spritzer

I'm not sure how that happened as the previous version was just fine. I blame the software...

Fairly sure it is Panasonic....

Doesn't matter which 2705 you use. Ahh so that's what those look like when the max height of any component isn't equal to the terminal blocks...

I haven't fired one of these up yet so there might be more gremlins hiding somewhere. It was a bit of a challenge to make it this small in the first place and then add the servo...

Just so you know, there is an error on that board... The lower A970 is the wrong way around. Ohh and take care with that empty hole in the top center. Very easy to make that go all arcy sparky.

There is a huge difference between the mosfets and the BJT's with the latter being far superior. That said I have the newest onboard PCB's should anybody need some.

It looks ok but I'd go for a single 30VCT@500ma winding instead of two separate windings.

This one would work with 500V capacitors (450V might be ok but we don't work well with "might") and larger heatsinks in the PSU. http://www.antekinc.com/as-1t300-100va-300v-transformer/

Given that we do the same to meat and it's the traditional Christmas meal I can't say I'm surprised. One of the breweries also used dried whale in one of their brews which caused a lot of ruckus. In other news, I've been experimenting with making beer bread where I take the barley after they are done with it and use it in bread. Tastes pretty awesome and I've also been making sour dough out of the barley by letting it ferment a bit. One of the most challenging mixes to work with was Surtur which is also one of the most non-PC names you can find. No other way of translating that except "blackie"...

Antek has a very nice transformer at 275V (i.e. close enough) and then you just need a 30VCT transformer for the +/-15V supply

Two 280VAC windings are perfect for +/-350V.

That socket should fit but these are the most common ones and what I used/designed for: http://www.ebay.com/itm/10-9pin-Ceramic-Tube-Socket-PCB-B9A-6DJ8-12AX7-ECC83-ECC82-12AU7-6922-6N11-Valve-/131242710847?pt=Vintage_Electronics_R2&hash=item1e8eacd73f Available from pretty much everywhere. The board also takes the teflon sockets such as this one: http://www.ebay.com/itm/NEXUS-MUSIC-Premium-9-PIN-9P-Gold-Plated-Teflon-Tube-Socket-12AU7-ECC83-2PCS-/181469359862?pt=Vintage_Electronics_R2&hash=item2a406aaef6 For tubes its the 6S4A that works and nothing else. There are really no tubes like it though some are similar but with incompatible pinouts. I got some 6973's on Friday which are a beam pentode and in current production which might work on a modified ST board triode strapped. As soon as I get my new board then I'll try it. As for matching, you need matched pairs for each channel but a matched quad doesn't hurt. Matched build is also what I'd go for so all of them RCA or GE, the actual branding doesn't matter. A note about the voltage though. After running my tiny 400V for some time now I would advice people stick to +/350V rail voltages as 400V runs too hot for comfort. It's the first amp I've ever had to use a fully vented cover for...

1W is more than enough and any combination will work as long as the end result is 580V. We chose those values as the board already had a lot of 150V units. If you jumper a trace, do it on the back to prevent arcing to the ground plane.

Here goes... R17 and R55 changed to 150K/2W R7, R8, R49 and R50 changed to 300K/2W R14 and R52 changed to 350K/0.5W R35, R36, R38, R39, R73, R74, R75 and R76 all changed to 100K/0.25W Also R37 and R42 changed to 0 ohm jumpers (or just jumpered by some wire) for the RCA/XLR switch rewiring. To fix the input section in general (you can leave the pot in place) do the following. Remove the RCA wiring and connect the wires directly to the XLR jacks. The XLR pinout is 1: ground, 2: + or hot and 3: is - or cold. You can see the numbers in the back if the socket and the pins go straight through and into the PCB. Ground goes to ground and the center conductor goes to + or hot. Now next step is to cut all the traces which lead from the XLR sockets to the pot, also all traces from the pot to the amplifier. Leave the ground all alone as it works just fine as it is. When the amps is shipped then the input goes to the center row (which is the wiper) and that will never work. You want the input for each phase to go to the row on one side which isn't grounded. Then take the output from the center row and feed it to the series resistors right next to the pot. Now the tricky bit about the RCA/XLR selector is how it is wired. It takes half of the traces off the pot and feeds it through the switch which means all those traces have to be cut (except those which lead to R37 and R42). Run a jumper wire from 500K resistor to each one of the input fets to replace the long trace running on the edge of the board. For the switch you take the - or cold legs off the XLR sockets (or the input to the pot) and run them to the center of the switch. Then the RCA side is grounded but the XLR side is left empty, not connected to anything. That way the switch does nothing in XLR mode but switch to RCA and the - side of the input is grounded. Easy as pie...

Well I bought it because it looked like it used the KGSS design without permission. Kevin gets so happy when I send him something to tear apart... The amp is a clear copy of the KGSS but doesn't work out of the box unless you count 1mm of volume control travel as a sufficient adjustment range. That repair work is mild compared to the Cavalli Liquid Lightning and it didn't cost less than 1k$....

I only sell them to bats.... it's a niche market

No need to change the chassis as it works just fine. I needed to properly mount the transformer and replace the feet but that is minor. I also had to tart it up a bit... It's nice to have power knob that actually tells me whether it is on or not. The input section is so fucked it's beyond belief. So the knob is wired the wrong way, the switch doesn't work and the RCA's are actually hooked up to the - input so any attempts to fix it means you have to start scratch. So just cut a bunch of traces and jumper. Here is my board as it is now: I changed the feedback resistors to 100K and the pull down resistors for the last stage changed to 350K RN60D's as they are run on the limit. The stock RN60D resistors changed to the correct values (300K and 150K respectively) and 2W Koa SPR. I also changed the resistors which ground the RCA/XLR switch to zero ohm jumpers.

If this was built with some sense then it would be a nice little amp but there is just so much wrong with the input section. I set to work on the RCA/XLR switch and found this: That would be the earth contact to the chassis, simply soldered to the steel screw. Fell apart when I simply pulled off the back panel so it has been done properly now. Also removed the stock feet (the chassis wobbled) and found out why. Nobody had thought to clean the mounting holes so they weren't flat. Discared the stock feet and put in some nice aluminum units instead. I have to open it again to find one trace on the input that is fucking it up for me.

To further add to this, I couldn't resist temptation and tore into it. It's very Cavalli in many ways, not really meant to be dissembled but at least semi decent soldering. Also something it has in common with Cavalli: Thermal paste is overrated anyway... So I ripped out all the underrated resistors and the volume control. The volume was an even bigger clusterfuck than anticipated, it was quite simply all incorrect. Channels are reversed, RCA switch doesn't work and naturally the input goes in the center for some unknown reason. I set to work... It took some doing but the TKD 4CP-601 managed to slot in there. Underneath all the traces are cut and replaced with wires but I was too lazy to fix the RCA's and the switch. Channel reversal is easy to fix too by simply crossing the outputs. So here is what it looks like now: So the feedback is fixed, the current source resistors for the third stage are the correct value and wattage, same for the pulldown resistors for it and the fourth stage. Sounds much better now and the volume can actually be used...

There is some value in there but also a lot of problems. How that balances out is a personal thing but this one will never be a full blown KGSS. Current is too low for that and no good way of changing it, at least not in the stock chassis.

The Megahertz stuff is a bad joke...

Just got the Chinese KGSS and the first job was to properly secure the transformer. The rest will need an awful lot of work. The volume control and the entire input section is fucked. Output stage has to be restored to proper working order, all transistors have to be remounted, all standoffs replaced with metal ones, output resistors fitted and all parts being run on the limit changed to properly rated units.

I had 0007 or something like that so they certainly started at 0001.