i'm on a roll... the kgsshv

[kevin gilmore]

for the onboard heatsink version we could do something like this with a set of jumpers.

But you will have to re-zero after moving the jumpers, and considering they would be

at -Vcc, any kind of external switch of relay would be a very bad idea.

 

Mikhail comes to mind with his arcing high voltage switch option.

[jgazal]

Is it possible to use a high voltage power relay for such function (http://www.mouser.com/ds/2/164/ftr-j2-967.pdf)?

[DouglasQuaid]

I was wondering how people would go about re-adjusting the offsets on their offboard amps.  Trying to adjust those pots with the board turned up 90deg seems far too easy to slip the screwdriver and cause some fireworks.  I'm assuming you'd need to lay the boards and sinks on their sides before doing any adjusting.
 

Seems like a set it and forget it attitude is best for those builds.

Edited April 24, 2014 by DouglasQuaid

[s_r]

I was wondering how people would go about re-adjusting the offsets on their offboard amps.  Trying to adjust those pots with the board turned up 90deg seems far too easy to slip the screwdriver and cause some fireworks.  I'm assuming you'd need to lay the boards and sinks on their sides before doing any adjusting.

 

Seems like a set it and forget it attitude is best for those builds.

 

Very carefully... Though I really haven't needed to adjust the bias/offset on my build. Last time I checked the farthest it was out once warmed up was about 2V.

[johnwmclean]

The current is fixed, I don't see why you couldn't use trimmers?

Dial it up for winter, dial it down for summer...

[kevin gilmore]

there are alternate versions of the pots with the adjust on the other side. So you could use a PLASTIC

adjustment tool on those...

 

without the servo the output voltage changes 7 volts going from 5 to 10ma

with the servo, it should work if adjusted carefully

Edited April 24, 2014 by kevin gilmore

[spritzer]

I was wondering how people would go about re-adjusting the offsets on their offboard amps.  Trying to adjust those pots with the board turned up 90deg seems far too easy to slip the screwdriver and cause some fireworks.  I'm assuming you'd need to lay the boards and sinks on their sides before doing any adjusting.

 

Seems like a set it and forget it attitude is best for those builds.

 

First off, never use a screwdriver to adjust the pots, that's what trimpot adjusters are for.  Non conductive too... 

 

As for the pots:

 

 

New version has the balance moved so the terminal block doesn't obstruct it. 

[demented badger]

I have reworked the 450V KGSSHV power supply to provide both pro and normal bias simultaneously. For pro bias I used the circuit as shown in the posted schematic with the voltage doubler and 10M90S followed by the 580V zener string. To add normal bias I tapped the circuit where the 500V option jumper connects and added a parallel duplicate of the circuit from the 10M90S to output, substituting a 280V zener string to get the lower voltage. With a little adjusting, I was able to get everything on an 8X8 board. Does anyone see any problem with this that I am missing? Thanks.

[spritzer]

So a question for the group.  As we all know high voltage BJT's that can take the full force of the KGSSHV are going away quite quickly.  None are currently being made and I quite frankly don't see them making a comeback, certainly not as TO220 devices.  There is a new mosfet on the horizon but it's a poor substitute for the real thing.  With this in mind I used the mafia network to try and source some NOS and I've hit paydirt.  Now 2SC4686A's are all but gone but I've been offered another part which was clearly meant to be used in the same role, all specs identical plus some (like the output capacitance) are even better and it's fully isolated package too.  These are also guaranteed genuine and made in Japan. 

 

Now the catch is that I have to buy a lot of them, thousands, and the end price is unknown though I hope it comes between 2-3$ when they would reach my hands.  The 4686A is currently at 4$ from Bdent and supply in Japan is all but run out.  I would naturally sell off my stash as needed to anybody building an HV so this would be enough for a few hundred new amps.  So does this make sense to anybody? 

 

I have reworked the 450V KGSSHV power supply to provide both pro and normal bias simultaneously. For pro bias I used the circuit as shown in the posted schematic with the voltage doubler and 10M90S followed by the 580V zener string. To add normal bias I tapped the circuit where the 500V option jumper connects and added a parallel duplicate of the circuit from the 10M90S to output, substituting a 280V zener string to get the lower voltage. With a little adjusting, I was able to get everything on an 8X8 board. Does anyone see any problem with this that I am missing? Thanks.

 

I see no problem with that. 

[demented badger]

Spritzer - It looked OK to me as well,. Thanks for the reassurance.

[Hopstretch]

So a question for the group.  As we all know high voltage BJT's that can take the full force of the KGSSHV are going away quite quickly.  None are currently being made and I quite frankly don't see them making a comeback, certainly not as TO220 devices.  There is a new mosfet on the horizon but it's a poor substitute for the real thing.  With this in mind I used the mafia network to try and source some NOS and I've hit paydirt.  Now 2SC4686A's are all but gone but I've been offered another part which was clearly meant to be used in the same role, all specs identical plus some (like the output capacitance) are even better and it's fully isolated package too.  These are also guaranteed genuine and made in Japan. 

 

Now the catch is that I have to buy a lot of them, thousands, and the end price is unknown though I hope it comes between 2-3$ when they would reach my hands.  The 4686A is currently at 4$ from Bdent and supply in Japan is all but run out.  I would naturally sell off my stash as needed to anybody building an HV so this would be enough for a few hundred new amps.  So does this make sense to anybody? 

 

Hell yes. And I for one would be happy to front you cash to secure the stash against a mafia "favor" somewhere down the line. 

[nopants]

seems like a totally normal thing to monopolize the remaining supply

[luvdunhill]

Interested but would want to see curve trace, etc. comparison of the parts.

[spritzer]

It would make sense to make sure this project doesn't go the way of the Dynahi and be killed by parts being discontinued.  It should also be noted that this is the part intended to be used in the KGSSHV all along.  The 4686A was just easier to source... 

 

Hell yes. And I for one would be happy to front you cash to secure the stash against a mafia "favor" somewhere down the line. 

 

No need for that.  The audio slush fund covers this... 

[insanity]

This makes sense to me. But what about parts like A1968? Doesn't seem easy to source aswell. 

 

Since I built my KGSSHV with the 1n100D: Why did you move away from it? Too high capacitance? Too much drift?

Edited April 25, 2014 by insanity

[kevin gilmore]

2sa1968 is a 900v part. the ixys part is a 1kv part.

Edited April 25, 2014 by kevin gilmore

[spritzer]

Yeah that and I have a huge pile of 2sa1968's and they are fully insulated which makes my life easier.

 

Spritzer - It looked OK to me as well,. Thanks for the reassurance.

 

One thing I missed, the zener string should be 230V as that is the max normal bias voltage.  A KGSSHV plus a higher normal bias voltage is not a good mix.  It's ARC time!!! 

[insanity]

So if you can get this new part, will you start selling it and the A1968?

[DouglasQuaid]

 

The current is fixed, I don't see why you couldn't use trimmers?
Dial it up for winter, dial it down for summer...

 

It's been a couple years since I've touched pcb software, but it shouldn't be much work to fit some pot legs between the current setting resistor.  That being said, it would probably be a heck of a lot safer to put a base resistor and have a trimpot after it.

 

This is all assuming that the resistors at those positions can be measured while the board is populated.  If they can't, then this all moot and I'd need to use resistors and jumpers.  Either a selector jumper to choose between two resistors, or two resistors in series added together and a jumper to short one of them for higher current.  Again, the board's already pretty tight, so this might make things larger.

 

It's been a while since I've dealt with transistor designs, but I don't believe that resistor sees more than 2V of potential difference and whatever current the resistor is set to.  If that's the case, then we might be able to get away with SMD resistors at these positions, perhaps the 3rd stage current source as well, and not have a significant change in the size of the board.  Please correct me if my assumptions are incorrect with these values, as my transistor analysis is rusty.

 

 

So a question for the group.  As we all know high voltage BJT's that can take the full force of the KGSSHV are going away quite quickly.  None are currently being made and I quite frankly don't see them making a comeback, certainly not as TO220 devices.  There is a new mosfet on the horizon but it's a poor substitute for the real thing.  With this in mind I used the mafia network to try and source some NOS and I've hit paydirt.  Now 2SC4686A's are all but gone but I've been offered another part which was clearly meant to be used in the same role, all specs identical plus some (like the output capacitance) are even better and it's fully isolated package too.  These are also guaranteed genuine and made in Japan. 

 

Now the catch is that I have to buy a lot of them, thousands, and the end price is unknown though I hope it comes between 2-3$ when they would reach my hands.  The 4686A is currently at 4$ from Bdent and supply in Japan is all but run out.  I would naturally sell off my stash as needed to anybody building an HV so this would be enough for a few hundred new amps.  So does this make sense to anybody?  

 

 

I see no problem with that. 

 

1. Are these alternate parts pin-compatible with the 4686A?

2. I'll be down for a few sets, especially if they end up being used in another project.

Edited April 25, 2014 by DouglasQuaid

[spritzer]

All of these BJT's have the same pinout. 

 

So if you can get this new part, will you start selling it and the A1968?

 

Yes on the new part but I'll never sell any of my 2SA1968's. 

[eggil]

I am in for may be 3 dozens of the new parts.

Thanks Birgir

[jose]

the "Spanish team" is interested  

 

Thanks Birgir

[insanity]

That is unfortunate. Will future KGSSHV board therefore feature an (isolated?) alternative for the A1968?

All of these BJT's have the same pinout. 

 

 

Yes on the new part but I'll never sell any of my 2SA1968's. 

[nopants]

Is there a problem with it being nonisolated? The board already has pads for isolated or otherwise

[kevin gilmore]

That is unfortunate. Will future KGSSHV board therefore feature an (isolated?) alternative for the A1968?

on board version already has this and had it for quite some time.

some off board versions already have this