inrush surge

[William Sommerwerck]

Back around 1980 I had to repair a ReVox deck. It kept blowing its fuse --
even though the fuse was the correct value and speed.

I watched the fuse as I turned on the unit. So much current was drawn that the
fuse heated up and bowed, then came back to normal as it cooled. After a few
cycles, the fuse wire grew sufficiently fatigued that it broke.

I don't remember how I "solved" this. I might have gone to a slo-blo, or
increased the value slightly. Anyhow, it didn't come back.

Thoughts? (Including "You're an idiot!" Well, there were no stories about the
owner's house burning down.)


"We already know the answers -- we just haven't asked the right questions."
-- Edwin Land

[John-Del]

It's unlikely on a ReVox deck, but some more advanced equipment I've worked on over the years has had multiple relays or SCRs in a timed power up sequence. One RPTV I worked on years ago would occasionally blow fuses on start (but never when running), and it turned out to be a stuck relay (prevented one SMPS from being power up until the first one did).

[N_Cook]

>> "We already know the answers -- we just haven't asked the right questions.."
>>
>> -- Edwin Land
>

Add an anti-inrush NTC thermistor made for the purpose, assuming no
actual fault situation .

[William Sommerwerck]

"N_Cook" wrote in message news:l6l9gt$gf7$1...@dont-email.me...

> Add an anti-inrush NTC thermistor made for the purpose,
> assuming no actual fault situation.

That, of course, was the issue. There wasn't anything obviously wrong with the
unit.

[Jeff Liebermann]

On Thu, 21 Nov 2013 06:35:37 -0800, "William Sommerwerck"
<grizzle...@comcast.net> wrote:

>Thoughts? (Including "You're an idiot!" Well, there were no stories about the
>owner's house burning down.)

Burning the house down is a bit extreme, but such burnt offerings have
been known to appease the Gods of Electronics.

When all else fails, a measurement or three might be a good idea. Put
a small value resistor (0.1 ohm) in series with the AC line, and
measure the inrush current and duration. Then compare with the fuse
ratings and see if the selected fuse type and value are sane.

I've run into cute tricks like a big Lambda 12V power supply, that had
a 10 ohm resistor in series with the AC power lead. When the power
supply output finally stabilized, a relay would short out the 10 ohm
resistor so that full power could be delivered. That works great
until the relay contacts arced closed, causing the PS to suck full
current on startup and blowing the fuse. It took a while to find that
one.

--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

[Jeff Liebermann]

On Thu, 21 Nov 2013 16:08:37 -0800, Jeff Liebermann <je...@cruzio.com>
wrote:

>I've run into cute tricks like a big Lambda 12V power supply, that had
>a 10 ohm resistor in series with the AC power lead. When the power
>supply output finally stabilized, a relay would short out the 10 ohm
>resistor so that full power could be delivered. That works great
>until the relay contacts arced closed, causing the PS to suck full
>current on startup and blowing the fuse. It took a while to find that
>one.

It's a Lambda LFS-48-24. Schematic is on Pg 6.
<http://www.surplussales.com/powersupplies/pdf/ps-lfs-48-48.pdf>

[Phil Allison]

"William Sommerwerck"

>
> Back around 1980 I had to repair a ReVox deck. It kept blowing its fuse --
> even though the fuse was the correct value and speed.
>
> I watched the fuse as I turned on the unit. So much current was drawn that
> the fuse heated up and bowed, then came back to normal as it cooled. After
> a few cycles, the fuse wire grew sufficiently fatigued that it broke.

** That is just what a FAST fuse does when subjected to current surges.

> I don't remember how I "solved" this. I might have gone to a slo-blo, or
> increased the value slightly.

** Fuses in the AC supply always need to be "slo blo" or "T" types.

There are darn few exceptions to this rule.

BTW

Fuses with a flat metal strip inside are inherently slow and made of tin
not copper too.


.... Phil

[Trevor Wilson]

**Premium fuses from known suppliers usually open at the specified
current and within specified times. Cheap, Chinese fuses are often
unknown quantities. Often-times a package of fuses will be all
incorrect. They don't care. They don't put their brand name on the fuse.

--
Trevor Wilson www.rageaudio.com.au

[mrob...@att.net]

Jeff Liebermann <je...@cruzio.com> wrote:
> On Thu, 21 Nov 2013 06:35:37 -0800, "William Sommerwerck"
> <grizzle...@comcast.net> wrote:
>
>> Thoughts?
>

> I've run into cute tricks like a big Lambda 12V power supply, that had
> a 10 ohm resistor in series with the AC power lead. When the power
> supply output finally stabilized, a relay would short out the 10 ohm
> resistor so that full power could be delivered. That works great
> until the relay contacts arced closed, causing the PS to suck full
> current on startup and blowing the fuse.

The version of this I've seen is on the DC input (274 V nominal, 100 A)
to a big variable-frequency drive. There are two relays; the "little"
SPST one that operates first has about a 10 ohm ceramic resistor in
series with the contacts. The "big" one that operates second has no
resistor, and is DPST, for both sides of the DC bus. You can hear the
"tick, tick" of both relays operating in a second or less when you power
on the machine, and the "tick" of the big one releasing when you power
it off.

There is also a 125 A fuse on the DC input. It is rated at 100 sec
max at 250 A (2x rated), 15 sec max at 375 A (3x), and 1 sec max at
625 A (5x).

Matt Roberds

[Jon Elson]

Jeff Liebermann wrote:


> I've run into cute tricks like a big Lambda 12V power supply, that had
> a 10 ohm resistor in series with the AC power lead. When the power
> supply output finally stabilized, a relay would short out the 10 ohm
> resistor so that full power could be delivered. That works great
> until the relay contacts arced closed, causing the PS to suck full
> current on startup and blowing the fuse. It took a while to find that
> one.
>

Or, the resistor is on the DC side of the bridge, and they have an
SCR triggered by the startup of the switching supply to short the
resistor. If the SCR fails to turn on, the resistor goes up in
flames.

Jon

[Mark Zacharias]

"William Sommerwerck" <grizzle...@comcast.net> wrote in message
news:l6l9lr$hcc$1...@dont-email.me...

Could it be the fuse was rated for 240 volt operation (at 1/2 the value) and
being used on 120 volts?

I often see this - 240 volt operation ha a fuse 1/2 the value of that for
120 volt operation.


Mark Z.

[Phil Allison]

"Mark Zacharias"

>
> Could it be the fuse was rated for 240 volt operation (at 1/2 the value)
> and being used on 120 volts?
>
> I often see this - 240 volt operation ha a fuse 1/2 the value of that for
> 120 volt operation.

** I see the reverse sometimes, mainly with 1960s Fender guitar amps.

The AC tranny has been replaced long ago to suit use in Australia - but the
fuse holder labelling is still for the USA.


... Phil

[Jeff Liebermann]

On Fri, 22 Nov 2013 16:14:02 +0000 (UTC), mrob...@att.net wrote:

>Jeff Liebermann <je...@cruzio.com> wrote:
>> On Thu, 21 Nov 2013 06:35:37 -0800, "William Sommerwerck"
>> <grizzle...@comcast.net> wrote:
>>
>>> Thoughts?
>>
>> I've run into cute tricks like a big Lambda 12V power supply, that had
>> a 10 ohm resistor in series with the AC power lead. When the power
>> supply output finally stabilized, a relay would short out the 10 ohm
>> resistor so that full power could be delivered. That works great
>> until the relay contacts arced closed, causing the PS to suck full
>> current on startup and blowing the fuse.
>
>The version of this I've seen is on the DC input (274 V nominal, 100 A)
>to a big variable-frequency drive. There are two relays; the "little"
>SPST one that operates first has about a 10 ohm ceramic resistor in
>series with the contacts. The "big" one that operates second has no
>resistor, and is DPST, for both sides of the DC bus.

That's the same idea as the one I described. It's useful for power
supplies that have relay driven power on/off circuits, where one
doesn't want the full AC power drain going through the on-off switch
as in TTL/CMOS on-off. Relays or an SCR are about the only way to
simulate the function of inrush current limiter. Were a big power
supply to use an NTC thermistor for the function, it would probably
explode. As mentioned, the problem is when the contacts short, the
full inrush current appears on turn-on, blowing the fuse.

>You can hear the
>"tick, tick" of both relays operating in a second or less when you power
>on the machine, and the "tick" of the big one releasing when you power
>it off.

The PS I was playing with only had one click. I think it was about
1-2 seconds after power was applied.

>There is also a 125 A fuse on the DC input. It is rated at 100 sec
>max at 250 A (2x rated), 15 sec max at 375 A (3x), and 1 sec max at
>625 A (5x).

Yikes, that's big. I don't think a fuse that big actually "blows".
More like "melts".

[mrob...@att.net]

Jeff Liebermann <je...@cruzio.com> wrote:
> On Fri, 22 Nov 2013 16:14:02 +0000 (UTC), mrob...@att.net wrote:
>
>> There are two relays; the "little" SPST one that operates first has
>> about a 10 ohm ceramic resistor in series with the contacts. The
>> "big" one that operates second has no resistor, and is DPST, for both
>> sides of the DC bus.
>
> That's the same idea as the one I described.

The more I think about it (and look at the wiring diagram), the more I
think the "big" relay is really two SPST relays operated at the same
time.

> It's useful for power supplies that have relay driven power on/off
> circuits, where one doesn't want the full AC power drain going through
> the on-off switch as in TTL/CMOS on-off.

In this case, the operator position is not convenient to where the
relays need to live, so the relays get switched by 12 VDC control
signals. There are some other safety functions between the operator and
the relays as well; the computer can decide that bad things are
happening and drop out the control signals to the relays.

> As mentioned, the problem is when the contacts short, the full inrush
> current appears on turn-on, blowing the fuse.

I think this system is set up to fail safe if one set of contacts gets
welded. The normal startup sequence is something like

1. Close main relay in - line.
VFD should not yet have DC.
2. Close pre-charge relay in + line.
VFD has current-limited DC. Input caps charge.
3. Close main relay in + line.
VFD has current-unlimited DC.
4. Open pre-charge relay in + line.

and the normal shutdown sequence is something like

1. Open main relay in - line.
The input caps should start discharging, and the voltage at the VFD
should start dropping.
2. Open main relay in + line.

If the pre-charge relay contacts or main relay + line contacts are
welded, then the VFD will see DC as soon as it does startup step 1. It
can then open the main relay - line and be safe.

If the main relay - line contacts are welded, it probably can't detect
it at startup, but it can detect it at shutdown when the voltage fails
to drop off; it can open the main relay + line and be safe.

>> There is also a 125 A fuse on the DC input. It is rated at 100 sec
>> max at 250 A (2x rated), 15 sec max at 375 A (3x), and 1 sec max at
>> 625 A (5x).
>
> Yikes, that's big.

Roughly 34.5 kW at nominal voltage. The motor that this thing drives
is rated at 33 kW, or just a touch under 25 hp.

Physically, the fuse is only about 1.5" (38 mm) long and maybe 0.75"
(19 mm) diameter, with big lugs sticking out of the ends to bolt it
down. In use, it lives behind a polypropylene cover that you have to
remove an interlock and a screw to get at.

> I don't think a fuse that big actually "blows". More like "melts".

Well, technically, all fuses just "melt". :)

Matt Roberds

[Jeff Liebermann]

In the Lambda LFS-48-24 power supply I mentioned, only one set of
contacts are required. The sequences is roughly:
1. Apply AC power (there's no on-off switch in the PS) with a 10 ohm
resistor in series with the AC line. BFC's (big fat caps) charge
slowly, limiting inrush current.
2. When the voltage across the BFC's is at full charge and stable,
short the 10 ohm resistor with an SPST (actually DPST with both
contacts in parallel) to provide full power.

Shutdown is something like:
1. Remove AC power. Relay remains closed until BFC's are nearly
discharged. This is to prevent momentary power glitches from
producing much longer duration power drops at the output.
2. Relay opens when BFC's are discharged.

>and the normal shutdown sequence is something like
>
>1. Open main relay in - line.
> The input caps should start discharging, and the voltage at the VFD
> should start dropping.
>2. Open main relay in + line.

Ummm... my guess is that it should be:
2. Open main relay in + line only when caps are fully discharged. In
this case, the idea is to NOT open the contacts under load to prevent
arcing.

>If the pre-charge relay contacts or main relay + line contacts are
>welded, then the VFD will see DC as soon as it does startup step 1. It
>can then open the main relay - line and be safe.
>
>If the main relay - line contacts are welded, it probably can't detect
>it at startup, but it can detect it at shutdown when the voltage fails
>to drop off; it can open the main relay + line and be safe.

If there was something to detect input current, it could detect a much
faster than normal rise in input current, and use that to shut down
the PS. The problem is that by that time, it would be too late, and
with welded contacts, there would be no way to shut it down. Argh.

>>> There is also a 125 A fuse on the DC input. It is rated at 100 sec
>>> max at 250 A (2x rated), 15 sec max at 375 A (3x), and 1 sec max at
>>> 625 A (5x).
>>
>> Yikes, that's big.
>
>Roughly 34.5 kW at nominal voltage. The motor that this thing drives
>is rated at 33 kW, or just a touch under 25 hp.

Ummm... that's a bit larger than the typical commonly available power
supply. Well, it's larger than anything I've ever seen. My closest
experience is with solar powered battery controllers and inverters.

>Physically, the fuse is only about 1.5" (38 mm) long and maybe 0.75"
>(19 mm) diameter, with big lugs sticking out of the ends to bolt it
>down. In use, it lives behind a polypropylene cover that you have to
>remove an interlock and a screw to get at.

Is the fuse cartridge filled with sand?

>> I don't think a fuse that big actually "blows". More like "melts".
>
>Well, technically, all fuses just "melt". :)

I think of melting as a slow motion explosion. I was playing with a
5V 300A computer power supply. It had a 250A fuse in the output. So,
I decided to see what would happen if I shorted the output with a
screwdriver. (Do not try this at home). There was very little spark
as the screwdriver welded itself to the copper bus bar contacts. I
watched the fuse slowly melt into a puddle of tin(?). It took about 5
seconds. The screwdriver was quite hot and had to be removed from the
contacts with a cold chisel and hammer. Apparently, the screwdriver
was too high a resistance to "blow" the fuse and NOT trigger the over
current detection circuit. I later tried to reproduce this with a
similar power supply, but couldn't find the correct resistance
screwdriver.

[mrob...@att.net]

Jeff Liebermann <je...@cruzio.com> wrote:
> On Mon, 25 Nov 2013 04:42:43 +0000 (UTC), mrob...@att.net wrote:

> In the Lambda LFS-48-24 power supply I mentioned, only one set of
> contacts are required. The sequences is roughly:
> 1. Apply AC power (there's no on-off switch in the PS) with a 10 ohm
> resistor in series with the AC line. BFC's (big fat caps) charge
> slowly, limiting inrush current.

Part of the reason the system I am describing has more than one set of
contacts is that switching the "line" power is also its job; it's not
like your power supply where the user is expected to provide the line
power switching.

>> and the normal shutdown sequence is something like
>>
>> 1. Open main relay in - line.
>> The input caps should start discharging, and the voltage at the
>> VFD should start dropping.
>> 2. Open main relay in + line.
>
> Ummm... my guess is that it should be:
> 2. Open main relay in + line only when caps are fully discharged. In
> this case, the idea is to NOT open the contacts under load to prevent
> arcing.

The VFD, here, has the ability (and authority) to command the motor to
zero speed, so it can arrange for there to be no load before it starts
trying to open the relays. Both + and - lines float with respect to
"ground", so it doesn't really matter which one it opens.

>> If the main relay - line contacts are welded, it probably can't
>> detect it at startup, but it can detect it at shutdown when the
>> voltage fails to drop off; it can open the main relay + line and be
>> safe.
>
> If there was something to detect input current, it could detect a much
> faster than normal rise in input current, and use that to shut down
> the PS.

Good point. There is a current transformer on the "line" side of the
DC input, before any relay contacts. I don't know if it uses that in
its startup/shutdown strategy, but there isn't any reason why it
couldn't.

>> Roughly 34.5 kW at nominal voltage. The motor that this thing drives
>> is rated at 33 kW, or just a touch under 25 hp.
>
> Ummm... that's a bit larger than the typical commonly available power
> supply. Well, it's larger than anything I've ever seen.

I can guarantee that you've seen one of these, just not at this level of
detail. They sold a lot of them in sunny California.

>> Physically, the fuse is only about 1.5" (38 mm) long and maybe 0.75"
>> (19 mm) diameter, with big lugs sticking out of the ends to bolt it
>> down.
>

> Is the fuse cartridge filled with sand?

Dunno. It's got a label or wrapper around the whole thing, and I've
never peeled it back. I've never blown one, and I've never unbolted one
from its holder to shake it. It has a 2,000 A "breaking capacity", if
that implies anything about its construction.

> I was playing with a 5V 300A computer power supply. It had a 250A
> fuse in the output. So, I decided to see what would happen if I
> shorted the output with a screwdriver. (Do not try this at home).

Charles Darwin never met *you*, apparently. :D

> I later tried to reproduce this with a similar power supply, but
> couldn't find the correct resistance screwdriver.

Somehow, that's not usually specified for screwdrivers, unless they
are deliberately sold with a very high resistance.

Matt Roberds

[Cydrome Leader]

Jeff Liebermann <je...@cruzio.com> wrote:
> On Thu, 21 Nov 2013 16:08:37 -0800, Jeff Liebermann <je...@cruzio.com>
> wrote:
>
>>I've run into cute tricks like a big Lambda 12V power supply, that had
>>a 10 ohm resistor in series with the AC power lead. When the power
>>supply output finally stabilized, a relay would short out the 10 ohm
>>resistor so that full power could be delivered. That works great
>>until the relay contacts arced closed, causing the PS to suck full
>>current on startup and blowing the fuse. It took a while to find that
>>one.
>
> It's a Lambda LFS-48-24. Schematic is on Pg 6.
> <http://www.surplussales.com/powersupplies/pdf/ps-lfs-48-48.pdf>

all the lambda stuff I've come across is bizarre in design and
construction.

I don't recall the model, but one was just a plain switching supply of a
few hundred watts with a gigantic wirewound power resistor in it. Nobody
else does that, nor do most places change their name every 5 weeks either-
lambda, nemic, veeco, tdk, wtf?