SVT CL Bias issue: Red LED behavior

[glebert]

I'm working on my SVT CL that I got as a project with a couple bad tubes. I'm trying to get the amp biased but have weird behavior on the red LEDs of the bias monitoring. No matter how many tubes I have in (2,4,6) with bias all the way down both LEDs are off, as I bring it up the red LED comes on first, pretty much at the point the tubes start conducting, and as the bias is increased then the green comes on along with it (green only comes on with 6 tubes). From watching the video from Ampeg the red shouldn't come on until an overbias situation, after the green comes on by itself. I never get the green by itself. Anyone had this situation?

I see on the schematic (page 5 https://www.thetubestore.com/lib/the...-Schematic.pdf) a revision to change some diodes from 1N3070 to 1N4007. I have no idea what is in my amp or if this change would affect it, but I do have it mostly apart so if I need to fix something I would rather do it now.

The amp does work as it is but I have only tested low volume. I don't have a fully matched set of tubes which is obviously not ideal, but since the red LED comes on even with two tubes it seems like something is wrong.

TIA,
Greg

[glebert]

OK, found the CL owners manual and per its troubleshooting chart it indicates mismatched tubes. Can't find a schematic that shows the bias monitoring LED circuit and also the power tube assembly to see how the current is being monitored and imbalance is being detected. I guess I get to go spend a couple hundred on tubes...

[Enzo]

Before you do that, verify the three small tubes. The two right behind the bias adjust controls are 12AU7, NOT 12AX7, make sure. And the one closer to the SLAVE OUT jack is a 12AX7. I have found many amps where someone mixed them up, and that can cause your light issue. Also, I find a weak/bad 12AU7 to be far more common than other issues when the lights won't zero in. Try different 12AU7s in there before investing.

Look on the power amp schematic (not the power tubes, I mean the page with the phase inverter.) The whole lower right area is the LED circuit, you can see the LEDs off on the right.

Look at the power tube page, and see all six power tube cathodes individually connect through a ribbon to the other page circuits. Note on that other page, each cathode circuit finds a 10 ohm resistor to ground, and the voltage dropped across that resistor feeds all those comparator circuits.

[nevetslab]

OK, found the CL owners manual and per its troubleshooting chart it indicates mismatched tubes. Can't find a schematic that shows the bias monitoring LED circuit and also the power tube assembly to see how the current is being monitored and imbalance is being detected. I guess I get to go spend a couple hundred on tubes...

I've attached that schematic here:

Power Amplifier PCB Schematics (419xxh2).pdf

The Bias LED window comparator circuit has a window that's between 14.7mA (before GRN comes on) and 29.4mA (where the RED comes on). With only two tubes in at a time, which is how I get basic plate current readings after pre-setting both upper and lower driver tubes, and measure the voltage drop across the 10 ohm cathode resistors, the BIAS LED's won't work, though the RED circuit does. And the Protect circuit also does, which saves a lot of damage if you have a tube that's conducting like crazy.

Enzo's absolutely right in being sure you have two good 12AU7 driver tubes in correctly. It does sound like you have an old mismatched set of power tubes. And, no doubt more than that, as you got this amp as a 'project'. Believe me, going thru this amp thoroughly to get it tuned up is a lot of work. I've posted numerous threads on the full procedure, which can be found in the Search archives....search for Ampeg SVT-CL and author being nevetslab. I'll get back to you later after getting home. Happy hunting!

[The Dude]

I've also had a few of these with burnt screen resistors, so check all sockets and make sure you have screen voltage on each. I had one amp come in with all but 2 screen resistors burnt

[glebert]

I've also had a few of these with burnt screen resistors, so check all sockets and make sure you have screen voltage on each. I had one amp come in with all but 2 screen resistors burnt

The first thing I did was check the screen resistors and they are all right on spec and I have screen voltage at each socket.

Thanks nevetslab for the schematic, I keep thinking I am losing my mind looking at the one I linked to and there are no power tubes in it.

I think my next step will be to measure the idle current with each of the tubes I have and see how much of a range I have. Looks like I should set it for 24mA per tube (0.24V across a 10 ohm cathode resistor) with a nominal tube?

[nevetslab]

I tend to go for 22mA nominal, so the variation in a Sextet of tubes will more than likely encompass 20mA thru 25mA. If you target 24mA, you'll be on the edge of the RED bias LED coming on. I rarely get a sextet of new Matched tubes having less than 3mA between min/max with the two driver tubes set for the same bias voltage. Over time, they of course drift apart. And, of course, driving a load, you ALWAYS see the RED lights on together with the GRN lights.

[glebert]

22mA it is then!

[nevetslab]

Can't find a schematic that shows the bias monitoring LED circuit and also the power tube assembly to see how the current is being monitored and imbalance is being detected. I guess I get to go spend a couple hundred on tubes...

I forgot to include the Power Tube PCB assy doc:

Tube Board PCB Schematics (41941h3_).pdf

Tube Board Parts List & PCB Layout (41941p5_).pdf

Technical Information Bulletin (TIB0003, June 2001) -- Power tube PCB modifications.pdf

[nevetslab]

While we're on the subject of getting tubes biased up and running, I have an SVT-CL over from our rental inventory, having the problem of it shutting down (goes into protect mode...Flashing RED LED). I tracked that down to the tube plugged into V3. The upper set was biased to around 14-16mA, while the lower set biased to 21-25mA. Then, I watched V3 start conducting, with it's plate current climbing past 40mA, RED bias LED came on quickly, then it hit Protect mode, pilot light began flashing as it turned the HT supply xfmr off. I did have to find which tube it was, not knowing right off until I monitored the six tubes with the open-chassis set up I use.



This new matched Sextet of J/J KT-88's just came in from Antique Electronic Supply in Arizona. so I set the amp up like this, only no power tubes installed. I checked the driver bias voltage for the upper and lower sets of tubes, both around -51.7VDC. Set both to -52VDC for a start, then plugged in the first two tubes into V3 & V4 positions to see what voltage I read across the cathode resistors. Both read around 110mV (11mA), so I decreased the bias to hit 220mV for both tubes, then checked the bias voltage change. Now both set at -48.7VDC. I let that first pair idle a couple minutes to see how much change there was. Increased to around 0.226VDC, so I marked the first to tubes per their position and moved on going thru the rest of the tubes from the matched Sextet. Got similar readings, spaning from 0.215V to 0.226V, until the last tube....I got a reading of 0.115V. Sigh.........

Let it idle for a while to see if there was any increase. It got to 0.120V and stayed there. Swapped the two tubes around (I tend to use V3 and V4 tube positions for all of the readings in single pairs for consistency sake. I had previously checked the readings of the Cathode resistors and screen resistors. The cathode resistors ranged from 10.00 to 10.26 ohms, so tight tolerance. Similar readings for the 220 ohm screen resistors. When i swapped the tubes V1 & V6 with each other, the readings moved with them, so I have one tube reading 10mA low in this new matched Sextet. RATS! I'll have to call Antique to see what else they have....if there's another tube from a similar reading set that could offer a better match, as this is not acceptable for a matched set. I've previously read the currents of a set having such an imbalance, and under load, there's also similar imbalance in the current sharing of the triad of tubes under drive conditions. I haven't tried that with this set, but having had high success with the J/J KT-88's in Matched Sextets from Antique, I'd like to continue that. Each tube has the Transconductance Value, used as the Matched Set code, along with the Plate Current used. Each tube's label has those to numbers, along with a seven-digit ID and a four-digit alphanumeric code, so there's info in the database associated with the Alpha Tube Matching system they use.

[glebert]

Yeah, the sextets really suck for trying to get six that all work and are tight match. I did like the Eden E300T that I had that had individual bias pot for each of the power tubes.

I did measure bias currents with the tubes I have and found a much wider range than I expected. There were enough that were close that I put them together in one of the triples and when I turned up the bias it went from no LED to green LED only, so I think the tube mismatch is the only problem, and now that I know better how the bias monitoring works it makes sense.

Thanks again for the help.

[nevetslab]

Having been maintaining a rental fleet of SVT-CL's and SVT-VR's for years here, that has yielded a number of pulls when I finally run out of spare pulls of a particular manufacturer (Sovetk, Svetlana, J/J, Electro-Harmonix, etc) of both KT-88's and 6550's. As time goes on, I often find one tube having gone bad, like I just found on this amp I'm working on. I was completely out of Sovtek's in transconductance range, so had to order a new Sextet, and end up with five more 6550's to live in the support tubes to hopefully keep some other amps alive and working that need Sovtek's in that range. I juggle tubes on a regular basis, using that set-up shown in the photos. The SVT-VR has a tighter range with their Balance LED. The SVT-CL has basically a 15mA range...the SVTVR is closer to 10mA, so a bit more touchy to balance. The residual hum you hear if you have a speaker connected while balancing the upper/lower bias tends to null around that 21-23mA range. I also find the power consumption/current draw to be similar from amp to amp where I bias them up. 224W is typical at idle @ 120VAC.

[glebert]

I was lucky enough to get an additional pair of JJ KT88s that are very close to my other tubes, so I can get a +/-1mA match in a trio. My Bias 2 red LED is still on though, and comes on before the KT88s are even really conducting, so I'm thinking that may be a 12AU7 problem, as Enzo described? I tried all of the 12AU7s I have but the LED is on with all, although in swapping them around some will also make the Bias 1 red LED come on as well.

[nevetslab]

I was lucky enough to get an additional pair of JJ KT88s that are very close to my other tubes, so I can get a +/-1mA match in a trio. My Bias 2 red LED is still on though, and comes on before the KT88s are even really conducting, so I'm thinking that may be a 12AU7 problem, as Enzo described? I tried all of the 12AU7s I have but the LED is on with all, although in swapping them around some will also make the Bias 1 red LED come on as well.

If you're not finding any of the six power tubes passing more than 29.4mA, then I'd suspect the Window Comparator IC for that section, or possibly resistors in that Window Comparator circuit. I don't think I've ever had to replace anything in those circuits other than one or more of the IC's. Of course that area just happens to be on the main board, with no easy access from the component side of the board when it's powered up. Somewhat easy to navigate from the foil side, though still a PITA. Keep us posted, as you may be in uncharted waters for all of us to learn from.

[glebert]

If you're not finding any of the six power tubes passing more than 29.4mA, then I'd suspect the Window Comparator IC for that section, or possibly resistors in that Window Comparator circuit. I don't think I've ever had to replace anything in those circuits other than one or more of the IC's. Of course that area just happens to be on the main board, with no easy access from the component side of the board when it's powered up. Somewhat easy to navigate from the foil side, though still a PITA. Keep us posted, as you may be in uncharted waters for all of us to learn from.

Played around with this more today. Powered up with two KT88 tubes, one per trio. No red LED at all on first try. Moved tube to other sockets. Only get red LED when the tube is in one particular socket. I think it is the V2, the one closest to the 12AX7 tube, middle of the row closest to the back edge of the chassis. Would it be so hard for Ampeg to actually put a picture with the schematic to say what socket is which? (rant over). Anyway, I confirmed that the current with the tube in that socket is never more than 26mA. Once or twice it seems like the red LED also flashed when I turned on the main power switch, so maybe that comparator is suspect?