Ampeg SVT Classic Issue

[Seawolf]

I'm working on an Ampeg SVT for someone. I've only been working on amps for about 2 years now so I'm certainly no expert. I know enough to fix most problems, keep from electrocuting myself, and keep from causing further damage to the amp. Anyway, if some of you aren't familiar with these, there are 2 separate sets of 3 6550 output tubes. Each set has its own bias network. These are about the easiest amps on the planet to bias. One of the output tubes is burned out. Doesn't light up, getting on top of the tube is white, not silver so it's legit shot. The guy plays 15-20 hours a week on the amp, but the tubes are only 6 months old. I suspect its just a bad tube, but wanted to rule out other things that might have caused it. One theory is the owner told me the previous guy was pretty incompetent. He had the guy re-tube and re-bias it when he serviced it. I'm wondering if he might have thrown an un-matched tube in there that was getting dealt more current than the rest and was biased way hotter than it should have been. Another theory is maybe a screen resistor had blown previously when the old tube shorted and instead of finding that and fixing it he just threw a new tube in there and the same thing happened? Looking at the schematic, the screen resistors are 22 ohm 1 watt resistors in case that helps anybody. Any other theories are welcome. The amp sounds fine as is, starts to get a bit lower in output after being pushed for a while and can get a little distorted after a while too, which are both things I would expect being down a tube. Another thing I wanted to ask is how reliable the bias pots/lights are on these amps. Are they kind of like the idiot lights in a car or can I safely rely on them to bias the amp? Am I wise to still verify the bias the old fashioned way?

[Stevem]

1). These amps have a specified way to set the bias as you may know.

2) the output stage has a 5 ohm 5 watt resistor on each tube that is there to blow out if a output tube shorts out yet still allow the amp to play on the other still functional tubes.

3) its critical that at least 2 of the 3 output tubes on each side of the OT be matched to within 5 ma and the 3rd tube be within 15% of those 2!

A tube that shows white has lost its vacuum which could be for several reasons , the most likley is physical abuse or getting wet by something while hot.

[Seawolf]

1). These amps have a specified way to set the bias as you may know.

2) the output stage has a 5 ohm 5 watt resistor on each tube that is there to blow out if a output tube shorts out yet still allow the amp to play on the other still functional tubes.

3) its critical that at least 2 of the 3 output tubes on each side of the OT be matched to within 5 ma and the 3rd tube be within 15% of those 2!

A tube that shows white has lost its vacuum which could be for several reasons , the most likley is physical abuse or getting wet by something while hot.

Thanks Steve. Yes, I'm aware of the bias setting. It's super simple and I wish it was always that easy. If I ever design my own amp, I'm going to borrow Ampeg's design for the bias circuit. I may be showing my inexperience here, but I thought the 5 ohm 5 watt resistor was the plate resistor and involved in setting the voltage gain, and that the 22 ohm resistor from pin 4 is the screen resistor which would be the one that blows if the tube shorts. Am I wrong in this observation? I figured the tube just being bad or got damaged at some point as the most likely scenario but I wanted to rule out other causes before I ordered a matched triplet for over $100.

[tubeswell]

The gassed out tube could have happened for any number of reasons.

But, the standard checklist for these is check for open Rg2s, and getting the bias right. So yank the output tube board out and replace any open Rg2s.

While you're in there, do yourself and your client a favour and remove the spade terminal connectors from the main heater wires (from the filament transformer) which connect to the output tube board, and solder these leads directly to the board spades. These connectors see a little over 11A in normal operation, but this is way higher at startup (when the filaments are cold), and added resistance from oxidation of these terminal connectors over time causes high voltage spikes at startup (which plays havoc with the other circuits fed from the filament PT), and which, coupled with the connectors working loose over time, results in charring around the connectors where they arc. Fixing this can save a whole lot of problems.

Then make yourself a bias probe where you can measure the plate, screen and grid voltage and the tube current for each tube. If you don't get the 6550 pairs arranged so that the overall tube current is the same on both sides, then the bias comparator window won't work properly.

When you put humpty back together again, don't mix up the wire connectors.

The SVT CL circuit is not the same as the SVT. There is no 5R resistor in the SVT-CL. (Also, refer to the technical bulletin recommending to change out the Rg2s to 220R 0.5W fusible)

[Seawolf]

The gassed out tube could have happened for any number of reasons.

But, the standard checklist for these is check for open Rg2s, and getting the bias right. So yank the output tube board out and replace any open Rg2s.

While you're in there, do yourself and your client a favour and remove the spade terminal connectors from the main heater wires (from the filament transformer) which connect to the output tube board, and solder these leads directly to the board spades. These connectors see a little over 11A in normal operation, but this is way higher at startup (when the filaments are cold), and added resistance from oxidation of these terminal connectors over time causes high voltage spikes at startup (which plays havoc with the other circuits fed from the filament PT), and which, coupled with the connectors working loose over time, results in charring around the connectors where they arc. Fixing this can save a whole lot of problems.

Then make yourself a bias probe where you can measure the plate, screen and grid voltage and the tube current for each tube. If you don't get the 6550 pairs arranged so that the overall tube current is the same on both sides, then the bias comparator window won't work properly.

When you put humpty back together again, don't mix up the wire connectors.

The SVT CL circuit is not the same as the SVT. There is no 5R resistor in the SVT-CL. (Also, refer to the technical bulletin recommending to change out the Rg2s to 220R 0.5W fusible)

Thanks so much. That's a ton of great info. Do you have a schematic or plans for the bias probe that you built? I currently just manually calculate everything by getting the voltage drop by subtracting plate voltage from the OT center tap then dividing the V drop by the resistance between the center tap and plates, then just do all the math from there. Multiply the plate current by the plate voltage to get the plate dissipation, which in this case seems it should be 29.4 watts if shooting for 70% of the max plate dissipation of 42 watts for a 6550. But I'm sure making a bias probe would make my life much easier. And I will definitely take your advice on the spade terminals and replacing the screen resistors with 220 ohm 1/2 watts.

When you say make sure the overall tube current is the same on both sides, I'm guessing you mean on both sets of 3 tubes? Also, I'm guessing I need to order a whole new matched triplet since that tube is shot? The tubes are only 6 months old so I'd hate to have to order 3 of them, but I'm not sure I see a way around it. Also, what would cause those Rg2's to be open? Just a combination of heat and cold solder joints? And from that bulletin, should I also remove the 1n4007 diodes as it suggests? What does the diode do exactly? Thanks again, you seem to be exactly the type of SVT expert I was hoping would find this thread. I'm looking forward to the day when I've got enough knowledge and experience to help other people out like this.

[tubeswell]

https://www.youtube.com/watch?v=Gf5yrt91vmc