[Eyediver]

Just built two Grubdacs and neither one is functioning correctly.  To make matters worse, I accidently tore out the "relief pins"  with a scratch awl (after trying solder wick) thinking they were bridges.  Eek!
 
Both DACs have larger than normal Wima Capacitors and neither currently have the USB braid grounded.
Both measure 3.3v at the test point.
 
1.  Will the larger than normal Wimas negatively effect performance?
2.  Could the lack of USB braid ground be causing some of the problems?
3.  Is there an easy enough way (such as bridging pins or running jumper wires) to repair the "relief pins"?


 
 

 
 
DAC 1:
1.  Does make music some of the time.  Sometimes registers when I plug it in and sometimes not.  Will intermittently unplug and re-plug itself in.
2.  LED is not lighting.  Measures 5v on leads.  Bad LED?
3,  Drilled out the output connections because my RCA wires (for a cable DAC) were too large.  I can connect the positve leads to the right and left capacitors but where should I connect the grounds?


 

 
DAC 2:
1.  LED does light but it does not make music.  When I connect it I get the message "USB device not recognized" and "Malfunctioning Device - Windows does not recognize it."
 
Any help is much appreciated!

 

[particleman14]

reflow both U1 and U3, use lots of flux!  Then do a big cleaning with alcohol 99.9% and a brush.  No recognition from windows usually means U1.  The intermittent playback usually means dirty chips.  Also switch led polarity to see if it lights up.

 

[jdkJake]

I also suggest you reflow X1 on the second DAC. It provides the clock for the chip. You may want to reflow it on the first dac as well.

As for the capacitors, they may have an effect on performance. They provide part of the low pass filter circuit on the output. Are those .22uf parts or .022uf parts? Their actual value will tell you if they will negatively effect performance.

By "relief pins", do you mean the through-board vias? The holes that appear to go through the board?

As for the output on dac1, per the schematic, the other end of the capacitor you referenced is tied to the ground plane. You should be able to use that as your ground reference. In the future, rather than drilling those out, use part of a resistor lead attached to the end of your wire to mount in the hole. Far less damaging to the board and allows future change/rework.

As particleman mentioned, LEDs are polarized. Try swapping the orientation first.

 

[jdkJake]

BTW, also check C1 on dac2. It might just be the angle, but, you might have missed soldering one of the leads.

 

[Eyediver]

Thanks for the responses guys.  They are much appreciated!
 
I switched the polarity on the LED for DAC 1 and it now works.  Yay!
 
Reflowed X1, U1 and U3 with lots of flux on both DACs and cleaned again but no change.
Checked C1 on DAC 2 an it appears to be fine.
 
The Wima caps are .22uf.
 
By "relief pins" I mean in reference to the 7th picture down on the build page 1 (shown below).  Apologies for not being more clear before.
 
Quoted from the build page:   [size=xx-small]http://www.diyforums.org/GrubDAC/GrubDACphoto-build-1.php[/size]
"Note the pins at left top, bottom left, and top right (in the pic) - this is the way Cobaltmute's PCB design "relieves" the pins attached to the ground plane - a trace runs perpendicular to the pins underneath. Those are not bridges!"
 
 
 

 
http://focus.ti.com/lit/ds/symlink/pcm2707.pdf

I scratched out the traces between pins 2 and 3, VCCP and HOST
and pins 25 and 26, [size=xx-small][size=xx-small]ZGND and [/size][/size][size=xx-small][size=xx-small]AGNDL in my first attempt to get these working and "clear the bridges".  (sigh)[/size][/size]
 
[size=xx-small][size=xx-small][size=xx-small]http://www.diyforums.org/GrubDAC/schematic/GrubDACschematic.pdf[/size][/size][/size]
[size=xx-small][size=xx-small]They look to be joined in the schematic so I should just be able to run a solder bridge between the two, yes?[/size][/size]

 

[jdkJake]

Well, hate to tell you, but, those wima caps Values pose a real problem. If they were .022uf you would have been fine, but not .22uf.

In this circuit, those caps along with the resistors at r8 and r9 form a low pass filter to filter out any high frequency noise in the signal. They are sized according to the desired corner frequency of the filter. From that frequency, you get a simple 6db per octave roll-off of all higher frequencies.

The design of the grubDAC uses 110R and .015uf to get a corner frequency of 96khz. Far above audible. Your use of .22uf puts the corner frequency at 6.6khz. Definitely audible. You will have no high-end to speak of.

Without inciting a religious war, IMO, this cap should be sized for at least 48khz or in this case .03uf.

So, you have two options. Replace the caps or replace r8 and r9. For 48khz, you could use 15R, for 96khz, you could use 8R. Personally, I would replace the caps with something closer to the intended design target.

 

[jdkJake]

In regards to the "relief pins", yes, you will need to reconstruct them. In this particular case, a soldier bridge is your friend, but, might be difficult to actually implement. Ideally, a wired connection would be better, but, it is extremely difficult to execute on such a small scale. For sure, either method will be a test of your SMD solder ability.

I suggest you try a well placed solder bridge first. If you have needle points (or similar) for your multimeter, you can verify the work by measuring the resistance between the pins in question.

 

[tomb]

jdkJake has given some great comments.  I had hoped you would post this in the GrubDAC thread, however, where cobaltmute (the designer) keeps an eye on things.  It may take him awhile to notice this thread (can't suscribe to something until you see it in the first place).
 
If he doesn't make a comment in a day or so, I'll shoot him an e-mail. (That's assuming Jake's comments don't fix it for you.)
 
P.S. Great pics, btw!!

 

[Eyediver]

Thanks for chiming in TomB and thanks for following up.
 
jdkJake,
I bridged the missing traces with solder on both DACs and grounded the USB braid on DAC 2 to the appropriate place.
I swapped out the oversize Wimas with appropriate substitutes and DAC2 is officially up and running.  I'm quite amazed at how good it sounds
without even being broken in.
 
Not sure if grounding the braid was one of the missing links but I will try it on DAC 1 in the next few days and report back. 
 
Thank you for your continued assistance jdkJake!

 

[jdkJake]

You are most welcome. Looks like you are well on your way to getting both DACs operational and that is good news indeed.

 

[Eyediver]

Grounded the braid on DAC1 today.  For some reason it is not even playing intermittently as it was before.  Any more ideas anyone?

 

[jdkJake]

EDIT: My previous version of this post was rather incomplete
 
So, my real question, was it working intermittently after you did the initial reflow but before you made the solder bridge mods?

 

[Eyediver]

Boy... I *think* that is correct.  I can take them off again and see what comes...

 

[jdkJake]

Yeah, tough call. You could always reflow one chip at a time and work your way towards "the beast".
 
Do you have needle points for your multimeter? You can verify a lot of the connections this way and localize the repairs. I did this with my gamma2. I isolated the reflows to a couple of pins on a single chip just by measuring continuity across the various connections and adjacent pins. It can actually speed things up considerably. 

 

[Eyediver]

Reflowed ALL the solder joints again and dug out the meter to check all the adjacent chip pins.  Everything seems to be in order.
 
Strangely, the DAC 1 is now working but only on my laptop.
DAC 2 plays on both my desktop AND my laptop.
 
Do you think it could be the 2706... somehow the part that communicates with USB bus is cooked by ESD or heat?