Gently winding up Variacs

[Cursitor Doom]

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

 

Ever hear of a dim bulb tester ? This is an incandescent lamp wired in series with the mains. If you have a blown AC fuse you can just jump it out because the DBT will limit the current. In the US, a 100 watt lamp will limit the current to under an amp. On like a solid state audio amp or alot of other things it may light up in the beginning but will dim out as the main filters charge up. If it never dims you start disconnecting things until it does and that is your short.

Some technologies do not like this, here and there, some depend on the steep rise in main B+ to start, sometimes through a cap to a lower source but that is used less often because it precludes certain methods of line isolation.Some TVs that had scan (LOPT) derived sources had problems with it, some SMPSes have a problem with it and most things that are microprocessor controlled for the power on have a problem when the standby is up, then the inrush of the main power goes on and kills the AC. Steps must be taken to rectify this problem if you want to use the DBT on such units.

Tube units also need to be handled differently. If it has selenium or copper rectifiers, they are hellishly inefficient and you might need a higher wattage bulb. In fact I have been considering building one that has multiple bulbs that can be switched in and out for different equipment.

If you have a tube rectifier, like a 5U4 or whatever, you can probably just fire it up as is, because that B+ will not be applied all at once. It only comes after the filament of the rectifier tube warms up. Usually, when an ALL tube (valve) unit blows the AC fuse immediately it either needs a power transformer or a small noise suppression cap across one of the windings. But you do have to be a little more careful if it has solid state rectifiers.. In that case, usually ohmmeter reading will tell you if you have a dead short.

Any caps that emit smoke at power up need to be replaced. A shorted power transformer stays shorted. Main thing is the smoke and smell really. Anything that smokes is bad, so mainly the DBT and variac protect DC coupled amplifier circuits and some regulator problems.

If you use the DBT in conjunction with the variac, you should be able to increase to full voltage in less than a minute unless the thing has massive filters in it. This IS an issues with some high fidelity amps that have like a half a Farad of main filtering. Some really do have that much, though 100,000 uF is more common. In any case you take the variac up a bit and the bulb lights up a bit and then dims down. When it dims down you give it a little more and a little more. When it stops dimming down (and this is with no signal and no load) THEN is when you look for the problem.

 

[N_Cook]

On 23/08/2015 18:41, Cursitor Doom wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

Its not a panacea, just gives better odds against an electro cap
failure, a totally dry one will not reform.
I don't know if its on the WWW but the Colossus at Bletchley Park is
wound up daily via a geared motor drive , over 40 minites IIRC, and also
down again in the afternoon. As a result they just loose about 1
thyratron a year, replacing caps would be no problem

 

[Trevor Wilson]

On 24/08/2015 3:41 AM, Cursitor Doom wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

**I start at zero when winding up my variable auto-transformer (aka:
Variac™ - Variac™ is a trade name, not the name of the device itself).
With modern DUTs (ie: anything manufactured within the last 30 years), I
don't bother too much. Older equipment will almost always have suspect
electros anyway. As other have stated, it's a crap shoot. I recently
refurbished a super 8 projector (European, 70s vintage, Beaulieu) for a
client. I started testing some of the electros and found that 70% had
either poor ESR or very low capacitance. I simply replaced the whole
lot. I didn't even bother testing them after the first half dozen.

I bet your old 'scope has a failed cap.

--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Samuel M. Goldwasser]

Cursitor Doom <curd@notformail.com> writes:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

This is fine most equipment with a linear power supplies. Switchmode power
supplies may be unhappy at significantly below minimum spec'd input voltage and
let out their smoke regardless of whether the caps are good.

--
sam | Sci.Electronics.Repair FAQ: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Sites: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: Anything sent to the email address in the message header above is
ignored unless my full name AND either lasers or electronics is included in the
subject line. Or, you can contact me via the Feedback Form in the FAQs.

 

[Trevor Wilson]

On 24/08/2015 3:41 AM, Cursitor Doom wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

**One more trap for the uninitiated: NEVER crank up an audio amplifier
fitted with V-FET output devices on a variable auto-transformer.
Destruction will be assured under such a test.

--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Phil Allison]

Cursitor Doom wrote:

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess.

** That would probably be very safe, but in most cases it can be done much faster. Filter electros are the troublemakers as they can require gradual reforming after a long period of idleness.

Had one like that the other day, a valve amp from the 1960s that had not been run in the last 10 years. With all valves removed, the AC supply current rose quickly with each voltage increases and then slowly fell back.

I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.

** That is what overheated electros do.


..... Phil

 

[gregz]

Cursitor Doom <curd@notformail.com> wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

Not much more, but all depends, of course. Then you got tube rectifiers. I
like using dim bulb AND variac. I don't know if it helps, but sometimes I
let sit at zero, after slow increases.

Greg

 

"**One more trap for the uninitiated: NEVER crank up an audio amplifier
fitted with V-FET output devices on a variable auto-transformer.
Destruction will be assured under such a test. "

You mean class D I assume ?

 

[N_Cook]

On 23/08/2015 18:41, Cursitor Doom wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

Tantalum caps failing, can go off with quite a bang, leaving an
"electrical" smell

 

[Phil Allison]

jurb...@gmail.com wrote:

"**One more trap for the uninitiated: NEVER crank up an audio amplifier
fitted with V-FET output devices on a variable auto-transformer.
Destruction will be assured under such a test. "

You mean class D I assume ?

** V-FETs were TO3 linear devices made and used by Sony and Yamaha in a number of hi-fi amplifers back in the mid 1970s. Common part numbers were 2SK82 and 2SJ28.

Then they suddenly stopped making them.

As a result, V-FETs and the few amps concerned are now rare and so have acquired a mythical status among the gullible.

BTW:

Do not confuse V-FETs with Hitachi's "Lateral Mosfets" that appeared later and were a huge success especially in high powered pro-audio amps.

Common numbers were 2SK135 & 2SJ50 or 2SK176 & 2SJ56.

Then they suddenly stopped in the late 1980s.

However, near identical TO3 MOSFETS are still made and carry numbers like BUZ901 and BUZ905.


.... Phil












Google " Sony V-FET " and learn more than you ever wanted to know.

 

[Cursitor Doom]

On Mon, 24 Aug 2015 09:24:38 +0100, N_Cook wrote:

Tantalum caps failing, can go off with quite a bang, leaving an
"electrical" smell

Probably not a tant, then. I don't recall any bang and certainly no
electrical smell. Just choking, acrid smoke. Oh, and when I looked
inside, there was a light film of fine, sooty, black carbon powder over
much of the innards.
I suspect whatever had failed was concealed or screened as I couldn't see
any evidence of charred components merely from splitting apart the
external cases.

 

[pfjw@aol.com]

On Sunday, August 23, 2015 at 1:43:24 PM UTC-4, Cursitor Doom wrote:

Hi all,

If a piece of faulty equipment hasn't been fired up for like many
*years*, what rate would it be advisable to slowly wind up the supply
voltage to it so as not to blow anything? I'd guess 24hrs for a 240VAC DUT
would do it but it's just a guess. I blew something in a scope up one day
after being too impatient; very annoying. Whichever component it was that
went phut filled the room with acrid smoke. I still have yet to track it
down.
And I'm also guessing one could get away with starting the process at
50VAC? What does the Panel think?

Mpffff.... a variac without proper metering is worse than useless. And a dim-bulb tester, while infinitely better than an unmetered variac is still a blunt instrument.

A variac needs an ammeter that is accurate within a single % of the scale it is measuring. Such that one may watch even tiny variations in how the current comes up as the voltage is raised. A few things:

a) Anything with a tube (valve) rectifier will not pass B+ until the applied voltage is somewhere between 70% and 80% of the nominal voltage. So, reforming caps by way of slow uptake on a variac for such a devices is a complete myth. Those caps will see, all at once, no less than 70% of the operating voltage - not hardly a soft start.

b) If any device is using as little as 5% more current than it should-if-optimal, that current will be expressed somewhere as heat. If in the winding of a power or output transformer, that will, eventually mean *POOF*. If in an IF transformer for a tuner or receiver, also, eventually, *POOF*.

c) Electronics, old and new, do not typically 'cure' themselves of faults, even if cajoled, massaged and seduced very slowly. Sure, caps _can_ reform, but that reformation is unlikely to be either reliable or permanent. Given the low cost of caps these days, even the thought of reforming vintage caps should give one the cold shivers at the very least.

d) One can, with the proper instrumentation, figure out what is going on with any given piece of equipment typically within several hours anyway. 99-44/100% of them, within several minutes. Sure, individual component thermal failures or intermittents can be a huge diagnostic problem, but as to general function, that should take little time at all.

So, cutting to the chase, a metered variac with a precise meter that is capable of reading *accurately* in very small increments is a very fine diagnostic tool, especially at the triage level. A simple, unadorned variac is useful only for dimming the lights - its original purpose. IF that is all one has, put it aside and start with the dim-bulb tester and a number of different 'bulbs' to get a very-ballpark S.W.A.G. at how much current any given devices is actually using.

For the record, if it takes me more than 5 minutes to conclude how well behaved any given device might be under full power, that would be a long time. HOWEVER!! Once any new-to-me device is at full operating voltage and running, I will watch it like a hawk for several hours at least before deeming it fit for polite society. And very few very vintage (pre-1960) devices get even that far without at least the power-supply caps being redone. It is simply not worth the risk.

Peter Wieck
Melrose Park, PA

 

[pfjw@aol.com]

On Monday, August 24, 2015 at 1:27:44 AM UTC-4, Phil Allison wrote:

Had one like that the other day, a valve amp from the 1960s that had not been run in the last 10 years. With all valves removed, the AC supply current rose quickly with each voltage increases and then slowly fell back.

How would that happen unless the unit had a solid-state rectifier? Typically that quick-rise, slow-fall is due to the "empty" caps accepting inrush, then as they charge becoming more 'resistant-for-lack-of-a-better-word'. And the only way the caps would be in the circuit is if the rectifier was in place - hence the question.

Put another way, it is a bad idea to allow a tube amp to make B+ without a load. Many marginally designed amps ran their filer caps at very near their ratings - and both vintage and marginally capable caps running unloaded might just go *POOF*.

Peter Wieck
Melrose Park, PA

 

>"** V-FETs were TO3 linear devices made and used by Sony and Yamaha in a >number of hi-fi amplifers back in the mid 1970s. Common part numbers were >2SK82 and >2SJ28. "

I'll take your word for it, but that doesn't explain why. I dug around a little and found the TA-8650 which is 80 watts per and uses three complementary sets of 2SK60 and 2SJ18 per. Seems like alot for 80 watts but who am I to bitch about something being a bit overbuilt ?

There are a couple of things about that amp that don't make a hell of alot of sense. First of all, they are current sharing without source resistors, at least separate source resistors. you said there aren't many that can do that, apparently these are among them. And they use two source resistors, damn, for an extra ten cents they could have just used three. I also caught an error where they claim to have -85 volts on the base of a transistors with the emitter grounded. But I am used to that shit.

Also, they are drawn as depletion mode devices, yet have a higher source for the drivers. (they ARE depletion mode right ?) OK, that might just be for better linearity but I bet those outputs REALLY don't like hard clipping. Yeah sure don't do that, but it should not be destructive. shit happens like when switching sources or whatever sometimes.

And neither should brownouts. I don't see how running off a variac at reduced voltage would fry this thing. But my search indicates that such devices were also used in some serious high end monoblocks but I did not get those model numbers. Any come to mind I can look up ?

Sony failure modes have been a bitch for a long time. Seems like they used to love selling alot of parts.

 

[Trevor Wilson]

On 24/08/2015 7:55 PM, jurb6006@gmail.com wrote:

"**One more trap for the uninitiated: NEVER crank up an audio amplifier
fitted with V-FET output devices on a variable auto-transformer.
Destruction will be assured under such a test. "

You mean class D I assume ?

**No.

--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Phil Allison]

pf...@aol.com wrote:

Phil Allison wrote:

Had one like that the other day, a valve amp from the 1960s
that had not been run in the last 10 years.
With all valves removed, the AC supply current rose quickly
with each voltage increases and then slowly fell back.


How would that happen unless the unit had a solid-state rectifier?

** It wouldn't.

Soon as silicon diodes became available in the 1960s, designers used them to replace the inefficient and troublesome valve rectifiers in TVs and amps of all kinds. A further efficiency was obtained by using four diode bridges or two diode voltage doublers and eliminating the centre-tapped secondary needed with valve "full wave" rectifiers.

Typically that quick-rise, slow-fall is due to the
"empty" caps accepting inrush, then as they charge
becoming more 'resistant-for-lack-of-a-better-word'.

** Inrush surges as you increment a variac are normally very brief, but in this case they were sustained, taking many minutes each time for the current to fall back to the residual being drawn by the primary of the transformer. I well knew this meant dodgy filter electros.

On my bench, the variac's output is followed by a 3.5 digit current meter with 1mA resolution and a (safety isolated) waveform output to view on a scope. 100Hz current pulses being drawn by filter electros appear very distinct from the steady 50Hz current draw of the power transformer.

With good electros and no valves installed (or the standby off), current pulses should be absent.

Put another way, it is a bad idea to allow a tube amp to
make B+ without a load. Many marginally designed amps ran
their filer caps at very near their ratings - and both
vintage and marginally capable caps running unloaded might
just go *POOF*.

** It's something to be aware of and one should check electro voltage ratings to make sure they are not being exceeded in this condition.

BTW: I have come across a few modern valve amps where electros further down the chain from the main filters were of such lower voltage that removing some of the preamp valves caused their rating to be exceeded.


..... Phil

 

[Trevor Wilson]

On 25/08/2015 9:32 AM, jurb6006@gmail.com wrote:

"** V-FETs were TO3 linear devices made and used by Sony and Yamaha
in a >number of hi-fi amplifers back in the mid 1970s. Common part
numbers were >2SK82 and >2SJ28."

I'll take your word for it, but that doesn't explain why. I dug
around a little and found the TA-8650 which is 80 watts per and uses
three complementary sets of 2SK60 and 2SJ18 per. Seems like alot for
80 watts but who am I to bitch about something being a bit overbuilt
?

**The 2SK60/2SJ18 are rated at a pitiful 5 Amps, which is why so many
are required. Here is what is printed on page 6 of the service manual of
the Sony TAN7:

"Note:
1. Apply the rated ac line Voltage to the set directly. Do not increase
the Voltage gradually by using a variable transformer or other such
instrument; this will cause a V-FET failure."

There are a couple of things about that amp that don't make a hell of
alot of sense. First of all, they are current sharing without source
resistors, at least separate source resistors. you said there aren't
many that can do that, apparently these are among them. And they use
two source resistors, damn, for an extra ten cents they could have
just used three. I also caught an error where they claim to have -85
volts on the base of a transistors with the emitter grounded. But I
am used to that shit.

Also, they are drawn as depletion mode devices, yet have a higher
source for the drivers. (they ARE depletion mode right ?) OK, that
might just be for better linearity but I bet those outputs REALLY
don't like hard clipping. Yeah sure don't do that, but it should not
be destructive. shit happens like when switching sources or whatever
sometimes.

And neither should brownouts. I don't see how running off a variac at
reduced voltage would fry this thing. But my search indicates that
such devices were also used in some serious high end monoblocks but I
did not get those model numbers. Any come to mind I can look up ?

**I recall my first V-FET amp repair. I did not pay attention to the
Sony Warning. A new set of V-FETs and I was good to go, after blowing up
the first set. Every V-FET amp since has been VERY carefully dealt with.
No bench failures since the first one. The most expensive and unique amp
I've ever seen was the Yamaha B1. It was a beast, with over-sized
V-FETs. They LOOK like To-3 devices, until you see one in the flesh:

http://www.thevintageknob.org/yamaha-B-1.html

http://upload.review33.com/avforum/201507/201507092155125540.jpg

Sony failure modes have been a bitch for a long time. Seems like they
used to love selling alot of parts.

**No different to any other company.



--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Dave Platt]

In article <d41uv0FspljU1@mid.individual.net>,
Trevor Wilson <trevor@SPAMBLOCKrageaudio.com.au> wrote:

"Note:
1. Apply the rated ac line Voltage to the set directly. Do not increase
the Voltage gradually by using a variable transformer or other such
instrument; this will cause a V-FET failure."

**I recall my first V-FET amp repair. I did not pay attention to the
Sony Warning. A new set of V-FETs and I was good to go, after blowing up
the first set. Every V-FET amp since has been VERY carefully dealt with.
No bench failures since the first one. The most expensive and unique amp
I've ever seen was the Yamaha B1. It was a beast, with over-sized
V-FETs. They LOOK like To-3 devices, until you see one in the flesh:

That feels to me more like a design flaw in the circuitry... something
which mis-biases the VFETs if the rail voltages don't come up and down
at the prescribed speeds. Possibly the gate voltages rise faster than
the drains, and the gate/source or gain/drain voltage limit is
exceeded?

A power amplifier that will fry its output stage if the mains voltage
bounces around (say, during a "brownout" of commercial power) really
seems like a poor design to market.

 

[Phil Allison]

Dave Platt wrote:

"Note:
1. Apply the rated ac line Voltage to the set directly. Do not increase
the Voltage gradually by using a variable transformer or other such
instrument; this will cause a V-FET failure."


That feels to me more like a design flaw in the circuitry... something
which mis-biases the VFETs if the rail voltages don't come up and down
at the prescribed speeds. Possibly the gate voltages rise faster than
the drains, and the gate/source or gain/drain voltage limit is
exceeded?

** Take a look at the power stage schem here:

http://www.angelfire.com/sd/paulkemble/sound8f.html

Note the unusual crosswise connection from the drive stage to the six V-FET followers.

The V-FETs used behave like J-FETs requiring reverse polarity gate bias voltage to turn them off, with zero bias they conduct heavily - the opposite of the case with transistors and MOSFETS.

So, when using a variac, a condition must exist where the needed reverse bias is missing but the V-FETs have enough DC supply available to fry themselves if held there.



..... Phil

A power amplifier that will fry its output stage if the mains voltage
bounces around (say, during a "brownout" of commercial power) really
seems like a poor design to market.