Jump to content

i'm on a roll... the kgsshv


kevin gilmore

Recommended Posts

finishing a project with a kintex-7 now so yeah whatever one has more than enough horsepower and compatible with the xilinx compiler.

the totaldac uses 74hc574, so yeah not so good either.

if someone with the fpga tools wants to do that part, fine by me. I definitely want a dsd to 192/24 thing inside.

 

 

Edited by kevin gilmore
Link to comment
Share on other sites

I haven't looked at KGSSHV boards but if it's anything like the other KG boards, it's used to enable/disable the servo.  You turn off the amp and disable the servo (remove the jumper) when you're making trimpot adjustments to set the offset and balance.  Once you are done, you turn off the amp and enable the servo (put the jumper in) again. 

Link to comment
Share on other sites

OfYlfSw.jpg

First of all, many thanks to Dr Kevin Gilmore and all contributors! I’m the lucky one to get KGSSHV… almost =) Well, here is the build, not my though. Long story short I bought the kit with all parts populated from someone I trust (and maybe he will help me, I’ll have to ship 30 pound unit a few thousand miles lol), PSU board was tested and amp boards weren’t. Then after dozen of refusals I finally found someone (highly qualified and reputable diyer) who agreed to do the casing and wiring, with no responsibility of how THAT may sound.

Funny thing to say, I’m the first one to listen to this unit. And it DOES sound tremendously resolution-and-soundstage-wise, needless to say. Except one thing. It’s not just me crazy, two more experienced listeners heard the same spot on. Compared to 007t  amp (as well as some other small diy amp) the bass is lacking big time. You can hear all the sub down to 20 Hz but dynamics and amplitude is missing. Barrel drums which are quiet on the recording are barely discernable (Shake off the Dust by Herlin Riley, to name just one)

 
 

big drums sound much smaller and drier with more kinda mid-bass impact. No hint of that manner not in a single owner impression!

The headphones are dyi, 3 models with different sensitivity and membrane thickness, no problem with them (they may hit hard like devil dare I say, close to HE-6 & beta22 rig). All parts populated are those recommended, though I can go into the details if needed. 20 hours warm-up.

I’ll have it measured and will get the readings is two days… But what may be the problem? Voltage swing? Volume pot (rk27 btw, original)?

Almost forgot to say, got the ground loop hum sound, pretty audible on sensitive headphone pair. The wiring is done strictly to that scheme https://docs.google.com/file/d/0B4NVeoQi85GgOTdvOGpxc1hyRFU/edit

 

SfrjYi0.jpg

yoAYkxg.jpg

vrN08uH.jpg

Link to comment
Share on other sites

I bet the hum is coming from the transformer.  That is literally the worst place possible to put it. 

As for the bass issue, what PCB is that hiding underneath the volume pot?  Not a chance the RK27 is the culprit and I can't even think what could possibly do that to the circuit.

Edit:  Also the diodes in the power supply could go boom at any moment...  They have to be insulated with this PCB layout. 

Edited by spritzer
Link to comment
Share on other sites

The hum disappears when I ground the case or rca banana, i.e. touch it. The hum increases when I touch the headphone cable. PCB underneath the volume pot is rca unbalanced to balanced converter. My dac is has rca outputs only, unfortunately.

Link to comment
Share on other sites

1 hour ago, spritzer said:

Ehhh why are you using a RCA to balanced converter?  Bypass that crap and just ground the - input of the amp. 

Grounding is simple, they all go to the PSU including the chassis. 

isn't it correct grounding? so there are I+ I- inputs of the amp and gnd

pw758hl.jpg

1 hour ago, spritzer said:

Edit:  Also the diodes in the power supply could go boom at any moment...  They have to be insulated with this PCB layout. 

thanks for your help

could you pls name the parts which should be insulated?

upd I guess I see them poor little diodes, jammed

lucky me 20 hours 'n no blow up?

k6CcZSL.jpg

I keep on digging...

all parts are the same as marked on the psu board 1.0

http://imgur.com/9wennYm - may have a look

Link to comment
Share on other sites

1 hour ago, spritzer said:

Ehhh why are you using a RCA to balanced converter?  Bypass that crap and just ground the - input of the amp. 

Grounding is simple, they all go to the PSU including the chassis. 

Well, my universe is upside down now and I'm in Borat mode so to say

So, no balanced, I remove that thing. Still using quad volume pot? A bit unsure how to +, - and gnd all that stuff.

Link to comment
Share on other sites

1 hour ago, *high said:

could you pls name the parts which should be insulated?

upd I guess I see them poor little diodes, jammed

No .......... the 8 large ones, high voltage, in groups of four with shiny metal on top in the front middle of your power supply.

that tab is HOT... look at these:

  STTH512FP
23 minutes ago, *high said:

Still using quad volume pot? A bit unsure how to +, - and gnd all that stuff.

 

 

Use half your volume pot, tie the minus inputs to ground at connectors on amp boards and drive the plus from volume pot, twisted pair or shelded cable from each section of pot to each board....easy!

 

Edited by congo5
  • Like 1
Link to comment
Share on other sites

When using RCA's you only use two of the four decks on the pot.  The output of those is taken to the + input of the respective channel while the - input of the each channel is connected directly to the ground tab next to it. 

Those diodes have to go and the board is clearly marked for the insulated units.  The person who built this board doesn't have a clue what they are doing as the tabs of the transistors is indeed live. 

Wrong heatsinks for the bias supply and I guess the other current sources as well.  Inadequate isolation as well since the tab of these sits at a very high voltage. 

  • Like 1
Link to comment
Share on other sites

I'm one of the guilty ones who didn't use isolated diodes; used a non-blessed BOM which specified non-isolated diodes.  Blew a lot of fuses before realizing the problem.  

Seeing the outlet wires go directly to on/off switch, no fuse on the 120V main line?  Should put one on in case of shorting (like non-isolated diodes).

Also, two-prong outlet plug?  Where's the ground/third prong or how do you ground the whole amp?  Is there a chassis ground?   

Link to comment
Share on other sites

I'll show you all the grounding again when I remove balanced pcb. I guess just one thing is missing here

https://docs.google.com/file/d/0B4NVeoQi85GgOTdvOGpxc1hyRFU/edit

nAyL6vEIXRe842.jpg

and therefore this build: 15V GND on the safety earth. On the other hand, that again may cause a ground loop.

So, transformer shield is hidden. 220V and 2-pin power socket, no ground plug.

Then there is

E2pWnERCB0ynkm.jpg

and

GrqdBjOIN0q1vA.jpg

So I guess my mistake is that volume pot GND is not on the star ground, as a theory says it should be on the spot with the weakest current, which is a volume pot gnd.

 

Link to comment
Share on other sites

9 minutes ago, spritzer said:

So the chassis is not earthed?  That is very, very bad idea. 

All the GND pads on the boards are connected together. 

It's grounded all right, as far as three-phase power transmission goes, but it's pretty common ungrounded / unearhed system standard. It sucks, I agree, but not that bad =)

So 15V GND and 450V GND and volume pot GND all connected together?

Link to comment
Share on other sites

I got a power amp from China and it was naturally ungrounded and it hummed no end.  I could actually feel the current flowing when I touched it.  One wire later and it was fine. 

Take one wire from each of the amp boards to the PSU and one from the pot to the PSU.  Then one wire to the chassis, nothing else.  That should fix the issue. 

Edited by spritzer
Link to comment
Share on other sites

It's time for awesome / horrific stories

Diodes are insulated, RCA single end is taken, the amp is star-grounded

So the reading of my 450V: bias TP 575V, bias 397V, R+R-  and L+ L- are balanced to 0V,

and horror time: R+ GND 136V and L- GND 122V, offset pot doesn't seem to change that

Huston, is it a big problem? 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.