Ampeg SVT blown tube & fuse

[Ampzone]

So we have a very clean EARLY MTI circa 1980 era SVT with a matching 8x10 cab that was re-capped power supply and preamp top to bottom last year. During the last couple sessions it developed a hum but played fine, then last week the mains fuse blew. Inspection of tubes showed one of GE 6550's had burnt inside the glass, gray plate burnt now black and the glass darkened. We checked the protection 5w resistors but none of them had cooked. After installing a new 10amp fuse and powering it back up it seems ok, a hum still exist, but I think a little less than before the blown tube.

Can we operate it with the five 6550's safely for another session while we await a new set of tubes?

Also what new brand 6550 tube out there currently can handle around 640v plates volts?

Thanks for any input.

[nevetslab]

It's best to run with 2 pairs (2 upper, 2 lower) on that so there's equal current sharing. It does raise the output impedance a bit, and you're not as powerful, but it will get you by.

J/J KT88's in a Matched Sextet work very well. That's what I've been replacing the eventual failure of the Sovtek & Svetlana 6550's in the rental inventory here at CenterStaging in Burbank, CA. I haven't found any particular 6550 or KT-88's that won't run on 640-660V plate supply.

The hum, if it wasn't there before, is due to the imbalance of the remaining power tubes.

If you have a multimeter, I'd recommend you check the remaining 5 tubes. You can plug and unplug the power tubes while in Standby, but wait a good minute before unplugging them after switching back into S/B. All that is present on the tube sockets in S/B mode is heater and bias. You'll want a load on the amp during the bias check/adjustment. If no dummy load, a speaker will suffice. Mark them per position (V4, V5, V6 for the lower set, V7, V8 & V9 for the upper set) Then test in pairs, using the test points K1 & Gnd, K2 & Gnd to check the plate current. With one pair installed, adjust the two bias pots to yield 24mV for the upper and lower tubes. Then, put in the next pair, and record what they read, and finally, removing one of that previous pair, measure the last tube and record it's plate current. Then, group the two pairs in the closest match per pair, put them back in (less the 'spare'), and adjust the bias on each side for 48mV.

If you don't have an oscillator available for the Balance Adjust, you can trim it to light the yellow LED, or to minimum hum out of the speaker. Check what the bias reading is after you adjust the balance, just to be sure you're still at/around 48mV on upper and lower halves of the output stage.

[Ampzone]

It's best to run with 2 pairs (2 upper, 2 lower) on that so there's equal current sharing. It does raise the output impedance a bit, and you're not as powerful, but it will get you by.

J/J KT88's in a Matched Sextet work very well. That's what I've been replacing the eventual failure of the Sovtek & Svetlana 6550's in the rental inventory here at CenterStaging in Burbank, CA. I haven't found any particular 6550 or KT-88's that won't run on 640-660V plate supply.

The hum, if it wasn't there before, is due to the imbalance of the remaining power tubes.

If you have a multimeter, I'd recommend you check the remaining 5 tubes. You can plug and unplug the power tubes while in Standby, but wait a good minute before unplugging them after switching back into S/B. All that is present on the tube sockets in S/B mode is heater and bias. You'll want a load on the amp during the bias check/adjustment. If no dummy load, a speaker will suffice. Mark them per position (V4, V5, V6 for the lower set, V7, V8 & V9 for the upper set) Then test in pairs, using the test points K1 & Gnd, K2 & Gnd to check the plate current. With one pair installed, adjust the two bias pots to yield 24mV for the upper and lower tubes. Then, put in the next pair, and record what they read, and finally, removing one of that previous pair, measure the last tube and record it's plate current. Then, group the two pairs in the closest match per pair, put them back in (less the 'spare'), and adjust the bias on each side for 48mV.

If you don't have an oscillator available for the Balance Adjust, you can trim it to light the yellow LED, or to minimum hum out of the speaker. Check what the bias reading is after you adjust the balance, just to be sure you're still at/around 48mV on upper and lower halves of the output stage.

This is a huge help, thank you. To confirm before I attempt this with a multi-meter.

Should I keep the remaining four tubes after mathcing on the inside sockets V5--V8?

With a stock Ampeg SVT 4ohm cab (reads about 3.7ohms) is the impudence shift significant enough to cause harm to the transformer?

What is the best mV reading range for a new full compliment of KT88's matched sextet?
Ive read measure mV from K1 to ground. The schematic states .072 volts,
so use the bias pot (VR1) to set it to .072 volts.
Then measure from K2 to ground and adjust VR2 until you see .072 volts.
Now go back and see if K1 has drifted. If so ,readjust it back to .072.

My meter should with this set of just 4 6550's read .048mV volts, and just two .024mV when set to mV?


For this tube match test put one tube on the V4-V6 side and the other on the V7-V9 side?

[Ampzone]

ok now with only 4 tubes in the bias is 48mV and 51mV on the meter. Adjusting the hum control (was set to fully clockwise) got rid of any hum going clockwise in fact made it very quiet but it was very scratchy while adjusting it.

[nevetslab]

This is a huge help, thank you. To confirm before I attempt this with a multi-meter.

Should I keep the remaining four tubes after mathcing on the inside sockets V5--V8?

With a stock Ampeg SVT 4ohm cab (reads about 3.7ohms) is the impudence shift significant enough to cause harm to the transformer?

What is the best mV reading range for a new full compliment of KT88's matched sextet?
Ive read measure mV from K1 to ground. The schematic states .072 volts,
so use the bias pot (VR1) to set it to .072 volts.
Then measure from K2 to ground and adjust VR2 until you see .072 volts.
Now go back and see if K1 has drifted. If so ,readjust it back to .072.

My meter should with this set of just 4 6550's read .048mV volts, and just two .024mV when set to mV?


For this tube match test put one tube on the V4-V6 side and the other on the V7-V9 side?

I don't recall the mechanical layout relative to the two driver tubes. I thought I might have some maintenance photos of one, but didn't find any. Yeah, V5 thru V8 would be fine. The minor shift in impedance won't harm the transformer. I had to do the same last week on a rental, where an SVT-VR needed a fresh set of tubes, and all I had was a matched Quad set of KT-88's (J/J's). It sounded fine, though the bass player did notice it wasn't as solid as the other SVT-VR in the stage setup.

24mV per tube translates to 24mA plate current, and 72mV is 3 x 24mV. That's the nominal biasing on all of the Ampeg SVT's, from what you have thru the current SVT-CL's and SVT2-Pro.

When I start with a fresh sextet of tubes, I'll pre-set the drivers to around -50V and pop in the first two tubes, one up, one down in the output stage, and see what the plate current shows. It usually takes a couple iterations of nudging the bias so each driver bias voltage is the same and I'm close to the 22-24mV reading. Then, with a common equal bias setting, you go thru and measure the rest of the tubes, marking them, or noting their matching code on your list. When you have all six tubes measured, you can then group them into the best two matching sets, pop them all back in accordingly to their upper/lower positions, and re-bias, setting to the 72mV reading on the test points. I let them idle after the first full biasing for about half an hour, and check them a few times. There will be some minor drift. Then, do a final tweak, and they should be good to go.

I don't know if you have hold-down clamps added to your SVT or not. Ampeg didn't start using them until the SVT-VR and SVT-CL, sometimes with base clamps latches, the others with the dual-spring & hold-down hat. I was getting a lot of tube breakage from the hold-down clamp/spring joints fracturing the glass at those junctions from all of our rental gear going out in the road cases, and being SO HEAVY, roadies really man-handle the cases, flipping them up and over onto their backs in the trucks. I've since added hi temp silicon rubber tubing, making a grommet to fit inside the 'hat', and tubing that slips over the clamp arm/spring joint, so there's no metal-to-glass contact...everything is cushioned. Also added short sleeves over the 'wings', so you end up with this:

[Ampzone]

I don't recall the mechanical layout relative to the two driver tubes. I thought I might have some maintenance photos of one, but didn't find any. Yeah, V5 thru V8 would be fine. The minor shift in impedance won't harm the transformer. I had to do the same last week on a rental, where an SVT-VR needed a fresh set of tubes, and all I had was a matched Quad set of KT-88's (J/J's). It sounded fine, though the bass player did notice it wasn't as solid as the other SVT-VR in the stage setup.

24mV per tube translates to 24mA plate current, and 72mV is 3 x 24mV. That's the nominal biasing on all of the Ampeg SVT's, from what you have thru the current SVT-CL's and SVT2-Pro.

When I start with a fresh sextet of tubes, I'll pre-set the drivers to around -50V and pop in the first two tubes, one up, one down in the output stage, and see what the plate current shows. It usually takes a couple iterations of nudging the bias so each driver bias voltage is the same and I'm close to the 22-24mV reading. Then, with a common equal bias setting, you go thru and measure the rest of the tubes, marking them, or noting their matching code on your list. When you have all six tubes measured, you can then group them into the best two matching sets, pop them all back in accordingly to their upper/lower positions, and re-bias, setting to the 72mV reading on the test points. I let them idle after the first full biasing for about half an hour, and check them a few times. There will be some minor drift. Then, do a final tweak, and they should be good to go.

I don't know if you have hold-down clamps added to your SVT or not. Ampeg didn't start using them until the SVT-VR and SVT-CL, sometimes with base clamps latches, the others with the dual-spring & hold-down hat. I was getting a lot of tube breakage from the hold-down clamp/spring joints fracturing the glass at those junctions from all of our rental gear going out in the road cases, and being SO HEAVY, roadies really man-handle the cases, flipping them up and over onto their backs in the trucks. I've since added hi temp silicon rubber tubing, making a grommet to fit inside the 'hat', and tubing that slips over the clamp arm/spring joint, so there's no metal-to-glass contact...everything is cushioned. Also added short sleeves over the 'wings', so you end up with this:

[ATTACH=CONFIG]50820[/ATTACH] [ATTACH=CONFIG]50821[/ATTACH]

very helpful tips! Thank you. I will do the same.

so I was able to get 48v on K1 bias knob & 48.3v on KT2 bias knob with my DMM probe in the provided common probe input on the amp chassis and the other probe switching back and forth. So it appears to be a close match with these 4 tubes. I did not check the other check of +-01 on the K2 adjustment setup instructions written on the chassis. I hope that's ok.

And the hum control was adjusted, one way on the pot and it gets a louder bassier hum, the other a buzzier higher pitched hum. I just settled on the lowest hum. It still has some hum but better than before.

Fingers crossed we can do this short session without killing something on the amp.

[nevetslab]

I presume that was -48V on one of the power tube sockets Pin 5, and -48.3V on the other. I forgot to mention where you'd find that. I'm so used to having an SVT-CL standing on it's end, with the bottom of the power amp PCB facing me, so I can just put the probe on Pin 8 of the driver tubes.

OH....I see what you're saying. 48mV on KT1 and 48.3mV on KT2 (translates to 48mA & 48.3mA plate current). You should be fine.

I've been buying the J/J KT88's (and all my other tubes) thru Antique Electronic Supply, using their Apex Matching system. When I have failures on one or two tubes from a set, I can often get replacements that match what I had bought, since the information is in their database. Easier than sorting thru 'pulls' that I accumlate, keeping a full sextet going.

[Ampzone]

I presume that was -48V on one of the power tube sockets Pin 5, and -48.3V on the other. I forgot to mention where you'd find that. I'm so used to having an SVT-CL standing on it's end, with the bottom of the power amp PCB facing me, so I can just put the probe on Pin 8 of the driver tubes.

OH....I see what you're saying. 48mV on KT1 and 48.3mV on KT2 (translates to 48mA & 48.3mA plate current). You should be fine.

I've been buying the J/J KT88's (and all my other tubes) thru Antique Electronic Supply, using their Apex Matching system. When I have failures on one or two tubes from a set, I can often get replacements that match what I had bought, since the information is in their database. Easier than sorting thru 'pulls' that I accumulate, keeping a full sextet going.

Yes thats it. There are probe taps on the back of this model "K1 & K2" beside the two bias pots then the common probe tap insert to the right. In addition a balance control (not sure how to set this) and hum balance control (very handy and works). Looking at my friends 1971 SVT and the current versions this thing is super duper well built, thick yellow PCB board, tubes are mounted on this thick metal chassis, not a pcb tube mount, wires are a heavier gauge, very neatly laid out and bundled inside, no plastic. circa 1980 MTI. I will order the KT88's as you suggest and then dial it in for 72ma on K1 - K2 test points.

Oh we made it through the session fine thanks to you and surprisingly not a huge power less as you noted, maybe 15-20% less oomph!

[nevetslab]

SVT Balance Adjust

Great to hear you got thru the session fine.

I've attached the schematic of the SVT II which in the Calibration notes step 3, it describes two methods of setting the Balance trim. Both methods use an oscillator set for 40Hz, and adjusted for 25V RMS out of the amp into a dummy load. Doing this with a 8x10 cabinet would do it, but would be louder than the dickens. Maybe lay the cabinet face down on the floor, propped off the floor by an inch or two, just to tame the SPL. I doubt if you have access to a distortion analyzer, but it can also be done using your DMM, and adjusting the balance pot to 0VDC between the two test points next to the bias adjust pots.

combined later power amp.pdf

[Ampzone]

Great to hear you got thru the session fine.

I've attached the schematic of the SVT II which in the Calibration notes step 3, it describes two methods of setting the Balance trim. Both methods use an oscillator set for 40Hz, and adjusted for 25V RMS out of the amp into a dummy load. Doing this with a 8x10 cabinet would do it, but would be louder than the dickens. Maybe lay the cabinet face down on the floor, propped off the floor by an inch or two, just to tame the SPL. I doubt if you have access to a distortion analyzer, but it can also be done using your DMM, and adjusting the balance pot to 0VDC between the two test points next to the bias adjust pots.

[ATTACH]50822[/ATTACH]

Great thank you. Just to confirm the signal generator set to 40hz output from this unit should be measuring 25v on my meter on the line out of the generator to the high channel 1 input of the svt? Volume set to?
Or are you talking about measuring off the speaker output of the svt being 25v on the DMM?

[g1]

25VRMS (40Hz) at speaker out into 4 ohm dummy load.
Adjust for 0VDC between K1 and K2.
Readjust hum balance control for minimum hum with no input.

[Ampzone]

25VRMS (40Hz) at speaker out into 4 ohm dummy load.
Adjust for 0VDC between K1 and K2.
Readjust hum balance control for minimum hum with no input.

The issue Im having is no dummy load, I cannot measure the output for 25v while setting the volume and removing the speaker cable from the amp to measure.

[nevetslab]

The issue Im having is no dummy load, I cannot measure the output for 25v while setting the volume and removing the speaker cable from the amp to measure.

Ah......my mistake. I didn't look at some other SVT schematics to find there's only one speaker jack on the back. That means you'd need a Y-Cord adapter Male Plug with two jacks.

[Enzo]

Well generally we'd just use the through jack on one of the cabs.

What is the problem taking a reading, spin the cover off the speaker plug and clip right to the wire connections inside.