Congrats Emooze

 

I'm using 35v caps and I have headinclouds to thank for that. He posted earlier in the thread about the 25v caps being run at close to its limit and I did the calculation and thought I'd do 35v just in case also.

Some time ago I did the new KGSShv board with Only IXYS parts. With power on both channels showed -380 V. Same thing both boards. Today I replaced 10n100d with c4686 and then it works.

 

Have anyone built KGSShv new board with all IXYS parts and have it working?

Yes it works. 

Some time ago I did the new KGSShv board with Only IXYS parts. With power on both channels showed -380 V. Same thing both boards. Today I replaced 10n100d with c4686 and then it works.

 

Have anyone built KGSShv new board with all IXYS parts and have it working?

Did you setup the board properly?  The new " I " labeled parts around the upper right hand corner of the board are for the IXYS parts but not the CCS. 

Edited October 21, 201312 yr by spritzer

The two 1K gate resistors in place, 240R as source resistors (not  250R – took what I had). No LED and its 175Ks and no A bridge. Is there more to consider. The lower source resistance couldn’t be the problem or…?

No change would cause such a dramatic difference.  Did you manage to balance the output at all? 

/

Edited November 11, 201510 yr by JoaMat

With the c4686 parts you have a 450V limit in the supplies With the IXYS parts you can go to 500V in the supplies.

No, that just applies to the current sources.  The 2SC4686A is the BJT output device, you are thinking of the 2SA1968 which is the PNP BJT used in the CCS. 

Right. Brain fade, as usual....

i still have the 5000pcs of 2sa1486 that i got from the highly suspect company...the ones where some of the leads are trimmed so thin they're about to fall off, about 5% of them test good for a very low voltage (but look identical), and they test positive for lead despite the box being labeled lead free. No idea where these came from but i sure am not going to use them. Luckily, i got the money back from my credit card company after they tried to stick me with store credit. Recently they sent me another email asking me when I'm shipping the parts back to them...yeah right...so they can re-inject them into the world supply? I'm thinking of making a video showing their destruction but need some creative ideas. Best i have so far is to dump them in a hot bucket of lye.

Perhaps a peace offering to Cavalli from Spritzer?

Is there some way to make a super long chain where they explode domino-style if you attach power? I would enjoy a gif of this

Thats what I just said.

don't invalidate my posts with intelligence and wit

well.... I guess my endeavour with the amp will not end so simply....

I took out the PSU and I found that the solder joints for the zener strings in the -500V rail are burnt.

 

Last time I measured, the output was stable and without any problem.

Replacing them will be easy, but that's not the point....

Using multimeter, each zener measured same forwarding voltage compared to the +500V rail's zener. Any two in series also measured the same. However, when measured all three in series together, the burnt ones have a little bit lower forwarding voltage than the one that's not burnt....

Zeners themselves don't show any burns.

It's amazing the PSU still worked under these conditions....

Anyone knows what might be causing this?

The prereg voltage probably rose due to the fubar transformer which means the zeners had to burn it off.  Normally they just split open when they fail but here they just cooked a bit. 

 

In other news we've worked on the HVk PSU a bit more to make it even safer to use.  Now there is nothing under the main caps but empty FR4 material so they can sit on the PCB.  The CT connections for the HV inputs have been removed plus a bunch of small fixes and alterations.  The ground plane clearance has also been increased so this one should be trouble free. 

 

Now a question, would somebody like us to do the same to the older style of the PSU boards?  This could be useful to have in amps that suffer damage etc. so the old mounting holes could be used but the board would be safer to use. 

 

Now a question, would somebody like us to do the same to the older style of the PSU boards?  This could be useful to have in amps that suffer damage etc. so the old mounting holes could be used but the board would be safer to use. 

 

I plan to use the k boards in the future, but I would like to have an old style board (or 2) with the improvements if it can be made available so I can be ready should I have any trouble.  I'm happy to help organize an order if that helps?

Edited October 25, 201312 yr by shipsupt

Why are you going to sue my lovely k boards...?    

 

I'll postpone the KGST PCB work and see if I can't find the latest large board version somewhere... 

You're right, too much litigation in this world already! 

 

Fixed it, and thanks!

Thanks Birgir. I guess I will replace them first and see what happens after one hundred hours of operationg.  

Improved old boards sounds great. If there is a GB, I'm definately in for that.

 I'm thinking of making a video showing their destruction but need some creative ideas. Best i have so far is to dump them in a hot bucket of lye.

 

I'm betting with a little assistance from Kevin you could cook up a batch of thermite which might make for a sufficiently impressive destruction film.  Especially if done at night.

I will order prototype boards as the design is pretty much ready... 

Here is an update, both PSU boards have been worked on and we dare to call them the final versions...     Here is the updated "k" version:

 

 

Red is the top layer while blue is the bottom.  The major changes are more clearance around the entire ground plane, less chance of a short around the rectifier diodes (different footprint), no ground plane or traces on the top layer under the main PSU caps and all traces on the top layer cut down as much as is possible.  There is now the provision for a fuse for the bias supply when running it off the prereg (500V version) as the voltage doubler naturally limits current for the 450V supply.  Also added  is a bias test point before the ballast resistor to check the bias supply.  No CT connections for the HV line as nobody used them.  Also we added markings to make the setup clearer, what should be used where and all that. 

 

Here is the new version of the large board:

 

 

Same changes as above but also moved the mounting holes to decrease the chance of shorts.  The design was also updated to the same spec as the "k" boards. 

 

I'll order up some proto boards later today. 

Edited October 27, 201312 yr by spritzer

Very nice Birgir.