Ampeg SVT 6146 open screen & Plate resistors

[DrGonz78]

So I am ready to fire this bad boy up and finish it off but everything is telling me to beware of these output tubes. My other thread involved reattaching the plate caps so they must have been cooked pretty good. Unfortunately these tubes are not that easy to test on my Hickock tube tester since the plug on the tester for the plate cap will not fit. My worry is that I just discovered that one of the 6146 plate resistors is open and two of the screen resistors are open. This does not make me feel very confident in the current set of output tubes at all. I have an order in for the screen resistors and I am going to replace any bias related resistors, including the 1 ohm cathode resistors too. Is it possible for a plate resistor to be open and the related tube to still be good?

I see a sextet of Chinese 6146b tubes on tubes and more. Are these things reliable? Probably the only way that I will be able to get a matching sextet is by ordering them new.
https://www.tubesandmore.com/products/6146b-chinese

[Chuck H]

Can we assume you've tested for proper bias voltage at the sockets? If it hasn't already been done I would probably replace the bias circuit capacitors as well.

[DrGonz78]

The amp has been fully recapped already, that was another thread. I ran the amp when I first got it and it sounded just awful. It sounded okay at 1 on the volume but turning it up higher was badly distorted. Right now I plan on soldering in 2 open screen resistors and I have already replaced the open screen resistor. I probably could test the tubes on my tester with alligator clip and wrap some wire around the plate cap. Definitely test the tubes that have open resistors associated with them.

[glebert]

Testers don't run at high plate voltages, do they? Seems like you could use the tester, still have some bad tubes, and redo all the damage that you just fixed. Wikipedia tells me Ampeg switched from the 6145b to the 6550 because the 6146b was unstable in that environment. Actually a fair amount of technical info about that issue in the wiki page, don't know how accurate it is.

Anyway, sounds to me like it needs new tubes. Is this yours or a customers?

[DrGonz78]

It's a customers. I have crossed my fingers the whole way through that the perhaps the output tubes might turn out good. I will probably order up some of those Chinese 6146b tubes from antique electronics supply. In fact I could even drive over to their Tempe office to pick them up.

[DrGonz78]

One more thing... none of the open resistors were exploded. They looked normal but just measured open.

[g1]

First off, you can't be positive that the tubes currently in the current positions are the ones that opened the resistors.
But, in the case that you did know this for a fact, I'd think the plate R's might be able to open in certain circumstances other than a bad tube. Even for screen R's, it usually means the tube is bad, but you never know if it was pre-stressed by a previous bad tube and got pushed over the edge by a good replacement tube.

[52 Bill]

I don't remember, do you have to run this with all 6 power tubes to do a start up test?

[nevetslab]

I don't remember, do you have to run this with all 6 power tubes to do a start up test?

I wlll pre-set the bias voltage on both upper/lower drivers, to around the nominal bias voltage, being about -45VDC. That still may be enough to cause too much current to flow, depending on the tubes' transconductance. I'll tweak the bias to be equal while getting 'into the ballpark of around 23mA per tube.

When I'm first powering up an SVT, of any generation, I begin with just one pair of power tubes installed, and monitor the plate or cathode current during powering up. It's a lot easier on SVT-CL's, which have a very effective protection circuit that will shut down the HT supply if a bad tube is encountered (sensing the excessive cathode current flowing in the bad tube). On others, I still do it with one pair at a time, letting them warm up in S/B. I use a small fuse value in the AC mains, which under normal operation, will hold and not blow switching out of S/B if all is ok. 3A Fast has worked in that mode. Usually when one of the tubes is faulty, like internal short, you'll see fireworks in the bottle, during which I either beat the fuse by switching back into S/B, or the mains fuse blow as it was quicker then me and pulling enough current to blow the fuse.

With each pair of tubes that I try, I'll record their plate current from the preset bias voltage applied, and continue until I've successfully powered up with all six tubes, in one pair-at-a-time. It also lets me juggle the tubes for a best-balance set, then I'll put all six tubes in and try again, to see what I get in operation mode. The bias will need to be tweaked again, as with all six tubes installed, they pull more current than in single pairs.

And, as g1 said, you don't know for certain the tubes in the new position are the ones that caused a plate or screen resistor to open. I always check the values of them, as well as look at their physical condition. It's always a nervous state switching them into operate for the first time, while your body is tense, awaiting an explosion that may or may not come.

[DrGonz78]

Another issue with these tubes is the repair I did on the plate caps. Most of them are reattached very well but one of the tubes the plate cap came off again. I did epoxy it up again and put it back together. Still it becomes a question of mechanical stability and of how many times can I safely remove the plugs from the plate caps. My customer is definitely fine with ordering up a sextet of new tubes and that will buy me piece of mind that we are setting the amp up right.

I should probably follow Nevetslab's procedure even with fresh new tubes. Two tubes at a time for testing each set to make sure they are truly good and there are no duds. My idea is to treat the new tubes in the same manner as the used tubes to make sure there are no bad tubes. The good thing is that my customer understands that the 6146b tubes in Ampeg were not the best suitable for the application. His intention with this amp is to play it at the gigs where he won't be pushing it hard. He has another amp that he'll use for those really loud gigs too.

Thanks for all the feedback so far in getting this amp back up to shape. All the advice will make the last stage of the repair a success I am sure.

[DrGonz78]

So I am powering up the amp with one set of new 6146b tubes. Plate voltage is @ 710v, which will probably come down a tad with all 6 installed. Anyway the first set of tubes are biased at 23ma. Bias set at -50.5v on one side and -51.8 on the other. I ordered these tube from AES and wrote a note to have the matched quad match the matched set. (That was their advice on how to order a matching sextet). Next set of tubes measure 40ma and 28ma. Last set was 36.4ma and 35.4ma.

So should I put the 28, 23 @ 23ma on one side and the other side will have the 39, 36 & 35 ma tubes? Then set bias for hottest tube to about 23ma on each side. Then install all tubes and bias each side to 72ma. Or should think about calling AES tomorrow to see if I can get a closer matching sextet? Even then who knows what I will end up with.

BTW G1 asked about the voltage at the 2nd node and it's around 440v, so that explains why that 450v cap held on before. That large dropping resistor was on point resistance wise so they got really lucky. Would have expected it to be higher though.

[DrGonz78]

Also note that I replaced all screen resistors, all plate resistors and the power tubes 1 ohm 1% cathode resistors. I replaced the resistors in the bias circuit and of course the bias voltage cap was done in the full cap job. So everything should pretty accurate.

[DrGonz78]

So I called AES with some questions about how they match these tubes. Apparently they are matched at the factory and sent to them labelled appropriately. He talked about the Apex matching that they do but was unsure if these tubes could even be matched on their system. I can always drive over next week and swap out the tubes to try my luck again. Or I can go with these tubes and match them side to side the best that I can.

[DrGonz78]

So does this sound like a viable solution? >>> "So should I put the 28, 23 & 23ma on one side and the other side will have the 39, 36 & 35 ma tubes? Then set bias for hottest tube to about 23ma on each side. Then install all tubes and bias each side to 72ma."

Not sure if this is what would be recommended by other's here that have experience working on these amps? My gut is telling me to go for it but I always have to question my gut. Especially when my gut has been growing bigger and bigger all my life.

[Chuck H]

I read your original post about trying to match the tubes you have for each half of the push/pull respectively. I don't have the answer to be honest. I've read that the SVT can really tear itself up when not bias adjusted properly. With that I would think starting with an imbalance is a bad idea. OTOH arranging the tubes for more of a balance per side comes with other issues that we know are not good for the tubes themselves. I just don't know. Maybe the wrong things are in place to solve this issue. Like asking whether 2+3 or 8-3 is closer to 4.