kevin gilmore

From what I've heard, the current sources for the tubes are the same cascode in the same exact physical place. Meaning the same screws you can't remove from the triple of small heatsinks. So all you have to do is dump the rest of the crap and drop in 4 tubes. Basically the same as the exstata tube version without the bipolar output stage. Very simple circuit. I could draw it up in a few minutes. The only issue is what the tubes are, and what the power supply voltages are. since evidently there are no other heatsinks on the board, likely its not grounded grid design. Probably a couple of hours of board layout max. I could turn something like that around in one day, and get pcbnet.com to deliver boards in 3 days. Its going to sound way better than the mosfet pile of crap. And be pretty much identical to tube based stax product of the middle 1970's. Still going to be built like shit.

LL mark 3 with tubes being shown at RMAF. all the mosfets in the amplifier are gone.

with those size paws, looks like its going to be a fair bit bigger than that.

how big is puppy supposed to get?

admin as in singular.

that schematic is for one channel plus a little bit. so there should be 8 power output cathode resistors total.

this matches the board http://gilmore.chem.northwestern.edu/megatron.pdf

60 seconds to let the tubes fully warm up same as the T2. you really want to do this at these current and voltage levels especially with expensive tubes. Ristar used 1k cathode resistors on the amp board instead of 1.5k resistors. The 10 big ones. Ristar changed the 15 ohm resistors on the power supply to 10 ohm an he left off the 2n3904.

needs t2 style power on delay for the high voltage. Or separate switch, one for filaments, one for high voltage. 150ma on the power supplies. about right with 1k resistors. room heater for sure. 135 watts of high voltage plus about 85 watts of tube filament power. no front end tube distortion, no limited output frequency response due to limited size center tapped inductor core etc... plenty of gain, low distortion. yeah baby.

If the power supply voltages are right and the offsets and balances are in the 10 volt Range you are done. Go and listen to the music. Leave the 15 ohm resistors alone for now, let me know The voltage across them. Leave the current limiters off for now. Will have modification soon. Too many projects reaching completion at the same time

yes the servo can be added to the older boards.

speaking of extraordinary http://gilmore.chem.northwestern.edu/dsc_0478.jpg an absolute perfect void free piece of red mallee and a chunk of absolutely black as can be ebony

I know 2 shills for ECP and we are both batshit crazy.

+2 coming soon LLmk3 Seriously this is the major safety issue. And its such an easy and cheap fix. Unless you make it external. (male and female plug with resistors inside) But with the mosfets it still sounds like crap. At least it would be a safe piece of poor sounding crap. I would love to see schiit do an electrostatic amp.

did I see anything about the lack of safety resistors on the outputs of the amp to limit current in fault conditions? I better go and read that again. No possibility of thermal runaway with bjt's in an amp like this, its pure class A, bias is fixed by the current sources! Not possible to build an AB1 amp for electrostats.

hmmm, same post as over there. pretty much same response.

No plans to do that

strange both of my boards go smooth from 120 to 200ma with the 75k resistor and the pot. have not tried the lower value resistor yet. kind of busy. northwestern network down for most of the day due to dns issues.

looking very nice

So let me understand this, you think the pictures are somehow faked? You think the hennyo class soldering job and all the cold soldering joints in a $5k product are somehow acceptable? You think that the screws on the current source triple and transistors are somehow replaceable by someone other than a magician? How about the transistors in the middle of the board? You think that a power dissipation in excess of 80 watts inside a box that size is safe? You think that a posted Thd+N graph that exceeds all possible theoretical calculations by at least 30db with J176 input fets designed for switching (not linear applications) with an input noise of a minimum of 10nv per root hz and an amplifier gain of 50db is real? when alex could have used the real thing instead (2sj74 or lsj74)? You think that the lack of safety resistors between the amplifier and the stax jacks used by every stax amp, every headamp product, every one of my diy products, etc is just a suggestion? Amos, you don't have a fucking clue what it is that you do not know about this. And Alex does not have a fucking clue, because if he did, he would not build a piece of shit like this. Cavalli, Beauty really is only skin deep. At least Mikhail's products sounded really nice when they did work. And when correctly repaired, they work for a very long time. This thing really did sound like a piece of crap. Mosfets don't belong in electrostatic amplifers! It sounds MUCH better now. We are not done yet. When we are done, we will present all the work to everyone including alex. Trust me, he won't do shit about it. Because it won't be liquid anymore.

some of the opamps have horrible offset specs themselves. as long as pin 6 of the opamp is in the range of +/-6v it is doing its job.

depends on which end the pot is at. 1.8k or 2.2k the 500k resistors are really a safety thing. best to simulate the pot for real, but they are ok, just a bit noisy.

although absolutely nothing can compare with Mikhail's build quality (well trevornetwork and hennyo).... The squarewave done right would have been a kickass amplifier. And the klone definitely is. extended use of stock LLmk1 and probably LLmk2 due to heat and other issues is going to definitely cause reliability issues, and out of warranty repairs are going to be nasty. LF are already known to be very hard to repair. Not allowing pictures to be shown elsewhere just makes it worse because the manufacturer has no reason to make better or more repairable products in the future.

1) with both boards connected the temps will drop a bit 2) you should change the 2k resistor to get the voltage to zero before you add the servo

With pot at minimum bias is 120 ma with pot at max it's 200 ma Probably should cut 75k resistor to 50k And 1 meg resistor to 100k. Servo takes 2 minutes to go from 200 mv to .2 mv with 1meg resistor. Thd and noise below baseline of my meter. Slew rate which is already Limited by input filter is very fast. Just great with hd800. Going to add pads to measure bias on next board rev