i'm on a roll... the kgsshv

[Kerry]

Nice fit.  Well done!

[eggil]

That is a beautiful build. Congratulations!

[spritzer]

I like that one a lot.  Something about the difficult ones... 

[kevin gilmore]

very nice, like the color

[headinclouds]

Yes thank you guys, it was a bit of a challenge!  Were a few moments before it all came together when I thought it might not be a good idea.

My main worry was that it would have noise and hum issues from the proximity of the transformer and inputs but possibly due to the balanced inputs no problems, it does not sound any different to previous builds.

 

Thanks Kevin I wasn't sure about the gold anodising but it turned out pretty.  My best offer in homage to your circuit.

[dziemian]

Hi Spitzer, Thank you for your post. Could you tell me how I can get these gerber files for on-board version to make my own boards? 

[JoaMat]

A very beautiful amplifier. Many congratulations.

[GeorgeP]

Yes thank you guys, it was a bit of a challenge!  Were a few moments before it all came together when I thought it might not be a good idea.

My main worry was that it would have noise and hum issues from the proximity of the transformer and inputs but possibly due to the balanced inputs no problems, it does not sound any different to previous builds.

 

Thanks Kevin I wasn't sure about the gold anodising but it turned out pretty.  My best offer in homage to your circuit.

Really cool build!

 

But how are the transistors connected to the heatsinks - did you use a bracket?

[headinclouds]

Thank you very much JoaMat,

 

GeorgeP,    They are screwed to the rather OTT 3/4" beams which I tap M3.  paradoxper posted pictures elsewhere of this arrangement.  The fourth picture here is looking from underneath and shows the beams and amps with servos.

 

I would add that I went down this route because I wanted to be able to get at the screws from above and also the beam carries standoffs to hold the top panel on.

Edited May 18, 2014 by headinclouds

[justin]

lol...i have a fedex 18 wheeler on the way here and i couldnt figure out what they were delivering. It's a pallet of Aavid PCB heatsinks i got at a good price from the netherlands (45 cents each) and paid a total of $23.99 shipping. still, 1400 of these needs a pallet and international air freight??

[spritzer]

100 of them take up a bit of space so I imagine 1400 could take up a pallet... 

[justin]

they just came. these are actually thinner 1/2" wide ones. 50lb of heatsinks on a pallet from the Netherlands. they must have union workers at arrow europe

[ervstil]

Hi!

 

I'm new to the forum, but not new to building amplifiers. A couple months ago I borrowed a set of SR-5 with adapter, hooked it up to my Pass F5 and ...wow. After that my dynamics felt so dull and distant. I immediately realized that i need to get some good electrostatics and build my self good amp. Since then i have been lurking around this forum, learning as much as possible about the different designs, especially Kevins. By the way, thank you all for sharing! Before i start to etching boards and winding transformers, i have a couple of questions.

 

- I have never built a tube amp, mostly because in normal amps solid state make more sense, but in the world of high voltage and high impedance the transformer disappears and it suddenly tubes make all the sense in the world. I'm very tempted to try the tube designs but: the BHSE design is not available, the standard Blue Hawaii is very hard to source parts for. The KGST looks very tempting, but I suspect that I will get more proven design and higher performance for by labor and bucks with the KGSSHV? So, can KGST compete with KGSSHV?

 

 

- I actually started to lay out a point to point schema for the KGSSHV, but the part count and the high voltage makes it rather bulky and somewhat dangerous...  Is there someone that can share a reasonable stable, compact and up to date layout of amp and PS of KGSSHV and KGST for etching? It would save a ton of time and money of trail and error.

 

- Regarding voltages for the KGSSHV, what do you recommend for performance and parts availability?

 

Thanks again // Christian Stille

[spritzer]

No problem supplying you with a stable board design Christian.  +/-400V means you can use current made parts and BJT's for the output devices, go any higher and you need to find out of production parts. 

 

The KGST vs. KGSSHV will be clear soon enough when I get some boards and can get building. 

 

they just came. these are actually thinner 1/2" wide ones. 50lb of heatsinks on a pallet from the Netherlands. they must have union workers at arrow europe

 

So these then?

 

http://www.aavid.com/products/standard/531102b02500g

 

It's Europe, everybody is in a union.  Though they not like in the US though...

Edited May 20, 2014 by spritzer

[ervstil]

Thanks spritzer!

 

I was under impression that the was up to date bom:s for 450v and 500v versions. I found one on an external site "D.I:Y Power", maybe i would have look up the validity of the substitute pars in that bom. Is there a big sonic difference between different versions?

 

I guess with the 400 version i can reuse the Power supply if (when) I get the urge to build myself an KGST.

 

Best Regards Christian

[Kerry]

I've got both the KGSShv (+/-450V) and the KGST (+/-400V) up and running now.  I haven't done any direct comparisons yet, but I will say that I could not stop listening to the KGST for the past week.  It has really drawn me in.

 

I don't think you could wrong with either.

[DefQon]

Ok  with some spare time on my hands I'm going to try best and complete my off board rev.04 (with rev.06 hv psu board).

 

Going to be using 2sa1968 for the ccs (so 450v build) and from what I've gathered with the older rev 0.4 offboard is:

 

The 2 50k resistors pads under the 2sa1486 is just one 50k crossed over to each side of the 50k pads, 2 of the silkscreened 50k's bridged across is a mistake?

 

2sa1968 will need the 2 175k resistors and the 2 500ohm resistors + led and

180ohms on top for 1968 whereas the IXYS fets will be using 250ohm instead, along 1k and 1.2k resistors instead of the 1968's 175k, 500ohm and led.

 

Will most likely be using other possible EOL parts for the fets:

 

2sc4686a, 2sa1486, 2sc2705, sc2240, sa970 and c1815. 

 

I can choose to either use 1 metal case lsk389 or 2 2sk170's as seen on the silkscreen of the board?

 

I can mix old value parts with modern day replacements with no problems (except the ccs).

 

Rev 0.6 offboards have the local and global feedback pads, mine does not, what should I do?

 

Any other things I should take note of?

 

Also with the older "bigger" hv board 4x 680uf/450v psu caps I have some 800 and 1000uf 450v caps lying around, is ok to use the slightly higher rated psu caps with a bit more capacitance? Same for the 100uf caps, I've got 150uf with the same voltage rating.

 

Haven't touched the PSU board for a very long time is there anything that I should take note of with the 0.6 version hv boards (possible mistakes etc)? 

 

Will have the PSU caps raised off the PCB about 2mm's as I read one owner had there traces shorted due to the close range.

Edited May 21, 2014 by DefQon

[spritzer]

On the 0.4 offboard amp boards the feedback is flipped so either cross the 50K resistors or the 1N914 diodes.  I crossed the diodes but it doesn't really matter... 

 

Correct on the parts marked with a A but not the 180R being changed for 250R.  That's for when you use IXYS parts as output devices plus some other changes.  This resistor sets the current through the output stage so you can play with it a bit.  Wouldn't really go lower than 120R though... 

 

1K and 1K2 are for the IXYS CCS.  You can't mix and match the old Japanese transistors and the new On semi stuff, different pinout. 

 

Use the global feedback as the amp was always meant to be used. 

 

Larger caps can be used but they also mean far more bang is something goes wrong. 

 

The 0.6 psu board works just fine but keep the caps off the board and made sure that all solder joints are very clean and precise.  Overhang even close to the ground plane can end very badly. 

 

Thanks spritzer!

 

I was under impression that the was up to date bom:s for 450v and 500v versions. I found one on an external site "D.I:Y Power", maybe i would have look up the validity of the substitute pars in that bom. Is there a big sonic difference between different versions?

 

I guess with the 400 version i can reuse the Power supply if (when) I get the urge to build myself an KGST.

 

Best Regards Christian

 

Kevin and I have never done any BOM's so I have no idea what's in them.  That said, any major sonic difference will come from changing the output devices.  The rest will be more or less the same assuming the parts are all of equal spec. 

[Pororo]

Guys, I need some help with troubleshooting.

 

I've got the compact PS and off-board Rev 0.6 amp boards.  A1968, 450V.

 

The PS looks good.  I get: +15.1V/-14.9V  +455V/-455V, bias is 585V.  Without the amp boards wired I'm seeing 475VAC and 18.5VAC from the toroid.

 

After I wired it up and turned it on, only 1 LED lit up, the one marked with a yellow arrow in the photo below.  The 3.15A slow blo fuse did not blow.

In less than 5 seconds the big heatsinks on the PS board got VERY hot and I had to turn it off.

 

There was no arcing or smoke in the 5 seconds that it was on.  I see no burnt resistors or bulging caps.

 

I disconnected the wires and turned on just the PS and everything was fine, heatsinks were cool.

 

I'm using all of the new Fets: STN9360, MPSW42, MPSW06, MPSW56.

 

I know the pinouts of these are different from the old Japanese ones.  You can see in the photos that I've crossed the legs (used 600V heatshrink).

 

Can I assume that the PS board is good?  Or is it possible that adding a load is causing something strange?

 

Any ideas?

Edited May 21, 2014 by Pororo

[kevin gilmore]

I'm sure the power supply is good, but you had better double check all the

rewired transistors. Something very wrong, and I know that board works

[DefQon]

On the 0.4 offboard amp boards the feedback is flipped so either cross the 50K resistors or the 1N914 diodes.  I crossed the diodes but it doesn't really matter... 

 

Correct on the parts marked with a A but not the 180R being changed for 250R.  That's for when you use IXYS parts as output devices plus some other changes.  This resistor sets the current through the output stage so you can play with it a bit.  Wouldn't really go lower than 120R though... 

 

1K and 1K2 are for the IXYS CCS.  You can't mix and match the old Japanese transistors and the new On semi stuff, different pinout. 

 

Use the global feedback as the amp was always meant to be used. 

 

Larger caps can be used but they also mean far more bang is something goes wrong. 

 

The 0.6 psu board works just fine but keep the caps off the board and made sure that all solder joints are very clean and precise.  Overhang even close to the ground plane can end very badly. 

 

 

Kevin and I have never done any BOM's so I have no idea what's in them.  That said, any major sonic difference will come from changing the output devices.  The rest will be more or less the same assuming the parts are all of equal spec. 

 

Thanks for that as I expected.

 

I've started sourcing the fet parts i.e the old Japanese Toshiba parts and for some of the fets such as the 1486 and 2705 there are BL and GR versions, do the different part codes matter for the fets?

 

I've learnt from my old SRM-1 amp that I repaired and asked for help here in the Stax thread that one of the 2SK160's were faulty and I replaced it with a GR versions which did not produce sound properly due to small idss range until I swapped out the 2sk part for a Y version with a higher idss rating and this fixed the problem so I'm making sure I make the right purchase on the fets on the first go.

 

Also the BOM I think pororo is referring to is the one on Lil Knights' website.

[kevin gilmore]

only the input transistors are jfets. Everything else is bipolar unless you use the

ixys parts which are depletion mode mosfets.

[DefQon]

only the input transistors are jfets.

 

So I assume this refers only to the 2 x 2SK170's/LS389 required on the board and the GR/BL variations with the other stuff doesn't matter?

 

So if I'm correct the idss drain current values only matter with the jfets and this explains to the GR/BL variations (for different drain current rating?)

Edited May 22, 2014 by DefQon

[Pororo]

I'm sure the power supply is good, but you had better double check all the

rewired transistors. Something very wrong, and I know that board works

 

I checked the datasheets for the transistors and this is what I've come up with for the original part and the sub.

Reading L to R with the front (flat side) facing you:

2SA1486 ECB     STN9360 BCE

2SC2705 BCE     MPSW42 EBC

2SC2240 ECB     MPSW06 EBC

2SA970   ECB     MPSW56 EBC

2SC1815 BCE     MPSW06 EBC

 

BUT, I've noticed that the silk screening on the board shows both 2SC1815 and 2SC2705 as ECB and this was the reference I used to originally rewire.

 

I removed the angle bracket and heatsink and pulled the 3 transistors and changed the pin orientation using the BCE reference.

 

Put the whole thing back together, connected the wires to the PS board and fired it up.

 

I had the exact same problem with the big PS heatsinks shooting up to high temps in under 5 seconds.  Immediately turned it off.

 

Is it possible that the 3 originally mis-wired(?) transistors were damaged the first time it was turned on and putting them in the right(?) orientation was pointless?  Would damaged transistors in the right orientation cause exactly the same problem though?

 

I double-checked the pin-outs above and they look right to me, but I'm not 100% sure that the datasheets are accurate.  Anyone see an error there?

[DefQon]

Maybe a short or cold solder joint somewhere? I had something similar happen with another amp that was based on opamp outputs, powering it on was messed up sound and a hot LM317 which ended up being the problem (and a short opamp) because I soldered it the other way around.