It's raining 10M45's!

[leadbelly]

Just wanted to throw this out for critique. With all the hoopla about the 10M45 and the great performance/parts count that it gives, I looked at a bunch of the amp designs out there with SS CCS's like SY, Gary P, Eli D, Tubelab, etc., and thought of this as an ultra cheap SET design. It has no chokes and uses a voltage doubler on a 120/240 isolation transformer, so iron would be very cheap, and the 10M45's are used to give high PSRR to negate the high ripple from the voltage doubler power supply. Is it workable? Is there a better way to apply the 10M45 for an SET?

[Shoog]

I personally think a Microwave Oven Transformer as a plate choke on the 6V6 is a better option. This is because you don't need the voltage doubler. Keep the 10M45 for the driver.

Isolation transformer, MOT and mains toroidal as parafeed output transformer and you have very cheap set of iron.

Shoog

[rdf]

The oven transformer also allows Vp to swing above B+ whereas the IXYS shuts off around ten volts beneath B+. Given the IXYS's PSRR as a driver load a dropping resistor to the power supply zener will probably provide the same performance.

[dsavitsk]

Originally posted by Shoog
Isolation transformer, MOT and mains toroidal as parafeed output transformer and you have very cheap set of iron.

Still, not as cheap as the single CCS chip (unless you have a source for $1 MOTs) plus you need a whole power supply which is going to add to cost a bit -- though I'm not sure the PSRR is that good that the above would work great. I think a pi filter would still be useful here, or you could use the Raleigh Audio scheme to throw in a few more 10M45's.

Also, LED's probably like a little more current through them than do 6SL7's, so you either need a current booster for the LED, or a tube that likes more current. If you want cheap, try a triode strapped 6J9p -- electrically the same as a 5842 and costs about $1 each.

[tubelab.com]

Been there, built that. I think that you could get by with a single 10M45 on the driver stage. As stated above the usual SE output stage can swing its plate voltage from near zero to almost twice the B+ voltage due to the stored energy in the OPT. The CCS loaded SE output stage can not swing the plate significantly above the B+ voltage. If it did the 10M45 would die due to reverse voltage. This limitation causes a limitation in output power to about 1/4 the expected value (half the expected voltage swing).

I built a similar output stage using a hybrid (tube and 10M45) CCS. It could handle over 900 volts. The idea was to just turn up the voltage to overcome this issue. I tested it at low voltage (400 volts) and found that it worked very well as a low powered amp (2.5 watts). Due to the complications related to hurricane damage, I never got to connect it to a high voltage power supply.

http://www.tubelab.com/Active%20load...E%20output.htm

If you want cheap, try a triode strapped 6J9p -- electrically the same as a 5842 and costs about $1 each.

What is a 6J9p, I can't find any info on it.

[dsavitsk]

Originally posted by tubelab.com
What is a 6J9p, I can't find any info on it.

The only links I have are http://tdsl.duncanamps.com/tubesearch2.php?TSD=6j9p and http://img81.imageshack.us/my.php?image=6j9phx7.gif

I was searching ebay for 6688's and found someone selling 6J9p's as Russian equivalents. I bought a batch, and they test good when tested as 6688's, and put into an amp that uses 6688's (actually 7788's, but close enough in this case) they seem to work okay. Email me your address and I'll send you a pair if you like.

[Shoog]

Still, not as cheap as the single CCS chip (unless you have a source for $1 MOTs) plus you need a whole power supply which is going to add to cost a bit

Your local dump/recycling center should be able to help.

Tubelabs goes into the real detail of why the CCS loaded approach. It's horribly inefficient (with SE, which is already inefficient). I did this using a tube CCS on top of a 807. I couldn't live with a 700V amp producing a few watts of output. I went over to the MOT loaded version, halved the plate voltage and didn't notice any significant sound difference.

Surely the 10M45 can be put to better use in the power supply section. I have read that a regulated power supply SE amp has tighter bass response.

If I was really going to try this approach again, I would consider something like a 6080 (two valves for a stereo amp) or a EL36 or EL509 run at about 100V 100mA per valve. The EL36 should be fairly happy with such low voltage high current and it gives you plenty of spare voltage to burn up in the power supply.
The high current presents little problems for the MOT, and because of the low output impedance a toroidal is easy to source and will perform excellently - better than most conventional SE transformers. You might be surprised at the killer performance of a peanuts build amp.

Shoog

[rongon]

I personally think a Microwave Oven Transformer as a plate choke on the 6V6 is a better option.

Wait a minute... You guys make it sound like this is an idea that everybody knows about, for years already...

I did a forum search for microwave transformer and didn't come up with anything. I admit my search was probably flawed (and hasty), but heck, I guess I could ask the experts here. Where is there a discussion of this? Does a microwave oven transformer ("MOT") make a decent plate choke? What about for a 12B4A, for use at line level?

I know a couple of DN2540's would be smaller and probably better, but I just like the idea of ripping apart a microwave oven for audio purposes.

--

[sheep]

What does a MOT look like?

[TheGimp]

The last MOT I pulled had a mass over 7KG.

I guess size doesn't matter?