My latest amp (845 / 211 SET A2)

[thehoj]

Hey guys, Here's my latest amp. It can run with either 211 or 845 power tubes, bias is a fixed bias configuration, and I have it set up to be able to provide appropriate bias for either the 211's or 845's.
It's a SET design, I'm using Rod Coleman filament regulators, and the amp uses mosfet source followers (Toshiba 2SK3564) after the driver stage to supply grid current for the power tubes.
HV B+ is tube rectified using a pair of 6CJ3 damper diode tubes. Final HV B+ is 1050V.

The driver is made up of CCS anode loaded #27 or #56 triodes feeding a CCS anode loaded 6BL7 (driver not pictured).
I've measured output power to be in the 32W - 35W per channel range (just before clipping) with both the 211 and 845 tubes, well into A2.

Because it's got so much heft to it, I have the driver in its own chassis, I also have the raw DC supply for the filament regulators in its own chassis. So it's a 3 chassis amp, and I'm only posting pics of the power stage at this point, just because I don't have the driver and raw dc supply put together in the proper chassis'.

I'm running a few custom wound Heyboer power transformers, a few Hammond power transformers, Monolith Magnetics output transformers, and a number of Hammond chokes, as well as a few Transcendar chokes.

I don't need to run this amp full tilt with my setup, it's just nice to know that if I'm approaching swinging the grid near or just over 0, it won't crap out.
Anyways, some of you may not agree with the sand in the signal chain, but that doesn't bother me, and I find this topology really interesting, and very cool to see working in practice exactly as I had expected it to.

The guts aren't exactly pretty, but it's very solidly wired up. I have no concerns about any part of the internals.
The amp as it stands (having the driver as well as the raw DC supply which feeds the filament regulators in separate chassis') weighs in at 96 pounds. The chassis was ordered from Landfall systems.

I'm currently running with some Klipsch Forte I's, and the amp is very quiet even with these fairly efficient speakers.

Sorry, the pics aren't the greatest.

 

[for_p1]

If you skip filament regulators (just use regular DC source) and convert to silicon carbide rectifiers for B+, you can fit everything in a single chassis and under 100lb weight. From my experience regulation is not necessary with 845/211 tubes - there will be almost no hum even without it. I would also replace green LEDs in current meters with yellow ones: yellow light from LEDs practically matches color of 845/211 filaments.

 

[thehoj]

If you skip filament regulators (just use regular DC source) and convert to silicon carbide rectifiers for B+, you can fit everything in a single chassis and under 100lb weight. From my experience regulation is not necessary with 845/211 tubes - there will be almost no hum even without it. I would also replace green LEDs in current meters with yellow ones: yellow light from LEDs practically matches color of 845/211 filaments.

I'm pretty happy with the filament supplies I'm using. I don't plan to change things now. I really don't mind having the raw DC supply in a separate chassis. Also, it's not the regulators that take up much space.. It is in fact the raw dc supplies that do.. 2 Hammond power transformers, 2 hammond chokes, a number of caps, etc.. I like having it outboard.

As for silicon diodes for rectifier, the tube rectifiers really don't take up much extra space, and they're working quite well, plus I like the built in slow ramp up of B+ that they provide (indirectly heated) - I know I could do this with solid state diodes as well, but I feel like things are running so well, I don't want to change anything.

The LED suggestion in the meters, I did actually consider amber or yellow diodes.. Not sure why I went green. I may change those out and see how I like it.
The meters display cathode current on the power tubes, to help with easy biasing.

 

[mrwood200]

Looks like an impressive hifi build! Interesting design, quality parts, and +1000V !
Now, let's get those photography skills up to snuff. More & better photos please!!

 

[santa]

Excellent.
How much capacitance are you using on the coleman regs?

 

[thehoj]

Looks like an impressive hifi build! Interesting design, quality parts, and +1000V !
Now, let's get those photography skills up to snuff. More & better photos please!!

Thanks. And I know I know! I have such poor lighting in my basement, and it's so so damn heavy to carry this thing upstairs.. I'm going to try to get some better pics soon.

Excellent.
How much capacitance are you using on the coleman regs?

Are you talking about the raw dc supplies feeding the coleman regs? I attached a schematic of those. I have one Hammond 20V @ 5A power transformer per tube, bridge rectified with Vishay MBR1045 schottky diodes, a choke input supply with a 10µF cap before the choke (so low it's effectively choke load. The reason for the 10µF film cap before the choke is apparently to control transient voltages at the choke input - according to the designer of the regulator). This feeds into Hammond 159ZJ 10mH / 5A chokes, and then 2 x 22,000µF Panasonic TSUP 35V caps in parallel. I measure about 100 - 150mV of ripple feeding into the regulators.. The regulators take that down to the µV range. Couldn't be happier with them.

 

[DeafbyKhorns]

Isn't 20v a little high for a 10v supply? You must be heating the room with those sinks!
Whose meters are those? I'm jealous my Simpson meters don't have a light!

 

[thehoj]

Isn't 20v a little high for a 10v supply? You must be heating the room with those sinks!
Whose meters are those? I'm jealous my Simpson meters don't have a light!

It's high because I'm using a choke input power supply..
Check out the difference between choke input and capacitor input here http://www.aikenamps.com/index.php/chokes-explained
Basically, better regulation, and lower rms current requirement for the transformer going with choke input.

I attached two PSUD2 sims showing the difference.. With the capacitor input you can see that I can get away with lower voltage power transformer, but the RMS current draw on that transformer is nearly 7A.. With the choke input you can see that I need the higher voltage, but the RMS current draw on the transformer is closer to the actual draw of the filament ~3.25A.

Also, for 10V output from the regulator, the designer recommends in the ~15V range of raw DC. PSUD2 shows more like 16.5V in the simulation, but in real world I'm getting about 15.5V.
I could have probably gotten away with 18V / 5A power transformers, but I figured I'd give myself a bit of breathing room, and could always drop it a bit with some resistors if need be.. Though I don't need to.

That all being said, yes the regulators make a bit of heat, and the raw dc supplies also do.. Though it's not that bad..

Here's a link to the regulators I'm using, as well as some instructions / design guides.
http://lyrima.co.uk/dhtreg/dhtRegIntro.html


As for the lights in the meters, I added LED's for that myself.. Just mounted some really really bright Green LEDs behind the faces inside the meters.. It took a bit of work, some holes drilled, etc.. I like it. And I use that as a power indicator of sorts.
Here are the meters I used https://www.fairradio.com/catalog.php?mode=viewitem&item=4071 Great price for some really excellent meters.

 

[RossW]

Very cool! I love big transmitter tube amps.

 

[for_p1]

How high is B+ supply for pre-driver (6bl7 tube)?

 

[thehoj]

How high is B+ supply for pre-driver (6bl7 tube)?

580V.

I attached a schematic. The tubes are biased such that they remain within their specs as specified on the datasheets.

I need some pretty high B+ for the driver, because the 6BL7 will swing nearly up to this high on the anode at full drive. My power stage requires nearly 400Vp-p to drive to full power in A2 when I'm using 845 tubes.

 

[thehoj]

I would also replace green LEDs in current meters with yellow ones: yellow light from LEDs practically matches color of 845/211 filaments.

I did it! I was on the fence.. But I prefer the amber LED's in the meters. Here are some pics.

I also attached a few pics of the process to put the LED in the meters.. It's a bit of work..

 

[thehoj]

Got the new driver built, and thought I'd upload a few pics of the proper driver and the power stage.. Pics are a bit better this time.

 

[Brice]

You've build a monster
Fantastic !

 

[thehoj]

Thanks! It is a monster.. Bigger than I had originally envisioned. I'm very happy with it for sure though.

 

[primosounds]

Very nice build. Looks like a labor of love. Could you post a schematic or schematics of the entire amp? regards, primo

 

[drtool]

What a wonderful 845. I hope to build one myself someday when I get up to speed. Meters are cool. Great job.

 

[thehoj]

Thanks guys. Here are some schematics.

Broken up into power stage, driver stage, and filament supply.

 

[Kegger]

Very NICE!

And I like your choices..

 

[tcqanh]

It's a SET design, I'm using Rod Coleman filament regulators,

Hello,
Rod board here causes a lot of heat. I do not like it for this position anymore. Why don't you take LCL circuit for this. Noise comes deadly quite and sound get fantastic.
Regards,
Quoc.