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

[Tinkerer]

So, looking more closely at the GRLV board and reading its thread, I realized the 16-0-16 carbon secondaries aren't going to cut it with a 4 volt drop when I want 15VDC. This means a 4th transformer. I actually found a nice used toroid for only $30 shipped with the mounting hardware and the right rating. It's trying to cram that in there might take a little work. Somebody please correct me if I'm being stupid.

Also, to get around the current limitations of the switches I'm using, I thought it might be good to hook up a 25A-rated relay with a heatsink. I'd have a keyed master power switch in the back for a small ungrounded 12VDC supply I made from a wall wart, and the circuit for that completed by engaging the front panel switch. I'd also run the LED's off this circuit and have the voltage divider to control brightness in front of them since it only take's about 3 VDC to engage the relay. Is there anything wrong with this setup for power?

Oh, got the front headphone socket mounts drilled and tapped as well.

And one last question, since I'll be drilling some back panel stuff next and I need to know this for jack options. Can this run single ended or only balanced? I know most of the other KGSS based designs can run single ended but I don't know how the output stage affects things. True push-pull is balanced only if I recall right.

[Pars]

For the 15v GRLV, you should have ~21.2 V ((16V * 1.414) - 1.4) raw after the rectifiers with your current 16-0-16 Vac transformer. I would think that would be plenty?


Sent from my iPhone using Tapatalk

[GeorgeP]

16-0-16 should work fine. Did you try it and then got lower voltages?

[Tinkerer]

No, was just looking at the board silk for 20V and the comments of the thread. If it works, that's great. I would assume I need to change some of the resistors though?

[mwl168]

I tested the GRLV drop out voltage when I built one for +/- 20VDC output and it was about 3VDC. the 16-0-16 secondaries should be plenty for the desired 15 VDC output as Pars has stated.

[GeorgeP]

1 hour ago, Tinkerer said:

No, was just looking at the board silk for 20V and the comments of the thread. If it works, that's great. I would assume I need to change some of the resistors though?

if you are building it for 15v you can't just follow the board silk.

[Tinkerer]

I know the R8 and R9 have to be changed even if I was using the silk AC input value. I assume with less, the R10 and R7 as well. I just don't know to what. Will try to figure it out when the group buy boards come in. I know a few people upgraded their Carbons with the GRLV. Since that's the changes I need to do, if that's specifically documented anywhere, I'd really appreciate anybody letting me know.

Edited November 25, 2016 by Tinkerer

[mwl168]

for V+: (R8 + R7) / R7 x 10 (reference voltage of D5 - LT1021-10), R7 = 1.5K, R8 = 750 for + 15VDC for V-: (R9 + R10) / R10 x 10 (reference voltage of D7 - LT1021-10), R10 = 1.5K, R9 = 750 for -15VDC   On the goldenreference4 board silkscreen, R7, R8, R9 R10 are all 1.5K. R7 and R8 are connected and R9 and R10 are connected. R8 and R9 are the 1.5K resistors closer to the DN2540 on each side of the board.

Edited November 25, 2016 by mwl168

[Pars]

^ As mwl168 said, R8 and R9 (the ones you leave alone for voltage changes *) are the ones that terminate at ground plane on one end.

* for some voltages, KG used 2K instead of 1.5K for these. I am using this same tactic on a supply for 35V (5K/2K).

[Tinkerer]

Getting some of the last back panel stuff done by hand before I give it a new coat of pain and labels. This is the relative position of everything near the back panel. I just needed to know if it's a problem to have a solid state relay and that little power supply that close to the volume pot.

 

[charlo89]

What is this pot ? It seems to be an old linear potentiometer

[Tinkerer]

Yeah, just an old helipot precision potentiometer. It tracks extremely well and I don't fool with volume too much so I figured I'd give it a try. I don't mind giving up log volume control for the price difference between it and an RK50. We'll see if it was a dumb gamble or not once I get it all together. I got the parts to replace it with a more standard 4 gang audio pot if I need to.

[Tinkerer]

Alright, getting ready to get all the stuff together. Casework minus labels is done. Secondary power circuit is together and works, powering the solid state relay and the case LED's. Left it on all day. Voltage divider circuit doesn't get too hot and controls LED brightness fine. Inputs up to the pot are all wired. Keyed bypass switch works in all three positions (normal, potentiometer bypass, and loop out only). Audio wire shields all connected to chassis ground. Audio grounds themselves kept to their separate channels. Tried to test my GRLV+ as my last PSU board but I was a dumbass and didn't put the NPN/PNP MJF TO-220 pair in. Didn't realize I used two MJF15031's until I started it up and got horribly lopsided voltages. So a couple days until that part comes in. Also, what's the pots for on the GRLV+ again? Looks like they balance between the two halves of each circuit, but I'm not sure where to attach the multimeter leads to measure it. On the side of the 1M resistors closest to the pot?

 

If anyone else is anywhere near completion on theirs and has run into any problems not mentioned in the thread. Now's the time to say something.

 

Otherwise, I just had a couple stupid questions. You can run this single ended since it's basically a carbon with a circlotron output stage bolted on the front, right?

EDIT:found the answer to CT transformer grounding. But now I'm wondering if having my grounding terminal block with two bolts to the chassis was such a good idea. Might have to isolate one of those screws.

 

I think if nothing explodes in the next few days, it should be playing music by this time next week.

Edited January 21, 2017 by Tinkerer

[Pars]

My GRLV didn't have the pot, but my impression was that it was used to balance out the +/- voltages, so just check outputs to ground? Or not use it...

[Tinkerer]

When I originally fired it up, the voltages seemed to automatically compensate for any adjustment I made. I would guess I because both halves are connected since I'm using a center tap setup.

 

With all the stuff minus the boards and brackets installed, this was getting a little too heavy to move around by myself. It's easily a hundred and twenty pounds but it's the bulk that's killer. So I went ahead and put it where It's going to stay, next to my desk and shuffled everything else around. I just thought it was worth showing the T1W next to it for a size comparison. Final assembly continues, but it's mostly just cosmetic stuff like lengthening the transformer leads and sleeving them to make them look pretty while I wait on the last couple parts. This monster is definitely going to need a friend to help me move it if it ever budges from this spot.

Edited January 22, 2017 by Tinkerer

[kevin gilmore]

I never said this was a tiny amp. Nor low power.

[Tinkerer]

Yep, that's what makes it fun.

 

[Tinkerer]

Just needed to ask, what's the max currents for the driver board and the output board?

 

Driver board was around 15mA default, and it will run 20mA+ with driver board based feedback, but I read earlier in the thread that the output board wouldn't like the input that high because of the opto relays?

 

Output board, around 25mA default, no idea what the upper limit is.

 

I have a lot of sink to play with here so just looking for the ranges I can deal with.

[kevin gilmore]

10 to 15  ma for the driver board, and 25ma for the output board.

But i have tested to 50ma and the heatsinks definitely get toasty that way.

No difference in sound after about 25ma

[Tinkerer]

Thanks. I'm waiting on one last GRLV part but I got this running on the 7815/7915 LV PSU temporarily to set it up. Nothing exploded. PSU's all good. LED's all on and good. Easy to balance. Currents adjusted.

 

So here's my problem. I've got +450VDC offset on both channels. Exactly the same on both so whatever it is, is mirrored. Adjusting the offset pot doesn't do much of anything either. Also, the output boards are putting voltage across the big 120ohm resistors at the right level but not really getting hot. Is that normal? Thoughts?

EDIT: 90% sure it's something on the modded driver boards since it's basically the B+. Will get a diagram up in a minute.

Edited February 2, 2017 by Tinkerer

[kevin gilmore]

you need to lower the value of the resistor in series with the offset pot.

[Tinkerer]

Lowered them to about 150 ohms. That did the trick.

Unfortunately, while I was setting the second one, one of the four HV900 supplies shorted and blew up. All the PSU stuff had run 48hours+ before under load so I don't know what happened. Going to take some work to get it all unwired, back out of the case and sorted.

 

EDIT: Looks like maybe a transformer lead was the culprit. Terminal screw got loose somehow, probably during cable routing. Looked snug because how close it was to its twin secondary but gave it a yank and it slid right out. That's a lot of exploded transistors and resistors for that kind of dumb mistake.

Edited February 2, 2017 by Tinkerer

[kevin gilmore]

all transformer leads need to be twisted and soldered.  and then cranked down good

[JoaMat]

@Tinkerer,Sorry for your accident. Did something broke on the output board?

[Tinkerer]

Left channel output board blew both mosfets and one of the big 120 ohm resistors literally into pieces. HV900 blew one 10M90S, the big mosfet, the big 12 ohm current resistor, and one of the CMF60 resistors. Might be more damage, but those were completely wrecked. This wasn't a pop and some smoke. It was a bang. No lifted traces or scorching though. Like Kevin said, a lot of power in this thing to be very careful of.