Linear power supplies for nixies

[gregebert]

[Forking a new thread from the partially-lighted nixie]

I've used 3 different linear (ie non-switching) HV supplies in my clock designs. I'll describe the basics of each in separate posts.

#1: Voltage doubler. My first nixie clock has no transformers, so I use the AC line directly (120V rms, 60Hz in USA). The real trick here is to make sure the design is fail-safe. When the AC-line is rectified, it produces about +170VDC. Nixie-tube datasheets, such as the Burroughs 5092, specify a minimum anode voltage of 170, so that leaves no margin. With a voltage-doubler, you get +340V and that gives plenty of margin, though at the expense of more wasted energy (heat). If your tubes run at 2.2mA, running at 340V instead of 170V will waste an extra 375mw per tube, or 2.25W for a 6-digit clock. You will get reliable tube ionization at 340V, and you will also have less variation in tube current as the tube ages, or line-voltage variation. It's important to keep tube current in the correct range, otherwise lifetime will suffer. To select the right anode resistor, 

calculate R = (AnodeSupplyVoltage - OperatingVoltage)/(OperatingCurrent).

For an IN-18, I've measured the operating voltage 137V at 5mA of current. Tube-voltage does vary with current.

In this example, R = 40.6K-ohms for a 340V anode supply.

Referring to the attached schematic, the half-wave voltage doubler uses 2 diodes (D6+D10, and D2), and 2 capacitors (C5 and C1). On the negative 1/2 cycle, C5 is charged to the line-voltage. On the positive 1/2-cycle, C5's voltage is "added" to the line-voltage, and you get twice the AC-input voltage charging C1. There are also triplers, quadruplers, etc that can be made.

Safety is very important, so that's why there are so many 'extra' components. Obviously fuses, and I put one on each AC line for redundancy, and also in case hot/neutral get swapped. Resistors R15 and R16 are actually 33 ohms each, and they have 3 functions: (1) limit inrush current so the fuses dont blow when the caps are initially charged, (2) create an RC low-pass filter along with C3 so that AC line-noise doesn't cause the clock to "skip" time, and (3) act as redundant fuses. Hard to explain without a picture of the PC board, but the fuse-clips are exposed metal and it's possible a short could happen "before" the fuse. The varistor and C3 suppress voltage spikes, and if spikes are particularly nasty enough to endanger the capacitors, the fuses will blow. C3 is rated at 1kV; C1 and C5 are rated at 450V.  So why are there 2 diodes (D6 + D10) ? In case 1 fails as a short, it will prevent AC from causing a catastrophic failure of C5. BTW, bridge rectifiers dont have this vulnerability; they simply short-out when a single diode shorts-out.

Unique for this clock, there's R21 to current-limit the AC line on the PCB to ~120mA before it's sent to the NE2 bulbs that I use for the colons; just extra paranoia in case something causes a short on the PCB. Those NE-2s each have a 220K series resistor. You can see 6 anode resistors (R7-R12). R3 & R4 form a voltage divider so I can get the 60Hz line-frequency reference; this gets clamped by diodes D3 & D4 before it goes to the 400-series logic for the clock divider. The remaining devices (D5, R2, Z1, Z2, C2) provide +10VDC for the CMOS clock logic. Why 10 volts ? It's plenty enough to drive the NMOS driver transistors for the nixie tubes (all 45 of them because it's direct-drive), and also improves noise-margin.

As I mentioned, this clock has no transformer. Part of the reason is to reduce power-consumption, but also because I had this silly idea that you can reliably operate CMOS devices off the AC line with sufficient circuitry. But, there's no isolation, so just about anything on the PCB can give you a shock.

[Dave]

very informative, thanks.

I look forward to the next two installments!

[taylorjpt]

...And I thought optimizing a switching power supply was supposed to be difficult!

[gregebert]

2. Instead of a voltage doubler, a 'boost' supply is another way to get higher anode-supply voltage. One version (hvsupply.pdf) 'adds' a few volts to the AC line before it's rectified, though it's not isolated. I use this in my big clock to get around +220VDC. The other version (neondr_pwrsupp.jpg)  accomplishes the same result, but it's isolated from the AC line and that's desirable for safety reasons.

For the non-isolated supply (hvsupply.pdf), I used a 36 V transformer. There's a full-wave section for 220VDC (D4 & C4). There's also a half-wave doubler (D1, D2, D3, C2, and C3) that generates +440V for the dekatron in this particular clock. R3, R12, and R4 are important for safety reasons: They discharge the capacitors. The role of R12 is subtle, but if you study the circuit you will see the discharge path thru the transformer. RESD2 is one of several high-value resistors in the overall design that provide a DC path between all supplies to reduce ESD susceptibility while the clock is being built.

The isolated supply (neondr_pwrsupp.jpg) is from the clock I'm currently designing. This circuit has not been tested in actual usage yet, but I have run quite a few simulations on it. I still need to run it with transformer winding resistance. The rectifier (XD101) and filter cap (C101) provide about +180V. This clock has fourteen IN-18 tubes, each running at 5mA, so it's a fair amount of current.

In order to get isolation, you need a dual-primary transformer. One of the primaries is connected to the AC line and supplies the energy; the other primary winding is used as a secondary winding. When using a transformer in this manner for isolation, you must be careful not to exceed the VA (volt-amps) rating and that means you must include the VA consumption for all secondary windings (which includes the primary winding that got re-purposed as a secondary). Assuming you use a diode-> capacitor rectifier, you will want to calculate the VA rating based on peak current, which is higher than the load current. If you pick a transformer with a VA rating that's too low, you will see more losses (heat). Some will be I-squared-R losses in the windings, and most will probably be from core saturation. The easiest way to determine peak current is with a circuit simulation (LTspice and ngspice are excellent simulators available for free). Or, you can just try it out and see if the transformer gets warm; if it does, get a higher VA rating.

I'll post another article about designing the correct filter cap value.

[chuck richards]

MPJA has some nice dual-primary type power transformers
for very affordable prices. An LP-430, which has 4 amp.
12-volt secondary works very nicely using one of the primaries
as a secondary, and then putting the real secondary in series
with it to add a bit more voltage.

Then, after rectification and filtering, that lashup feeds to a zener
diode shunt regulator which maintains 170 volts within a volt or so.

Since I am not much up on switching supply design, and since
I kind of like big, dumb, simple sorts of things, this
HV supply seems quite workable. It's being tested now in a nixie
clock which is spread out on solderless boards. It uses 6 of the 8422
tubes and it uses 6 of the 74HC160 counters.

Chuck

>>--
>>You received this message because you are subscribed to the Google
>Groups "neonixie-l" group.
>>To unsubscribe from this group and stop receiving emails from it,
>send an email to neonixie-l+...@googlegroups.com.
>>To post to this group, send an email to neoni...@googlegroups.com.
>>To view this discussion on the web, visit
>https://groups.google.com/d/msgid/neonixie-l/a7b6bab1-eb4f-4b67-b19c-
>69e5528d9f57%40googlegroups.com.
>>For more options, visit https://groups.google.com/d/optout.
>>



$4.95/mo. National Dialup, Anti-Spam, Anti-Virus, 5mb personal web space. 5x faster dialup for only $9.95/mo. No contracts, No fees, No Kidding! See http://www.All2Easy.net for more details!

[marta_kson]

Using a double primary as a substitute for a secondary is a big no-no! You won't have the appropriate insulation to the mains. You will be just the thickness of two layers of enamel from grabbing the mains supply if any parts of the circuit can be touched. Scarry, isn't it?...

Compromizing safety to save on money is never an alternative. Order a properly made transformer with a safe insulation distance to the mains on the HV winding before something lethal happens...

[gregebert]

It's not the cost, but the availability. Ever since the demise of vacuum tubes in the 1970's, high-ish voltage transformers have been getting scarce. The only exception being transformers used in microwave ovens, which produce way too much voltage. I'm leery of salvaged transformers because they can degrade with age, and it's unlikely you will find 2 or more of the same item.

Hooking 2 transformers back-to-back is another method to get isolation, though less efficient.

As long as nothing is electrically exposed, it should be fine to use (or should I say, misuse) a dual-primary transformer.

[Charles MacDonald]

On 15-09-30 10:11 AM, gregebert wrote:
> It's not the cost, but the availability. Ever since the demise of vacuum
> tubes in the 1970's, high-ish voltage transformers have been getting
> scarce.

If you are willing to PAY for them, there are reasonable units form
folks like Hammond. They have crept up in price, although the recent
drop in the Canadian Dollar has probably put some downward presure on
the price in the States.

these guys for example
https://www.hammfg.com/electronics/transformers/classic/261-262 cover
off a lot of Nixie applications, using the "filament" windings for the
digital stuff.

Hammond has been making similar units for about 70 or more years so no
surprises, You an get them through places like Mouser or Antique
electronics. they make many of them in Guelph Ontario so even if they
run low on stock the back order should be manageable.

--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
No Microsoft Products were used in sending this e-mail.

[gregebert]

I did some research on UL/CSA approved transformers, and there is a requirement that all windings withstand a minimum breakdown voltage, even if they are intended to be connected together, such as dual-primaries. Depending upon the VA rating and the voltage, the breakdown must be between 1050 and 4000 V RMS according to how I read the spec (UL5058-2 / CSA C22.2 #66). The test is conducted between 1 winding, and all other windings and the core combined and at elevated temperature. There are copies of the spec online.

I knew there had to be some amount of isolation, but I did not realize it was that high. While I would never expose or touch anything that is supposedly "isolated", it does reassure me there is decent insulation.

[chuck richards]

Yeah. I just take a very simplistic view of it all.

Those guys at UL test that stuff under extremely harsh conditions.
If it does not break down and fail on those tests, it is very likely
that any of those windings can be used in any way anyone wants them to
be, as long as common sense is used as far as excessive current,
voltage,
power, and heat.

One of my rules of thumb: If it runs hot enough that I can't
hold my hand on it indefinitely, it's running too hot.

Chuck


---- Original Message ----
From: greg...@hotmail.com
To: neoni...@googlegroups.com

Subject: [neonixie-l] Re: Linear power supplies for nixies
Date: Wed, 30 Sep 2015 21:14:11 -0700 (PDT)

>I did some research on UL/CSA approved transformers, and there is a
>requirement that all windings withstand a minimum breakdown voltage,
>even
>if they are intended to be connected together, such as
>dual-primaries.
>Depending upon the VA rating and the voltage, the breakdown must be
>between
>1050 and 4000 V RMS according to how I read the spec (UL5058-2 / CSA
>C22.2
>#66). The test is conducted between 1 winding, and all other windings
>and
>the core combined and at elevated temperature. There are copies of
>the spec
>online.
>
>I knew there had to be some amount of isolation, but I did not
>realize it

>was *that* high. While I would never expose or touch anything that is

>
>supposedly "isolated", it does reassure me there is decent
>insulation.
>
>

>--
>You received this message because you are subscribed to the Google
>Groups "neonixie-l" group.
>To unsubscribe from this group and stop receiving emails from it,
>send an email to neonixie-l+...@googlegroups.com.
>To post to this group, send an email to neoni...@googlegroups.com.
>To view this discussion on the web, visit

>https://groups.google.com/d/msgid/neonixie-l/49ff59a7-f8f7-4ccc-bd7c-
>5969e6fd2514%40googlegroups.com.

[NeonJohn]

you are correct. it is perfectly safe to use dual primaries (or
secondaries) for isolation. The wire is triple-coated with enamel,
plastic and another layer of enamel.

While I have a multi-thousand dollar Topaz 10kVA Ultra-isolator on my
development bench, we require something cheaper for production testing
of our inductin heaters. Dual primary dry type transformers fill the
bill just fine. Each new transformer is HiPotted at 2500 volts winding
to winding for 10 minutes before being put into service. Never had one
fail.

John

On 10/01/2015 12:14 AM, gregebert wrote:
> I did some research on UL/CSA approved transformers, and there is a
> requirement that all windings withstand a minimum breakdown voltage, even
> if they are intended to be connected together, such as dual-primaries.
> Depending upon the VA rating and the voltage, the breakdown must be between
> 1050 and 4000 V RMS according to how I read the spec (UL5058-2 / CSA C22.2
> #66). The test is conducted between 1 winding, and all other windings and
> the core combined and at elevated temperature. There are copies of the spec
> online.
>
> I knew there had to be some amount of isolation, but I did not realize it

> was *that* high. While I would never expose or touch anything that is

> supposedly "isolated", it does reassure me there is decent insulation.
>
>

--
John DeArmond
Tellico Plains, Occupied TN
http://www.fluxeon.com <-- THE source for induction heaters
http://www.neon-john.com <-- email from here
http://www.johndearmond.com <-- Best damned Blog on the net
https://www.etsy.com/shop/BarbraJoanOriginals <-- please visit
PGP key: wwwkeys.pgp.net: BCB68D77

[Dekatron42]

Can anyone direct me to a document that says that it is allowed to sell an electronic apparatus that uses a primary winding as a secondary winding - I spent a lot of time Googling this and I can't find anything. I am also concerned about safety and what an insurance company would have to say if a fire breaks out and the culprit is the home built equipment which uses a primary winding as a secondary winding.

/Martin

[JohnK]

One of the 'requirements' here in Australia is/was for the primary and secondaries to be on separated parts of the bobbin. I haven't looked for the 'rule', I just remember that when it was introduced the blurb with various electronic kits [and advertisements] highlighted the fact that the mains transformer met xxxx.  I prefer the split bobbin windings because they 'should' be safer if the transformer gets very overheated. I do not know what transformers are required to have the inbuilt thermal fuse.  I also do not know if separating the windings on the bobbin [as is done on the small transformers] is efficient enough to be used on larger transformers. 

There was a debate in a local electronics magazine about safety and transformers. It was pointed out there that toroidal transformers had a history of failing when hot due to mechanical pressures between the layers. I don't know how well-built the offending transformers were. Might have been 'cheapies' for an amplifier kit.

 

Lots of "I don't know" from me there - but it might contribute to talking points :-)

 

AND, the topic of RCDs. There are RCDs built in to mains plugs available. They might be good for extra safety.

 

One aspect of RCD useage which is often overlooked: they can shut off the power if there is a fire.  It can be worthwhile taking an earth/ground into "2-wire" equipment or installations to assist with protection against water ingress and fire. It is a long story, but I was in a position to analyse faults of PIR [passive infra red] detector installations and it is amazing what actually goes wrong. Plan for the worst thing to happen - it will somewhere ! 

 

John K

Australia

----- Original Message -----

From: Dekatron42

To: neonixie-l

Sent: Friday, October 02, 2015 4:37 PM

Subject: [neonixie-l] Re: Linear power supplies for nixies

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/a6a3a865-ae1b-4e3d-879f-e7dc87318c03%40googlegroups.com.

[gregebert]

If you are going to sell finished products, your best option is to use an *external* agency-approved power-supply to produce ~12-18V. Laptop computers are a good example of this practice. Of course, you'll need an internal DC-DC converter so it's no longer a linear power supply :-( 

Regulations will vary by region, and there are organizations (UL, CSA, VDE, etc) that can certify your product. If you go that route, it's going to cost a lot of money and probably not worth the expense for something like nixie clocks. I'm certain that the transformer construction is a lesser issue; there are all sorts of things that affect approval.

Another option is to do what I do:  chicken-out and make things only for yourself.  I take enough risks at my day-job because my employer has cost and schedule constraints developing their bleeding-edge products. When I make nixie clocks, I do it at my own leisure with no regard for cost, complexity, or schedule. I keep working on it until it's perfect. Then I plug it in for everyone to enjoy.

BTW, at work we call 1.3 volts "high-voltage", and 1.5 volts is "extremely-high-voltage".  Nobody discusses 1.8 volts anymore because that's insanely too high to deal with....

[chuck richards]

.... Reading your last post, I would suppose then,
that 5 volts would be considered "high tension" :)

>
>
>---- Original Message ----
>From: greg...@hotmail.com
>To: neoni...@googlegroups.com

>>--
>>You received this message because you are subscribed to the Google
>Groups "neonixie-l" group.
>>To unsubscribe from this group and stop receiving emails from it,
>send an email to neonixie-l+...@googlegroups.com.

>>To post to this group, send an email to neoni...@googlegroups.com.

>>To view this discussion on the web, visit

>https://groups.google.com/d/msgid/neonixie-l/1f38e9e2-3bb0-47fa-a6fe-
>94abf42387a6%40googlegroups.com.

>>For more options, visit https://groups.google.com/d/optout.
>>

[Charles MacDonald]

On 15-10-02 03:07 AM, Dekatron42 wrote:
> Can anyone direct me to a document that says that it is allowed to sell
> an electronic apparatus that uses a primary winding as a secondary
> winding - I spent a lot of time Googling this and I can't find anything.
> I am also concerned about safety and what an insurance company would
> have to say if a fire breaks out and the culprit is the home built
> equipment which uses a primary winding as a secondary winding.

depends on the jurisdiction. Generally to sell something you need it to
pass an electrical inspection to be 100% kosher. In fact technically
here in Ontario, you need an inspection to be able to legally plug it in
yourself!

That is what the c(UL)US or C(CSA)us stickers you see on manufactured
goods are all about.

[NeonJohn]

On 10/02/2015 03:07 AM, Dekatron42 wrote:
> Can anyone direct me to a document that says that it is allowed to sell an
> electronic apparatus that uses a primary winding as a secondary winding - I
> spent a lot of time Googling this and I can't find anything.

I'm truly amazed that anyone would think that this sort of triviality
would be written down somewhere. Just as you wouldn't find, for
example, an official document saying that one can use a 2n3904 as a very
fast avalanche pulse generator. Or any of the millions of other
techniques engineers use to accomplish the job.

If you pay close attention to the NEC, you'll see that most requirements
can be modified or over-ridden by "good engineering judgement". That
is, a competent engineer can look at a specific situation, determine
that the cook-book requirement doesn't fit and design a solution
specific to the situation.

Of course, the engineer assumes liability for any subsequent
malfunctions, just as an architect does for new building techniques.

I can assume with reasonable certainty that you don't know how the
agency approval (UL, ETL, etc) works. There's no massive tome on high
that is consulted to determine if a given gadget works.

If you take your new design for a transformer-based wall-wart to ETL,
they will consult their files to see if there is a testing procedure
already on record. If so, you pay them about $2500 and they test your
gadget against the procedure.

If your gadget covers new area, then you pay them something starting at
about $10k for them to develop the testing procedure. THEIR engineers
use good engineering judgement based on experience when determining the
testing procedure. "Is 2500 VDC for 1 minute enough of an interwinding
potential or should it be 4500?" Based on their collected body of
experience and data, as well as any applicable standards, they'll select
an appropriate value.

Contrary to popular belief, at least here in the US, a product does NOT
have to have agency approval to be marketed and used, except for a few
malignant jurisdictions such as NYC.

At Fluxeon, we decided at start-up not to waste the money on agency
approvals for our portable induction heaters. We lose a sale here and
there but all in all, that has turned out to be the correct decision.

As Chief Engineer and as a member of the Board, the onus for product
safety falls on my shoulders. My qualification requirements are vastly
tougher than any agency would require.

A prime example is the output transformer. It has about 1200 VAC on the
primary and about 60 VAC on the other. It is also the life safety
barrier separating line voltage from the user. An agency might require
a safety factor of 5 and require a HiPot test of perhaps 6kVDC for a
minute.

Every transformer is tested at 8kVDC for one minute primary to
secondary. The prototypes and random samples pulled from production are
tested at 12KV high frequency AC for 12 hours. High frequency AC is a
much tougher test than DC because the HF generates dielectric losses and
other effects not seen with DC.

> I am also
> concerned about safety and what an insurance company would have to say if a
> fire breaks out and the culprit is the home built equipment which uses a
> primary winding as a secondary winding.

I don't know where this widely believed fiction originated from but at
least in every state I've lived in, homeowner's insurance doesn't look
any further than whether arson was involved.

When I place burned after a computer monitor caught fire in the night,
the adjuster made a copy of the fire marshal’s report, cut me a check
for the policy limit, wished me good luck and left. All over with in 30
minutes.

John

[Jeff Walton]

Agency approval only becomes a big deal if there is a lawsuit and the attorneys are trying to find negligence.

In business purchases, many larger companies demand agency approvals by policy. The company I am working with manufactures a device for field testing of thermal conductivity, which involves large heaters. Selling into the European block requires the CE approval process, which is a self-administered and documented testing. While it is demanded by many customers, the self-test and approval process leaves a tremendous amount of leeway and flexibility. We do have CE markings. As NeonJohn implied, there is not always a black and white answer but you should make the effort to eliminate known hazards.

In a legal contest, if you do something that can be shown to be a willful "bad" engineering practice and is found as negligence, you lose. A clock burning up is probably not going to trigger a lawsuit - unless it kills someone. At that point, the outcome depends on how good your legal representation is.

Beyond the input isolation and materials used to build the clock, there probably isn't much to be worried about from a strictly legal standpoint. Over the years, I've lost a few projects to transformer and other line-connected component failure and have come to love a well designed, agency-tested power brick on the front end of anything that stays plugged in and runs unattended. They aren't perfect either...

I had a small fire from a nixie clock I designed in the late 60's, which ran just fine for over 30 years until a cap failure in the voltage doubler caused overheating, followed by a short, and caught the wood/plexiglass housing on fire. For some reason, the fuse did not fail soon enough. The insurance company paid the claim with no problems. Fortunately, smoke damage only and no one hurt.

Jeff

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/560F2381.1020001%40neon-john.com.

[Jim_Z]

Having worked for UL as an engineer I have a few comments.

UL does not "approve" anything- they List, Recognize and Classify items. "Approval"  implies that an item can be used without regards to its limited testing by UL. The UL lawyers hate the term "Approved By UL" and routinely  send out letters to the corporate abusers of that terminology.

In the case of power supplies they are usually "Recognized" components meaning that they are tested to certain parameters but are dependent on the actual environment they are applied to (that is the backwards UL symbol). For instance, if a  previously UL "Recognized" power supply is installed in a computer that is submitted for "Listing" and it is used within its UL defined "Recognition" than limited investigation is required.

Also note that in the case of power supplies it's not important to UL whether they really work or not- only that they perform to the UL requirements which don't typically specify performance values.

Jim 

On Wednesday, September 30, 2015 at 9:14:11 PM UTC-7, gregebert wrote:

[Charles MacDonald]

On 15-10-05 03:44 AM, Jim_Z wrote:

> Also note that in the case of power supplies it's not important to UL
> whether they really work or not- only that they perform to the UL
> requirements which don't typically specify performance values.
> Jim

Thinking of this thread bought back a converstaion I had in the 1980s
with one of the senior people at what was Commodore Canada. They were
just coming out with the Vic-20 and the topic of the conversation got on
to CSA.

Basically he told me that they were expected to submit two production
samples of each product. CSA would take these and literally torture them
to death. for a pwer supply for example, short the output and make sure
the fuse blows or it otherwise shuts down. This after drawing twice the
rated current for a while and getting neither smoke nor flames. He also
talked about the lab dousing the unit with a flammable substance and
making sure it did not catch fire.

(at that time before NAFTA items needed a CSA sticker in canada and a UL
sticker in the USA. Now several labs can do the tests and put a "C" or
"us" or both depending on whose standards are met)

I am sure he was exaggerating somewhat, but his voice was exasperated
when he mentioned that they them charged extra if he wanted the test
items, which were now totally unusable even for parts - back

I can see the sort of thinking that going into passing those tests when
I overworked a blender and it stopped working. The maker had buried a
thermal fuse inside the motor. which had opened. not serviceable but it
would not catch fire in the UL/CSA tests.

[David Forbes]

I once had to fix our neighbors' coffee grinder at Burning Man. I found that it
had a blown thermal fuse, probably because they were using a household coffee
grinder for an espresso bar which served dozens of cups a day.

Solution:
1. Bypass thermal fuse
2. Write on the case "For playa use only. Safety systems disabled"
3. Safety third!

On 10/5/2015 1:32 PM, Charles MacDonald wrote:
> I can see the sort of thinking that going into passing those tests when I
> overworked a blender and it stopped working. The maker had buried a thermal
> fuse inside the motor. which had opened. not serviceable but it would not catch
> fire in the UL/CSA tests.

--

David Forbes, Tucson, AZ

[Dekatron42]

I do not know anything about this topic, that's why I asked.

I was told by a friend that the Low Voltage Directive in the EN61010 standard was the right place to look for answers, so I did and found some details but I do not understand it all. I found this http://www.ni.com/white-paper/2827/en/ which talks about instruments and isolation, which is interesting as it also mentions transformers and the categories used. So I emailed two transformer manufacturers and the Swedish National Electrical Safety Board and asked them about the use of transformer winding usage and also asked my insurance company.

One transformer manufacturer answered me that as long as the combined voltage of the secondary plus primary winding is within the isolation category, and voltage for the primary isolation barrier, it would of course work and be within the approval limits but they did not recommend it for a few reasons, the foremost reason being that a transformer is not tested for this connection and it would probably fail (not due to the voltages used but more so because there is no test for this condition) in an approval test under this condition and/or at least need an expensive new way of testing to be performed for an approval - being within the approval limits is not the same as being approved when used in a way that the approval testing has not tested nor taken into account they told me. One other reason for not using this way of connecting the windings is because the heating of the transformer windings in this condition is not taken into consideration when the transformer is designed and that could affect performance.

When it comes to the insurance question it would not be ok if it did not have an approval and was the starting cause for the fire, if it had an approval it would be just fine.

So, it will probably work ok but it might be hard to get an approval and therefore not good to use if it catches fire, at least in Sweden and with my insurance company. So I'll stick with transformer bought from reliable transformer manufacturers for now.

/Martin

[gregebert]

I dont know if it's still common practice, but many radios and small TVs in the US in the 1960's & 1970's had a 'hot-chassis' where one side of the AC line was connected to the metal interior chassis. Obviously, these devices had non-conductive (usually plastic) cases. Polarized cords supposedly ensured the neutral side was connected to the chassis, but extension cords etc left it a 50-50 chance it was electrically hot. I cant recall if any of these were UL listed, or not.

You would think that if hot-chassis devices were "safe", then usage of a pseudo-isolation transformer would be safer. I was probably 12 years old when I opened-up my first hot-chassis radio (yes, it was unplugged) and couldn't believe my eyes when I saw the power-connector tied to the large all-metal chassis. 

And for you vacuum-tube enthusiasts, you may have wondered why there was a tube with a 35-volt filament (35W4). That's because the filaments were wired in-series to the AC line, and the remaining tubes had 12.6-volt filaments.

[Instrument Resources of America]

On 10/6/2015 3:50 PM, gregebert wrote:

I dont know if it's still common practice, but many radios and small TVs in the US in the 1960's & 1970's had a 'hot-chassis' where one side of the AC line was connected to the metal interior chassis. Obviously, these devices had non-conductive (usually plastic) cases. Polarized cords supposedly ensured the neutral side was connected to the chassis, but extension cords etc left it a 50-50 chance it was electrically hot. I cant recall if any of these were UL listed, or not.

Those chassis IIRC were insulated from the cabinets with nylon/plastic bushings.   I think even the control shafts were also insulated from the chassis just in case one of the control knobs came off.

You would think that if hot-chassis devices were "safe", then usage of a pseudo-isolation transformer would be safer. I was probably 12 years old when I opened-up my first hot-chassis radio (yes, it was unplugged) and couldn't believe my eyes when I saw the power-connector tied to the large all-metal chassis. 

And for you vacuum-tube enthusiasts, you may have wondered why there was a tube with a 35-volt filament (35W4). That's because the filaments were wired in-series to the AC line, and the remaining tubes had 12.6-volt filaments.

Except for the audio output tube which had a "50" volt filament, e.g. 50C5 or 50B5.   Ira

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/cd819a31-c745-4ac8-91a6-a9118b094ffd%40googlegroups.com.

[threeneurons]

On Tuesday, October 6, 2015 at 3:50:51 PM UTC-7, gregebert wrote:

I dont know if it's still common practice, but many radios and small TVs in the US in the 1960's & 1970's had a 'hot-chassis' where one side of the AC line was connected to the metal interior chassis. Obviously, these devices had non-conductive (usually plastic) cases. Polarized cords supposedly ensured the neutral side was connected to the chassis, but extension cords etc left it a 50-50 chance it was electrically hot. I cant recall if any of these were UL listed, or not.

Those are the "All American 5" (AA5) tube radios, which first came out in the late 30's. Various tube line ups, which changed as more modern tubes became available. Still being made into the early 70's. There were also TV sets made with hot chassis and series heater strings. Usually "portables". "Portable" being a 35 lb box with a handle on top.

UL standards get superseded by newer ones, continually. I remember one time when there were 3 standards covering consumer products. I contacted UL, to ask which one I should use for a new product. Their reply was "pick one". But they also said, that the older ones would be phased out, and eventually there would only be one. So the AA5 probably met some long defunct UL standard. UL is not a government regulatory agency, but a private organization setup by the insurance industry.