[tomb]

 

 
If there is a way to debug a PCM2707, I'm not aware of it.  I once built a pupDAC and used too-high a soldering temperature on my Hakko.  It fried the PCM2707, but it was h*ll figuring that out.
 
Despite that statement above, I can't seriously suggest that's what happened to your DAC - sight unseen.  I still think something is shorting out.  In my case above, every single voltage on the PCB was right on.  It just didn't work.  Yours seems to be a different issue.

 
 
Ah snap! That could very well be it! My iron was set very hot, as I was using it for large point to point soldering. I did not think about lowering it as I figured it wasn't gonna be an issue if I left the iron on the pins only a second. But with the number of pins on those chips, I could have over-heated it.
 
How did you come to the conclusion the chip was fried? Any definitive test I can run?
 

 
I only mentioned it because you were asking about the PCM2707.  Remember that I said in my case, all the voltages were dead on (and there are many on the PupDAC) - it just wouldn't connect without a PC USB-connect error, no matter what I tried.
 
Yours seems to be a true short somewhere with the voltages off and the tremendous heating.

 

[KimLaroux]

Okay. Is there a netlist already available for this board? I'd make one but since I have no clue what it looks like I wouldn't know where to start.
 
I'll use the scans on the website as a starting point to make a continuity check. It gives me an idea of what's connected to what, though the vias makes it confusing for a newbie like me.

 

[KimLaroux]

In the meantime, here's the promised pictures:
 

 

 

 

 
Don't mind the cut trace by C10. I did that to figure if the problem was on the PCM or DAC chip. It seems to be a problem with the PCM chip, as opening that circuit didn't change anything.

 

[Avro_Arrow]

 
It looks like a picture perfect (pun intended) soldering job.

Dis-mount L3 and see if U2 functions normally.
If so, replace U1.

 

[tomb]

   
It looks like a picture perfect (pun intended) soldering job.
 
Dis-mount L3 and see if U2 functions normally.
If so, replace U1.

Agreed.

 

[KimLaroux]

Yay well I spent the evening verifying that the continuity of the circuit matched what I read on the schematic. I didn't find anything out of order.
 
I guess the next thing is to change the PCM chip.
 
But uhm, how to I desolder such a chip without ruining the pcb? =X
 
Also, the BOM lists the PCM2706 but I got a PCM2707. I'm guessing the 07 is simply an update to the 06 and any of the two would work on the Grub?

 

[Avro_Arrow]

 
But uhm, how to I desolder such a chip without ruining the pcb? =X
 
 

There are a few different ways...
The best is a hot air desoldering station.
 
Failing that, cut all the leads from the chip and then
just wipe off the leads with your soldering iron.
Be careful not to cut the board.

 

[Avro_Arrow]

   
 
Also, the BOM lists the PCM2706 but I got a PCM2707. I'm guessing the 07 is simply an update to the 06 and any of the two would work on the Grub?

They have some different features from each other, but for your purposes, they are interchangeable.

 

[tomb]

  Yay well I spent the evening verifying that the continuity of the circuit matched what I read on the schematic. I didn't find anything out of order.
 
I guess the next thing is to change the PCM chip.
 
But uhm, how to I desolder such a chip without ruining the pcb? =X
 
Also, the BOM lists the PCM2706 but I got a PCM2707. I'm guessing the 07 is simply an update to the 06 and any of the two would work on the Grub?

It will probably result in destruction of the chip, but I like to make a large solder blob over all the pins on one side of the chip. If you have a dental pick, you can lift the pins while the solder blob is melted on one side.  Do this by placing the pic in the space that exists between where the pins exit the chip and the pads on the PCB.  Repeat for the other four sides and you should have the chip removed.  Clean up with de-soldering braid and install the new chip as if it were a new PCB.
 
Use whichever method sounds easiest for you.

 

[wakibaki]

When I'm replacing a SMT chip, I cut the defunct chip off the board, pin by pin, using a scalpel. You put the scalpel tip on top of the pin, as close as you can get it to the chip body and press hard, using one hand to hold the scalpel proper, and pressing with two fingers of the other hand further down the shaft. You hear a sharp click as the pin parts and the scalpel blade drops down a fraction of a millimeter and comes to rest on a tiny ridge on the chip body. When you've worked your way through all the pins the chip will drop off. Then you pick up all the severed pins off the pads with a hot soldering iron and clean with braid.
 
You have to be careful when doing this because you have to apply quite a lot of force, and you must take great care that the blade does not slip (or break) and damage either you or the board, but it's quick, positive and in my experience runs less risk of damaging a pad than prying at the pins.
 
w 
 
Oh, 2706 supports an external ROM. Missed that A_A had already said cut it off.

 

[MisterX]

 ~~But uhm, how to I desolder such a chip without ruining the pcb? =X

 
Chip Quik.
 
 

 
Or if you can stand to watch the videos from that EEVblog guy....
 
http://www.youtube.com/watch?v=UmD7F0

 

[Avro_Arrow]

   
Chip Quik.
 

That stuff works great...the only problem is it's fifteen to twenty bucks for the kit and if you only have to remove one chip...

 

[KimLaroux]

  When I'm replacing a SMT chip, I cut the defunct chip off the board, pin by pin, using a scalpel. You put the scalpel tip on top of the pin, as close as you can get it to the chip body and press hard, using one hand to hold the scalpel proper, and pressing with two fingers of the other hand further down the shaft. You hear a sharp click as the pin parts and the scalpel blade drops down a fraction of a millimeter and comes to rest on a tiny ridge on the chip body. When you've worked your way through all the pins the chip will drop off. Then you pick up all the severed pins off the pads with a hot soldering iron and clean with braid.
 
You have to be careful when doing this because you have to apply quite a lot of force, and you must take great care that the blade does not slip (or break) and damage either you or the board, but it's quick, positive and in my experience runs less risk of damaging a pad than prying at the pins.
 
w 
 
Oh, 2706 supports an external ROM. Missed that A_A had already said cut it off.

 
Oh, that works well.
 
I used my pocket knife, which I sharpen like a razor blade. I put the rounded edge on the row of pins, then pushed down while rocking the knife back and forth. It slowly dug into the row of pins and cut them from the body. I then wiped the pins away using my iron.
 
 
But now it gets even weirder. I plugged the board in, without U1, and got the exact same message from dmesg:
 

 [30536.808216] usb 6-1: new full-speed USB device number 2 using uhci_hcd [30536.921122] usb 6-1: device descriptor read/64, error -71 [30537.135224] usb 6-1: device descriptor read/64, error -71 [30537.338090] usb 6-1: new full-speed USB device number 3 using uhci_hcd [30537.450086] usb 6-1: device descriptor read/64, error -71 [30537.663111] usb 6-1: device descriptor read/64, error -71 [30537.866219] usb 6-1: new full-speed USB device number 4 using uhci_hcd [30538.275207] usb 6-1: device not accepting address 4, error -71 [30538.377220] usb 6-1: new full-speed USB device number 5 using uhci_hcd [30538.787205] usb 6-1: device not accepting address 5, error -71

 
I mean, WHAT?!? How can my computer detect anything when there's no circuit to communicate with? 

Anyways I got a bunch of PCM chips on order, we'll see when they arrive, I guess.

Thanks everyone for your help.

 

[DingoSmuggler]

But now it gets even weirder. I plugged the board in, without U1, and got the exact same message from dmesg:
 

 [30536.808216] usb 6-1: new full-speed USB device number 2 using uhci_hcd[30536.921122] usb 6-1: device descriptor read/64, error -71[30537.135224] usb 6-1: device descriptor read/64, error -71[30537.338090] usb 6-1: new full-speed USB device number 3 using uhci_hcd[30537.450086] usb 6-1: device descriptor read/64, error -71[30537.663111] usb 6-1: device descriptor read/64, error -71[30537.866219] usb 6-1: new full-speed USB device number 4 using uhci_hcd[30538.275207] usb 6-1: device not accepting address 4, error -71[30538.377220] usb 6-1: new full-speed USB device number 5 using uhci_hcd[30538.787205] usb 6-1: device not accepting address 5, error -71

 
I mean, WHAT?!? How can my computer detect anything when there's no circuit to communicate with? 

That's probably a good sign - if the chip was shorting out it's supply voltage, then strange that it would also be able to communicate at all via USB.
The USB host must just be picking up the fact there's a load on its power lines. The USB host actually controls the current available, it's not just hooked up to a 5V line in your PC. By default there should be 100mA available, high power USB devices need to negotiate with the host to get 500mA.

 

[tomb]

 

But now it gets even weirder. I plugged the board in, without U1, and got the exact same message from dmesg:
 

 [30536.808216] usb 6-1: new full-speed USB device number 2 using uhci_hcd[30536.921122] usb 6-1: device descriptor read/64, error -71[30537.135224] usb 6-1: device descriptor read/64, error -71[30537.338090] usb 6-1: new full-speed USB device number 3 using uhci_hcd[30537.450086] usb 6-1: device descriptor read/64, error -71[30537.663111] usb 6-1: device descriptor read/64, error -71[30537.866219] usb 6-1: new full-speed USB device number 4 using uhci_hcd[30538.275207] usb 6-1: device not accepting address 4, error -71[30538.377220] usb 6-1: new full-speed USB device number 5 using uhci_hcd[30538.787205] usb 6-1: device not accepting address 5, error -71

 
I mean, WHAT?!? How can my computer detect anything when there's no circuit to communicate with? 

That's probably a good sign - if the chip was shorting out it's supply voltage, then strange that it would also be able to communicate at all via USB.
The USB host must just be picking up the fact there's a load on its power lines. The USB host actually controls the current available, it's not just hooked up to a 5V line in your PC. By default there should be 100mA available, high power USB devices need to negotiate with the host to get 500mA.

Agreed.  It sounds like the USB PCM chip (2707) was unable to negotiate with the host - probably because it was shorted somewhere.  So, maybe replacing it will fix the problem.