Yes, it does look nice. Though, I received a new panel made from bamboo in the mail yesterday with my 2nd O2 amp. The bamboo panel with the black case looks sweet!
I'll try to get a picture posted this weekend so you can see if you agree.
Start-up Problem
I completed the O2 and passed the initial checks with no apparent problem, but on the initial test drive it was pretty obvious that my right channel was awfully distorted while the left seems to be OK.
Would the short already mentioned between via hole and gain switch near R21 cause this?
I completed the O2 and passed the initial checks with no apparent problem, but on the initial test drive it was pretty obvious that my right channel was awfully distorted while the left seems to be OK.
Would the short already mentioned between via hole and gain switch near R21 cause this?
Forget about my last post...I was frustrated...still am tho.
We have pinned down the problem....a faulty board
It seems that the path between R16 and U1 pin 6 is the culprit, as result we had -10.6 volts or so in that pin.
What to do next? We can add that R16 and C19 cap off the board to overcome the issue but to be honest I don't like it, it does the job but I hate to have components floating around.
I bought the PCB from JDs Labs, I have thought on returning it (if that can be done) but its already populated with all the bells an whistles
We have pinned down the problem....a faulty board
It seems that the path between R16 and U1 pin 6 is the culprit, as result we had -10.6 volts or so in that pin.
What to do next? We can add that R16 and C19 cap off the board to overcome the issue but to be honest I don't like it, it does the job but I hate to have components floating around.
I bought the PCB from JDs Labs, I have thought on returning it (if that can be done) but its already populated with all the bells an whistles
That does sound like the gain switch problem. It affects the right channel and causes distortion. Just take a small (jewelers) screwdriver and carefully bend the two tabs on the switch that are over the via back under the switch. The shell of the switch is soft metal and bends pretty easily. The tabs can look like they are not touching the via but actually are.
As for your voltage on U1 pin 6 (pin 6 is on the left channel), pull U1 out of the socket and check if one of the pins bent underneath U1 rather than going into the socket properly. Easy to happen. The pins on chips are bent out slightly to stay put in automatic insertion equipment. I always place an IC on one side on an anti-static bag and carefully roll the pins a few degrees so they are exactly perpendicular to the body, then repeat on the other side. Goes right in the socket properly that way.
If you don't have any bent pins on U1, what are the voltages from all 8 pins to ground (the metal shell of the gain switch) with nothing plugged into the input jack and the gain switch not pressed in?
Last edited:
I did check for continuity between the gain switch body and the via hole (bottom of PCB) and they do not appear to be in contact with each other.
To be on the safe side I´ll bend the legs *** suggested.
After that I´ll try to describe briefly what we did, what I recall and hope I´m not mixing the results from this morning tests.
We checked the proper sitting of U1 with its socket and all legs were bent prior the mounting in the socket, we even tried an different IC, same result. After that we started checking for resistance... between pin2 and pin1 we got 1.5K witch matches R22, but when we checked between pin 6 and 7 we didnt get any reading, we should have measured the 1.5k from R16, we checked the solder and the resistance itself out of the board and was OK.
After that we "bypassed" the board by soldering one end of the resistance directly to the IC pin6 and the other end to the board (yes we leaved out C19 out of the game with this test) , mounted the IC in its socket an voila! the -10 volts did not appeared in readings anymore.
With the above tests we concluded that some how the path between R16 and pin6 of U1 was the problem. We cant think of any other reason that could have caused the problem or that could have escaped the tests we did. The only thing that comes to mind is a non visible solder bridge somewhere but I cant figure out where a short would develop the symptoms and evade the test we did.
At some point we did check all the voltages from U1 pins and seemed OK, cant recall them now but tomorrow will try to recheck them and post the results....maybe you guys can help us to find a different outcome.
BTW I talk in plural since APASSGEAR (aka my father) old time member in the forum has help me a lot with the project.
It would be relatively rare, but you may be right about a PCB problem there. The PCB trace going from U1 pin 6 out to R16 and C19 is on the top of the board, partially underneath the U1 socket. The trace relies on a through-hole at pin 6 to connect the bottom to the top of the board. Plated through holes can be bad, either break or not get fully plated in manufacture.
The first thing to try is to re-heat the solder joint on pin U1 pin 6 if you haven't already. If that doesn't do it you can test easily enough. Pull U1 out of the socket. Then with your ohmmeter on a low ohm range put one lead on the bottom of the PCB at pin 6 and the other on the end of R16 (on the top of the PCB now) closest to the board edge on the top. That should be nearly 0 ohms. If not the through-hole is bad and/or that hidden trace under the socket is broken.
Like sofaspud says you can do a fix with a jumper wire to the IC pin. Not pretty, but it works.
The first thing to try is to re-heat the solder joint on pin U1 pin 6 if you haven't already. If that doesn't do it you can test easily enough. Pull U1 out of the socket. Then with your ohmmeter on a low ohm range put one lead on the bottom of the PCB at pin 6 and the other on the end of R16 (on the top of the PCB now) closest to the board edge on the top. That should be nearly 0 ohms. If not the through-hole is bad and/or that hidden trace under the socket is broken.
Like sofaspud says you can do a fix with a jumper wire to the IC pin. Not pretty, but it works.
Last edited:
We have pinned down the problem....a faulty board
It seems that the path between R16 and U1 pin 6 is the culprit, as result we had -10.6 volts or so in that pin.
Can you post a picture of the problem... is this a short or an open... I've got several boards I'd like to inspect prior to putting the parts on them...
Thanks
db
If it's a tricky via, it would be easy enough to flow a little solder into it - an option for one who wants to avoid conspicuous repairs. I could understand how someone might prefer not to trust that, though.
Worth a try! I usually just solder the suspect ones on both the top and bottom of the board, but can't get to the top of this one since the socket completely covers it.
I have 4 boards I bought from JDS around Xmas time... just ohm'd them out and they all tested fine...
When I look at the board, if it had a via (thru-hole) issue, it would be very easy to just push a resistor wire into the via and solder on both sides.. then cutting level with the surface.. but, if its a trace issue... probably the same fix... just put the resistor wire into the via.. solder on both sides, then bend the wire over to the connection point where the trace is no longer open... then clip both ends... easy fix and a good fix...
I will say that of the many that I have built, I have only had one that gave me any issue.... the issue was a mis-balance between the left/right headphone... I trouble shot it down to the headphone jack and started to replace the jack. After I removed most of the soldier I tested it one more time and it worked perfect... so I can assume that the jack had too much soldier and was bridging.. didn't replace the jack and it still works perfect today... now just waiting on the ODAC so I can start to build the other boards...
It would be relatively rare, but you may be right about a PCB problem there. The PCB trace going from U1 pin 6 out to R16 and C19 is on the top of the board, partially underneath the U1 socket. The trace relies on a through-hole at pin 6 to connect the bottom to the top of the board. Plated through holes can be bad, either break or not get fully plated in manufacture.
The first thing to try is to re-heat the solder joint on pin U1 pin 6 if you haven't already. If that doesn't do it you can test easily enough. Pull U1 out of the socket. Then with your ohmmeter on a low ohm range put one lead on the bottom of the PCB at pin 6 and the other on the end of R16 (on the top of the PCB now) closest to the board edge on the top. That should be nearly 0 ohms. If not the through-hole is bad and/or that hidden trace under the socket is broken.
Like sofaspud says you can do a fix with a jumper wire to the IC pin. Not pretty, but it works.
Solder joins seem OK at pin6, we even changed the IC socket at one point during test. We did check continuity between R16 an pin6 a you suggest but we did it at the top of the board, we will try as you mention, checking with one probe on top (R16) and the other on the bottom at pin6...you never know LOL.
What I want to check is the via hole(s) in the area, now I recall having a bit of solder in one of them around there, I dont see how would this cause the issue since adding a drop of solder only "reinforces" an already formed path between top and bottom layers, we were sure there was no unwanted bridges forming and did a visual check on this...unless the bridge is so small that isnt to be seen....ah! lots of things to check, isn't this fun
I´ll try to upload a pic tonight, but is fairly easy to check, if you have a DMM check for continuity between U1 pin6 and R16 (you should have continuity between R16 hole closest to the front of the board an pin6)
Thanks for the suggestion, I´ll check the pros and cons
Hi guys,
Yes, it's an open connection all right and we should try to repair it as dixter sez by adding a piece of lead on the vias and solder and see if this solves the open connection, or the lead beneath the board as someone else suggested. No big deal anyway.
On another subject altogether, it really calls my attention that you guys are using ceramic caps on a feedback path, of course, they are small and suits well this portable design but from my book this are no no and avoid them even on the power supply circuit. I'm suggesting you try other types like polyester, poliprop or styrene... specially this last two
Cheers, Tony
Yes, it's an open connection all right and we should try to repair it as dixter sez by adding a piece of lead on the vias and solder and see if this solves the open connection, or the lead beneath the board as someone else suggested. No big deal anyway.
On another subject altogether, it really calls my attention that you guys are using ceramic caps on a feedback path, of course, they are small and suits well this portable design but from my book this are no no and avoid them even on the power supply circuit. I'm suggesting you try other types like polyester, poliprop or styrene... specially this last two
Cheers, Tony
If you think those comments settles the issue be my guest... other seasoned fellows and designers will digress, and yes mostly is a subjective issue if you will.
Blue Rob will be building his with styrene caps at the feedback. BTW, we briefly tested the AD712 at U1 with very good results.
Cheers, Tony
Just curious - which caps are you replacing, C19 and C20 or also C11 and C12? What part numbers have you picked for the replacements? Thanks.
The AD712 does look like an interesting part! The gain/phase plot still has a break at 100kHz like the NJM2068 (good audio chips). The phase plot is more linear. And a thousand times less input bias current and input offset current being FET vs. bipolar (vs. the NJM chip)! The noise specs are better, too. The quiescent current is no higher and the THD is 10x better, although the O2 output stage would likely cover up any improvement there.
Hmmm.... I'm going to have to order some of those in and give the chip a try.
http://search.digikey.com/us/en/products/AD712JNZ/AD712JNZ-ND/751006
Last edited:
agdr
AD712 might help when using a hi z source since it will load it less but some of the benefits could be lost at the output as you say. I understand U1 is the voltage gain amp or front end if you will so I suspect it could be better but we have made no comparos yet. Being using Sennheiser's ref for the brief auditions since the Etymotic's are not at hand.
Yes, only C19/20 for the moment.
Cheers, Tony
AD712 might help when using a hi z source since it will load it less but some of the benefits could be lost at the output as you say. I understand U1 is the voltage gain amp or front end if you will so I suspect it could be better but we have made no comparos yet. Being using Sennheiser's ref for the brief auditions since the Etymotic's are not at hand.
Yes, only C19/20 for the moment.
Cheers, Tony
Did you read the documents posted in the comment that sgrossklass linked to? They indicated that the harmonics of the ceramic caps that are used in the O2 are only bested by polystyrene by a few dB, and that's down in the -120dB territory.
I do like your DIY approach and willingness to fiddle around with things and try out improvements. Let us all know how it turns out. If you can get it measured on a scope, that'd be awesome!
The caps that Rob used are from my box of parts that I have accumulated during the years.
Looking at Mouser inventories they have this axial one at 0.35$ ea.
23PS122 Xicon Film Capacitors - Additional Technologies
if you look other suppliers they could have radial ones which might be better for the available space and lead footprint. Beware tho from non branded parts that could have lousy lead material
Cheers, Tony
Looking at Mouser inventories they have this axial one at 0.35$ ea.
23PS122 Xicon Film Capacitors - Additional Technologies
if you look other suppliers they could have radial ones which might be better for the available space and lead footprint. Beware tho from non branded parts that could have lousy lead material
Cheers, Tony