Homebrew Gibson GA-5T stopped working, blowing fuses

[rayandkerry]

Hi all.

I built a sweet little Gibson GA-5T amp, detailed here: https://el34world.com/Forum/index.php?topic=26237

It just died last week.  It's my favorite amp.  Waaaaaaaaah!!!!!!!

Ok, now that I got that out, I wanna fix it.  ;-)

When I turn the amp on, with the standby switch in the off/standby position (i.e. no high voltage going to the tubes, only the heaters getting power), the amp light goes on for a half-second and the fuse blows. 

So I removed the tubes and one of the power tubes i was using had a rattling in it.  So I tried one more time without any tubes in at all and same result...the amp light goes on for a half-second and the fuse blows. 

So I built myself a light bulb current limiter, and sure enough a 200w incandescent bulb glows bright when the amp is on standby with no tubes.

So I disconnected the wires from the PT...which was pulled from an old Webcor 333-2 record player...and took some resistance readings. 

120v primary wires  = 12.1 ohms
390V secondary wires = 168.7 ohms
6.3v secondary wires = 0.7 ohms

I also checked the two solid state diodes I'm using as a rectifier.   They check high resistance in one direction and no resistance in the other, so they seem good.

Does this mean my PT is the problem?  I don't have any original resistance specs for the PT...it's an ancient thing pulled from a $15 Goodwill find.

Any help/guidance you can offer will be greatly appreciated!  This is the perfect home amp and I miss it! ;-)

Ray

[thetragichero]

lift each secondary pair, cap off, and see which one makes the light burn bright. measure resistance between the two heater strings with pt disconnected

[shooter]

Agree ^
or just disconnect all PT secondary and test.
guessing you have a "touch" somewhere; a stray strand of wire between pins, to ground

[rayandkerry]

Thanks! 

But not sure I understand what you're both telling me to do.

lift each secondary pair, cap off, and see which one makes the light burn bright. measure resistance between the two heater strings with pt disconnected

The resistance readings listed above are with all leads from the PT disconnected from the circuit.  I measure 0.7 ohms between the two heater wires on the PT.

Additionally, are you saying i need to:
1. hook up the primary and cap off all the secondary
2. Plug into light bulb limiter and turn amp on
3. If the light glows is indicates there's a short in one of the primaries?

Agree ^
or just disconnect all PT secondary and test.
guessing you have a "touch" somewhere; a stray strand of wire between pins, to ground

Ok, with the PT completely disconnected I measure the resistance between the two heater wires that run to all the tubes.  I measured 3.8 ohms, and thought that must be the problem.  But then I pulled the indicator bulb from the socket and measure the bulb resistance ant 3.8 ohms.  With the bulb out, I measure a completely open line when testing between the heater wires going to the tube.  So there is apparently no short between the heater wires in the amp when the PT is disconnected.

[shooter]

Additionally, are you saying i need to:1. hook up the primary and cap off all the secondary2. Plug into light bulb limiter and turn amp on3. If the light glows is indicates there's a short in one of the primaries?

yes
(Note: you're eliminating the "amp" as problem, which part of the PT is moot since it's not really a field repairable item)

[rayandkerry]

Ok.  With each individual secondary wire capped off, the current limiter bulb does not glow.

So I hooked up the heater wires and tried again (still no tubes).  The amp's indicator light comes on as it should and the current limiter bulb again does not lightup.

This seems to infer the issue is in the high voltage chain?

So...at this point do we know that the PT is good (because the current limiting bulb doesn't light when the secondaries are capped off)? 

What do I do next?  If I hook up the high voltage secondaries now, I'm back where I started.

Thanks in advance for any thoughts!

[thetragichero]

time to find the short

[shooter]

I also checked the two solid state diodes I'm using as a rectifier.

guessing you are not using a tube rectifier?


replace both diodes, but DO NOT connect them "downstream" Isolate them, but put your meter + lead there and - lead on ground
power up with limiter

[rayandkerry]

time to find the short

Ok.  But have we eliminated the PT as a problem?  At this point it's gotta be somewhere in the power supply starting at the PT high voltage secondaries, right?

I already tested Tthe rectifier diodes.  Next is the first filter cap with a 220k safety/bleeder resistor across it.  So do i just individually pull each capacitor and test for a short?

[rayandkerry]

guessing you are not using a tube rectifier?

replace both diodes, but DO NOT connect them "downstream" Isolate them, but put your meter + lead there and - lead on ground
power up with limiter

Correct on the rectifier.  And will do.  Gotta dig some new diodes.  What should I expect to see?  And is this testing the diodes or the PT?  When I tested the diodes individually, they both test 0 ohms in one direction and 1.19M and 2.13M ohms in the other direction

[rayandkerry]

replace both diodes, but DO NOT connect them "downstream" Isolate them, but put your meter + lead there and - lead on ground
power up with limiter

Ok got my diodes out.  Am I measuring AC or DC on from downstream side of one diode to ground?  AC? 

[shooter]

    both, but DC is fine, without a filter it's just full wave AC.  I would expect something close to your original no tube value when you built it.


once you prove that;
disconnect all the "taps" so you have just the PS connected  R's and C's
apply power watch light

[rayandkerry]

   both, but DC is fine, without a filter it's just full wave AC.  I would expect something close to your original no tube value when you built it.


once you prove that;
disconnect all the "taps" so you have just the PS connected  R's and C's
apply power watch light

Ha.  I was thinking you could view it either way.  In any case, I just tried both.  Interestingly, the AC was not really working...it read .6v and 2.78v. 

On the DC reading I got 97.8v and 98.2v.  I never took a no load reading when I built it, but the first node of the power supply was 259v when I built it with tubes in.  So I'm not sure how I could get 259v loaded with tubes from two unloaded halves at 98v each.  Is my PT blown?

[shooter]

once you prove that;disconnect all the "taps" so you have just the PS connected  R's and C'sapply power watch light

[rayandkerry]

once you prove that;disconnect all the "taps" so you have just the PS connected  R's and C'sapply power watch light

Working on it, LOL...juggling other life stuff which was rudely getting in the way of this. 

Thanks for the guidance so far.   The reason I ask at this point was because you mentioned "once you prove that"...and I'm not sure what "that" is.  Meaning, if I get similar voltage reading now compared to the reading I got when I built this amp, then I've proven the PT is ok?  Did I interpret you correctly?   Forgive me for being dense...can you clarify what "once you prove that" means

In any case, I didn't think of taking that measurement at the time.  I only measured voltages after hooking everything up and putting tubes in.  So I measure about 200v unloaded now and measured about 260v loaded after the initial build. 

Did I prove "that"?  If I did interpret you correctly, then I think I didn't prove the PT is ok.  I was under the impression that unloaded voltages are higher than loaded ones.  So again, I could not imagine getting 260v loaded from a post rectified, unloaded voltage source of 200v.  Just trying to understand as i go.

But working on the next step... 

[shooter]

I've proven the PT is ok?  Did I interpret you correctly?

you have something drawing too much current, so we're slow walking to that Issue


the PS isn't really a load to the PT, but, if there is a problem in the PS, it will load the PT down


so an "unloaded" (good) PS (no connections to the amp) should read higher than your original "loaded" working values

[rayandkerry]

you have something drawing too much current, so we're slow walking to that Issue

the PS isn't really a load to the PT, but, if there is a problem in the PS, it will load the PT down

so an "unloaded" (good) PS (no connections to the amp) should read higher than your original "loaded" working values

Ok, that was very helpful, thanks!  I did as you said and the current limiter bulb did not light.  Furthermore, each of the points on the power rail range from 301v to 298v...all higher than the 260v loaded reading.  So this means everything's fine up to this point?

I suppose now I hook up the first power node to the OT?  Ugh, I'm scared.   

[thetragichero]

start connecting until one lights up the bulb

[shooter]

+1
although a PS can be bad and you won't know till you LOAD it
so I prefer to start with a small load, the pre amp.  your "unloaded" volts shouldn't drop much

[thetragichero]

i can get on board with that reasoning

[PRR]

...I already tested the rectifier diodes. ...

With a meter, right? 1.5 Volt stress at most. It may be breaking down at the 500 volts it experiences in use. At this point I would really throw the dollar and get two new 1N4007. (Or ten; they are universal in 6V6 and small 6L6 amps.)

I have not read every post, but it is a "weak link in a chain" problem. Disconnect all, then re-connect one at a time. If you connect two parts at once, and it's bad, you don't know which part is bad.

The 190-0-190 winding is 175 Ohms on the sheet and 169 Ohms in your hand. Close enough for mass production. 12 Ohm primary seems likely. Heater windings mostly work, or not. What matters more is conductivity *between* windings.

[rayandkerry]

Ok.  Everything is hooked back up (with replaced diodes), no tubes yet, and the current limiter bulb does not light.

...I already tested the rectifier diodes. ...

With a meter, right? 1.5 Volt stress at most. It may be breaking down at the 500 volts it experiences in use. At this point I would really throw the dollar and get two new 1N4007. (Or ten; they are universal in 6V6 and small 6L6 amps.)

I have not read every post, but it is a "weak link in a chain" problem. Disconnect all, then re-connect one at a time. If you connect two parts at once, and it's bad, you don't know which part is bad.

The 190-0-190 winding is 175 Ohms on the sheet and 169 Ohms in your hand. Close enough for mass production. 12 Ohm primary seems likely. Heater windings mostly work, or not. What matters more is conductivity *between* windings.

Hmmm. I must have lost that page of the webcor documentation for the 333-2.  I have the part number and voltages/amperages...but not the resistance readings.  Thanks for that :-)

And I replaced the diodes at the start base on shooter's suggestion.

+1
although a PS can be bad and you won't know till you LOAD it
so I prefer to start with a small load, the pre amp.  your "unloaded" volts shouldn't drop much

Ok...what now?  Place the first preamp tube in and fire it up to see if the current limiter bulb lights?

[shooter]

put everything up to the PI, leave that and the PA tubes out.


If the light is real dim, plug into the wall and measure the last tap (connected) voltage DC

[rayandkerry]

put everything up to the PI, leave that and the PA tubes out.

If the light is real dim, plug into the wall and measure the last tap (connected) voltage DC

Awesome...will do.

By the way, I've not had the amp hooked up to a speaker so far.  That's not an issue for the OT if I'm not placing any signal through the amp or keeping it on for more than a few seconds, is it?

[shooter]

if that seems good,
Kill power/discharge,
spend quality time with a bright light looking over EVERYTHING that's NOT hooked up, take your time, think to yourself;
"If i was a stray hair, metal shaving, un-trimmed, soldered wire........." where would I hide 

[shooter]

you need a speaker when you have PA tubes in, otherwise you're fine

[rayandkerry]

Ok.  I tried both 12ax7s in the first position.  Each time the heaters got red like normal and nothing bad happened.  :-)

But the current limiter bulb did not light at all.  I'm assuming that's because it's a 200w bulb and were just talking about a single 12ax7?  I tried a 100w also, but no dim glow.

The voltages are still high, but the last node of the power supply  (which feeds the preamp tube) has dropped about 30v.  Although I did not plug directly into the wall.  I was still running through the 200w bulb. 

if that seems good,
Kill power/discharge,
spend quality time with a bright light looking over EVERYTHING that's NOT hooked up, take your time, think to yourself;
"If i was a stray hair, metal shaving, un-trimmed, soldered wire........." where would I hide 

Hahaha.  Now I'm really scared...metal gremlins running through my amp! Ok, putting it away for tonight...im going to need to concentrate clearly for that check.  Need food now ;-)

you need a speaker when you have PA tubes in, otherwise you're fine

Thanks, that makes sense.

Also, thanks so much all the help so far! 

[shooter]

I'm going by this schematic


If you have "other" PA tubes, I'd start with them.
If you can't find anything lingering, loose, touching;
put in the PI and PA tubes
connect speaker
put meter across the cathode resistor for the PA tubes, set to read Volts DC
power up with limiter and get reading, watch light

[rayandkerry]

Thanks Shooter.  I ran through and checked for stray metal pieces...didnt find anything.

I'll put in the PI tube and test first to see nothing blows up and the current limiter bulb doesn't glow.  If that's good, I'll hook up a speaker and try again with the power tubes.  I'll record the voltages and report back.

Thanks again for the help!  Fingers crossed ;-)

[rayandkerry]

Ok.  Got back to this today and things are looking pretty good.  Here are my voltages with all tubes in.

Reading taken - initial build reading - current reading
B1 - 259v - 252v
B2 - 254v - 250v
B3 - 209v - 210v
PI/trem , pin 6 - 155v - 153v
PI/trem, pin 1 - 105v - 86v
Preamp, pin 6 - 142v - 159v
Preamp, pin 1 - 124v - 141v
Cathode resistor - ? - 14v

I played a little bit and everything sounds alright.  The tremolo functions as it did before.  I think I'm good!   

Thanks so much for walking me through this trouble shooting!   Now let me see if I can do a post mortem on what happened...

SUMMARY
First the amp just died and would blow fuses immediately.  Even after I removed all tubes.  One of the 6K6GT power tubes i was using had a rattle in it when shook.  It's a grey smoky glass so I can't see what's rattling around in there.  When hooked the amp up to a current limiter with no tubes, the bulb glowed bright, indicating some kind of large current draw.

At the suggestion of shooter and PRR, I replaced the two solid state 1N4007 diodes I'm using for rectification.  It should be noted that they tested fine using a multimeter (i.e. under low voltage conditions). Then we tested the power transformer and the disconnected heater string for shorts, finding none.

So we tested the voltages for the unloaded power supply using the current limiter.  Voltages were high as expected for an unloaded power supply, and the current limiter bulb did not light, indicating we were ok so far.

Then we moved on to hooking up each individual section to the power supply, one at a time, checking voltages and making sure the current limiter didn't glow at each step.  Once the entire supply was hooked up and proved to be ok, we scoured the amp for stray metal gremlins that might have caused a short somewhere.

Since we didn't find any, we added the first tube (preamp), and tested again.  Voltages looked high for the preamp, but that made sense.  Since the current limiter bulb did not light, we repeated the process with the PI tube added.  Finally I powered up with the power tubes in and got good readings and the amp seems to function fine.

Did I summarize the process correctly, or miss anything?

Ok, now to the cause. 

POSSIBLE CAUSE
I'm guessing my 6k6gt had a catastrophic failure.  Since i can't reall see what's broken, I'm guessing it was either the heater or the plate...but I don't really know.  Then failing power tube drew a bunch of current quickly causing the fuse to blow. 

Based on the fact that the only thing changed were the rectifier diodes, I'm guessing the current rush from the failing tube burned the diodes in such a way as to become a shorted wire under high voltage conditions, thereby basically causing a short right at the start of the power supply, before the first filter cap.

So...can all you gurus comment on whether I've correctly understood what happened and how we located the problem?

And, again...thanks to everyone on this site who helped me (and all the other folks like me on this forum)!

[rayandkerry]

Bump.  Just wanted to see if my post fix summary was correct.  Any thoughts?

[shooter]

You did well
I fixed broke systems for a living, after awhile I quit trying to "understand" all the causalities, I simply fixed it and went rock climbing. best stress, brain clearing medicine ever!   

[rayandkerry]

You did well
I fixed broke systems for a living, after awhile I quit trying to "understand" all the causalities, I simply fixed it and went rock climbing. best stress, brain clearing medicine ever! 

Hah.  Ok fair enough.  I'll still refer back to this thread the next time an anp starts blowing fuses at startup.

Thanks again for the help!

[shooter]

As you should, this is for knowledge, enjoyment, life.  I love the challenge of anything broke, when you do it for $ though, you do the happy dance, take the $'s and live life.  Which brings you here,