SVT clone power supply high B+

[idlefaction]

Hi team

I've completed my SVT build, and it sounds amazing!!







I used the specs for the original 70's power transformer; it's resulted in massive B+ voltages and I was wondering if anyone could chime in on whether it's OK to run 6550's on 730VDC...? They have a 600V max anode voltage, but even the modern SVT amps seem to run them at 640-660V. Not sure what the story is here, hopefully someone can chime in?

Cheers,
--
Darren

[Tom Phillips]

Beautiful build!

The original 1970 SVT schematic lists 695V B+ for the unit that used 6550 power tubes. Looking at my personal notes I have confirmed B+ readings in that range on SVTs with 120VAC line input. Since it's common for line voltage to exceed 120V you could easily reach the 730V you measured. There are probably lots of opinions about how well modern tubes could handle that B+ but it is common for musical instrument amp designs to run tubs beyond their published ratings.

One additional factor is that you are using only 4 of the standard 6 SVT power tubes. With a stock spec power transformer you will therefore have higher voltage due to the lower loading on the power supply.

[Jazz P Bass]

I believe the key here, that is allowing that sky high B+ voltage on the plates, is the Screen Voltage being held low (345 some volts).

Bias current being low will also let the B+ rise.

The SVT CL indicates -45Vdc on the bias pins.

[idlefaction]

Beautiful build!

Thanks Tom!! 4 years in the making. It sounds, in a word, amazing.

Since it's common for line voltage to exceed 120V you could easily reach the 730V you measured. There are probably lots of opinions about how well modern tubes could handle that B+ but it is common for musical instrument amp designs to run tubs beyond their published ratings.

Yeah, here in NZ we are on 230VAC but it usually measures 240VAC+ so that will be contributing. Are there any rules-of-thumb on how a high B+ might affect tube life? I haven't spotted anything online in my Google searching.

I've only had it running for a couple of minutes at a time, so I'll give it twenty minutes or so tonight and see where things settle to. And will post some more photos of the amp with everything wired up, as I can see a few missing things in that guts shot above!

Will also post the turret board layouts in case anyone else wants to build one.

Just noticed I joined this board in 2011, and that was my first post. *heh*

[idlefaction]

I believe the key here, that is allowing that sky high B+ voltage on the plates, is the Screen Voltage being held low (345 some volts).
Bias current being low will also let the B+ rise.

IIRC Screen voltage is also high at 495V, but I'll re-check that tonight. That is probably measured on the wrong side of the screen resistors, too. Did you mean that when you use a very high B+ you should use a lower screen voltage to compensate?

I only have the output tubes biased at 50% plate dissipation at the moment, so I'll crank that up to 65% and see if it affects the B+ any.

[idlefaction]

Okay! The good news is that the B+ does come down from 730V to 705V, and the heater filament voltage also comes down from 7.35V to 7.10V. The wall was reading 238VAC, and I adjusted the bias up to 74mA/side, which is 62% max plate dissipation.

The not so awesome news is that the plate voltage is 505V. As indicated above, really for optimum tube lifespan this should be a lot lower, but it's about what the original 70's SVTs had, so cest la vie!

Some more photos of the finished build...









And as promised, here's my layout, in case anyone else is foolish enough to attempt to build one of these.

[hylaphone]

What a great job -- takes some berries to scratch build one of these!
May I ask what you used for the mids inductor?

What power tubes are you using?

FWIW, I had a 70's SVT in for repair last month. It had the wrong driver tubes and was biased at around 25%, putting plate voltage upwards of 750V.
It had been played hard this way for several years, the tubes were SED 6550's. They tested fine, and once the driver/bias was corrected they all idled happily within a few mA. Tough tubes.

[hylaphone]

Also FWIW, I'd be more concerned with your idle dissipation than plate volts. It's a super dynamic power amp, and Ampeg wisely advised cool bias.

[überfuzz]

Judging by the pictures it seems you missed out on 2 power tubes that would have helped soaking some juice out of the power transformer. This is probably keeping the B+ a tad high.

[MarkusBass]

Wasn't Fender 400-PS amp using 6550s @700V? The schematic says "SPECIAL DESIGN AND SELECTED 6550 TUNGSOL TUBES".
Why is the voltage too high if you used the spec for the original 70's power transformer?

Mark

[patlaw]

Who is the manufacturer of the blue perf board used for turrets? I like that idea a lot. The green stuff looks too home made.

[g1]

Why is the voltage too high if you used the spec for the original 70's power transformer?

1) The line voltage is higher now than it was in the 70's
2) He is only running 4 power tubes when the original PT was running 6 power tubes.

[Tom Phillips]

I believe the key here, that is allowing that sky high B+ voltage on the plates, is the Screen Voltage being held low (345 some volts)...

I agree. The Ampeg design used 350V on the 6550 screens. That’s from 1970 service information.

IIRC Screen voltage is also high at 495V, but I'll re-check that tonight. That is probably measured on the wrong side of the screen resistors, too. Did you mean that when you use a very high B+ you should use a lower screen voltage to compensate?

The Ampeg SVT power transformer had a separate winding for the screen poser supply. Does your amp use that approach? Anyway...that’s how they generated the 350V screen supply. The resistors we call “screen resistors” (one for each power tube) are there to limit the screen current under high power and overdrive conditions. If you were to insert a large voltage dropping in the screen supply to substantially lower the screen voltage at idle then I think the resulting saggy supply would adversely affect the amp. The original SVT design used 22Ω screen resistors. If you are using larger screen resistors, such as 1kΩ then you may be OK with the high screen voltage. Note, however, that the “Design Maximum Rating” for the 6550 screen voltage is 440V and the “Design Maximum Rating” screen dissipation is 6W. Of course we don’t really know what the ratings are for currently available 6550s since the few spec sheets that are published seem to be clones of the legacy spec sheets.

I only have the output tubes biased at 50% plate dissipation at the moment, so I'll crank that up to 65% and see if it affects the B+ any.

As hylaphone pointed out , Ampeg used a cooler bias. The original published “Calibration procedure” for the SVT specified 72ma per side of three 6550s. With 695V on the plates that gives 16.68W plate dissipation per tube which is 40%. I always thought that sounded fine and the amp produced 300 to 330W of clean power. I was never tempted to bias the SVT hotter. Give cooler bias a try and let your ears decide. I’m sure the tubes will like it.

Okay! The good news is that the B+ does come down from 730V to 705V, and the heater filament voltage also comes down from 7.35V to 7.10V. The wall was reading 238VAC, and I adjusted the bias up to 74mA/side, which is 62% max plate dissipation.
The not so awesome news is that the plate voltage is 505V. As indicated above, really for optimum tube lifespan this should be a lot lower, but it's about what the original 70's SVTs had, so cest la vie!.

You lost me there regarding what caused the changes and why voltages seem to be all over the place. Anyway...the original SVT version (after they switched from 6146B power tubes to 6550s in 1970) used 695V B+, 350V on the screens, 6.3V for the heaters and was biased at ~17W per power tube.

There are many tradeoffs that can be made but I’d recommend against simultaneously pushing the limits on multiple parameters.

By the way, what version of the SVT schematic did you use as the basis for your build?

[g1]

The not so awesome news is that the plate voltage is 505V. As indicated above, really for optimum tube lifespan this should be a lot lower, but it's about what the original 70's SVTs had, so cest la vie!

I think you mean the screen voltages are 505V. You said your plates are in the 700V range. I don't know which "original SVT" voltage you are referring to here, but as was mentioned, screens in the original were around 350V.

[Jazz P Bass]

Here are the reissue Power schematics.

Ampeg SVT CL Output Tube PCB Schematic.pdf

Ampeg SVT CL Power Amplifier PCB Schematic (Late).pdf