Salas DCG3 preamp (line & headphone)

Member
Joined 2017
Paid Member
Solder-blob on 2 legs @ M1 (G+D, left + center)

Is there a chance it survived my gaffe?

Measured R, results are unconsistent across M1 - M4, so unless you recommend something else I will reconnect it and see what happens.

Since the board is off the sink, I recleaned it with IPA and have redone the keratherm-pads. I‘ll leave it for today to let it dry.
 
  • Like
Reactions: 1 user
Member
Joined 2017
Paid Member
Good morning!
I have good news: channel two seems to be working, bias is down to around 1mV (immediately after powering it on, it is moving in slow waves from 10mV down to 0mV).

[Edit: voltage across R10 (10r) is 1.34V]

Should any further tests be done?

I must go and do some dayjob, unfortunately, so won’t be able to give it an ear until tomorrow or so!
 

Attachments

  • 097CAE25-26F4-4604-9ABC-6EF5AD805566.jpeg
    097CAE25-26F4-4604-9ABC-6EF5AD805566.jpeg
    441.8 KB · Views: 105
diyAudio Chief Moderator
Joined 2002
Paid Member
By bias mV you mean output DC offset without the servo op-amp in yet I believe. Yes it looks normal. Also 134mA output stage bias current is normal (1.34V/10Ω).
The pdf instructions set up steps will suffice for tests. Does not look it has a special issue after you spotted and removed the solder blob short from the M1 Mosfet.
 
  • Like
Reactions: 1 user
Member
Joined 2003
Paid Member
Thank you, Salas and cyberpup!

The 17.02vdc is on the output (line/G, with input shorted)…

It’s the channel that has the i-select stealing power from DCSTB btw. can‘t imagine this is the problem, as i-select is working and DCG3 turns on and measures correcly 17.3 vdc on the V+/GND and V-/GND.
Power wiring was correct (I color-coded the wires + verified with multimeter)

Since I was quite carefully during the build, and verified everything on scm and dmm, I am confident that at least the valuss on the resistors are correct—so I firstly will re-verify visually…

Same for the Mosfets, I took great care to have a good, plane, burr-free surface 😇, and I used the plastic-washers. Will of course recheck for nonconductivity!

I had to leave the workshop, so will make further measurements tomorrow.

Still, maybe these descriptions point to something specific?

Thanks again!
David
Conductivity around the MOSFETs is touchy. Too tight, I got shorts when measuring. Slightly loosened, problem went away. Test first, then go. Mine measured fine, but then before fire up, "I made sure they were tight" and this shorted them, so when I fired up, problems.

Russellc
 
Member
Joined 2017
Paid Member
Ah, those are the bad moments.
I seem to have been lucky though, it seems the DCG3 is behaving as supposed.

(One channel's offset is behaving funny, though, but I can't judge wether it's fine or not: it goes up and down ~10mV in, like, 10-15 seconds-cycles, after some 15 minutes or so, I believe it calmed down a bit? I will let it run some 30' and look after that, and then will be the great moment: In goes Opamp, first listening!)
 
  • Like
Reactions: 1 user
Member
Joined 2003
Paid Member
Ah, those are the bad moments.
I seem to have been lucky though, it seems the DCG3 is behaving as supposed.

(One channel's offset is behaving funny, though, but I can't judge wether it's fine or not: it goes up and down ~10mV in, like, 10-15 seconds-cycles, after some 15 minutes or so, I believe it calmed down a bit? I will let it run some 30' and look after that, and then will be the great moment: In goes Opamp, first listening!)
If you haven't installed the 8 legged device, that may be normal, Salas can confirm...

Russellc
 
  • Like
Reactions: 1 user
diyAudio Chief Moderator
Joined 2002
Paid Member
Can be normal hovering the offset mV before applying the servo chip. When all Mosfets work correctly they will measure comparable Vgs each one vs the other ones on the always been good channel. That's the DC voltage between pins 1-3 on each Mosfet.