current feedback electrostatic amp

[starcat]

The tread on the Aavid 581202B02500G failed completely on me. Absolutely impossible to tighten the bolt properly and there is also no space for a nut from behind. Or are you guys using different sinks for TO-220?  

[Pars]

If you noticed in the datasheet, these are threaded for a 6/32 Imperial screw, which will not fit thru the insulator that you are using. I think most people use a longer screw with a nut/washer on the back (long enough that it is outside of the fins). Or use a 6/32 nylon screw with no insulator, but I don't know if that is verified to be tight enough or not.

[starcat]

Exactly, thread wasn't 4-40 UNC nor M3 and anything else do not fit through the insulator. Will use a longer screw, didn't thought about that, thanks. I was thinking about milling out a small portion of the two inside fins to fit in the nut. 

Edited March 30, 2021 by starcat

[Pars]

They do make another variant of that heatsink (2100G) which is not drilled/tapped, but I don't know if anyone stocks it. You could drill and tap that one for whatever reasonable fastener you wanted to use (4-40 or M3).

[jose]

The volt result may vary depending of your local supply (Congo explained it over there) but I don't think you have many differences. Order the one that makes you feel the most comfortable.

Be very careful with capacitors. 

6 hours ago, starcat said:

The tread on the Aavid 581202B02500G failed completely on me. Absolutely impossible to tighten the bolt properly and there is also no space for a nut from behind. Or are you guys using different sinks for TO-220?  

 

I recommend that you remove the metal washers.

 

 

[starcat]

Thanks, Jose.

 This looks like the final setup using stainless steel hardware and non conductive EK-TIM Ectotherm thermal compound on both sides of the Aavid isolator.

[starcat]

Powering the ES CFA with 400VDC from the GRHV, is there any need for 550V electrolytic caps in the GRHV or would 450V do fine as well, I mean even in the T2 the caps are 450V? 

PS: I am building the GRHV as well and will decide later on which PSU to leave with the amp. 

Edited April 1, 2021 by starcat

[jamesmking]

1 hour ago, starcat said:

Powering the ES CFA with 400VDC from the GRHV, is there any need for 550V electrolytic caps in the GRHV or would 450V do fine as well, I mean even in the T2 the caps are 450V? 

PS: I am building the GRHV as well and will decide later on which PSU to leave with the amp. 

if you are referring to the high voltage lines on a diy t2 - those psus use two 450V caps in series for double the voltage rating of one cap at the cost of halving the capacitance (except for rails bellow 300VDC output). The golden reference does not use series capacitors like this. the golden reference uses a single input and output cap per rail. So the each cap has to withstand the full voltage by itself.

if the output is 400V then 450V is ok for the output cap giving you 10% or so margin. the input cap will need to have a higher rating than this because all regulated power supplies require more input voltage than they provide output voltage so they can maintain regulation.   I think, if I remember correctly, the golden reference is about 330VAC input for 400VDC output. 330VAC once rectified through the diode bridge will give 330* squareroot(2) ~ 466Vpeak. Plus main voltage can vary, plus transformers can vary so to be safe you would want at least a 500V input cap. You may be able to get away with a lower input VAC if the golden reference can maintain regulation and your line voltage does not sag. Also consider that the higher the current draw the lower the output of a transformer and visa versa. So if you over spec your transformers current capacity the output voltages can be higher than you expect...

Edited April 1, 2021 by jamesmking

[starcat]

Thanks, James, appreciated.

Going then with 550V caps. The only one however that is kind of orderable atm at Mouser is a 1000uF 550V one, the KEMET ALC10A102EP550 - possible to use this one as the board says up to 680uF... Physically it will just fit perfectly in the case. 

[jamesmking]

20 minutes ago, starcat said:

Thanks, James, appreciated.

Going then with 550V caps. The only one however that is kind of orderable atm at Mouser is a 1000uF 550V one, the KEMET ALC10A102EP550 - possible to use this one as the board says up to 680uF... Physically it will just fit perfectly in the case. 

it seems like at the moment the only thing in stock is out of stock notices.... 😞

check the height of the 1000uF cap... at 105mm tall its a beast which will not fit into a 2u case. (around 65mm depending upon the height of the standoffs for the pcb is about as tall as you can go for a 2U (80mm tall case)... are you going 120mm tall 3U case?

the issues with very large capacitance input caps is that they will create a very large inrush current on switch on so you may need a larger value fuse than you would otherwise. Also the larger the cap the more diode switching noise it will cause because the caps will pull more current for shorter periods of time from the diode bridge rectifier. For the output cap some psu can become unstable if the output cap has a too high (or too low) capacitance. You may want to send a private message to kevin gilmore - the designer of the golden reference psu to get his input. Personally I use 470uF most of the time on the golden reference because I could not find any 680uF caps that would fit into the 2U cases I use.

Edited April 2, 2021 by jamesmking

[starcat]

This is exactly what I have been fearing from. Other than that the case fits it perfectly, with just some 5-6mm or so clearance to the top plate (it is not a 2U case, so more room). 

Some of the smaller ALC10 series caps, Mouser is selling only in quantities of 1000x, kind of nice! 

@kevin gilmoreWhat would be your thought about using a 1000uF 550V cap in the GRHV?

Appreciated all the feedback!

[kevin gilmore]

I think that caps that big can be very dangerous. So if you are going to use them, find a way to make sure you can never come in contact with either of the terminals.

[spritzer]

I would place that as top tier stupidity

[starcat]

6 hours ago, kevin gilmore said:

I think that caps that big can be very dangerous. So if you are going to use them, find a way to make sure you can never come in contact with either of the terminals.

Thanks Kevin for your feedback. Well, 680uF might be less dangerous, but still dangerous enough and for either precautions need to be taken. The question was if the circuit will be working properly with 1000uF as well or any side effects are to be expected if going above 680uF.  

[kevin gilmore]

the guy in vietnam is using even bigger caps. so no reason why it should not work.

[starcat]

Who is this guy from Vietnam?

I would actually put 680uF but can't find any orderable in 550VDC... 

[kevin gilmore]

ULTRA REFERENCE electrostatic headphone amp KGSSHV Carbon for Stax SR-007 SR-009 | eBay

they are 1500uf caps as i remember.

[Pars]

Those things aren't film capacitors as he states in his ad are they? Electrolytic would be more like it.

[starcat]

I would say electrolytic, yes. Interesting where he has got them as they aren't off the shelf. 

Edited April 4, 2021 by starcat

[tee]

I asked him before, those are 600uf film caps.

[Pars]

Ahhm, damn, they're these guys: https://www.mouser.com/ProductDetail/Cornell-Dubilier-CDE/947D601K901DCRSN?qs=INK05T%2FwFA13oHAmtB1Cnw%3D%3D

Crap, those are a bit pricey, but in stock.

[starcat]

900V and only 10%. 75mm high.

Edited April 4, 2021 by starcat

[mwl168]

You can buy the 2000uf ones for a cool $700 a pop.

[starcat]

For that money one is better of connecting it to a battery. Much better and cheaper. 

Or if you like it modular: 1025uF 1800VDC 115A Modular DCR

Check out the 600V Kemet 2100uF 10% and the 900V version for a much more realistic price. 

Edited April 4, 2021 by starcat

[starcat]

Laying down the modules for my CFAE. I decided to go with the GRHV instead of the BH Mini. Custom making the front and back panels. Back panel will have on the left/right symmetrically laid out and flush mounted the XLR ins/loop-outs and in the middle the IEC inlet. Fabricating the PCB boards for the XLR in/outs atm. Looking to use the CP4-2500 pot on a board as well together with two Stax teflon jacks on the front, symmetrically positioned left/right from the pot. Will use the 2.54mm Amphenol PV wire to board connectors between the XLR and volume pot PCBs as connectors are gold plated and directional, crimping a shielded Sommer Cable Cicada SO-D14, similar to the Canare L-2B2AT. Using 600V 300°C teflon wire for the HV. And... waiting for remaining components and transistors for the GRHV.

Edited May 8, 2021 by starcat