KGSSHV Carbon Build Thread

[sorenb]

BTW there are (at least) two versions of version 1.8 ...

One has the CPC wrongly hooked up, but can easily be fixed: pics are of a skinny 

[Superberg]

Thanks Soren,

It looks like I am lucky enough and have the version 1.8 that is messed up!

I had a hard time finding the "correct" cicruit diagram and gerber files for my boards. In additiona to that, I didn't check the layout properly and I was getting confused because of the resistor values since the output current source in my case also uses 600R insted of 1k. I thought that 600R was for the CPC! Only once you told me to look at the CPC I noticed the layout issue!

I removed the CPC and now it works. I am getting 409 V at the output now. I have to tidy up the board a bit cause I burned the 3k resistor in the process and don't have a replacement.

Thanks a lot again, I should have posted earlier! At least now I understand how the whole PSU works ?

Martin

[MartinC700]

Hi All,

Build is near completion all seems to be working. One question regarding offset adjustment. After setting the offset to around 17v, after I engage the opto servo, what should the offset read? It reads around 15v, is that normal? I was expecting it to null but that's probably me not understanding the circuitry fully.

Regards

 

OK, stupid me! - Read through the thread and found that the OP27 needed to be installed for the Opto Servo to engage, all working now. Time for adding the attenuator  and input selector and cleaning up the wiring.

[MartinC700]

 

Firstly many thanks to all the contributors to this forum whos posts and direct assistance helped immensely with my amp build, there were a few issues on the way, but the amp is functioning perfectly and is a huge improvement over the STM-727II it replaces.

There are many far more seasoned builders than I, but hopefully by mentioning a couple of issues I encountered it may help any future builders.

1) Check and double check your resistor values - I missed one and it resulted in quite a few replacement parts being needed.

2) If you are going to check voltages using your probe on the top of the screw terminals, tighten the screw down first (changed out the original 7815 regulator thinking it was faulty as a result)

3) You need the OP27 opamp installed even if you are using the opto servo to control offset.

4) The posted instructions mention adjusting the left channel balance and offset - you also need to perform the same operation on the right channel

5) I bought an inexpensive set of high voltage 12kv linesmans gloves from Amazon. Probably overkill but for the initial power up of the boards gave me more confidence in working with HV supplies.

6) Mentioned in another thread but Arrow Electronics offer most parts that Mouser have at lower cost and with free overnight shipping

7) If you need circuit boards, Easy EDA is pretty good at circuit layout and integrates with board manufacturer JLCPCB who are really inexpensive (read $2 for 5 small boards)

Differences from what I would consider as standard build included:

Replaced the 7815 and 7915 with AMB Labs Sigma 78 and Sigma 79 drop in circuits - soldering on the miniscule SM components is a challenge but aside from that they work really well.

Used a ladder attenuator from Acoustic Dimensions, made a custom PCB to connect to it (I have 3 spares if anyone needs one).

Installed input switching between single end and balanced, with also a feed through setting where the KGSSHV is totally bypassed. The SE input setting also connects the SE ground to the amps -ve balanced input.

Picture is the KGSSHV Carbon with it's sibling a TPA Buffalo ESS9035 DAC built last year.

 

Cheers

 

Martin

[Blueman2]

Congratulations, @MartinC700!!!  Really nice build.  And thanks for given back to the community with the learnings.  Enjoy!

[johnwmclean]

Well done Martin it looks a beauty!

[MartinC700]

Thanks guys. I was a little too quick in thinking everything was working perfectly . SE input is fine, but when I hooked up via balanced, there is a noticeable distortion on one channel. It almost sounds like overloading the input, but is there regardless of the volume setting. I checked the source and that's fine when using balanced into the STAX amp. I assume that the issue must relate to the input circuity on the affected channel. Disassembling the map to get to the boards will be fun. Not too familiar with balanced circuitry so it will be a learning experience.

[MartinC700]

Quick Update: Fixed the issue. Not 100% sure what the issue was, tightened up all connections and changed out the XLR leads, and working fine now. Internal shot of the amp included if anyone is interested. Internal signal wiring is Teflon insulated silver (bought a bunch years ago at a surplus store). Custom PCB for the attenuator and 3D printed brackets for the switching and attenuator. 

Cheers

 

[s_r]

Don't think I've seen a failure mode like this, but while testing this carbon's amp boards the 5.1k resistors let out the magic smoke. I remember spritzer mentioning that a steel screw should be fine for the crees. Could a short from the screw be possible though? As far as I can tell nothing else blew up. I did spot a stupid mistake which was not soldering the base & emitter legs of the STN9360 nearest to the 2SK170s on one of the boards. I don't think that would cause the same failure in both boards though.

Edit: I realized I had the HV supply hooked into the outputs of the board. Talk about stupid mistake. Though I guess that explains why the output resistors were the only components to cook.

Edited May 25, 2019 by s_r

[Hirsch2k]

Hi all,

 

I would love to start building the KGSSHV Carbon. I am not an complete beginner but certainly have never tried a project in this scope. After reading for a few days in the threads I have a few questions and hope I can ask them here.

 

I have the boards at home already and would first order the transformer. I thought about audiohonics.fr, does anyone have experiences with their custom transformers?

I would use the GR HV supply with the LV regulator on board. On the first page I found:

Quote

• 2× 360VAC @350mA
• 16-0-16VCT (anywhere from 15-0-15VCT to 18-0-18VCT should work fine) @500-700mA (the higher amperage is way overkill)

Would this this be the transformer to use with aforementioned supply?

Futhermore, the BOM on the first page was updated 2015 I think. Is it still usable or are there newer versions?

Additionaly I think I will need at least one Stax socket, does anyone know where to source them?

 

Thanks a lot in advance!

[shaihulud]

As for the Stax socket, I bought some from this seller and they seem to be well built.

https://www.ebay.com/itm/Stax-socket-5-pin-Pro-Bias-WHITE-Teflon-custom-made/264309639005

 

[gepardcv]

That transformer looks like overkill to me. Maybe it's for a GRHV cranked to 450V? I think most people build at 400V. I ordered mine from Toroidy, with the following specs:

Primary windings:
2x 115V (60cm leads)

Secondary windings:
2x 335 VAC at 200mA (60cm leads)
16-0-16 VCT at 500mA (60cm leads)

Works well.

You should review the BOM carefully and check it against the boards to make sure you know what you're buying. You may need to replace some outdated components, but this is easy, and will help you make sure you understand exactly what you need.

As for Stax sockets, I think most of us buy them from HeadAmp. You can use the eBay store: one, two, or five pieces.

Happy building!

[s_r]

Been busy listening to this, just replacing the output resistors was enough. I suppose lesson learned is that as soon as you stop looking for rookie mistakes is exactly when they happen.

[Hirsch2k]

@gepardcv Thanks a lot!

Yes, that would be the specs for the 450V supply, but as you said, 400V should be sufficient.

Here is my Mouser BOM for the GR HV: https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=eefe01ae3c

Replaced 1N4007 with N4007RLG and 3N254-E4/51 with KBP201G as they are EOL.

Replaced MF1/2CC2002F with CMF6520K000BEBF at R7.

Was wondering if I can replace RN60D3903FB14 with YR1B392KCC 392K .1% since it is only rated for 250V?

Thanks a lot!

 

Edit:

13 hours ago, s_r said:

Been busy listening to this, just replacing the output resistors was enough. I suppose lesson learned is that as soon as you stop looking for rookie mistakes is exactly when they happen.

Looks super clean, which potentiometer did you use?

[gepardcv]

8 hours ago, Hirsch2k said:

Was wondering if I can replace RN60D3903FB14 with YR1B392KCC 392K .1% since it is only rated for 250V?

I'm not sure how much voltage that part of the circuit sees. In any case, I had better luck ordering a bunch of extra RN60D resistors and hand-matching them to <0.1% than relying on the manufacturers' matching process. YMMV, of course.

[s_r]

9 hours ago, Hirsch2k said:

Looks super clean, which potentiometer did you use?

It's a TKD 4CP-601 (50kOhm).

[Hirsch2k]

15 hours ago, s_r said:

It's a TKD 4CP-601 (50kOhm).

I heard it is a reasonably priced replacement for the ALPS. Does anyone who lives in Europe have a spare PCB for it laying around ?

15 hours ago, gepardcv said:

I'm not sure how much voltage that part of the circuit sees. In any case, I had better luck ordering a bunch of extra RN60D resistors and hand-matching them to <0.1% than relying on the manufacturers' matching process. YMMV, of course.

Thanks, I will think about that. Ordered the toroidal from audiophoncics yesterday since it was quite a bit cheaper than toroidy but they have delivery times of about 4 weeks ?

[gepardcv]

On 6/8/2019 at 2:32 AM, Hirsch2k said:

Does anyone who lives in Europe have a spare PCB for it laying around ?

Spritzer probably does, though the Mjolnir Audio DIY page doesn’t seem to list them any more. You might want to ask him. I found his XLR input and Stax jack PCBs quite useful also.

[Pars]

As I may have mentioned, I am going to be building a Carbon for another user, and he would like to do a universal PSU that could work for the Carbon, a GG, and I don't know what else.

These are the GRHV boards he sent me and I'm trying to make sense of them and what I would need for the UPSU. I will also be using a GRLV for the LV.

Also, for heaters, I presume the F+/F- are for that? 12V heaters I presume?

1. 
2. 
3. 

Thanks for any enlightenment!

PS: with the complexity and all the varieties of these (Fat? Skinny? Dual?), there really should be a separate thread for the GRHV and all of its variants. I recall reading this thread (and wherever else they are discussed) and thinking to myself "God, I'm glad I'm not wading into that cluster...". Famous last words...

 

[mwl168]

The first and third boards in your pics are what commonly referred to as the GRHV dual boards. First board contains B- and 7815/7915 LV regulation. Third board contains B+ and bias (+580VDC) supply. Both boards have the built-in HV delay circuit. For Carbon, GG and BH they are all you need but you can replace the onboard LV regulator with the GRLV. Most builders seem to be using ac filament supplies (myself included).

The second board looks to me like what Kevin has laid out for a different purpose. IIRC, it contains a B+ supply and regulated filament supplies. It’s what I plan to use for JoaMat’s mini T2 build for front end tubes plate B+. You can use it for the same purpose on the Megatron but many simply use a drop resistor from the B+. 

The other GRHV boards are basically different dimension boards Kevin laid out to accommodate different chassis layout options. Other than that some earlier version boards do not have the built-in HV delay circuit, they are all the same identical circuit. 

[jose]

3 hours ago, mwl168 said:

... For Carbon, GG and BH they are all you need but you can replace the onboard LV regulator with the GRLV. Most builders seem to be using ac filament supplies (myself included).

 

I also used a GRHV "standard" without delay and AC for the heating filaments. I decided to use 450Vdc for all since at the time Kevin commented that the high voltage was better for the BH / GG.

You can replace the part of the LM7X15 with a GRLV but I do not know to what extent you will notice differences. IMO, the most important you have to keep in mind is the voltage with which you want to feed those amplifiers 400 or 450Vdc and the consumption to modify the value of the resistance of 5.1R

7 hours ago, Pars said:

...I recall reading this thread (and wherever else they are discussed) and thinking to myself "God, I'm glad I'm not wading into that cluster...". Famous last words...

That always happens to me.

[mwl168]

14 hours ago, mwl168 said:

... Both boards have the built-in HV delay circuit...

I need to correct myself. The above statement should be "Both boards have the built-in HV soft-start circuit...". 

The soft-start circuit is intended for HV delay but the "delay" part needs to be implemented separately off board. There are many such delay modules available on eBay or you can simply wire a manual switch like I did. 

[jose]

I use an arduino controlling an SS relay on the primary  of HV transformer. You can also use a circuit similar to T2 (with a LM555). It's very easy to do.

[Pars]

Thanks guys!

Michael: So for a complete supply (universal), I would need to use both the 1st and 3rd boards, plus a GRLV if desired instead of the 7815/7915 onboard? This would give me B+ (+400V), B- (-400V), Bias (+580V) and the +/-15V? For AC heaters, do you just take the AC volts right off the transformer windings? Or do most people offset this by a certain voltage (can't remember what that is called, but they have it riding on a +/- DC voltage IIRC).

And does anyone have an actual full schematic for these?

goldenreferencehvsic-2.pdf

Edited June 12, 2019 by Pars

[Pars]

double post 

Edited June 12, 2019 by Pars
double post