Ampeg SVT 6146 open screen & Plate resistors

[g1]

I'd think the current being balanced for the 2 sides of the OT would be preferred.

[DrGonz78]

So I think I will probably call AES Monday to swap these tubes out. I will ask them to kindly to put aside these 6 tubes in case I need to swap any out. It's nice living 20 minutes away from that place.

[DrGonz78]

So I took a trip over to AES today and got 6 more of these tubes. I set the bias supply voltage to -52v for both sides and started testing a pair at a time. The first sextet that I got last week measured at 16.7, 19.5, 25.1, 26, 30.1 & 33.5ma. The sextet that I just picked up today is 15.5, 17, 19.6, 32.3, 32.5, & 33ma. So obviously neither set was matched but now I can select a sextet out the twelve tubes and return the other six that I don't use.

So I have this sextet where the bias is 26, 30, 32.3, 32.5, 33, & 33.5ma. The other sextet is 15.5, 16.7, 17, 19.5, 19.6, & 25.1ma. Which one should I choose? I am leaning towards the first set.

Also, AES does actually match the tubes on their Apex tester. So that part about the factory doing the matching and putting labels on the box was not accurate. As ususal the true tube tester is the amplifier. I have to give a shout out to AES for having the greatest customer service and for taking great care to make things right.

[Chuck H]

Are these tubes being tested for current in the amp and in the same socket? Each tube should be plugged into the same socket one at a time. Just wanting to be sure there's nothing un-adjusted (or a leaky cap) about the amp that may be skewing the results.

[Leo_Gnardo]

Which one should I choose? I am leaning towards the first set.

Yes, the percentage spread is lower on the first set, and with lower Pc readings less likely that one will "run away".

Implementing their tube testing system was a master stroke for AES, it upped their game considerably. However, there is one more thought, at what plate / screen grid voltage are they testing? We know 700V and 350V are what the tube is going to see in an SVT, so testing at those voltages would be primo. I know of only one tube dealer that will go to the trouble of testing at specified operating voltages, Jim McShane in Chicago, but I doubt he has a quantity of 6146B to offer. For other tubes, he sticks with ones he feels he can trust. Apparently Jim got bitten by some lousy JJ's earlier in the century, and he still avoids them to this day. No JJ power tubes to be found there last I checked.

[nevetslab]

Yes, the percentage spread is lower on the first set, and with lower Pc readings less likely that one will "run away".

I likewise concur with Leo. DrGonz78 did state he had preset the bias on both upper/lower halves, so I presume the same TWO tube sockets were being used in getting the current readings on each pair. That's how I do it when I'm getting my initial readings. I'll select the triad of tubes for each half from that, put them all in, then will do a final biasing for equal average current, let it burn in a while, and do a final trim afterwards. I forgot which Ampeg amp you're working on. Most of our inventory here are SVT-CL's, so each tube has its' own 10 ohm cathode resistor. On SVT-VR's, I made a clear plexiglass test panel with probe openings to get across the plate resistors with all six tubes in place, or in pairs as well (using long EZ-Hook probes). Earlier SVT's also have plate resistors, accessible from the bottom.

[DrGonz78]

Are these tubes being tested for current in the amp and in the same socket? Each tube should be plugged into the same socket one at a time. Just wanting to be sure there's nothing un-adjusted (or a leaky cap) about the amp that may be skewing the results.

I agree completely. So I have had each set in the amp and have recorded each socket being tested. The socket I use is one on each side where the socket tested with is the one closest to output transformer. I think tonight I am going to record all the reading once more but reverse each tube. So by the end I will have readings for each tube on both sides at the same socket to compare any differences. I replaced the coupling caps C5 & C6 since they were old and original. So I know that there is no DC leaking onto the grid where the bias voltage is fed to the output tubes. The bias voltage is rock steady!! I also changed any resistors related to the bias circuit which included 1 ohm 1% cathode resistors for testing bias. With all that said, one more pass of all twelve tubes is a good idea. Then I will have bias readings of all tubes, not only from one socket but two!! Like Enzo always says never think of a reason not to test or check something, just do it. I will report back later how all that pans out and by then I think I will be pretty much ready to pick a sextet.

[DrGonz78]

Yes, the percentage spread is lower on the first set, and with lower Pc readings less likely that one will "run away".

Implementing their tube testing system was a master stroke for AES, it upped their game considerably. However, there is one more thought, at what plate / screen grid voltage are they testing? We know 700V and 350V are what the tube is going to see in an SVT, so testing at those voltages would be primo. I know of only one tube dealer that will go to the trouble of testing at specified operating voltages, Jim McShane in Chicago, but I doubt he has a quantity of 6146B to offer. For other tubes, he sticks with ones he feels he can trust. Apparently Jim got bitten by some lousy JJ's earlier in the century, and he still avoids them to this day. No JJ power tubes to be found there last I checked.

While chatting with customer support on the phone they had questions for me from their supervisor. You know they gotta make sure I am not doing something wrong. Towards the end of talking about this whole situation I had stated that perhaps the plate voltage used on the Apex system is much lower than what this ampeg is putting out. I think that led to their supervisor finally agreeing with the results.

[Leo_Gnardo]

I will report back later how all that pans out and by then I think I will be pretty much ready to pick a sextet.

With all that testing, you will have identified the "safe sextet"

[DrGonz78]

With all that testing, you will have identified the "safe sextet"

So for curiosity sake I should at least test a couple of pairs reversed from their original sockets to see if there is a difference. I had checked all the original six tubes last week. Last night I ran a test of those same six tubes again along with the new sextet. So yes I have been pretty thorough.

[DrGonz78]

I likewise concur with Leo. DrGonz78 did state he had preset the bias on both upper/lower halves, so I presume the same TWO tube sockets were being used in getting the current readings on each pair. That's how I do it when I'm getting my initial readings. I'll select the triad of tubes for each half from that, put them all in, then will do a final biasing for equal average current, let it burn in a while, and do a final trim afterwards. I forgot which Ampeg amp you're working on. Most of our inventory here are SVT-CL's, so each tube has its' own 10 ohm cathode resistor. On SVT-VR's, I made a clear plexiglass test panel with probe openings to get across the plate resistors with all six tubes in place, or in pairs as well (using long EZ-Hook probes). Earlier SVT's also have plate resistors, accessible from the bottom.

The 6146 version has 1 ohm resistors for the cathode resistors on this model. Yes I had thought about having one for every tube but that would be hard to implement with the circuit board. I don't really feel like hacking the traces on the circuit board or anything like that. All plate resistors and screen resistors have been replaced as new ones. There is also K1 and K2 test points on the chassis for easily taking readings across the 1 ohm cathode resistors.

https://ampeg.com/support/files/Sche...20Version).pdf

[DrGonz78]

So I finally got 6 tubes in the amp and biased to 72ma each side. Gonna start audio testing tomorrow and hopefully this beast of an amp will stay off my bench for the rest of my life.

Quick question about the 10A fuse. It calls for a GLH 10A fuse which an obsolete part. The damn fuse in there was a 32v 10A fuse, of course! I think GLH is a fast acting type so I put an AGC fast acting 10A 250v in place. Is this correct?

[Leo_Gnardo]

I think GLH is a fast acting type so I put an AGC fast acting 10A 250v in place. Is this correct?

So called "normal" fuse, that is not a slow or delayed type is what I've been using in SVT for many years. Measuring current while running the amp full square wave into rated load, IIRC 7 amps, maybe a pinch more. These days with a proliferation of fast, super fast, time delay, slow acting, etc the choice isn't clear sometimes, what is equivalent to the "normal" fuses of yesteryear. I think you're doing as well as you can with your choice. Hope your SVT passes the audition!