:30:33 -0700, David Nebenzahl wrote:

> On 4/11/2011 8:25 AM Smitty Two spake thus:
>
>> In article <inv5oj$nkt$1...@news.albasani.net>,
>> Jeff Thies <jeff_...@att.net> wrote:
>>
>>> My first encounter with you
>>
>> Jeff, since you don't hang out on s.e.r., I'll clue you in about Phil.
>> Regardless of what he knows or doesn't know, he calls everyone a
>> fuckwit, and worse, all the time. Virtually every sentence of every
>> post is chock full of derision. If you choose to dialogue with him,
>> you'll get great heaping portions of his demented abuse. So don't take
>> it personally.
>
> That's true, and it's also true that he often acts--seriously, no
> joke--like he's off his meds.
>
> But it's also true that he knows more about electronics than probably
> all the "experts" here (a.h.r.) put together.

That's debatable. Phil is a troll with decent knowledge in electronics.
He sets traps like a spider spins a web then comes back to pounce on
those in the trap. I think Phil gets a lot of his knowledge on Google.

--
Live Fast Die Young, Leave A Pretty Corpse

The Daring Dufas

unread,
Apr 11, 2011, 3:08:47 PM4/11/11
to

Google? Is that some new kind of mind altering drug? ^_^

TDD

Meat Plow

unread,
Apr 11, 2011, 3:18:52 PM4/11/11
to

Sometimes it seems that's the case.

Jeff Thies

unread,
Apr 11, 2011, 3:28:09 PM4/11/11
to
On 4/11/2011 2:46 PM, Meat Plow wrote:
> On Mon, 11 Apr 2011 11:30:33 -0700, David Nebenzahl wrote:
>
>> On 4/11/2011 8:25 AM Smitty Two spake thus:
>>
>>> In article<inv5oj$nkt$1...@news.albasani.net>,
>>> Jeff Thies<jeff_...@att.net> wrote:
>>>
>>>> My first encounter with you
>>>
>>> Jeff, since you don't hang out on s.e.r., I'll clue you in about Phil.
>>> Regardless of what he knows or doesn't know, he calls everyone a
>>> fuckwit, and worse, all the time. Virtually every sentence of every
>>> post is chock full of derision. If you choose to dialogue with him,
>>> you'll get great heaping portions of his demented abuse. So don't take
>>> it personally.
>>
>> That's true, and it's also true that he often acts--seriously, no
>> joke--like he's off his meds.
>>
>> But it's also true that he knows more about electronics than probably
>> all the "experts" here (a.h.r.) put together.
>
> That's debatable. Phil is a troll

he does have the outward symptoms.

with decent knowledge in electronics.
> He sets traps like a spider spins a web then comes back to pounce on
> those in the trap.

I have no time for playing games. Anyone sending me down that path is
not worth the time consumed.

I think he is more than a little looney about the whole 1mF cap
business. Not that I'm an expert on appliances but there are plenty of
good reasons why such a thing is not used in other commercial and home
electronics. I've been inside thousands of different line driven
electronics and I've never seen anyone reflect his thinking. It makes me
doubt his practical background, and that is an understatement.


I think Phil gets a lot of his knowledge on Google.

Not a clue on that. I'll leave him in your group where he can foster
as much contempt as he desires.

Jeff
>
>
>

Meat Plow

unread,
Apr 11, 2011, 3:53:59 PM4/11/11
to

The sad part is Phil usually is right with things involving pro-audio
gear. I would take his word hands down on guitar and power amps. But you
cannot know everything about any kind of electronic device like he would
like others to believe he does. And why he traps people with one decent
reply then comes back scathing is beyond me.

John Robertson

unread,
Apr 11, 2011, 3:58:45 PM4/11/11
to
Stormin Mormon wrote:
> What is this "old light bulb trick"?
>

Wire a 25W - 75W incandescent light bulb in series with the load
(transformer primary). If the light is bright then there is a short in
either the primary or secondary side output. If it is bright for a
second then dims right down (power supply caps charging up) then there
is no serious short.

Picture of one such setup:

http://www.flippers.com/images/OvercurrentTestJigFused.JPG

Note I recommend an isolation transformer. The fuse holder is there so
the light can act as an indicator that the fuse has blown, or remove the
fuse for "the light bulb test/trick".

John :-#)#

--
(Please post followups or tech enquiries to the newsgroup)
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
Call (604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

Phil Allison

unread,
Apr 11, 2011, 7:02:18 PM4/11/11
to

"Jeff Thies"
Phil Allison

>
>> ** What sort of know nothing JERK thinks that capacitors dissipate
>> energy
>> ??
>>
>> Then proceeds to calculate the reactive impedance and treat it the same
>> as
>> resistance ??
>>
>> Big bad.
>
>> Zero out of ten.


** Jeff - you are much worse than merely a know nothing idiot.

You are one colossally narcissistic fool and a public menace.

God knows what it is you DO know something about - but certainly
electricity and electronics are not among them.

Piss off and stop TROLLING .


.... Phil

Phil Allison

unread,
Apr 11, 2011, 7:04:21 PM4/11/11
to

"Smitty Shitty Two"

> Jeff Thies


>
>> My first encounter with you
>
> Jeff, since you don't hang out on s.e.r., I'll clue you in about Phil.


**As if you have the metal capacity for any such thing.

> Regardless of what he knows or doesn't know, he calls everyone a
> fuckwit, and worse, all the time.

** Blatant lies.

I only call fuckwits fuckwits.

Shame is, the whole of usenet has been taken over by them.


.... Phil


Phil Allison

unread,
Apr 11, 2011, 7:10:30 PM4/11/11
to

"Jeff Thies is a Moron "

** Piss off - you fuckwit WANKER !!!


> I think he is more than a little looney about the whole 1mF cap business.

** FFS IMBECILE -

I suggested a capacitor of " 1uF rated for continuous use across the AC
supply"

Standard practice with inductive loads for decades.


> Not that I'm an expert on appliances ...


** Only thing a fool like YOU is expert on is puling your tiny cock in
public.


FUCK OFF - DAMN TROLL !!


Don Klipstein

unread,
Apr 11, 2011, 7:26:45 PM4/11/11
to
In <intmki$cmm$1...@news.albasani.net>, Jeff Thies wrote:

<SNIP to getting a nonpolarized 1uF cap suitable for 120 VAC>

>If not then go buy a 200V non polarized 1uF cap. It is no easy chore.
>

>http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=131081

200V DC rating is not enough, despite 120 VAC having a peak voltage
of only 169 volts.

AC does make things worse, by rapidly repeated severe voltage swings
combined with some development of heat.

The capacitor needs to have an actual AC rating, and one that includes
your AC voltage with a comfortable safety margin. If it has a DC rating,
chances are that will be around 400 volts DC. And not every cap rated
400 VDC is dsafe for use with 120 VAC, even if non-polarized.

Preferably, it should be "UL recognized" or the like, for some assurance
that it is reasonably reliable against failure, or at least an unsafe
failure.

Back in the early 1980's, in an experimental sodium lamp ballast, I have
blown an 800VDC cap and two 600 VDC ones with 240-260 VAC 60 Hz with less
than 10 operating hours of this combined among the three of them. One of
those capacitor blowups was a spectacular one that left a major oil stain
on the ceiling above. I learned the hard way that actual AC ratings are
required here.

===================================

One more thing: If there is a switch upstream of a capacitor across a
power line, then the switch may be in for severe contact pitting.

Then again, I doubt a voltage spike lasting long enough to burn out a
transformer will be absorbed by a 1 uF cap across a 120V AC line.

I would lean to looking for intermittent overload by the transformer's
load, the transformer's load intermittently drawing a large amount of DC
(such as by failing-open one diode in a bridge rectifier).

Or, the transformer being connected incorrectly.

(Such as a 120V/240V one with 2 primary winding sections needing both
primary sections to be connected in parallel with each other for full
power handling at 120V, but only 1 of them is being used.)

--
- Don Klipstein (d...@misty.com)

Phil Allison

unread,
Apr 11, 2011, 7:34:52 PM4/11/11
to

"Don Klipstein"

> Then again, I doubt a voltage spike lasting long enough to burn out a
> transformer will be absorbed by a 1 uF cap across a 120V AC line.


** You missed the point entirely.

A voltage spike ( or a series of them) can easily cause insulation failure
in the enamel winding wire of the primary - then the energy to explode the
lead in and lead out wires comes from the 120 volt AC supply.

.... Phil


Stormin Mormon

unread,
Apr 11, 2011, 7:37:11 PM4/11/11
to
OK, that makes a bit of sense. Seems like a bit of bother
for what you can track down, with a VOM and some common
sense.

Reminds me of the time I got asked to change a thermostat.
New one needs common, old one did not. After several trips
to the roof, I got me some common sense. Still, it was a
rather long and tiring job. The first level of roof was
maybe 16 feet, and there was about a 4 foot rise to the next
level, so I ended up buying two ladders for that job. Worked
out, I've used both ladders several times since then.
Learned a lot, too.

--
Christopher A. Young
Learn more about Jesus
www.lds.org
.


"John Robertson" <sp...@flippers.com> wrote in message
news:hMCdnUAET_PowD7Q...@giganews.com...

Don Klipstein

unread,
Apr 11, 2011, 8:00:43 PM4/11/11
to
In article <90f8dc...@mid.individual.net>, Phil Allison wrote:
>
>"Jeff Thies"
> Phil Allison wrote:
>
>>>>> I suggest you provide the next replacement for that vulnerable tranny
>>>>> with
>>>>> some "protection" - firstly an in-line fuse of say 1/4 amp AND a
>>>>> capacitor wired across the primary of say 1uF rated for continuous use

>>>>> across the AC supply.
>>>>>
>>>>> If there is an overload on the tranny, the fuse will blow.
>>>>>
>>>>> The 1uF capacitor should suppress spike voltages enough to save the
>>>>> tranny from harm.
>>>>
>>>> 1 uF sounds a little high.
>>>
>>> ** No it ain't.
>>
>> I thought you had simply misspoke and that this was an honest error.
>>
>> The reactance of a 1uF cap at 60Hz is: 2652 ohms (1/(2*pi*F*C)
>> Online calculator:
>> http://www.kusashi.com/reactance-c.php?f=60&c=1&stage=results
>>
>> V^2/R = W
>>
>> Assuming primary, as why would you put it on the secondary:
>>
>> 120^2 / 2652 = 5.43 W
>>
>> Does that not seem wrong to you?
>
>** Completely.
>
>Such a cap dissipates no energy at all.
>
>What planet do you come from ??

With AC, volts times amps is not necessarily watts.

If determination of volts and amps are both "true RMS", (non-"true RMS"
meters are usually OK for line voltage and for current through a cap
across an AC line),

then volts times amps is "volt-amps". Ratio of power consumption (in
watts) to VA is power factor. Capacitors have very low, ideally zero,
power factor.

>> If not then go buy a 200V non polarized 1uF cap. It is no easy chore.
>

>** 1uF caps for use across the AC supply are cheap and plentiful.

Digi-Key has 3 easy enough to find 1uF ones with X1 or X2 rating, all
$3-plus plus shipping with minimum order requirement to avoid a surcharge.

Among those 3, my favorite is FC233820105, which is by Vishay/BC
Components. Its datasheet refers to 2 UL standards and 2 CSA ones, and has
a link to an application note making a claim that the referenced UL ones are
sufficient for achieving the "UL Recognized" ("backwards UR") mark.

>Typical examples are metallised polypropylene " class X2 " types and sell
>for $1 or $2 each.
>
>http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=399-5466-ND

That one needed me to delve into its datasheet to see that it claims to
meet the two relevant UL standards and that it has X2 rating. Its AC
voltage rating is 275 volts. 91 cents each plus shipping, with a minimum
order requirement to avoid a surcharge.

Any of the 4 capacitors above (2 mentioned specifically) look good, in
the unlikely event what is needed is a capacitor across the line.

>What planet do you come from ??

--
- Don Klipstein (d...@misty.com)

Phil Allison

unread,
Apr 11, 2011, 8:03:24 PM4/11/11
to
"Don Klipstein"


>
> Any of the 4 capacitors above (2 mentioned specifically) look good, in
> the unlikely event what is needed is a capacitor across the line.


** You are one smug, fucking PITA aren't you ?

.... Phil


Don Klipstein

unread,
Apr 11, 2011, 8:10:06 PM4/11/11
to

PF of a capacitor suitable to connect across an AC line is usually less
than .01.

>My first encounter with you, but no doubt others have had the same
>reaction.

<Make that longish story shorter>

Phil Allison is well known in sci.electronics.design to be quick to
get brash, even calling people names and sometimes a little worse. If
only he would avoid that, a lot more people would think a lot more highly
of him than they do now, since he does fairly well know what he is talking
about when it's electrical.

Don Klipstein

unread,
Apr 11, 2011, 8:12:48 PM4/11/11
to
In article <invbun$64h$1...@news.albasani.net>, Jeff Thies wrote:
>On 4/11/2011 11:25 AM, Smitty Two wrote:
>> In article<inv5oj$nkt$1...@news.albasani.net>,
>> Jeff Thies<jeff_...@att.net> wrote:
>>
>>> My first encounter with you
>>
>> Jeff, since you don't hang out on s.e.r., I'll clue you in about Phil.
>> Regardless of what he knows or doesn't know, he calls everyone a
>> fuckwit, and worse, all the time. Virtually every sentence of every post
>> is chock full of derision. If you choose to dialogue with him, you'll
>> get great heaping portions of his demented abuse. So don't take it
>> personally.
>
> Thanks. That's about what I surmised. A North American version of Rod
>Speed.

I think Rod Speed is much worse. Meanwhile, it appears to me in my
experience in s.e.d. that P.A. is, like R.S., an Aussie.

Don Klipstein

unread,
Apr 11, 2011, 8:51:55 PM4/11/11
to

Is the O.P. having other things in his house blowing from voltage spikes
severe enough to blow transformer primary winding insulation?

I have seen lots of things blow from line voltage surges, but no
transformers indoors blowing when line voltage spikes blow other things.

My experience is that usual 120V-primary step-down transformers can
produce pulses of 2 kilovolts when used on pulses in reverse. In fact, I
have done that with about 8 different transformers dozens of times each,
and none of them lost their ability to do that. (I am aware of line
voltage spikes being noted to get even higher.)

There is also the issue of line voltage surges not easily being loaded
down by capacitors for whatever reason. I have experience with them
blowing things that had capacitors, including a CFL that had probably a 22
or 47 uF capacitor across the output of its internal bridge rectifier as
they usually do. No transformer failures in the same house from the same
event, though there were electronics failures.

The usual solutions to absorb line voltage spikes are MOVs and other
devices that absorb voltage surges by becoming conductive in response to
excessive voltage.

Phil Allison

unread,
Apr 11, 2011, 9:24:33 PM4/11/11
to

"Don Klipstein"

Phil Allison wrote:
>"Don Klipstein"
>>
>>> Then again, I doubt a voltage spike lasting long enough to burn out a
>>> transformer will be absorbed by a 1 uF cap across a 120V AC line.
>>
>>** You missed the point entirely.
>>
>>A voltage spike ( or a series of them) can easily cause insulation failure
>>in the enamel winding wire of the primary - then the energy to explode
>>the lead in and lead out wires comes from the 120 volt AC supply.
>
> Is the O.P. having other things in his house blowing from voltage spikes
> severe enough to blow transformer primary winding insulation?


** Totally irrelevant.

The furnace unit and the transformer in question are all we are discussing.

You have clearly not bothered to read my first or my other posts in this
thread.

Eg:

" High voltage spikes on the primary could also cause insulation failure
leading to the damage seen in the pics - lightning does this sort of thing.
So also could back emfs from the blower fan if the is a bad connection in
the AC supply feed."

Whenever AC power to that furnace is disconnected, the blower fan will
deliver a back emf spike - meanwhile that poor, little tranny is wired in
parallel with it. Other devices in the house are NOT involved.

Most AC supply transformers can tolerate repeated 2kV spikes on the primary
till the cows come home - but a badly wound one cannot. This is a specific
and fairly recent problem with small transformers made in China and
elsewhere in Asia where the makers are not fully aware of the issue of
insulation failure in the enamel windings.

Once there is a layer to layer insulation failure ( between adjacent wires)
in the primary of that tranny - the AC current draw will jump up to many
amps and may cause the feed in and out wires to explode - as seen in the
pics.

If the spike voltage is suppressed, the tranny will likely survive.

A 1uF cap provides a low impedance path for such a back emf spike, virtually
shorting it out.

OTOH a varistor provides no conduction path until its breakdown voltage is
exceeded, but is also a means of suppressing the spike voltage to a safe
value.


.... Phil


John Robertson

unread,
Apr 11, 2011, 9:31:58 PM4/11/11
to
Tony Miklos wrote:
> On 4/9/2011 10:06 PM, John Robertson wrote:
>> Steve Turner wrote:
>>> On 4/9/2011 7:29 PM, Jim Yanik wrote:
>>>> "robb"<so...@where.on.net> wrote in
>>>> news:WfKdnQOvu_d2bz3Q...@earthlink.com:
>>>>
>>>>> "Steve Turner" wrote in message news:inoakf$qin$1...@dont-email.me...
>>>>>
>>>>>> A continuation of the "Why does the 115V->24V transformer keep
>>>>>> blowing on my Trane XB80?" discussion I started on 04/02/2011.
>>>>>>
>>>>>> Yep, My A/C unit blew another transformer. Pictures (and wiring
>>>>>> diagram) here:
>>>>>>
>>>>>> http://www.flickr.com/photos/bbqboyee/sets/72157626457562742/
>>>>>>
>>>>>> Trane's manual for the unit is here (for perhaps better viewing of
>>>>>> the wiring diagrams that I also copied to my above flickr site as
>>>>>> jpg images):
>>>>>>
>>>>>>
>>>>>>
>>>>> http://www.trane.com/webcache/un/furnaces%20%28furn%29/product/22-1666-
>>>>>
>>>>> 07_04012009.pdf
>>>>>>
>>>>>> As you can probably see in the pictures, there is visible charring
>>>>>> of the 115V leads going into the transformer, and of course the 115V
>>>>>> circuit is open (again). If you didn't see my first thread, this is
>>>>>> the third transformer the unit has blown. In the previous
>>>>>> discussion, it was discussed that perhaps the first one just blew
>>>>>> because of old age (6 years), and the second blew because it wasn't
>>>>>> a proper replacement (poor quality, made in China, etc.). This
>>>>>> third unit is most certainly a proper replacement, and it's most
>>>>>> certainly indicative of a real problem I have somewhere else in the
>>>>>> unit. I didn't see any such charring on the previous two units, at
>>>>>> least not like this.
>>>>>>
>>>>>> I never got a real chance to test out the system after installing
>>>>>> this third transformer. We had cool weather for several days, and I
>>>>>> never tried to force the system to come on so I could monitor it;
>>>>>> that was probably a mistake. Unfortunately, I was also absent from
>>>>>> the premises during the extended times when the unit was most likely
>>>>>> operational, so that didn't help either. However, my family tells me
>>>>>> that it WAS working and cooling the house rather nicely, for at
>>>>>> least a day, perhaps two. I'm getting 115V in all the right places,
>>>>>> so it doesn't look like an over-voltage condition to me. Perhaps
>>>>>> it's an overheating condition? It looks to me like the only real
>>>>>> load on this circuit is the blower motor; could the motor be causing
>>>>>> this? The blower spins freely when I turn it by hand. Start
>>>>>> capacitor on the motor maybe? Relay on the control board perhaps?
>>>>
>>>> a bad start cap would not affect the 24v control transformer,that just
>>>> powers the controller board and relays.
>>>> either something is loading the transformer or a problem with input
>>>> voltage.
>>>>>>
>>>>>
>>>>> Just a thought,
>>>>>
>>>>> When you put the new transformer in does the "Diagnostic Light" LED
>>>>> turn on ?
>>>>> Does the LED blink or flash in one of the patterns indicated in the
>>>>> "diagnostic codes" section that you posted a link to ?
>>>>>
>>>>> The "diagnostic codes" imply that the control board can detect
>>>>> several of the problems mentioned by others.
>>>>> Maybe a good starting point as there is not enough time to test with
>>>>> test equipment.
>>>>>
>>>>> robb
>>>>>
>>>>>
>>>>
>>>> a line voltage monitor may be needed to see if there's some short-term
>>>> overvoltage applied to the transformer,and a oscilloscope may
>>>> determine if
>>>> the input waveform is sinusoidal or otherwise,because 60hz iron core
>>>> tranformers don't like extreme distortion on their input,it gets
>>>> converted
>>>> to heat,not output voltage. a DMM will not show line distortion or
>>>> short
>>>> term overvoltages.
>>>> are the wires burned close to the transformer,or over their entire
>>>> length?
>>>
>>> I peeled the insulation back from the hot and neutral wires and the
>>> burning only occurred right at the terminals, no more than 1/4" into
>>> the wiring.
>>>
>>
>> So the problem is the joint of the transformer wiring to the lugs. They
>> may have cold solder joints, or the lug may be slightly loose - enough
>> to crack the solder connection over time. Your transformer may still be
>> good!
>>
>> When you are examining the transformer wire make sure it is stripped
>> enough to get good solder coverage. Burn off the insulation (enamel) -
>> do not sand, scrape, or try to erode it because if you scratch the wire
>> it will break at the scratch.
>>
>> Re-solder the transformer connectors after first making sure the wire
>> has at least two complete wraps around the solder lug and the solder
>> makes a good meniscus joint to the wire and lug. Also use a good grade
>> of solder, not regular 60/40, find some 63/37 solder (Kester is best) -
>> this solder resists cold solder joints due to movement when cooling by
>> solidifying very fast.
>>
>> John :-#)#
>
> Hi John, fancy meeting you here! Wait, I didn't realize this is cross
> posted to sci.electronics.repair Now I know why you are here. Glad you
> liked the package.
>
> Tony

Hi Tony,

Yes, we enjoyed the peanut brittle...

I do not think the original poster noticed my suggestion above, that
failure mode is really not going to be too much load or some weird AC
situation - it's either a short to the metal case the transformer is
mounted in or a problem with the wire to solder lug joint.

Would be nice to know if that idea was checked out...

cl...@snyder.on.ca

unread,
Apr 11, 2011, 9:46:25 PM4/11/11
to
On Mon, 11 Apr 2011 15:48:43 +1000, "Phil Allison" <phi...@tpg.com.au>
wrote:

>
><cl...@snyder.on.ca>
>
>>>** The primary appears to be EXPLODING !!
>>>
>>> You ridiculous wanker.
>>>
>> Phil - your mamma should wash your mouth out with soap.
>
>
>** You need to get your hands off of it.
>
>
>> I SAID the other two scenarios were long shots ..
>
>** And I said they were ridiculous drivel.
>
>> - but so is everything else that has been suggested.
>
>** Your opinion is based on your ignorance only.
>
>> The windings of the trasnformer do not
>> APPEAR to be overheated - looks like just blackened at the connections
>> between the winding and the connecting wires.
>
>
>** See the vaporised metal coating deposited on the plastic cover next to
>the tranny?
>
>That is a damn EXPLOSION !!
>
>It happened very suddenly and made a loud bang too.


>
>I said:
>
>" High voltage spikes on the primary could also cause insulation failure
>leading to the damage seen in the pics - lightning does this sort of thing.
>So also could back emfs from the blower fan if the is a bad connection in
>the AC supply feed."
>

>If the insulation on the enamel wire of the primary is punctured by a HIGH
>VOLTAGE SPIKE, effectively shorting out most of the primary - then the
>120 AC supply ( no fuse exists remember ) will easily turn the two exposed
>wire ends into metal vapour !!!
>
>Cos they just became the fuses.
>
>It just so happens that many small transformers made in China, India & Sri
>Lanka etc are very prone to this sort of failure - due to bad manufacturing
>practices.
>
>
>
>.... Phil
>
>
So you are agreeing with what I said before - most likely problem is
cheap crap component.

We can be relatively sure it wasn't lightning 3 times.
If the back emf from the motor is causing the problem, the bad
connection to the motor should have made itself VERY evident by now.
Not saying it is NOT part of the problem - like you, I wasn't there
watching it fail - and neither you nor I heard the noise you speak of.

And IF the problem is what you say it is (and I'm not saying it is
not), then probing the old primary and getting to the winding beyond
where the solder joint "exploded" you should be able to measure a
significantly lower than normal primary resistance.

That measurement has not, as far as I know, been made and reported.

If it was mine, or if I had the transformer at hand, I'd have it apart
and analysed in no time. If the primary is shorted, I'd know, within
an hour or two of the failure..
And with that second primary, it would not be hard to determine if the
110 volt primary is shorted without even dissassembling the
transformer..

At this point no-one has actually posted FOR SURE what the primary
configuration is. Is it a mult-tapped primary, set for 115, 208, and
230 volts, or is it 2 independent primaries, or is it 2 primaries that
need to be connected either in series or parallel depending on the
voltage (115 or 230)

I don't know this, and you don't know either unless you are
clairvoyant, because it has not, to this point, been established and
reported.

So we are all guilty of the same thing - making ASSumptions.
>

cl...@snyder.on.ca

unread,
Apr 11, 2011, 9:57:58 PM4/11/11
to

I have to agree with Don. DC power component on secondary OR primary,
or not connected right are the 2 most likely problems after "cheap
crap component"

David Nebenzahl

unread,
Apr 11, 2011, 9:59:57 PM4/11/11
to
On 4/11/2011 4:37 PM Stormin Mormon spake thus:

> OK, that makes a bit of sense. Seems like a bit of bother
> for what you can track down, with a VOM and some common
> sense.

As usual, being an ignoramus, you completely miss the point.

The light bulb not only allows you to detect if there's a short causing
too high a current draw, but it also *saves the device under test* from
releasing the magic smoke if there is such a smoke.

If you use some other method of measuring current--break circuit and
insert ammeter, or use clamp-on-meter, the meter will tell you there's a
short, all right, at about the same time the transformer erupts in a
ball of smoke and fire.

cl...@snyder.on.ca

unread,
Apr 11, 2011, 10:01:51 PM4/11/11
to

That would go a long ways to explaining his potty mouth.

Phil Allison

unread,
Apr 11, 2011, 10:22:09 PM4/11/11
to

<cl...@snyder.on.ca


> If the back emf from the motor is causing the problem, the bad
> connection to the motor should have made itself VERY evident by now.

** Not true at all.


> If it was mine, or if I had the transformer at hand, I'd have it apart
> and analysed in no time. If the primary is shorted, I'd know, within
> an hour or two of the failure..


** What is clear is that the primary has suddenly drawn a very large current
and that means there are SHORTED turns inside the primary.

Go away wanker.


..... Phil


Phil Allison

unread,
Apr 11, 2011, 10:23:56 PM4/11/11
to

<cl...@snyder.on.ca>

>
> I have to agree with Don. DC power component on secondary OR primary,

** You are both wrong.

> or not connected right are the 2 most likely problems after "cheap
> crap component"

** Which is no help to the OP.

.... Phil


Jeff Thies

unread,
Apr 11, 2011, 11:07:50 PM4/11/11
to

Perhaps that goes all the way. Hey, I learned something about Oz today.
Funny the guys from NZ I know seem quite nice. I suppose it was a surly
lot they are descended from... or is it the climate? No matter.

Jeff

Don Klipstein

unread,
Apr 12, 2011, 12:57:40 AM4/12/11
to
In article <90hnur...@mid.individual.net>, Phil Allison wrote:
>
>"Don Klipstein"
> Phil Allison wrote:
>>"Don Klipstein"
>>>
>>>> Then again, I doubt a voltage spike lasting long enough to burn out a
>>>> transformer will be absorbed by a 1 uF cap across a 120V AC line.
>>>
>>>** You missed the point entirely.
>>>
>>>A voltage spike ( or a series of them) can easily cause insulation failure
>>>in the enamel winding wire of the primary - then the energy to explode
>>>the lead in and lead out wires comes from the 120 volt AC supply.
>>
>> Is the O.P. having other things in his house blowing from voltage spikes
>> severe enough to blow transformer primary winding insulation?
>
>** Totally irrelevant.
>
>The furnace unit and the transformer in question are all we are discussing.
>
>You have clearly not bothered to read my first or my other posts in this
>thread.
>
>Eg:
>
>" High voltage spikes on the primary could also cause insulation failure
>leading to the damage seen in the pics - lightning does this sort of thing.

Why would that be, in your words, "totally irrelevant" to whether or not
the O.P. had any electrical/electronic failures elsewhere in his home
attributable to voltage surges?

As opposed to 3 transformers blowing in the same appliance with nothing
else anywhere in the home running into trouble from voltage surges?

While I have experienced roughly 8 transformers surviving repeated
abusive pulse-in-reverse-direction developing about 2 KV across the
primary without any degradation against ability to do so?

>So also could back emfs from the blower fan if the is a bad connection in
>the AC supply feed."
>
>Whenever AC power to that furnace is disconnected, the blower fan will
>deliver a back emf spike - meanwhile that poor, little tranny is wired in
>parallel with it. Other devices in the house are NOT involved.

Why would the solution be deploying a capacitor, rather than repairing
the poor connection or deploying a voltage-dependent spike-absorbing
device such as an MOV?

>Most AC supply transformers can tolerate repeated 2kV spikes on the primary
>till the cows come home - but a badly wound one cannot. This is a specific
>and fairly recent problem with small transformers made in China and
>elsewhere in Asia where the makers are not fully aware of the issue of
>insulation failure in the enamel windings.

So, why should we hear about problems about that from only one customer
blowing 3 of them?

>Once there is a layer to layer insulation failure ( between adjacent wires)
>in the primary of that tranny - the AC current draw will jump up to many
>amps and may cause the feed in and out wires to explode - as seen in the
>pics.
>
>If the spike voltage is suppressed, the tranny will likely survive.
>
>A 1uF cap provides a low impedance path for such a back emf spike, virtually
>shorting it out.

Why should 1 uF protect the trannies in question while 22-47 uF fails to
protect a compact fluorescent lamp from a line voltage surge that blows
even other electronics in 2 houses but did not blow any in-home trannies?

>OTOH a varistor provides no conduction path until its breakdown voltage is
>exceeded, but is also a means of suppressing the spike voltage to a safe
>value.

Certainly protects against applying for even a microsecond more than
roughly 250-300 volts across transformer primaries, where it appears to me
that we agree that transformer primaries usually survive 2,000 volts pulse
voltage?

I seem to think that the trannies are probably blowing from either
improper wiring (connecting only 1 of the 2 primary winding sections
possibly noted for 120V usage), or from secondary load malfunction
including bridge rectifier failure in manner of a diode "failing open".

Or, extreme-oddball trouble such as sticking a magnet to the tranny.
But that's grasping-at-straws, like line voltage irregularities that blow
3 trannies in 1 piece of equipment but draw no other complaints such as
blowing of electronics downstream of the tranny in question, or elsewhere
in the house where one appliance blew 3 trannies.
--
- Don (d...@misty.com)

Phil Allison

unread,
Apr 12, 2011, 1:14:41 AM4/12/11
to

"Don Klipstein the Troll"

> Why would that be, in your words, "totally irrelevant"

** You posted under my words and it was not relevant to them.


> Why would the solution be deploying a capacitor, rather than repairing
> the poor connection or deploying a voltage-dependent spike-absorbing
> device such as an MOV?

** A suitable MOV is also a possible solution.

You still have not read all my posts in this thread.


> So, why should we hear about problems about that from only one customer
> blowing 3 of them?

** Stop making silly stuff up - you pathetic wanker.


> Why should 1 uF protect the trannies in question ...

** Already answered.

( snip bullshit story)

Keep on ignoring the evidence in the photos and you will never get close to
what sort of failure has occurred.

Facts are just like water off a ducks back to you - Don.

Co you fools like you believe whatever they fucking feel like.

.... Phil


Don Klipstein

unread,
Apr 12, 2011, 1:29:53 AM4/12/11
to

Please add helpfulness, as opposed to adding nothing but saying who
is wrong.

I chimed in explaining stuff every time. At this point that I respond
to, you are doing nothing but claiming who is wrong.

It appears to me, posting nothing but who is wrong is even lower than
my grasping-at-straws bit of any magnets on the transformer in question.

Don Klipstein

unread,
Apr 12, 2011, 1:51:20 AM4/12/11
to
In article <90i5eb...@mid.individual.net>, Phil Allison wrote:
>
>"Don Klipstein the Troll"
>
>> Why would that be, in your words, "totally irrelevant"
>
>** You posted under my words and it was not relevant to them.

You snipped the relevance.

>> Why would the solution be deploying a capacitor, rather than repairing
>> the poor connection or deploying a voltage-dependent spike-absorbing
>> device such as an MOV?
>
>** A suitable MOV is also a possible solution.

So, why should I get merely "also a possible solution" after part
numbers for capacitor part numbers were digged up?

> You still have not read all my posts in this thread.

I have endured doing that, though I prefer to not waste time at
contaminating Usenet with wasting time to responding to even the posts of
yours that I don't find deserving a response.

>> So, why should we hear about problems about that from only one customer
>> blowing 3 of them?
>
>** Stop making silly stuff up - you pathetic wanker.

Do you claim that the transformer customer blew fewer than 3? Do you
claim that the disappointed transformer customer is not alone?

If so, post support to your claims!

>> Why should 1 uF protect the trannies in question ...
>
>** Already answered.
>
>( snip bullshit story)
>
>Keep on ignoring the evidence in the photos and you will never get close to
>what sort of failure has occurred.

However, it was already explained to me clearly that the photo'ed tranny
showed charring evidence of severe primary overcurrent. What do you
expect to gain from me, or against me, if I take the time to look at the
photos that this thread gave me graphic descriptions of?

>Facts are just like water off a ducks back to you - Don.

>Co you fools like you believe whatever they fucking feel like.

--
- Don (d...@misty.com)

Phil Allison

unread,
Apr 12, 2011, 3:37:15 AM4/12/11
to

"Don Klipstein = one Looney Troll"

>>> Why would that be, in your words, "totally irrelevant"
>>
>>** You posted under my words and it was not relevant to them.
>
> You snipped the relevance.

** Blatant lie #1.


>>> Why would the solution be deploying a capacitor, rather than repairing
>>> the poor connection or deploying a voltage-dependent spike-absorbing
>>> device such as an MOV?
>>
>>** A suitable MOV is also a possible solution.
>
> So, why should I get merely "also a possible solution" after part
> numbers for capacitor part numbers were digged up?

** Wot brain dead drivel.


>> You still have not read all my posts in this thread.
>
> I have endured doing that,

** Blatant lie #2.


>>> So, why should we hear about problems about that from only one customer
>>> blowing 3 of them?
>>
>>** Stop making silly stuff up - you pathetic wanker.
>
> Do you claim that the transformer customer blew fewer than 3? Do you
> claim that the disappointed transformer customer is not alone?

** More brain dead drivel.

Don is right off with the fairies and pixies and TROLLS.


>>Keep on ignoring the evidence in the photos and you will never get close
>>to
>>what sort of failure has occurred.
>
> However, it was already explained to me clearly that the photo'ed tranny
> showed charring evidence of severe primary overcurrent.

** Nothing you have posted explains that fact at all.

Got a fucking clue how much current it takes to vaporise copper wires like
that ?

Obviously fucking not.

Facts are just like water off a duck's back to you - Don.

Cos fuckwits like you believe whatever they fucking feel like.


FOAD

- you useless, autistic, PITA turd.

.... Phil


Phil Allison

unread,
Apr 12, 2011, 3:38:13 AM4/12/11
to

"Don Klipstein"


** FOAD - you stupid, ridiculous arsehole.


Stormin Mormon

unread,
Apr 12, 2011, 5:38:03 AM4/12/11
to
There you go, again. Completely mistaken about what I'd
reccomend in the way of diagnostic. Just making up garbage,
and then trying to assign it to me. Well, you should do that
more often. Then, everyone can see what your true character
is like. Please write often, and lengthy. The more you
write, the better I look.

--
Christopher A. Young
Learn more about Jesus
www.lds.org
.


"David Nebenzahl" <nob...@but.us.chickens> wrote in message
news:4da3b1e4$0$15512$8226...@news.adtechcomputers.com...

Tony Miklos

unread,
Apr 12, 2011, 10:46:41 AM4/12/11
to
Yes, the thread got too long. If you saw the photo, and know the
primary went open on 3 transformers, it sure looks like a problem on the
primary side. If the load was too much, those wires on the secondary
side would most likely have been at the very least discolored from the heat.

Tony

Tony Miklos

unread,
Apr 12, 2011, 11:06:30 AM4/12/11
to
On 4/11/2011 7:26 PM, Don Klipstein wrote:
> In<intmki$cmm$1...@news.albasani.net>, Jeff Thies wrote:
>
> <SNIP to getting a nonpolarized 1uF cap suitable for 120 VAC>
>
>> If not then go buy a 200V non polarized 1uF cap. It is no easy chore.
>>
>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=131081
>
> 200V DC rating is not enough, despite 120 VAC having a peak voltage
> of only 169 volts.
>
> AC does make things worse, by rapidly repeated severe voltage swings
> combined with some development of heat.
>
> The capacitor needs to have an actual AC rating, and one that includes
> your AC voltage with a comfortable safety margin. If it has a DC rating,
> chances are that will be around 400 volts DC. And not every cap rated
> 400 VDC is dsafe for use with 120 VAC, even if non-polarized.
>
> Preferably, it should be "UL recognized" or the like, for some assurance
> that it is reasonably reliable against failure, or at least an unsafe
> failure.
>
> Back in the early 1980's, in an experimental sodium lamp ballast, I have
> blown an 800VDC cap and two 600 VDC ones with 240-260 VAC 60 Hz with less
> than 10 operating hours of this combined among the three of them. One of
> those capacitor blowups was a spectacular one that left a major oil stain
> on the ceiling above. I learned the hard way that actual AC ratings are
> required here.

Yes, many moons ago I was playing with adding a mylar cap in series with
the primary of a neon sign transformer. It worked great for making the
neon dimmer which was my attempt, but I recall measuring a much higher
voltage across the cap due to the reactence and impedence of them in
series. I'm guessing but I think it was up in the 400vac or so range
maybe higher. Not good for the transformer winding either. I didn't
blow them but stopped using them after reading the voltages.

cl...@snyder.on.ca

unread,
Apr 12, 2011, 12:34:28 PM4/12/11
to
On Tue, 12 Apr 2011 17:37:15 +1000, "Phil Allison" <phi...@tpg.com.au>
wrote:


>


>Got a fucking clue how much current it takes to vaporise copper wires like
>that ?
>
>Obviously fucking not.

How about "more than 15 amps"?

And it didn't blow the circuit breaker or fuse.
So the current came from somewhere other than the mains?
OK - the only other source is back emf from the motor -
Not saying it cannot be.
But why does the insulation on the transformer fail, and not the
insulation in the motor windings??
Most likely, if your scenario is correct, because there is something
wrong with the quality of the transformer windings.

And what is the solution????? Other than a transformer that can stand
up to real world abuse?
Find and repair the power connection problem that is causing the
back-emf problem before it causes other problems (like a fire from an
overheated power connection to a 1/2 HP motor)

NOT by putting a capacitor which may or may not have any effect across
the primary of the transformer. And MAY have the wrong effect.

Now, you need to locate the bad connection (if that is what it is)
Where do you start looking? It pretty well needs to be in the part of
the circuit that is always connected to both the motor and the
transformer - in other words upstream from wherever the transformer
and motor are connected (in common) to the mains, or else the back emf
would be "sunk" by the load on the mains, and the grid itself. - so
you are talking bad breaker, bad connection at the panel, bad shut-off
switch, or bad connections between the panel and the switch - not bad
relays, connections at the relays, or connections at or in the motor.

You would THINK there would be other symptoms - like the furnace
intermittently not running, noises, smoke, etc from the arcing
connection?

Again, not saying you are wrong - but you are not being very helpfull,
and you are sure being obstinate and dissagreeable

John Robertson

unread,
Apr 12, 2011, 1:40:02 PM4/12/11
to

Yes, I looked at the pictures (rather poor resolution). The failure
looks like a connection issue because there is no evidence of windings
overheating. If the output was shorted the transformer would burn up -
assuming it could sink enough current, but it would be unlikely to flare
at one spot. This is a pretty standard 115VAC transformer, I've seen
many of them...

If someone tried to pump 5000 volts into the transformer then possibly
you could get an arc like that to the transformer frame, so that would
mean a serious miss-wiring of the furnace where the igniter somehow was
connected to the primary of the transformer and then turned on.

Hard to see that happening as the transformer powers the logic board
that would enable the igniter to fire up!

Meat Plow

unread,
Apr 12, 2011, 1:54:14 PM4/12/11
to

A friend traveled to NZ then to OZ in 2004. He said when he got off the
plane in NZ he thought the pilot had landed in China because 99% of the
people around were Asian. I have relatives living near Sidney and they
said this isn't the case there.

--
Live Fast Die Young, Leave A Pretty Corpse

David Nebenzahl

unread,
Apr 12, 2011, 2:51:43 PM4/12/11
to
On 4/12/2011 2:38 AM Stormin Mormon spake thus:

> There you go, again. Completely mistaken about what I'd
> reccomend in the way of diagnostic. Just making up garbage,
> and then trying to assign it to me. Well, you should do that
> more often. Then, everyone can see what your true character
> is like. Please write often, and lengthy. The more you
> write, the better I look.

In your mind, perhaps.

Don Klipstein

unread,
Apr 12, 2011, 5:28:35 PM4/12/11
to
In <tru8q6hhouec3fkd0...@4ax.com>, cl...@snyder.on.ca wrote:
>On Tue, 12 Apr 2011 17:37:15 +1000, "Phil Allison" <phi...@tpg.com.au>
>wrote:
>
>>Got a fucking clue how much current it takes to vaporise copper wires like
>>that ?
>>
>>Obviously fucking not.
>
>How about "more than 15 amps"?
>
>And it didn't blow the circuit breaker or fuse.
>So the current came from somewhere other than the mains?

I don't remember any V-A rating stated for the tranny in question since
the number of posts in the 2 threads is huge. Though, I have the
impression it's around 40 or 50 or so.

A 120V one this size usually has its primary around AWG 24 or 26, and a
240V one or a dual-primary 120/240V one usually has primary winding size
around AWG 28. Properly connected for 120V, that means 2 roughly-AWG-28
strands in parallel, effectively around AWG 25.

A transformer winding this size can gradually get charring hot at as
little as 1 or 1.5 amps or so.

Leads of this size get charring hot within a minute at several amps.

The CRC Handbook says fusing current for copper wire is 20 amps for AWG
26, 30 amps for AWG 24. Though these figures are only approximate, I
usually hear that these fusing current charts are for bare wire in free air.
Fusing current will be less for a wire next to another turn of overheating
wire.

Meanwhile, I doubt a 15 amp slow-blow fuse or slow-acting breaker will
keep a 26 or 24 AWG wire from fusing. One thing to keep in mind is that a
15 amp fuse or breaker normally passes 15 amps without opening, and may
take a few seconds or more to open at 20 amps.

I think the overcurrent that caused charring here came from the power
line without opening the breaker/fuse. Back EMF won't do that, though it
could cause insulation failure. However, I doubt that explains 3 blown
transformers and nothing else blown.

Don Klipstein

unread,
Apr 12, 2011, 6:06:47 PM4/12/11
to
In article <90j6ug...@mid.individual.net>, Tony Miklos wrote:

>Yes, the thread got too long. If you saw the photo, and know the
>primary went open on 3 transformers, it sure looks like a problem on the
>primary side. If the load was too much, those wires on the secondary
>side would most likely have been at the very least discolored from the
>heat.

This makes me think the most likely causes are:

* Improperly connecting the transformer (such as using only 1 of the 2
primaries of a 120/240V dual primary transformer)

* DC flowing through the secondary. That can occur if the tranny's load
has a fullwave rectifier with one diode open. If the fullwave
rectifier has discrete diodes or a dual diode, the problem may be a bad
solder joint at one of the diodes.

DC through the secondary is well known to be able to cause transformer
cores to saturate. That can lead to the primary drawing excessive
current.

* Less likely - there is a magnet on the transformer core. That
combined with the tranny's normal magnetic flux can cause core
saturation, similar to the effect of DC in a winding although this
usually only causes part of the core to saturate.

However, a magnet on the core appears to me "probably unlikely to be
an actual problem" unless the tranny is of marginal design, line voltage
is on the high side, or the magnet is especially powerful.

There is the matter of voltage spikes on the primary blowing the
insulation. However, I consider that unlikely if nothing else has been
blown while 3 transformers blew.

Dave M

unread,
Apr 12, 2011, 7:43:26 PM4/12/11
to
Don Klipstein wrote:
> In article <90j6ug...@mid.individual.net>, Tony Miklos wrote:
>
>> Yes, the thread got too long. If you saw the photo, and know the
>> primary went open on 3 transformers, it sure looks like a problem on
>> the primary side. If the load was too much, those wires on the
>> secondary side would most likely have been at the very least
>> discolored from the heat.
>
> This makes me think the most likely causes are:
>
> * Improperly connecting the transformer (such as using only 1 of
> the 2 primaries of a 120/240V dual primary transformer)
>
> * DC flowing through the secondary. That can occur if the tranny's
> load has a fullwave rectifier with one diode open. If the fullwave
> rectifier has discrete diodes or a dual diode, the problem may be a
> bad solder joint at one of the diodes.
>

That's not true. an open rectifier does not allow DC current to flow
through the secondary. It's just 1/2 of the power line cycle. During the
opposite half of the cycle, no current flows in the secondary. True, it's
unidirectional current, but it's an intermittent current, not constant DC.
If your statement were true, then half-wave rectifiers wouldn't be feasible.

--
David
dgminala at mediacombb dot net

Phil Allison

unread,
Apr 12, 2011, 7:57:33 PM4/12/11
to

"Dave M"

>
>
> That's not true. an open rectifier does not allow DC current to flow
> through the secondary. It's just 1/2 of the power line cycle. During the
> opposite half of the cycle, no current flows in the secondary.

** So the average value is non zero and that means there is a DC component
to the current.

> True, it's unidirectional current, but it's an intermittent current, not
> constant DC.

** It will show a nice, steady reading on a DC current meter.

The interesting thing is how there is no corresponding DC component in the
primary current.


.... Phil

Michael A. Terrell

unread,
Apr 12, 2011, 7:58:30 PM4/12/11
to

Stormin Mormon wrote:
>
> OK, that makes a bit of sense. Seems like a bit of bother
> for what you can track down, with a VOM and some common
> sense.


The 'Dimbulb tester' will show a transformer that is in saturation,
or one that has a shorted turn. Your VOM can't.


--
You can't fix stupid. You can't even put a Band-Aid™ on it, because it's
Teflon coated.

Don Klipstein

unread,
Apr 12, 2011, 8:49:41 PM4/12/11
to

A unidirectional pulsing waveform has a DC component.

That analyzes to a sum of DC, fundamental frequency AC, and AC at
harmonic frequencies. The average as averaged over a whole cycle is the
DC component.

It is fairly well known that a transformer driving a halfwave rectified
load can run into core saturation problems from the DC component in the
unidirectionally pulsing current waveform.

cl...@snyder.on.ca

unread,
Apr 12, 2011, 10:05:51 PM4/12/11
to
On Tue, 12 Apr 2011 10:40:02 -0700, John Robertson <sp...@flippers.com>
wrote:


>>>
>> Yes, the thread got too long. If you saw the photo, and know the
>> primary went open on 3 transformers, it sure looks like a problem on the
>> primary side. If the load was too much, those wires on the secondary
>> side would most likely have been at the very least discolored from the
>> heat.
>>
>> Tony
>
>Yes, I looked at the pictures (rather poor resolution). The failure
>looks like a connection issue because there is no evidence of windings
>overheating. If the output was shorted the transformer would burn up -
>assuming it could sink enough current, but it would be unlikely to flare
>at one spot. This is a pretty standard 115VAC transformer, I've seen
>many of them...
>
>If someone tried to pump 5000 volts into the transformer then possibly
>you could get an arc like that to the transformer frame, so that would
>mean a serious miss-wiring of the furnace where the igniter somehow was
>connected to the primary of the transformer and then turned on.
>
>Hard to see that happening as the transformer powers the logic board
>that would enable the igniter to fire up!
>
>John :-#)#

And the FACT that the furnace in question uses a hot plate ignitor
further removes that scenario from any semplance of possibility

cl...@snyder.on.ca

unread,
Apr 12, 2011, 10:14:16 PM4/12/11
to
On Wed, 13 Apr 2011 09:57:33 +1000, "Phil Allison" <phi...@tpg.com.au>
wrote:

>

Unless the half wave rectifier is on the PRIMARY side.
I don't see any evidence of that on any of the diagrams I saw, but is
there something else on this furnace circuit? By code there cannot be
- but we don't know know this to be the case - or what is on the
circuit if there is.
>
>
>
>
>
>
>
>
>
>
>
>

Mark

unread,
Apr 12, 2011, 10:32:32 PM4/12/11
to
To the OP...
another long shot..

it looks like the primary of the transformer is wired in series with
the door safety switch. Is it possible that the door safety switch
is loose and arcing. This could cause a high voltage to appear at the
transformer and cause the insulation to fail...


Mark

Tony Hwang

unread,
Apr 12, 2011, 11:03:39 PM4/12/11
to

> Hi,
Lots of ripple. If it were on audio circuit, lots of hum!!!!
>
>
>
>
>
>
>
>
>
>
>

Don Klipstein

unread,
Apr 13, 2011, 12:23:08 AM4/13/11
to
In <rj1aq6lpg08u5m3l8...@4ax.com>, cl...@snyder.on.ca wrote:

>On 13 Apr 2011 09:57:33 +1000, "Phil Allison" <phi...@tpg.com.au> wrote:

>>"Dave M"
>>>
>>> That's not true. an open rectifier does not allow DC current to flow
>>> through the secondary. It's just 1/2 of the power line cycle. During the
>>> opposite half of the cycle, no current flows in the secondary.
>>
>>** So the average value is non zero and that means there is a DC component
>>to the current.
>>
>>> True, it's unidirectional current, but it's an intermittent current, not
>>> constant DC.
>>
>>** It will show a nice, steady reading on a DC current meter.
>>
>>The interesting thing is how there is no corresponding DC component in the
>>primary current.
>>
>>.... Phil
>
>Unless the half wave rectifier is on the PRIMARY side.

There won't be a rectifier feeding the primary of a transformer with DC.
A transformer does not pass DC from one winding to another the way it does
with AC.

Changing AC that transformers work with to DC that electronics work with
has to occur downstream of the transformer.

>I don't see any evidence of that on any of the diagrams I saw, but is
>there something else on this furnace circuit? By code there cannot be
>- but we don't know know this to be the case - or what is on the
>circuit if there is.

I have yet to look at these diagrams, but is the circuit board
powered by this tranny shown to "board level" as opposed to "component
level"? If so, then the board can have, probably does have, a rectifier
not shown in the diagram.

David Nebenzahl

unread,
Apr 13, 2011, 3:07:56 AM4/13/11
to
On 4/12/2011 9:23 PM Don Klipstein spake thus:

> I have yet to look at these diagrams, but is the circuit board
> powered by this tranny shown to "board level" as opposed to "component
> level"? If so, then the board can have, probably does have, a rectifier
> not shown in the diagram.

With all the pontificating you've been doing here, Don, I would've
thought you'd at least had glanced at the wiring diagrams the OP posted,
way up there somewhere. Sheesh.

The controller is shown as a block. It most certainly has at least one
rectifier on it, as it contains electronics that no doubt requires DC
power to operate. Thought you'd have figured it out. (Not just a relay
board.)

tra...@optonline.net

unread,
Apr 13, 2011, 7:22:15 AM4/13/11
to
On Apr 12, 8:49 pm, d...@manx.misty.com (Don Klipstein) wrote:

> In article <9fOdnZZQwKACfjnQnZ2dnUVZ_rmdn...@giganews.com>, Dave M wrote:
> >Don Klipstein wrote:
>  - Don Klipstein (d...@misty.com)- Hide quoted text -
>
> - Show quoted text -

Interesting theory. One helpful piece of info which it's quite
amazing we
still don't have is what the currents are on both sides of the
transformer.
For saturation to occur I would think the transformer would have to
be close
to fully loaded. Which it could be, given the trends to lower cost,
cheaper
components, etc.

cl...@snyder.on.ca

unread,
Apr 13, 2011, 5:46:30 PM4/13/11
to
On Wed, 13 Apr 2011 04:23:08 +0000 (UTC), d...@manx.misty.com (Don
Klipstein) wrote:

>In <rj1aq6lpg08u5m3l8...@4ax.com>, cl...@snyder.on.ca wrote:
>>On 13 Apr 2011 09:57:33 +1000, "Phil Allison" <phi...@tpg.com.au> wrote:
>
>>>"Dave M"
>>>>
>>>> That's not true. an open rectifier does not allow DC current to flow
>>>> through the secondary. It's just 1/2 of the power line cycle. During the
>>>> opposite half of the cycle, no current flows in the secondary.
>>>
>>>** So the average value is non zero and that means there is a DC component
>>>to the current.
>>>
>>>> True, it's unidirectional current, but it's an intermittent current, not
>>>> constant DC.
>>>
>>>** It will show a nice, steady reading on a DC current meter.
>>>
>>>The interesting thing is how there is no corresponding DC component in the
>>>primary current.
>>>
>>>.... Phil
>>
>>Unless the half wave rectifier is on the PRIMARY side.
>
> There won't be a rectifier feeding the primary of a transformer with DC.
>A transformer does not pass DC from one winding to another the way it does
>with AC.

I fully understand that, byt what happens if, say, the Mrs plugs her
cheap blow drier into the same circuit as the primary of the
transformer is plugged in to, and she puts the blow drier on low (it
uses a half wave rectifier to drop the power to the heater) - and you
end up with a DC component on the primary - which saturates the
primary without any increase in secondary (load) current or power.


>
> Changing AC that transformers work with to DC that electronics work with
>has to occur downstream of the transformer.
>

I am very aware of that.


>>I don't see any evidence of that on any of the diagrams I saw, but is
>>there something else on this furnace circuit? By code there cannot be
>>- but we don't know know this to be the case - or what is on the
>>circuit if there is.
>
> I have yet to look at these diagrams, but is the circuit board
>powered by this tranny shown to "board level" as opposed to "component
>level"? If so, then the board can have, probably does have, a rectifier
>not shown in the diagram.

Again, fully aware that the circuit board WILL have a rectifier which
is not shown on the diagram, and better than 95% chance it is full
wave bridge rectifier. There are precious few solid state control
devices that run on AC.

cl...@snyder.on.ca

unread,
Apr 13, 2011, 6:06:05 PM4/13/11
to

As I posted in an earlier post (not sure what values I used, but I'll
run a scenario anyway), assuming a 2 ohm winding, a 2 volt DC
component will cause 1 AMP of current to flow in the winding. IF that
coil happens to be the primary of a 40va transformer, The full load
current on that transformer is about 350ma, so the probability that 1
amp of current in the primary, with no AC voltage applied, would have
more than saturated the core is pretty good.
Add the quiescent current of the primary, and you have a saturated
core. And that's just a 2 volt DC component on a 115 volt primary.

If the primary is 4 ohms instead of 2, you have half an amp with 2
volts, or 1 amp with 4 volts.

Doesn't take much to put 2 volts DC across a 115 volt line.

See http://sound.westhost.com/articles/xfmr-dc.htm for more
information from someone who may have a bit more credibility than you
guys may give me.

Stormin Mormon

unread,
Apr 13, 2011, 7:22:37 PM4/13/11
to
So, it's cheaper to put a half wave rectifier in, instead of
a second coil of heater wire? I rather doubt that.

--
Christopher A. Young
Learn more about Jesus
www.lds.org
.


<cl...@snyder.on.ca> wrote in message
news:h36cq692alhsro2ju...@4ax.com...

Phil Allison

unread,
Apr 13, 2011, 8:48:40 PM4/13/11
to

<cl...@snyder.on.ca>

> As I posted in an earlier post (not sure what values I used, but I'll
> run a scenario anyway), assuming a 2 ohm winding, a 2 volt DC
> component will cause 1 AMP of current to flow in the winding. IF that
> coil happens to be the primary of a 40va transformer,

** A 40VA tranny designed for 120 volt 60 Hz operation has a primary
resistance of 16 ohms.

Getting a 2 volt DC offset on a 120 volt AC supply takes some doing.

> Doesn't take much to put 2 volts DC across a 115 volt line.

** Like hell.

For a 2 volt DC offset to exist, the peak voltage in one polarity must be
6.3 volts higher than the other.

With a typical impedance at the outlet of 0.25ohms, this equates to 25 amps
peak load in one polarity and none in the other.


> See http://sound.westhost.com/articles/xfmr-dc.htm for more
> information from someone who may have a bit more credibility than you
> guys may give me.


** I helped Rod write that article.

Toroidal trannys are very sensitive to DC offsets while regular E-core types
are hardly bothered by them - the difference is that while the former has
no air gaps in the core, the latter is full of them.


.... Phil


cl...@snyder.on.ca

unread,
Apr 13, 2011, 10:39:31 PM4/13/11
to
On Wed, 13 Apr 2011 19:22:37 -0400, "Stormin Mormon"
<cayoung61**spamblock##@hotmail.com> wrote:

>So, it's cheaper to put a half wave rectifier in, instead of
>a second coil of heater wire? I rather doubt that.

I don't remember the details, but I know quite a few DO work that way
- and they also used 12 volt DC blower motors, without a transformer,
by basically using the heater element as a rheostat..

Phil Allison

unread,
Apr 13, 2011, 10:45:03 PM4/13/11
to

<cl...@snyder.on.ca>
"Stormin Mormon"

>
>>So, it's cheaper to put a half wave rectifier in, instead of
>>a second coil of heater wire? I rather doubt that.
>
> I don't remember the details, but I know quite a few DO work that way
> - and they also used 12 volt DC blower motors, without a transformer,
> by basically using the heater element as a rheostat..


** My 1600 watt "Black & Decker" hot air gun is made exactly like that.

.... Phil


John Robertson

unread,
Apr 13, 2011, 11:51:43 PM4/13/11
to

Steve,

Not to be rude or anything, but a LOT of talk has gone on about your
problem and I have not seen any further posts from you on the matter. No
comments about any suggestions at all.

I'm not going to spend any more time on this thread until we hear back
from the original poster with more information otherwise we are simply
blowing smoke (sorry - couldn't resist).

Steve Turner

unread,
Apr 14, 2011, 12:32:56 AM4/14/11
to

Just mentioned elsewhere that I have a new transformer on order and I can't
really do anything until that arrives. Rest assured I appreciate all the input
and I'm taking it all into consideration. Thanks.

Vic Smith

unread,
Apr 14, 2011, 12:41:24 AM4/14/11
to
On Wed, 13 Apr 2011 23:32:56 -0500, Steve Turner
<bbqb...@swtacobell.net.invalid> wrote:


>
>Just mentioned elsewhere that I have a new transformer on order and I can't
>really do anything until that arrives. Rest assured I appreciate all the input
>and I'm taking it all into consideration. Thanks.

You bought the fuses, right?
I DON'T want to hear about another blown transformer.

--Vic

Don Klipstein

unread,
Apr 14, 2011, 12:43:29 AM4/14/11
to
In <4da54b91$0$26215$8226...@news.adtechcomputers.com>, D Nebenzahl wrote:
>On 4/12/2011 9:23 PM Don Klipstein spake thus:
>
>> I have yet to look at these diagrams, but is the circuit board
>> powered by this tranny shown to "board level" as opposed to "component
>> level"? If so, then the board can have, probably does have, a rectifier
>> not shown in the diagram.
>
>With all the pontificating you've been doing here, Don, I would've
>thought you'd at least had glanced at the wiring diagrams the OP posted,
>way up there somewhere. Sheesh.
>
>The controller is shown as a block. It most certainly has at least one
>rectifier on it, as it contains electronics that no doubt requires DC
>power to operate. Thought you'd have figured it out. (Not just a relay
>board.)

With such indication even as described here giving low indication as
to rectification having a 1-component bridge rectifier, 4 discrete diodes,
a fed-with-center-tap 2-diode fullwave scheme, 2 separate diodes or a
single 3-lead dual-diode used for that...

If the controller is only shown as a block, how well does it show the
rectifier scheme, as in whether the rectifier's diodes are discrete
individual diodes or integrated into one rectifier package?

If anyone here sees that noted to such extent that it's not a waste of
my time to take a look there, please pipe up!

It is loading more messages.
0 new messages