Filament problem: help please!

[EFK]

Gibson GA42RVT w/ 6L6GC - has started to cut out while playing, sound fades out for @ 5 or 6 seconds, then comes back. Was initially very intermittent, now more frequently. This amp has DC on filaments. Heater lines from transformer tie into - for lack of a better way to describe it - a little black square @ 1" square w. 4 legs, one from each corner. I have 6.5 V AC into it and this 'black square' apparently transforms this AC to DC. AC power in from transformer is rock solid - does not change. W/ all tubes out (pre and power) the DC current out of the DC legs is 8.3 DC but if I let it run for @ 5 minutes, it slowly climbs to 8.6 DC (and beyond?) Little black box (can you tell I am not an electrician?) does not heat up unloaded. W/ all tubes in, current DC out is 6.3 but the black box starts heating up and gets HOT. Too hot to touch and in fact if I leave my test leads hooked on, the little plastic jackets around the alligator clips almost seem like they may start to melt if they're touching it. Should this thing get hot like this? What is this thing? As it heats up and I let the amp run, the DC current out to the tubes slowly drops: 6, 5.9, 5.8, 5.7, etc. etc. etc. Will sometimes settle around 5.4, sometimes around 4.8, I can watch it for a while and I'll hear a louder hum from the transformer (that seems where it is coming from) and it will cut out completely and the heaters die, then come back... I have swapped tubes in all positions, no change. I have checked for shorts and cannot find any. Re-soldered some connections which looked suspect, still can't find any bad connections. Jiggling tubes, jiggling wires - there does not seem to be any rhyme or reason as to what causes the DC heater voltage to go up and down. I am losing my hair!!!!!! If I turn the amp off and let cool down for 5 minutes or so, turn it on and DC voltage is back to 6.3 and the whole slow death process starts all over again. Any help will be VERY much appreciated. Thank you!

[oc disorder]

The little black box is a bridge rectifier its just 4 diodes in the one package
ready to go...First thing to check is the soldering on the pins of the bridge rectifier. From the schematic you ? posted last year there is an electrolytic
capacitor C43 a 15000uF 10 volt (replaced a heap of these in a Mac board...
talk about difficult to remove but thats another story) which may be worth changing ... do the relays work ok ? 'cause they also run from the supply.
May be worth measuring the heater voltage while turning them on and off.
The marshall DSL +TSL had this problem Enzo reccomends removing the bridge rectifier and cleaning (scraping) the pins and resoldering it back.
Search for "Marshall DSL401 - Fading Out" .......sounds similar to me..

"Education is what you get when you absorb Enzo's posts"

[stokes]

Disconnect the cap that is filtering the dc out of the rectifier (little black box).Sounds like it is leaky.If the dc voltage stays steady that is the problem.

[EFK]

I thought about the 'black box' (I shall henceforth call it a rectifier - thanks for informing me) being loose as I noticed a few times when the heaters would die, if I pushed on it or tapped it w/ a stick the voltage would come back to @ 4.8 or 5. However, as I mentioned, it did the same thing when I tapped just about everything else too. So, I did resolder the two DC pins (the 2 AC side looked really good and like I mentioned, the AC current coming in and on the legs was solid) but it seemed to make no difference. The DC lines, positive and negative, run directly to the tabs on V7 power tube socket; there is nothing between the rectifier DC pin turrets and the power tube. I'll have to find this cap - I think it's the huge brown one under the standby switch right next to the rectifier. The terminals on the board read Fil. DC - and Fil. DC+. If I disconnect it should I try running the amp w/o it, or just check voltage on standby? Tubes in or out? Is it normal for this rectifier to get so hot? Thanks for the quick reply guys!

[EFK]

Hmm. I'll be darned if I can find a C43 anywhere on that board. The big brown cap right next to the rectifier is C30. In the photo above, the yellow line out of the rectifier is direct to pin 2 of V7 and the green line is direct to pin 7 of V7. From there the heater wires daisy chain directly to V6, then to the pre tubes V5 through V1 right down the line - absolutely nothing else in the circuit as far as I can tell!

[oc disorder]

Its not normal for it to run hot, warm maybe...
Just wondering if it has become damaged..
A replacement can be obtained from mouser this one is more than adaquate
Mouser Part #: 583-BR102
Manufacturer Part #: BR102
Manufacturer: Rectron
Description: Rectifiers - Bridge 10A 200V
Might as well replace the capacitor as well
If it is the "brown one" its an axial type electrolytic capacitor that is the leads come out either end as opposed to a radial type where the leads come out one end.
A mouser radial type
Mouser Part #: 647-UVZ1A153MHD
Manufacturer Part #: UVZ1A153MHD
Manufacturer: Nichicon
Description: Radial Aluminum Electrolytic Capacitors - 105 Degree 15000uF 10V 105c 18x35.5 20% 7.5LS
A Mouser axial type
Mouser Part #: 594-2222-138-14153
Manufacturer Part #: MAL213814153E3
Manufacturer: Vishay/BC Components
Description: Miniature Axial Electrolytic Capacitors 15000uF 10V 20%

The main difference apart from the appearance is the price and a 13 week wait for the axial type.

The radial type would be fine you just have to be creative how you mount it.

These should be less than 10 bucks including postage.

One other thing have you tried different preamp tubes ?

Try removing them one by one while measuring the low voltage replacing them after the measurement and moving to the next one.

I had a regulator that supplied some pre tubes which was shutting down
One of the valves supplied had some weird fault...This was not a Gibson amp but was also made in China.

Just read your additional post either they forgot it or it is being elusive.
Sometimes due to revisions the schematic component numbers do not match the product.

Look for a capacitor 15000 10v only one lead to the positive of the supply the other goes to ground. Probably a small Radial type then.

From what you say about "tapping it" it does look like a dry joint to me or
a cracked joint which is aggravated by heat.

I would resolder all the heater connections no matter how good they look.

Possible the fault is in the Bridge rect as it has been running so hot.

Well I would waste 10 bucks just to eliminate those possibilities.

[EFK]

That big brown sucker is a radial, which I guess will make my life easier!

Yep, I have swapped all sorts of tubes in and out - that was my first thought, a funky tube - but nothing seemed to make a difference.

Just for curiosity's sake, is there some reason the 6.3 AC lines from the transformer aren't running directly to the V7 (and thence down the line) socket filament tabs? Why go through the hassle to convert to DC? Won't the heaters run on either?

[Enzo]

The heaters will run on either, sure. But the AC current offers more chance of injecting hum into a high gain input stage than DC does. Many amps use DC for the heaters of the input stages of tube preamps. It is one method of reducing hum.

[EFK]

So could that be a valid method by which to isolate in which part of the circuit the problem lies? Temporarily hook the transformer 6.3 heater lines to the filament lines directly (after desoldering the rectifier) - amp may hum (although this sure is not a high-gain amp, nor do I care for it to be so!) but if the heater voltage remains steady then the prob. can't be in the tube socket wiring nor the tubes. ??? Any problems or cautions, damage etc by doing this?

[oc disorder]

Go for it !

[Enzo]

The initial stage is relatively high gain whether the amp is overall or not.

You could try that, but really, I would check your bridge. You getting the AC across its AC terminals? If not, find out why. You getting DC across its DC terminals? If not, then it is bad. Check the four sides, one is open I'd bet. You have DC there, but none at the tube? Check continuity from the + side of the bridge to the tube socket. ALso check the - side of the bridge to the tube socket. If one side or the other lacks continuity then there is the problem. There may be a series dropping resistor or something, but that is still not an open.

Check the tube socket, if one of the heater pins is spread open, it may not be making contact with the tube pins.

[EFK]

Well I just hooked the AC directly to the heater lines. Absolutely perfect voltage - no flucuations at all. Noisy though. So, I just ordered replacement cap and rectifier from Mouser. Thanks guys! I'll see how this goes...

[EFK]

To follow up, I replaced the rectifier first and tested that out and it's working beautifully, steady voltage. No need to replace the cap as the rectifier was obviously the problem, but I've got extras now just in case.... Thanks again everyone.