Build Thread - A New Take on the Classic Pass Labs D1 with an ESS Dac

opc

Member
Joined 2004
Paid Member
That'll do it!

Insulate those guys and you should be good to go. Let me know if you get it powered up.

I would still say that you ought to take the DAC off the I/V when you power it up. Even if it's not powered up, you could still damage it if the voltage is too high on those nodes.

Cheers,
Owen

P.S. Kudos to you qusp!!! I generally jump to the more complicated problems, but I'm glad you started there!
 

opc

Member
Joined 2004
Paid Member
Did you try to adjust things? If all the fets were connected to the heatsink, that means all the drains were connected together. If that's the case, then you would have run a little more power through the two drain resistors, but that's about all. The mosfets themselves should not be damaged since that would only happen if a large voltage was applied to the gate.

Take your time and go through everything again. Maybe it would be best to sleep on it for the night and take a fresh look tomorrow.

Regards,
Owen
 
Hey Opc, while ive got you here, while i'm waiting for my new transformers, in order for one of my test setups to fit i have to use some that are a bit too low voltage, 32vac secondaries 50va, with the lt1085/1033 and the diode drop i can only get +/-36.5v under load. with this voltage i can only get 1.5v at the source before i run out of gas, which position can i change to get it back, say if i put a leaded resistor across the 30k9 would this work? or should i just forget about it and run at 1.5 until the new iron gets here? i have other tx to run at higher voltage, like 60v and ive used that no problem, but its too large physically for what i need to do.

yeah Felipes symptoms suggested to me that it was something pretty elementary, since he had the right psu voltage, no smoke and the save readings across the board. then i remembered nearly doing that myself and it fit.

i always try and exhaust the simple explanations first.
 
Member
Joined 2009
Paid Member
More measurements:

R13 199R
R29 553R

R12 812R
R30 100R

Values are with resistors soldered at pcb & I/V not powered.

Right resistors numbers & measured values:

R13 199R
R29 553R

R14 812R
R30 199R

Seems R14 & R29 are wrong value, have I to change? Have I to measure also R16, R17, R25 & R26?

Owen fortunately you sent me 6 extra Caddocks MP930 200R with the PCB
 
Last edited:

opc

Member
Joined 2004
Paid Member
Hi Guys,

qusp:

If I were you, I'd just run it at 1.5V and wait for the new transformers. I measured the I/V set to both 0V and 1.65V at that point, and there wasn't any difference in performance from a distortion standpoint. The only thing it affects is the idle current of the DAC, which should be 0 at 1.65V.

Your project looks superb by the way! Thanks for posting up the pictures. It's going to be one hell of a DAC when you get it all finished up! Can you post a picture or two of the chassis you're using? It looks pretty cool with the curved corners in some of the shots. Did you fab it up yourself?

Merlin:

Measuring resistors in circuit can be a little tricky and could be giving you strange results when nothing is actually wrong. It's exceedingly rare that a resistor would fail and change resistance to a value near the original one. they usually either fail open, or go to relatively high resistance.

If possible, disconnect the power supplies and try measuring again, this time from the drain of Q2 to the drain of Q4 and from the drain of Q1 to the drain of Q3. You should get 400 ohms.

Next measure from source to source as above, and you should get 800 ohms. Measuring R30, R12, R11 and R27 in circuit will be impossible and will give you incorrect values. One trick you can try is to measure each resistor with the leads one way and then the other (swap which side of the resistor the red lead is on). If you get different values depending on the direction, then you're not actually able to measure that value in circuit, and the part will need to be removed for measurement.

Can you verify that you've removed the DAC from the I/V? You can't set things up with the DAC in place and turned off. You need to start with the DAC removed, adjust everything to the correct voltage, then shut everything down, mount the DAC and power everything up together including the DAC. Once it's up to temp, adjust it again with the DAC powered up and you should be good to go. Having the DAC mounted and not powered will result in odd behavior.

Like I said before, having the drains shorted together (by not using the mica insulators) should not have damaged the mosfets, and at worse, would have dissipated about 5W through the two 200R resistors, which should have been fine if they were properly mounted. I really doubt you damaged anything unless there's something else wrong that we're missing.

Regards,
Owen
 
Last edited:
ermm, not sure what you are trying to say, measuring them in circuit means they are in parallel with other resistors and components on the board including the pots, doesnt really help us with anything. those caddock resistors ARE 200 ohms and the 30k9 are 30k9 if theyve been correctly placed of course, if you used the caddock, or whatever to220 resistor you used and it was 200R, thats what it is, its very unlikely that something has changed that.

in saying that, those positions on my board are all in the Mohms and coming down, because the fets and caps are in parallel. one strange thing i've never encountered before happened though. i measured across the 30k9 positions in circuit of course and in one direction with the leads i got over 5Mohms, but switch the leads around and its open circuit. it was the same for that resistor in all 4 quadrants; how is that possible? what mechanism is at work here? is being in parallel with a fet enough to stop any current from flowing through there at all? i would have thought it would still take the path of least resistance instead of being stopped in its tracks.

whats happening there opc? do you know
 
haha funny that you talk about the exact mechanism. i have experienced it being different, that makes sense due to polarity, but fully open circuit is a new one to me.

opc. thanks for the kind words, indeed its looking pretty sweet, you should see it all loaded up, titan is a pretty impressive piece of kit, thats the blue board over in the back corner vertically mounted, its a 8 x 192khz, or 4 x 384khz usb-i2s convertor with 3 x TCXO. i'll get some pics further out for you tomorrow

finished haha, i need to draw a line under it at some point and declare it finished. once i get the second dac done, separate prereg chassis, get the new transformer in use and the prereg pcb fabbed i'm calling it a day.

opc's post wasnt there when i started my reply, thats why it repeats a lot of what he said
 
Last edited:
Member
Joined 2009
Paid Member
Hi Guys,

Merlin:

If possible, disconnect the power supplies and try measuring again, this time from the drain of Q2 to the drain of Q4 and from the drain of Q1 to the drain of Q3. You should get 400 ohms.

Next measure from source to source as above, and you should get 800 ohms. Measuring R30, R12, R11 and R27 in circuit will be impossible and will give you incorrect values. One trick you can try is to measure each resistor with the leads one way and then the other (swap which side of the resistor the red lead is on). If you get different values depending on the direction, then you're not actually able to measure that value in circuit, and the part will need to be removed for measurement.

Can you verify that you've removed the DAC from the I/V? You can't set things up with the DAC in place and turned off. You need to start with the DAC removed, adjust everything to the correct voltage, then shut everything down, mount the DAC and power everything up together including the DAC. Once it's up to temp, adjust it again with the DAC powered up and you should be good to go. Having the DAC mounted and not powered will result in odd behavior.

Like I said before, having the drains shorted together (by not using the mica insulators) should not have damaged the mosfets, and at worse, would have dissipated about 5W through the two 200R resistors, which should have been fine if they were properly mounted. I really doubt you damaged anything unless there's something else wrong that we're missing.

Regards,
Owen

See pics, now I start to change the resistors to see what's measure before power on the I/V
 

Attachments

  • DSC00051_mini.jpg
    DSC00051_mini.jpg
    24 KB · Views: 367
  • DSC00052_mini.jpg
    DSC00052_mini.jpg
    20.6 KB · Views: 358

opc

Member
Joined 2004
Paid Member
Good to hear! Make sure you take your time putting it all back together, and make sure all the parts on the bottom are making good contact with the heatsink, including the resistors!

You'll need to use a piece of FR-4 PCB material to shim the resistors to the correct height (make them as tall as the mosfets) and make sure there is sufficient mounting pressure on all parts so they make good thermal contact with the heatsink. If those two resistors were properly mounted, they should not have failed like that.

Regards,
Owen
 
yeah i'm thinking the resistors werent in good contact as well and the added power dissipation through them blew them, but in a strange half-assed way. myself i would actually desolder them and wick out the solder; mount them to the pcb with the screw through them directly to make sure they all have good even contact. I used arctic silver on the resistors mounted directly to the pcb without a washer, as there is no electrical connection even on the zfoil metal tab.

there is enough space for a screw between them and the pcb, so i actually mounted everything with low profile hex screws to the sink first and then added the pcb over the top. It was a pita wiggling it to get all in the right place, but it assured good thermal contact. You'll have to then carefully do one turn at a time on each screw on each device if you ever want to remove it from the sink so you dont stress the legs, but if you do it right you shouldnt ever have to do it.

now of course if that sounds like too much trouble, do it with the fr4 to keep at the same level, but do not rely on eye to get that right, a fraction of a mm could make the difference between good contact and none at all. a tip if you do it my way, add a tiny dab of solder in the right angle bend of each resistor leg to add strength.

the above goes especially for the guys who got zfoil resistors from me, because the legs are pretty thin pure copper and if you have made a sharp bend in them, they are a bit weakened and wont like being wiggled around too much, take it from me i busted one, luckily i bought 2 spares.

so take from that, dont use sharp edge needle nose pliers and make a really sharp bend in the leads of TX220Z, make it a bit rounded by using round needle nose pliers, or by just gripping it a bit loosely.

anyway looks like we are getting somewhere, glad things are getting back on track for you Felipe, if nothing else, think of yourself and your experience in a positive light, this is good info for the build thread all because of you :cheers:
 
Felipe, i still think you didnt get proper contact first time for it to fail in that way and only one side, regardless of your torc wrench, if one device was soldered at a slightly different level it will stop the others from getting good contact, but you will still have good tension on the wrench which makes you think it does. you might try even reflowing the solder on each one as you tighten the wrench if thats possible with only 2 hands. i prefer soldering them after its mounted for that reason

do watch you dont strip the threads in the heatsink though by using too much pressure
 
Last edited: