Another One Bites The Dust: SDS-DOA

[kevin gilmore]

OK, perfect for the filaments

LAMBDA LQ-530 0-10V, 0-14A DIGITAL DC POWER SUPPLY - eBay (item 320264048971 end time Jul-10-09 19:32:45 PDT)

Great for the high voltage. Still overkill.

I have more than a dozen of these, various voltages and currents.

They can't be killed.

SORENSEN [email protected] REGULATED VARIABLE DC POWER SUPPLY:eBay Motors (item 350188032602 end time Jul-06-09 12:50:09 PDT)

[n_maher]

i'm sure the Menace sounds very good, but i think that getting Al's amp up and working as close to the original sound signature as possible is the way to go, as that amp was really, really good.

I agree completely, I'm hopeful that fixing the power supply is all that will be required.

[kevin gilmore]

There is absolutely no reason to touch the amplifier chassis in any way.

Replace the power supply chassis with a couple of power supplies in the meantime

while you decide the ultimate fate of the old power supply. Won't change the sound

in the slightest.

You can go back to the old supply later. Or just change out the fet and diode

and use it till the filament transformer finally burns up completely.

[Icarium]

Hrm I should figure out something to do with the supply you sent me.. or send it back to you.

[spritzer]

The filament transformer would be easy enough to replace with a larger Hammond.

Zero to 35 amperes... A little bit of overkill

...just a tiny bit

[kevin gilmore]

Hrm I should figure out something to do with the supply you sent me.. or send it back to you.

The rules on surplus stuff i send to people is that i never ever want it back.

Not a lot of room left in that sds supply to add a bigger transformer.

OH hell, just tie the wires in, and cram it in there somewhere. Use hammer

to make sure it fits.

[spritzer]

I can send Al a Thorshammer if it helps. It's supposed to shoot lightning sparks so that could just power the amp for him.

[Voltron]

Thanks Birgir, but I might get a little carried away with this thing.

[Voltron]

KG:

The Heathkit idea seems ok, although I really want the HP beastie if I were honest about it, but what would have to be done to hook that up to the amp?

Are you to the point where you have a schematic and/or an idea of how to recreate a similar PS in my existing case? I think that is what I should do, unless the current parts can be used with a new trafo and solid state parts.

[kevin gilmore]

I would really need high resolution pictures of the amp pointed to the neutrik

connectors from a couple of different angles to determine which wire is which.

As far as i can tell, 2 of the 5 wires total are filament, one is ground, one is

front end B+ and the other is output tube B+

A simple harness to the heathkit is all that would be necessary. Ironbut should

clearly be able to figure out which wire is which and to make up a fast harness.

There is literally thousands of really outstanding surplus power supplies out there.

The heathkit is nice because it has both supplies that you need.

But anything that meets the following specs should work.

something that makes 6.3 VDC or 6.3 VAC at up to 10 amps.

something that makes around 150 VDC at about 150ma.

(best guess at total high voltage current)

So anything in the 0 to 300 volt range with 250ma or more is

perfect.

Certainly not as pretty as your current supply. But will likely give

you 20 years of trouble free power.

The black gate output caps are your limiting factor given the WCF

design, so absolutely no more than 350 volts max.

Now as to how to fix that current mess and still have the gas tubes light

is another story entirely. Unless you want the gas tubes to be a night light

but otherwise electrically not part of the circuit. Mikhail was all about show

anyway. Not the least of which is that unproperly filtered gas tubes are

noisy anyway. Filament transformer will have to be replaced at some point

in the future anyway.

Neil found someone good to fix his amp (although its going to be expensive,

lots of man hours of work there) maybe you can contact that guy and show

him my current best guess on the power supply schematic and then show

him pictures of the thing. If he doesn't run away screaming, you might be

able to get him to fix the thing, at least temporarily.

I've been playing with a few different ideas using the transformers and other

parts in there i can salvage. Currently i don't like what i've come up with.

Its kind of stupid if what you thought you were getting really was an all tube

power supply. Lots of solid state all over that thing, none of the caps across

the diodes needed to shut them up...

There is another version of this power supply out there that actually has

6l6 pass tubes, that would be a better idea, but now we are talking about

more holes in the top chassis, no reasonable way to wire the mess...

[kevin gilmore]

OK, i have something that i really like. It is completely in the spirit of the original

power supply, fixes absolutely everything, can't burn up as there is no sand in

the high voltage section, no dangerous anything, nothing out of spec... Best

part, no modification of the chassis. Well almost, need 9 pin to 8 pin adapters

for the 6au6's. Keeps all the iron. Otherwise adapter plates for the 6au6's.

Uses only one gas tube per channel, 1 socket is for the 6au6 and the other

is for the 6l6.

http://gilmore.chem.northwestern.edu/sdsgaspsfix.jpg

unfortunately, still sand in the filament supply, but absolutely nothing can

be done about that.

WARNING: pinouts of the 6au6 completely wrong. And the picture of the 6l6 is

not correct either, but the pinouts are right.

Cute part 2. The switch that was the voltage selector is now the voltage pot.

It can be a dual section in which case both supplies adjust at the same time, or

if you do it as 2 seperate knobs on the same shaft, you can adjust the front

end gain power supply seperate from the output power supply.

Now all you need is an ultra trendy analog meter to stick into the front panel.

(ok, now you have to do some chassis work) do it eddie current style and it will

look real cool.

[Voltron]

Now I am intrigued but I don't understand your design at all because I am schematically challenged. What are the new tubes doing that the old gas tubes were not, and how does the voltage pot idea work? i.e., wtf does this mean:

"It can be a dual section in which case both supplies adjust at the same time, or if you do it as 2 seperate knobs on the same shaft, you can adjust the front

end gain power supply seperate from the output power supply."

Thanks man, and I look forward to figuring this out.

[nikongod]

unfortunately, still sand in the filament supply, but absolutely nothing can be done about that..

AC heaters. Bias them 75V above ground. Tubes that hum with "biased" AC heaters are not worth using in an amp 1/8 as expensive as this one.

People like to blame the amp when tubes hum with AC heaters, and you know MFR's NEVER like to use AC for this reason. The old hum-bottles some dude had that never ran noisy on DC are now noisy on AC heaters (durr) any hum is bad, and these tubes have been in his family for generations: his grandfather built them and smuggled them into the country! How could they be anything but great? The cheap SET amp with no global feedback he just bought hums like whoa with his priceless communist tubes. Its terrible customer service to tell these people that any of their gear is crap, save the phone call and complaint thread on the web, spend $5 on a rectifier and a cap, and use DC heaters. Whether DC heaters are any better with tubes that dont hum is of course a topic of long heated threads all over the place, but you have to be truly special to make DC heaters hum and thats all that matters when your trying to make a buck.

The darkvoice threads on head-fi are great examples of this: my amp hums with some tubes but not others. Why does my amp suck? Its your tubes (OK darkvoice grounds their heaters which dosnt help the situation at all), but nobody wants to hear that because they were expensive NOS tubes....

Now I am intrigued but I don't understand your design at all because I am schematically challenged. What are the new tubes doing that the old gas tubes were not

The new design is an ACTIVE regulator with some feedback built in to get low output impedance, where the old one was either a source follower/capacitance multiplier or a shunt based regulator. The schematic KG posted up is 100% tube which I also think was the intent of the original PS, except his schematic is like 10milion times nicer on the patrick82 scale than even my optimistic guesses of whats in there.

and how does the voltage pot idea work? i.e., wtf does this mean:

"It can be a dual section in which case both supplies adjust at the same time, or if you do it as 2 seperate knobs on the same shaft, you can adjust the front

end gain power supply seperate from the output power supply."

It means that the adjustments to voltage are no longer done in steps. Rather than what amounts to a "high/medium/low" switch you have a knob that goes from low to OMFGWTFBBQ!!!111 without steps. Unless Im mistaken you can also adjust voltage while the amp is on with this setup (cool as hell!).

As far as the dual pot thingy, you have to find a pot like what is sometimes used for EQ controls for separte channels on cheap stereo amps: no biggie. They are abundant in old junk gear. You could also just use 2 totally separate knobs/pots.

On that note, Im not sure that separate *adjustable* regulators for Gain and output are a GREAT idea unless care is taken to limit the range (and then why even bother?) for the regulator for the gain tubes. I can see hardcore failures if the plate voltage on the gain tubes comes up higher than B+ on the output tubes, or when the same plate voltage sets the idle current for the output stage unsafely high for 6sn7.

Why not use something like this:

The PS KG drew up, with the lower limit of voltage at 340V with a simple SS regulator LOCKED IN at 320V (or whatever mikhail used in your amp and 20V more from the tube regulators to the output stage, my numbers as example) fed off of KG's regulator. The whole mess could be built dual-mono and would never have the risks of funky/incompatible voltages outlined above. This adds some silicon to the circuit as a regulator, but its somewhat protected by the tube ( LOL ) AND makes the current draw from both regulators identical. with separate regulators for output & gain you have one regulating 20mA and one regulating 50-80mA. This would light up at 30-60ma per regulator (same current on both regulators, variance is the use of different output tubes ) no mater what.

[kevin gilmore]

There are lots of ways to do this, i think this is the best.

Some notes.

voltron really liked the sound of 2 x 75 volt tubes... i.e. 150 volts for the power supply.

The WCF amplifier he has is AC coupled between first and second stages

so you can definitely have 2 different voltages if you want to.

The design as introduced works with just about all gas tubes and has a

intentionally limited voltage range.

With a 75 volt gas tube the output is about 90 to 150

with a 90 volt gas tube the output is about 110 to 180

...

with a 150 volt gas tube the output is about 170 to 300

And yes you can adjust the voltages with the amp on.

You would want to turn the output section voltage up

and down slowly otherwise you will get a slight bass thump.

Edited June 28, 2009 by kevin gilmore

[kevin gilmore]

updated schematic with correct 6au6 pinout

http://gilmore.chem.northwestern.edu/sdsgaspsfix.jpg

[kevin gilmore]

actual measured voltages with quickly built hacked together power supply

with

0a3 (75) output voltage range 123-216

0b3 (90) output voltage range 140-257

0c3 (105) output voltage range 159-298

Looks like 150ma no problem.

520 volts at the plate of the regulator tube.

[Voltron]

Thanks for all the effort Kevin. I am traveling today and will be at OHare if you W nt to stop by

[Frihed89]

This is slightly off-topic (and slightly on).

I was digging through My MPX3 Slam SE last night and sure enough i came upon an FET that was screwed into the side of the case. No, i don't know it's function, but shouldn't it be insulated against the case? What else should I worry about with these?

[Dusty Chalk]

No! He's using the case as a heatsink; if anything, redo it so that it is heatsunk more appropriately.

[John E Woven]

I think what he's refering to is that the FET is screwed to the case without a means to insulate the FET from the case, resulting in possible issues with grounding / current running through the case.

[kevin gilmore]

depends on whether the fet is in a fully isolated case or not. In any case

thermal grease or a silicon grease washer is a real good idea to keep the

board from burning up.

[Frihed89]

I think what he's refering to is that the FET is screwed to the case without a means to insulate the FET from the case, resulting in possible issues with grounding / current running through the case.

yes.

Thank you Kevin. I'll need to unscrew it at the very least, but i don't think so. (I forgot to even check what type it was). The washer is a good idea.

[neilvg]

Damn... I just caught on to this thread. Sigh. My repair guy thinks its largely out of scope to work on my amp and doesn't really agree with the design philosophy and so would rather just build a new amp not repair this one.

Not a good option.

We need a solution.

If I can get a replacement PSU that works, I can get the Amp section going again I am sure.

Neil

[penger]

Too much madness to sort out?

That is quite unfortunate neilvg, I hope you find a suitable solution soon.

[kevin gilmore]

If your build 2 of my schematics

as in 2 filament supplies and 2 of the high voltage supplies tweeked for 400 volts

(and one of them grounded upside down to make -400)

plus a bias supply you would have one kickass power supply.

And a small C- sully.

Sort of like that 12 tube thing mikhail built except mine would be built with the

right parts and actually work instead of blowing up.

(yes one of those has died recently too, looks even worse on the inside)

So here is a story. Out of college 34 years ago i went to work for zenith radio

and learned lots from the true tube masters that worked there. Most of which

were at least 30 years older than i was at the time. So recently i've contacted

a couple of them, as most have already passed into that great DHT heaven in

the sky, and i found one of them who was willing to work on the SP stuff after

i told him of people' woes. That is until i showed him pictures of the stuff.

At that point he said no fucking way, it would be easier to strip it to the bare chassis

and start over wiring things correctly. The 3 pictures i showed him was this

SDS mess, neil's amp, and spritzers amp.

So i'm back to trying to find someone who will work on this stuff.

If i do find someone, it ain't gonna be cheap. Unless i find a high-school

kid that wants to learn.

For neil's amp, the 4 pots on the front panel HAVE TO CHANGE. Plus

they have to be wired differently. Schematics of the modifications

have indeed been tested and work fine. But spritzer needs to test

his way on his amp.

Edited July 1, 2009 by kevin gilmore