kevin gilmore

12 volt SLA battery. Works great. The 7ah battery will run it for about 18 hours.

Getting rid of Yamas is a really good thing. Incompetent morons. I recently saw the work they did (and charged $2k for) to try and repair a T2. Someone serious should pick up the dealership.

Hand wiring takes too much time...

If you do it with this case, there is plenty of room for everything and you can lay the transformer down. Its 3.5 inches in diameter and about 2 inches high, so on the power supply board that is 8.65 inches long there is plenty of room for the transformer.

What i ment to say for the T2 like dynamic amp, was BALANCED OUTPUT only. It would still be balanced or unbalanced input. The thing is that the servo will make both sides of the output zero volts, but with respect to ground it may still be 1 volt or more. (same as the nelson pass things)

SEE it goes like this.... The higher the power supply voltages (everything else staying the same) the faster the slew rates. The word OVERKILL does not seem to be in my vocabulary. It would take me a while to find it, but i think ti kan said the B22 did something like 75 volts per microsecond. I'm not sure i believe that, but then an electrostatic amp should be able to do 10 times that. Or 750 volts per microsecond. A non-trivial task for sure. Last nutty idea for this year... Anyone interested in a dynamic version of the T2 with 6as7's as output tubes?? Sort of a re-do of my original MIG amp, except completely circuit board based, with + and - 75 volt power supplies... Balanced ONLY. Although one of these days i'm going to do a DC coupled circlotron...

That is going to be a bit of a problem with the neutrik jacks... They are designed for 6.35 mm. Might fit anyway but they are going to bend a bit.

This will likely go down the same way as the T2, except i'm not doing chassis this time. (well except for me) So in a month or two after i have all 3 pots, and have measured 5 times or more to make sure everything will fit, then i will do a board run. Say 20 pieces of each. The boards are much cheaper than the T2 so its no big deal if it takes me a couple of months to get rid of them all...

added the tkd pot. adding the alps pot might be a bit tricky. Since they are on extenders like the T2, you can put it into any chassis you want. http://gilmore.chem.northwestern.edu/kgsshvamp.jpg first cut at power supply schematic http://gilmore.chem.northwestern.edu/kgsshvpower.pdf and bom (missing among other things the xlr jacks) but close http://gilmore.chem.northwestern.edu/kgsshvbom.xlsx The power supply costs more than the amplifier. (which is usually the case for high end stuff) You can certainly go cheaper and unregulated i you want... Board pair estimate about $50 (1 each stereo balanced amp board, and 1 power supply board)

Need the layout diagrams for any/all pots...

first cut at the schematic http://gilmore.chem.northwestern.edu/kgsshv.pdf The boards definitely will fit together in one chassis http://gilmore.chem.northwestern.edu/kgsshvamp.jpg http://gilmore.chem.northwestern.edu/kgsshvps.jpg depending on the thickness of the transformer which will mount to the back panel i'm pretty sure that if you do it as one box, it sends up 14 across by 11 deep with 1 inch between the 2 boards. I might have to make it a bit longer to accomidate the transformer.

Yes you can certainly change the part values and add the current sources as outboard devices on justin's board. In fact that is exactly what my prototype is. Power supply is obviously different. The pot i was talking about was the input pot. Something that is available. Like the rk25, or some such, because evidently the rk50 is soon to be discontinued. Starting on the schematic now.

The heatsinks are the rest of the heatsinks for the T2's. All gone. Because this thing is going to be less than 12 inches, standard heatsinks should be available, and without all the extra machining...

2 boxes it is http://gilmore.chem.northwestern.edu/kgsshvamp.jpg board is 11.0 x 6.975 inches designed to fit in a box 11.2 x 7.0 inches inside with .25 thick front and rear panels. i have the dact in there as the pot because the rk50 would cost more than the rest of it combined. power supply hopefully the same exact size. What other pot should i add... I have actually cleaned up the house a serious amount... All i have left is 30 of the heatsinks. Soon i will be taking those to work. All the T2 chassis parts will be packaged and leave from work. So no more 600+ lbs of aluminum stashed everywhere. You can swap out the parts on your KGSS to turn it into the high voltage version. You can leave the resistors at the top of the 3rd stage, or build the current sources on addin boards... The original kgss design is TEN years old... maybe a bit more.

No way CJ preview. No box for the amp. NO BOX FOR THE POWER SUPPLY. 1 killovolt flying around with enough energy storage to kill. Sort of like the BHSE prototype i have, only much more dangerous. I'm not going to even supply pictures because i do not want to encourage this kind of behaviour. The company supplying insurance for the meet would probably go non-linear. Overengineered? I would never do such a thing. Headphone amps are supposed to weigh at least 50 lbs right?

So krell style matched pair with heatsink on the left side for the amp, and on the right side for the power supply ??

So for whatever reason a number of different projects are coming together at the same time. This is the latest. http://gilmore.chem.northwestern.edu/kgsshv.jpg Its the high voltage version of the kgss with higher voltage pnp transistors, a cascode current source for the 3rd stage, the lsk389 part (or singles) and much higher voltage output transistors. The result is something that runs on +/-500 volt power supplies, and is as close as i know how to do to the T2. (except fully solid state, and easy to build) The transistors that go on the heatsinks are already insulated, which makes that part easy. Not a lot more money than an exstata, and much better sound. So do i make it a single box, Do i make it a two box affair with the power supply in a seperate box... What are people interested in??

sanyo has announced that the 2sc3675 will be end of life'd in december of 2010. So if you want the exact same sound buy your parts now. Although there should still be tens of thousands of parts in stock from what i'm told. KSC5042 is also a replacement, but has been end of life'd too. A Replacement is a 2sc4686A part from toshiba which is still in active production and is a higher voltage part i'm using in the new KGSSHV. (+/-500 volt power supplies, same as T2)

Those are pretty miserable connectors, i have a few. But they are better than nothing. Soon i will have some teflon units made up. The problem is that the pins are smaller in diameter than the stax pins, so that plug mates with the amphenol connector perfectly, but not real stax connectors.

The Xa5400es works really well with the diy T2, so that is what i'll be bringing to CJ. Otherwise i would bring the krell, but the sony is a bit more sturdy and a lot cheaper to replace if it gets trashed.

The question of whether or not you are a federal employee is probably critical to this. If you are a contractor that works for a company that supplies product to the federal goverment under contract, are you bound by the same rules as if you worked for say Sandia or Fermi labs. I know for a fact that a couple of people have side businesses and work for Fermi at the same time, whether or not they are breaking the law is something else entirely. For my T2 chassis, i pay the machine shop the same exact rate that any other external customer would. Many but not all of the outside customers of the machine shop are other universities, but definitely not all of them. The biggest outside customer is definitely not associated with any university or college. Clearly the stuff inside the machine shop was paid for in part, and probably in full by the federal goverment. Or suppied directly by the federal goverment. Evidently it is clearly legal, as i get an official university invoice. This is the kind of stuff that the university would never make a mistake on, as serious amounts of grant money would hinge in the balance. In any case, i'm definitely NOT making a profit on anything i do.

The circuit design software is a seriously hacked version of the public domain SPICE fortran code compiled for multi-processor linux boxes. I'm the one that did the hacking. And i continue to hack on it as necessary. I could clearly take it home and run it on my own linux farm if i wanted to. 100% legal. Anyone anywhere could download the source code and do what i did. The windows interface that feeds it was written by me.

It is nice that the only components that get destroyed during the prototyping phase are virtual ones. I write the simulations at night, or on the weekends. No work time goes into any of this. I set the priorities low enough so that if there is something else to do the computer stack does that first. One of the many benefits of my occupation. And it is all absolutely legal. Ray's engineer does the same thing with circuit board layout software that was paid for on a goverment contract. Mikhail used the skunkworks to machine up his chassis at prices significantly less than what i have to pay to get the same kind of thing done. And WAY less than justin pays to get his stuff done. This kind of cheating is where it is at. If i was actually doing this kind of thing for a living, things would be a lot different.

If pete is doing it for $50 an hour, he is still way to cheap. He, like craig are doing it for different reasons including building stuff in ways that no one else would do. As far as prototypes, i certainly cheat using stacks of computers to do 99% of the work for me. I know how it is going to work and what it is going to sound like, long before i ever build one. Finally i now know of a singlepower unit with leaking black gate output caps. Had to happen sooner or later, especially when you push them over their rated specifications. Luckily the headphones were not damaged. Another unit i worked on a while ago, has dying electrolytics in the filament circuit. Another issue i knew was going to happen sooner or later. The singlepower chassis were actually kind of nice till mikhail decided to go the massive modifications route. Its all the extra add on's that turned the things into something that could not be repaired. Plus the fact that very few really wanted the ppx/mpx things compared to the initial run of chassis he made. Same sort of thing with the fully machined monster chassis put together with 6-32 screws that snap. Todd's prices are fair for what you are getting. Ask Frank how much his stuff would sell for if he actually wanted to sell any of it.

neutrik makes a stock part for about $6 each.