KGSSHV Carbon Build Thread

[hhobeika]

Turned out the transformer is big because I requested encapsulation.  SumR uses few encapsulation container sizes.  The transformer diameter with same specs goes down in diameter from 135 mm for the encapsulated version to 115 mm for the non-encapsulated version.  The height goes down from 73 mm to 54 mm.  This size is more reasonable.  The transformer is still shielded but not encapsulated.  I am wondering if encapsulating the transformer is of any usefulness.

[mwl168]

@hhobeika

The secondaries spec should be fine. You can dial back the current on the 15vac, 100mA ought to be enough for the Carbon front end which draws less than 20mA IIRC.  B+/ B- draws around 80 mA both channel combined. 250mA should provide ample headroom.

Where are you located? If you are in USA, you can check out Antek transformers which are well built, reasonably priced and sized.

 

 

[hhobeika]

Thank you Rodeodave and mwl168 for the info,

I am located in Canada and just checked Antek.  It's good to have options.  In my case, going this route though will require a separate transformer for the low voltage from what I could see on their website.  This combined with shipping will get close to the price of a custom designed single transformer from SumR.

I will drop the encapsulation option to reduce the size and will check with SumR if dialing back the current on the 15VAC will result in further reduction in size.

[Blueman2]

I will also put in a recommendation for Antec.  Here is what I used for my Carbon:

 

AN-0215 - 25VA 15V Transformer  $11.00 USD

AS-2T300 - 200VA 300V Transformer  $38.50 USD

[muskyhuntr]

A couple of years back, I had a need for a pair of transformers that had to be built to fit in a 1U chassis.

SumR was able to modify the transformer and  the encapsulation to fit my needs.  Transformer was also shielded. 

Kinda what he does, custom trannies.

Me

[Pars]

Getting ready to order the PSU parts and transformers for the Carbon I'm building. These are the transformers I was thinking about:

AS-0515 - 50VA 15V Transformer

AS-3T325 - 300VA 325V Transformer

Let me know if these are appropriate or not.

I know the 50VA 15V is big for the requirements, but it is shielded whereas the 25VA are not. It will be going in a separate PSU enclosure, so probably doesn't make much difference, but I prefer the shielded, particularly if it were to be reused for something else.

I also don't know if going with the 300VA HV transformer is massive overkill or not. It looks like some use the 200VA, but those are only $1.50 cheaper, are 350V instead of 325V, and are out of stock. The 325V should theoretically give me a raw 459Vdc to work with.

[Blueman2]

On 9/6/2020 at 10:58 AM, Pars said:

Getting ready to order the PSU parts and transformers for the Carbon I'm building. These are the transformers I was thinking about:

AS-0515 - 50VA 15V Transformer

AS-3T325 - 300VA 325V Transformer

Let me know if these are appropriate or not.

I know the 50VA 15V is big for the requirements, but it is shielded whereas the 25VA are not. It will be going in a separate PSU enclosure, so probably doesn't make much difference, but I prefer the shielded, particularly if it were to be reused for something else.

I also don't know if going with the 300VA HV transformer is massive overkill or not. It looks like some use the 200VA, but those are only $1.50 cheaper, are 350V instead of 325V, and are out of stock. The 325V should theoretically give me a raw 459Vdc to work with.

I had the same thoughts come up when I ordered mine.  I ended up going with smaller (in voltage and in VA) simply because that is what would fit in my case.  If space is not an issue, I think your choices are good.  

[waltzingbear]

Greetings

In the process of gathering parts and am down to psu and transformer choices. Blueman2, have you gotten your transformers or are they just ordered. I was wondering what the actual performance is (voltage with and without load)

TIA

Alan

[Blueman2]

Just now, waltzingbear said:

Greetings

In the process of gathering parts and am down to psu and transformer choices. Blueman2, have you gotten your transformers or are they just ordered. I was wondering what the actual performance is (voltage with and without load)

TIA

Alan

I did this build about a year ago.  I have used Antec for all 4 of my builds so far and have been very happy with them.  Antec is very conservative with their VA rating, so I would not worry so much about that.  As for the voltage, I think they spec their voltage at 115, but not sure.  It will say on their site for each transformer.  You then have to look at your normal line voltage.  For me, I typically am at 122 to 124 volts AC here in San Francisco area with PGE.  So my 300V transformers typically give me 320VAC under load.  But YMMV depending on local AC voltages.  

[waltzingbear]

Thanks

One of my concerns was deciding the voltage rating of the input caps, safety vs physical size. I'm used to having the SRM1 MK2 on my workstation and want to fit the new amp in as well as I can.

Cheers and thanks for the info. It is indeed what I was looking for.

[Pars]

GRHV question: I'm building a pair of the singles intended for 400V output. On the BOM I have, there are 2 680uf caps for each board; 1 is a 550V caP, the other a 450V cap. I presume the 550V would go right after the rectifiers in the schematic, and the 450V on the output.

The BOM that was posted by jamesmking uses 4 x 470uf 550V Kemets. Is there much difference between the 680uf and 470uf in terms of sound or anything?

BTW, I think I had mentioned this before, but the GRHV really deserves its own thread. Confusing to have to weed through the Carbon thread, and I don't know how many other threads with GRHV posts in them.

Edited September 18, 2020 by Pars

[jamesmking]

4 hours ago, Pars said:

GRHV question: I'm building a pair of the singles intended for 400V output. On the BOM I have, there are 2 680uf caps for each board; 1 is a 550V caP, the other a 450V cap. I presume the 550V would go right after the rectifiers in the schematic, and the 450V on the output.

The BOM that was posted by jamesmking uses 4 x 470uf 550V Kemets. Is there much difference between the 680uf and 470uf in terms of sound or anything?

BTW, I think I had mentioned this before, but the GRHV really deserves its own thread. Confusing to have to weed through the Carbon thread, and I don't know how many other threads with GRHV posts in them.

Pars, I entirely agree about the seperate thread for GRHV.

Almost all 680uf 550V and even 500V caps are 80mm high - too high to fit into a 2U chassis (for example mouser lists some 500V less than 80mm high but none are stocked items, for 550V 680uF ALL are 80mm) . I also like the kemets for their life expectancy (although they are not cheap). I put 550V into both positions simply because at some point I might reuse the power supply and change its output. If you are budget conscious the output caps can be 450V or 500V for 400V output.

This is one of the cases where moar does not always equal better. The kemet caps are very low ESR and so the inrush current from the GRHV quite considerable - with 470uF caps its enough to weaken a 3.15A slow blow fuse to the point where after 10-30 switchons the fuse can fail. 680uF is going to be even worse inrush... This also puts more strain on the power supply transformer. For example, on my mini T2 with GRHV supplies the high inrush chokes the HV transformer and results in it humming for quite some time, I don't get this problem with the kgsshv on my full T2 (despite the fact the full T2 has 25% more idle power consumption) because kghsshv in it has even less capacitance and so less in rush than the GRHV...

I have not done a listening comparison between 680uF and 470uF. For me the only considerations was space and reliability.

Edited September 18, 2020 by jamesmking

[Pars]

Thanks James! Yes, I had noticed the cap size; the 60mm is appealing for a 2U chassis. It will have inrush limiting; initially CR60 or 90 thermistors. Four of the 470uf are actually slightly more than the two pairs of 680uf, but the price isn't an issue.

[gepardcv]

I built my Carbon with 680uF Mundorf/FT caps from this group buy from 2015. The 3.15A fuse I put in originally has never failed, after far more than 10-30 on-off cycles. No thermistors 🤷‍♂️. No problems after years of use.

The 76mm height restriction was a bit annoying, but I made a fully-custom case slightly larger than 2U (100mm, using Fischer SK 56 100 SA heatsinks).

Edited September 19, 2020 by gepardcv

[Pars]

Thanks! I was wondering if there was a sonic difference between the 470uf and 680uf. At 550V, these are huge caps energy wise. The 60mm height and a 2U enclosure would sway me to using the 470uf, but I'll have to ask the guy I am building this for what his preference is.

Paging @mwl168, but I think he used something similar to a Modushop Slimline, with beefy sides but not heatsinks for his universal power supply case? IIRC, he had told me that the Carbon amp boards toss a bunch of heat, but the PSUs aren't that bad? Since this will be a 2 box build, with a universal tube/SS PSU, if that is the case, I still think I would go for the 470uf caps. We'll see...

[mwl168]

Hi Chris:

For my universal PSU (GRLV + GRHV), I use a Galaxy 3U chassis  which comfortably accommodates the 680uf/550V capacitors. The chassis has no true finned heatsink, it has what the vendor calls the "quasi heatsink".

I use this PSU to power my Carbon, Blue Hawaii, Grounded Grid and the CFA. I bolted the transistors of the GRHV to a L angle which is then bolted to the "quasi heatsink". Even powering the Blue Hawaii, which draws much more current than the Carbon does, the "quasi heatsink" only glows warm to the touch.

I believe even if you just bolt the GRHV transistors to the bottom aluminum plate it should be fine - this is exactly what I did with my Megatron PSU which draws even higher current I believe.   

[jamesmking]

4 hours ago, mwl168 said:

Hi Chris:

For my universal PSU (GRLV + GRHV), I use a Galaxy 3U chassis  which comfortably accommodates the 680uf/550V capacitors. The chassis has no true finned heatsink, it has what the vendor calls the "quasi heatsink".

I use this PSU to power my Carbon, Blue Hawaii, Grounded Grid and the CFA. I bolted the transistors of the GRHV to a L angle which is then bolted to the "quasi heatsink". Even powering the Blue Hawaii, which draws much more current than the Carbon does, the "quasi heatsink" only glows warm to the touch.

I believe even if you just bolt the GRHV transistors to the bottom aluminum plate it should be fine - this is exactly what I did with my Megatron PSU which draws even higher current I believe.   

I agree about the low temperatures of the GR psus. In my experience the GRHV seems to generate very little heat (compared to the T2 psu whos 3W resistors are inferno level hot), For my blue hawaii I managed to fit everything including a GRLV, GRHV, transformer and amp boards in a single modshop 400mm deep disapante. For the mini T2 I had to go separate box for the transformers but I still managed to fit the dual GRHV, separate GRHV for the 250V, GRLV boards and amp boards in a single 400mm disapante. In both cases the majority of the heat was from the amp boards and the overall GR transistor temperature levels very acceptable.

[Pars]

Good to know. Thanks guys!

[Pflugshaupt]

Hello forum. First of all thank you all a great deal for creating this forum and of course heaps of thanks to Kevin Gilmore and the other designers for creating these wonderful amps and to everybody here for sharing their knowledge for everyone to learn. I read through a lot of this thread and recently finally finished my own kgsshv carbon. It is based on chinese ebay amp PCBs, the BOM from this thread, is running on a 400V kgbh-PSU I already had and was sounding great for several weeks until I messed up something this morning and now have a very hard time troubleshooting. I might have mechanically damaged a part or maybe temporarily not connected things correctly, but now I'm facing very strange output and thought maybe someone came across this before.

My right channel is fine, but the left one suddenly produces some hum as well as weird static noises that change over time and no more music.. just some sort of occasional crackling. For some reason my B- now measures -430 V (was -407 V before) and I measure over 100V voltage drop on the 20kohm resistor connected to B- which is a lot more than on the right channel.

Now the strange thing I randomly came across: If the -15V input is disconnected, I get music and it seems to even sound somewhat ok. Does this make any sense? 

What I did before things broke was causing vibration to the board, so maybe something is loose. I also might not have -15v and 15v connected correctly initially when starting up again.. could that cause damage?

Thanks for any help, I'll go on comparing my two sides to hopefully figure it out, but any pointers would be very welcome!

 

[Pflugshaupt]

Answering myself - I figured the -430V is a PSU problem unrelated to the sound/crackeling/voltage drop after the 20K resistor.

[Pflugshaupt]

My best guess is that I accidentally destroyed one of the 2SK170 on the left board. It unfortunately got bent once and I bent it back and after that it was still working, but maybe now the vibrations have killed it. Could that explain the strange behaviour I'm seeing? (running half-ok with -15 V disconnected, but only strange static with +15 and -15). As the 2SK170 are hard to obtain, I have no spare one to try (yet). I also get a few seconds of sound when turning on or off, but as soon as voltages are stable, only hum/static/crackles and a whole lot of silence in between.

[Pflugshaupt]

I desoldered and tested my 2SK170s and they seem to be fine. In the meantime I figured out the problem with the PSU. One of the voltage limiting zeners was broken and B- was no longer limited to -400V and went to about -450V, but probably not stable? Tomorrow I'll get some replacement parts and that will hopefully fix the PSU and give me options to replace things on the left channel. If B- goes wild, what is the most likely part to break? Nothing looks slightly burned besides maybe one of the PZTA42, but that may be flux from the build. Maybe it'll start working again once the B- is correct again. That would also explain why I get sound during turning on/off as during that time the voltage drops.  On the other hand, the right channel didn't break so something has got to be off on the left.

[waltzingbear]

its entirely possible that one channel had more leeway before breaking than the other. Hence one still working. The hum probably came from a combo of the failure (whatever it is) in the L channel from the unregulated (broken) -400 supply.

I would get the LV 15V supplies working next after you get -400V regulation back.

I have no idea which transistor maybe bad, but most likely one is. Once you get the supplies working correctly you should measure the voltages on both channels, that may point to the problem.  The observation that it passes audio while turning on or off leads me to think that is happening when it passes thru working bias points. And stabilizes with everything biased off.

Check your board well for missed solder joints, cold solder or arcs.

As an aside, you can use almost any correct type of FET to trouble shoot. The 170 is for low noise and distortion, others will work just fine..

 

do you have a scope?

[Pflugshaupt]

Thanks a lot for your help. I'll report back once I got my mouser order that went accidentally to Canada instead of Belgium :o. So maybe not tomorrow :). That will most likely allow me to fix the power supply and maybe this will take care of the issue already. LV +-15V is still working fine. I do have a scope, but I don't really know what to measure to diagnose the issue.

[Pars]

OK. Getting ready to order parts for the GRHV and ran into these issues:

1. No one seems to have any C2M1000170D Cree SiCs
2. Mouser no longer stocks the Kemet 680uf/550V caps (nor does anyone else that I could find).

For 1, does anyone have a couple of the Crees that I could buy? For 2, Mouser still stocks the 470uf/550V Kemets. Should I use a pair of those on the input, and use the 680uf/450V Kemets on the output, or just use 4 of the 470uf/550V?