[Blanko]

Unfortunately I'm experiencing the same issue as Mullet. When Volume is low or no music is playing
I sometimes hear a pulsing noise lasting one or two seonds.
There also is a frequent crackling sound. Those sounds aren't very loud but quite anoying. Only the left channel is affected.
I double checked every chip for bridges though my magnification isn't sufficient. I held the PCB against a bright LED lamp.
There was nothing suspicious.
I used 99.5% isopropyl alcohol for cleaning. As Mullet indicated his problem was solved after he had replaced U9 with LMH6643
and swapped U5 plus U7.
Could this help in my case too? It's hard for me to get replacement parts. A Mouser shipment to Germany costs me 30$.
 
Somehow I managed to measure DC offset. It reads -1.5 mV on the left and 1.2 mV on the right channel.
This is really frustrating since I put so much effort in it but I got excited to early.
I have no experience with desoldering so I fear I end up with a bricked pup in the end.
Any suggestions?

 

[Misterrogers]

Hey Blanko - I'm sure tomb or cobolmute will provide 'next steps', just wanted to encourage you to hang in there. DIY is a journey that teaches some of its very best lessons the hard way. Push through the troubleshooting, learn as much as you can about the 'why' the change/fix addresses the issue, and work carefully. 
 
As to desoldering, you'll need a couple of things in your kit. Firstly, good quality solder braid. There's a world of difference between the effectiveness of the good vs. the cheap stuff. If you're going to continue doing SMD work, you'll need to at least pick up a reasonable second Iron. Early on, I removed 2 pad components with two irons, then cleaned up with braid. A cheap solder sucker can also help, though more so with through hole work.
 
I've since picked up some SMD Tweezers, power suction, hot air pencil, reflow oven... yeesh, I've spent some money. And while I have, it's all seriously lowered the frustration level and increased the enjoyment. It's no big deal to lift parts (without lifting pads .
 
Bottom line - hand in there!

 

[Avro_Arrow]

I noticed I get some noise when running the Pup from my laptop but it's perfectly quiet when used with my desktop.
Can you try a different source?
 
 
Quote:

Unfortunately I'm experiencing the same issue as Mullet. When Volume is low or no music is playing I sometimes hear a pulsing noise lasting one or two seonds.
There also is a frequent crackling sound. Those sounds aren't very loud but quite anoying. Only the left channel is affected.
I double checked every chip for bridges thoug my magnification isn't sufficient. I held the PCB against a bright LED lamp. There was nothing suspicious.
I used 99.5% isopropyl alcohol for cleaning. As Mullet indicated his problem was solved after he had replaced U9 with LMH6643 and swapped U5 plus U7.
Could this help in my case too? It's hard for me to get replacement parts. A Mouser shipment to Germany costs me 30$.
 
Somehow I managed to measure DC offset. It reads -1.5 mV on the left and 1.2 mV on the right channel.
This is really frustrating since I put so much effort in it but I got excited to early.
I have no experience with desoldering so I fear I end up with a bricked pup in the end.
Any suggestions?

 

[Mullet]

It could be the same issue. I'm still not sure whether my issue was with U5 and U7 or if it had more to do with U9. I still haven't gone back to a different U9 and I probably won't because it's not worth the time.
 
I'd advise to buy something like chipquik, which will help remove the chip(s). Either way, you're going to have to spend more money to get this working. Overall, you should be able to get this working properly like I did.

 

[tomb]

Sorry to hear this, Blanko.  The good thing is that from your description, it sounds like the opamp, U9.
 
Here are some questions/suggestions:

• I can't tell from the single photo you posted, but did you solder the R6, R9, R17, and R21 resistors on the back side of the PCB?  If on the top of the PCB, it's possible the extra lead length is causing the opamp to become unstable.
• It's possible the mica capacitors may be contributing to this, also.  Can you squeeze them further down into the PCB?  I've built over 10 now and the OPA2836 is absolutely stable on my builds.  However, I squish those mica's down so far into the PCB that some of the leads and hole-plating come off in strips.  You certainly don't want to break anything by forcing it too hard, but I'd try to get the cap body as close to the surface of the PCB as possible.  You can heat one of the holes/leads up, then try to pull it through with pliers from the other side.  It depends on much you have to grip on the bottom after trimming the leads, though.
• In the whold scheme of things, removing the opamp is pretty simple.  Add enough solder to one side of the opamp - try to cover all four pins on one side.  The soldering iron should be able to keep the whole solder blob melted.  At the same time, use a small knife/dental pick or similar to carefully pry the chip up from the PCB on that side.  Add a solder blob to the other side and remove the chip.  Clean up the PCB with de-soldering braid.  It works an awful lot simpler than it sounds.
  

 
If none of that works, then maybe Mullett is correct that the problem could be the charge pump or one of the regulators.   Let us know what you can about the questions/suggestions.

 

[Blanko]

First of all thanks for encouraging me.
What I have done so far:
 
1. Tried a different source. I plugged it in the USB hub of my raspberry pi. Same problem.
 
2. Desoldered the mica cap and got it one mm deeper into the PCB but the lead space on this one is simply too big to get it flush.
    I could push a few resistors deeper into the PCB. Yes, R6, R9, R17, and R21 are mounted on the backside.
   So far no change. all measurements where as before.
 
3. Desoldered U9 with two irons, wicked all excess solder up, cleaned pads and put LMH down.
   The 5V are both lower than before. ca 4.8V. DC offset is very high. R: 3.5mV L: -4.5mV.
   Still recognized by the PC, still playing audio. Pulsing and crackle are gone but I have a much higher noise floor than before.
    Before right channel was silent. Now both sound like a small waterfall.
 
I guess I have still the same problem as before only the two OPA's behave different in this matter.
At least I could compare both OPA's and I have to say that the 2836 is the better one not only in measurements.
There are some audible differences especially in the heights and upper voices. The 2836 has more headroom and crisper heights than the LMH.
The LMH sounds a bit richer and warmer in the mids. It also has a more punch in the lows but the bass of the 2836 goes a bit deeper and is more defined
 
Depending on whether I get spares for U5 and U7 or not I'm going to replace them. But for now I don't know what to do else.
Could I have added to much Capacitance? C2 and its brothers are 68uF, C10 is 220uF and C26/27 are 470uF.

 

[Mullet]

I have a few extra 2835s and wanted to give the 2836 a shot. Anyone want to trade?

I think I already know the answer to this... being that the PCM2707 is only capable of 16/48 does that mean that is what is coming out of the PCM1794? Or does the 1794 upsample or oversample to 24/192?

 

[Avro_Arrow]

You only get what the PCM2707 gives you...
 
 
Quote:

I have a few extra 2835s and wanted to give the 2836 a shot. Anyone want to trade?

I think I already know the answer to this... being that the PCM2707 is only capable of 16/48 does that mean that is what is coming out of the PCM1794? Or does the 1794 upsample or oversample to 24/192?

 

[tomb]

Quote:

First of all thanks for encouraging me.
What I have done so far:
 
1. Tried a different source. I plugged it in the USB hub of my raspberry pi. Same problem.
 
2. Desoldered the mica cap and got it one mm deeper into the PCB but the lead space on this one is simply too big to get it flush.
    I could push a few resistors deeper into the PCB. Yes, R6, R9, R17, and R21 are mounted on the backside.
   So far no change. all measurements where as before.
 
3. Desoldered U9 with two irons, wicked all excess solder up, cleaned pads and put LMH down.
   The 5V are both lower than before. ca 4.8V. DC offset is very high. R: 3.5mV L: -4.5mV.
   Still recognized by the PC, still playing audio. Pulsing and crackle are gone but I have a much higher noise floor than before.
    Before right channel was silent. Now both sound like a small waterfall.
 
I guess I have still the same problem as before only the two OPA's behave different in this matter.
At least I could compare both OPA's and I have to say that the 2836 is the better one not only in measurements.
There are some audible differences especially in the heights and upper voices. The 2836 has more headroom and crisper heights than the LMH.
The LMH sounds a bit richer and warmer in the mids. It also has a more punch in the lows but the bass of the 2836 goes a bit deeper and is more defined
 
Depending on whether I get spares for U5 and U7 or not I'm going to replace them. But for now I don't know what to do else.
Could I have added to much Capacitance? C2 and its brothers are 68uF, C10 is 220uF and C26/27 are 470uF.

I'm guessing now that it's somewhere in the regulators or the charge pump, as Mullett also suggested (U5 and U7).  However, keep in mind that's only a guess.  It's very strange that things were fine at first, but then deteriorated to this extent.
 
I doubt the capacitance could be any issue.  I've built several with higher value capacitors - so have other people.  My latest builds use 1500uf at C22 and Panasonic FM's at C26/C27 - same 330uf, but the impedance and ripple are much, much higher.  There's no issue.

 

[Mullet]

I'd think its the charge pump that is generating the pulsing noise. Once I changed that out things seemed to work out.
 
What's the benefit of using higher capacitance C22 and higher impedance/ripple for C26/27?

 

[tomb]

Quote:

I'd think its the charge pump that is generating the pulsing noise. Once I changed that out things seemed to work out.
 
What's the benefit of using higher capacitance C22 and higher impedance/ripple for C26/27?

I hope that's good advice and thank you for it - you have more experience getting a problem pup to work than me.

 
C22 is the primary power cap.  My experience with the Bantam and Alien seem to indicate that the higher capacitance you can manage at the power level, the more stable the DAC USB connection is - plus, more bass may result because power is more stabilized throughout the board. A similar thing is true with C26/C27.  Those are the "power caps" for the opamp output.  So, the more capacitance - or higher quality capacitance - you can put there, the better the bass.  Bass usually needs more power or cleaner power to really come through.  Think putting two 470uf caps in a CMoy circuit instead of two 220uf as originally conceived.
 
'Course, the above is all un-proven empirical opinion.  It seems pretty likely that it's not going to hurt, though, unless the capacitance is so high that the DAC draws enough start-up current to get knocked off the USB circuit.

 

[geofftnz]

I put together my pupdac today, but it hasn't gone particularly smoothly... It recognises in Windows, all voltage test points pass, but I'm getting a very large offset of -0.5V on the left channel and -2.4V on the right. Hooking up the scope shows some audio signal, but at a very low level. I'm not sure where to start debugging - I've reflowed half the connections on the board, concentrating around the resistors surrounding the opamp, but not joy.
 
Any ideas?

 

[tomb]

  I put together my pupdac today, but it hasn't gone particularly smoothly... It recognises in Windows, all voltage test points pass, but I'm getting a very large offset of -0.5V on the left channel and -2.4V on the right. Hooking up the scope shows some audio signal, but at a very low level. I'm not sure where to start debugging - I've reflowed half the connections on the board, concentrating around the resistors surrounding the opamp, but not joy.
 
Any ideas?

 
The opamp controls offset.  Ensure that you have the resistors soldered on the bottom of the PCB (R6, R9, R17, R21)*, that all resistors are correct, and that they're flush to the surface of the PCB (minimizes parasitic effects from leads longer than they should be).
 
If none of that works, the kit still comes with the OPA2835 in addition to the OPA2836 that I manually add to the kit.  De-solder the OPA2836 and install the OPA2835 - see if that makes a difference.  I'll send you another OPA2836 it that turns out to be the case.
 
 
* For that matter, ensure that you haven't mixed up R6 with R5.  The silkscreen for R6 on the bottom of the PCB is defective and it looks like "R5."

 

[geofftnz]

   
The opamp controls offset.  Ensure that you have the resistors soldered on the bottom of the PCB (R6, R9, R17, R21)*, that all resistors are correct, and that they're flush to the surface of the PCB (minimizes parasitic effects from leads longer than they should be).
 
If none of that works, the kit still comes with the OPA2835 in addition to the OPA2836 that I manually add to the kit.  De-solder the OPA2836 and install the OPA2835 - see if that makes a difference.  I'll send you another OPA2836 it that turns out to be the case.
 
 
* For that matter, ensure that you haven't mixed up R6 with R5.  The silkscreen for R6 on the bottom of the PCB is defective and it looks like "R5."

 
hmmm... I was wondering if it was maybe a resistor mix-up. Both signals out of the opamp are very stable non-rail voltages. I'll check that when I get home. FWIW I installed the OPA2835 first - I'll swap with the OPA2836 if the resistors don't show up anything.
 
This project hasn't been my best soldering effort either, so I can't rule out bad workmanship.

 

[geofftnz]

   
The opamp controls offset.  Ensure that you have the resistors soldered on the bottom of the PCB (R6, R9, R17, R21)*, that all resistors are correct, and that they're flush to the surface of the PCB (minimizes parasitic effects from leads longer than they should be).
 
If none of that works, the kit still comes with the OPA2835 in addition to the OPA2836 that I manually add to the kit.  De-solder the OPA2836 and install the OPA2835 - see if that makes a difference.  I'll send you another OPA2836 it that turns out to be the case.
 
 
* For that matter, ensure that you haven't mixed up R6 with R5.  The silkscreen for R6 on the bottom of the PCB is defective and it looks like "R5."

 
I checked the resistors and they were all in the right place. (I was being super meticulous, so it was a bit worrying to think that I'd maybe messed that up). I measured a bunch of point to point resistances around the opamp and they all summed to expected values. In the end I cut out the 2835 and replaced with the 2836. It's now showing +130mV offset on both channels - not ideal, but better than the whole volts I was seeing before. On my ****ty scope with no audio playing I'm seeing a 50Hz at around 3mV, plus what looks like a ringing pulse up in the MHz range at a similar amplitude. Audio looks good on the scope when I play something, so I'm going to be brave and hook it up to my (AC-coupled) amp.
 
Edit: audio sounds good to my untrained ear. I'm a much happier lad compared to last night - didn't really want to have killed 2 out of 3 cobaltmute/tomb kits