nopants

Kevin Gilmore Circlotron Output, Large Enclosure [high-slew] Electrostatic Amplifier

162 posts in this topic

Hope you guys don't mind, but I'm taking the liberty to collect some info, links etc. here. 

 

900V floating supplies

350V/us

 

the cole slew  (couldn't resist) seems to be happening in bits and pieces in multiple threads on multiple forums and there are some implementation issues which seem to be worth discussing...namely fitting all of this somewhat daunting amplifier into one case.

 

I guess at some point Kevin decided to split the input and output stages-

output board

output schematic

 

The output stage features the floating supply

floating supply fboard

floating supply schematic

 

the input/driver board

 

 

As of today the aim is to fit this all in a single modushop 5U chassis, the floating supplies make it prohibitive to split the build across multiple boxes.

 

Transformer Specs:

 

the one would be a 450v version kgsshv transformer

 

the other 2 are 4 x 380vrms @ 100 ma each

 

 

Supposedly trivial to get an EL34 version of this working but SiC version will be first

 

Just wanted to hopefully direct discussion here instead of dragging it across the kgsshv/GB/headfi threads. 

 

I'll update this when I get around to fishing around for the appropriate information. Thanks to Kevin for making the design process somewhat transparent

 

----------------------------------------------------------------------

 

Figured I'd put this here too, since Kerry speaks for all of us:

 

I just saw the schematic for this - It's official - I've shat  :D

 
Edited by nopants

Share this post


Link to post
Share on other sites

Don't know how else to organize this ATM...

 

Some items:

 

Marc suggested using DC Link? capacitors in the floating supplies-

 

 

Do you really need that much capacitance, if not what about using DC-Link capacitors? At least do the math from a cost point of view.

 

 

go and check the price of the dc-link capacitors, they start at $65 each

 

2 x 680uf in series is 340uf, and you really don't want much less than that.

 

there are no dc-link caps that are rated at 1kv

 

anyone find something else that works, let me know. The board with 6 caps

will not fit in the box.

 

and when did the stax mafia ever care about cost.

 

 

I'm not 100% sure about each of the appropriate pieces that compose the entire amplifier, but  the first image that formed in my mind the F5X enclosure described here. Transformers on the bottom, amplifier + output psu on the sides, standard psu on the back/over the transformers.

 

Seems like something to leverage here would be the way EUVL chose to mount the capacitors...maybe another set of daughter boards? Allows more options for aspect ratio, and you're not subjecting these honking capacitors to as much gravity-

 

post-3212-0-74125200-1435918289_thumb.pn

 

I will try to be productive and see how the actual measurements stack up. One bonus of using "standard" enclosures  is that we also have access to riser panels which seem to be made with the larger cases in mind, There is also the possiblity of standardizing the I/O, faceplates, board mounting holes blah blah blah

 

 

figuring on the transformers at least 20% bigger than the t2 transformers so say 4 inches diameter and 2.5 inches thick

 

the power supply boards are 7.45 x 6.48 and the height will be that of the power caps

 

the amplifier board and output board should stay at 6.28 x 3.37

 

the kgsshv power supply is 6.92 x 5.16 and the same cap height

 

chassis is pesante 5U monster amplifier box

 

any other ideas, and i will certainly listen

 

 

Edited by nopants

Share this post


Link to post
Share on other sites

how to get it in the box, I don't know for sure yet.

 

what I do know is that I have the opto bias working  perfectly and

a stereo prototype that I can listen to. (well at least on the bench,

no way to move it)

 

its flat out amazing.  As much fun as listening to the T2 for the first time.

 

did some testing at 1kv peak to peak into 50k(50w) in parallel with

470pf.  just fantastic. till the cap exploded at 20khz.

 

switched to the 1kv rated terminal blocks which are a fair bit bigger,

but not much more money.

 

 

as far as I know the only push pull solid state electrostatic amp

ever. (parts for togte are not available)

 

output board schematic

http://gilmore.chem.northwestern.edu/circlotronoutput.pdf

board size 5.45 x 4.05 inches

 

driver board same as kgsshv-circlotron with a few resistor changes

picture

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

board size 5.45 x 4.05 inches

 

picture with the 1kv rated terminal blocks

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

likely going to use the other opto to make a safe current monitor

 

new power supply picture with the 1kv rated terminal blocks

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

board is 7.45 x 6.48 inches

 

also you need the 2 x fat power supplies

 

thinking of 3 transformers, 2 of which have 5 hv windings

(4 for the 900v and 1 for the hv supply)

 

and then one small transformer for the +/-15 supply

Edited by kevin gilmore

Share this post


Link to post
Share on other sites

a good candidate for the nfb preamp in front, less 5u face plate clutter haha

 

If you are willing to emulate the F5X layout you should be able to open up a significant amount of heatsink area if you're planning on mounting the PSU's vertically by offboarding the caps (need to check x dimensions). then it can be something like

 

[insert sink layout here later]

 

mounted on the sink, assuming the output board spits out the most heat? based on the floor space/clearance maybe some of these can be turned horizontal via brackets, like either the driver or the output, depending on which of these needs more sink area.

 

transformers on the bottom, not sure where to put the fat PSUs if you need 2 of them...

 

For reference: http://www.diyaudio.com/store/amplifier-chassis/deluxe-amplifier-chassis-enclosure-5u-217.html

 

  • Width: 450mm with 360mm usable internal space (14.1")
  • Depth: 400mm (15.748")
  • Height: 210mm (8.26")
Edited by nopants

Share this post


Link to post
Share on other sites

you need

 

4 x 900v floating power supplies (could split into singles if it makes more sense)

 

2 x 450v ground referenced power supplies (likely the fat power supplies)

 

2 x 15v ground referenced power supplies (on one of the fat power supplies)

 

1 x bias  (on one of the fat power supplies)

 

board files posted

 

circlotronhvpower.zip

kgsshvcarboncirclotron.zip

kgsshvcarboncirclotrondriver.zip

Share this post


Link to post
Share on other sites

As in you need 2 of your currently pictured floating supply boards, right? 2 900V outputs per board? Splitting the boards would make things nice and symmetric if you position the boards on each side of the output board.

 

Also footnote for stupid people like me, but opt for 3 oz cu if you're going to fab the boards

Edited by nopants

Share this post


Link to post
Share on other sites

the current hv power supply is a dual board designed for one channel

 

depending on how you assemble, you can put some of the terminal

blocks on the back of the board, especially if mounting vertically.

 

so you could put the output stage board right next to the power supply board

and put the power connectors on the output stage board on the back for

very short wiring

Edited by kevin gilmore

Share this post


Link to post
Share on other sites

ok got it, I was spacing out in board layout land

 

so one "standard" fqp hv supply can just go over the transformers if the driver board doesn't need more current than that

Share this post


Link to post
Share on other sites

This reminds me of the nuclear arms race in the 50s between the US and the USSR...except there is no USSR.   Comparing KG's amps to any other designer's is like comparing the US nuclear arsenal to, say, France. ;D

Edited by JimL
1 person likes this

Share this post


Link to post
Share on other sites

it probably makes more sense to move the opamp stuff from the output board

back to the driver board, makes it easier to wire, going to look into that now.

Share this post


Link to post
Share on other sites

This reminds me of the nuclear arms race in the 50s between the US and the USSR...except there is no USSR.   Comparing KG's amps to any other designer's is like comparing the US nuclear arsenal to, say, France. ;D

 

FrenchMilitaryVictories.jpg

Share this post


Link to post
Share on other sites

We could make those boards a bunch smaller without sacrificing safety.  :)  Also routing around the crazy high voltage parts to make sure they can't spark might be a good idea. 

 

I don't see a ground plane being a problem if we pull it back to say 60mill or even more.  We could block out the area around the output devices too just for an extra level of safety.  Also goes without saying that these have to be quality boards, made by people who understand HV boards.  No super cheap PCB's here. 

 

To think there are people on HF who think I'm going to build these...  :palm:

Share this post


Link to post
Share on other sites

If you are confident about the shrinking (I'm already convinced haha) then I guess it's a question of how much sinking is really required....

 

btw have you had any issues with bracketing the output SiCs on the carbon? After reading the F5X article I was wondering if it is worth the effort to to clamp the devices to the sink, as opposed to just using a single screw/washer/etc.

 

After cobbling together the Megatron I can't imagine having to do it "professionally", let alone this beast.

 

The more I think about this amplifier the more I'm becoming part of team stax power amp. What knob could possibly do a circlotron justice

Share this post


Link to post
Share on other sites

No problems using angle brackets with the Carbon but they have been dropped on the newest version of the board.  The devices will be mounted directly to the sinks from now on. 

 

The plan is to build four Megatron's but who knows if I'll go through with it. 

Share this post


Link to post
Share on other sites

ok that's good to hear, maybe it'll be worthwhile to bracket one of the circolotron driver/output/psu so it doesn't occupy a ridiculous amount of space in two dimensions

I should start looking at eBay for used forklifts so I can tote the circlotron to all those meets

Share this post


Link to post
Share on other sites

Wow, massive madness here !

Can't wait to see it built.

Share this post


Link to post
Share on other sites

i have a number of things at work that use these to246 parts.

 

all of them are spring loaded to a thicker than normal silicone pad, no hole.

i posted a picture of my original prototype, has the same things.

 

this thing runs at about 2.5kv, so it must work.

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

 

 

what i will be doing for the carbon is the same thing, except i will have a bracket over the

to246 devices, with 3 holes, and the 2 x to246 in the middle, then tighten from below.

 

i also don't know where to get sheets of that silicone shown above

 

unless i can find the numbers on the spring retainers, i have a few, but don't know where to buy them.

 

same thing would definitely make sense if you are going to attach directly to heatsink.

Edited by kevin gilmore

Share this post


Link to post
Share on other sites

I just drill and tap individual holes but something like this would work.  Not a whole lot of room though as I'm right up against the end of the sink. 

Share this post


Link to post
Share on other sites

/

Edited by JoaMat

Share this post


Link to post
Share on other sites

 

btw have you had any issues with bracketing the output SiCs on the carbon? After reading the F5X article I was wondering if it is worth the effort to to clamp the devices to the sink, as opposed to just using a single screw/washer/etc.

 

 

I used large washers (~16mm M3 hole) when I build my F5X, got the idea from some of Nelsons amps, they work really nice spreading the pressure evenly.

Share this post


Link to post
Share on other sites

messed with the feedback a bit and moved all the stuff back to the driver board

board files updated

output board schematic updated

 

 

just in case you people think this is a monster...

 

we could do a megatron with 845's  ::)

its super simple. But brutally expensive and

huge.

 

at some point a se 845 with transformer, my take on franks beast

a dynamic amp, same thing

Edited by kevin gilmore

Share this post


Link to post
Share on other sites

Might skipped my attention, but has the schematic for the input been posted?

Had a look at the floating psu .... Seems very similar to the old kgsshv psu.... Would that pcb work using offboard caps?

Share this post


Link to post
Share on other sites

its the old power supply because the voltage exceeds the capabilities of the new supply.

16 offboard caps is not going to be easy. plus possible issues with the breakdown voltage

of the insulation of the capacitors when mounted in capacitor mounts.

 

driver board is modified kgsshv-carbon board with stuff missing.

 

newest posted boards, aligned the opto connections between the 2 boards, designed

to wrap around from board to board

Edited by kevin gilmore

Share this post


Link to post
Share on other sites

if you don't think that mounting the caps horizontally will be a problem then leave it

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now