Okay, just powered up my board and followed the instructions here
NwAvGuy: O2 Details
"Briefly Check The Supply Voltages" was all right. But after installing U2 I got stuck. The above mentioned test was still fine as before but measuring from negative of BT1 to pin 4 of empty U4 socket my DMM showed nothing. Well, it displayed a minus sign but not -11.8 as it should be it just said -0.00 but if I switch the probes (red on minus battery and black in the socket) i get 25V as on the battery tests above. And measuring between BT1 minus and pin 8 of U4 shows 25V and not 11.8.
My battery tests also showed 25V. I have a really cheapish DMM so I was thinking maybe it can't measure lower than 25V and no minus voltages?
And by the way, I've cut the ground traces at J2 to prepare for installing the ODAC. Maybe this is why? I don't know...
NwAvGuy: O2 Details
"Briefly Check The Supply Voltages" was all right. But after installing U2 I got stuck. The above mentioned test was still fine as before but measuring from negative of BT1 to pin 4 of empty U4 socket my DMM showed nothing. Well, it displayed a minus sign but not -11.8 as it should be it just said -0.00 but if I switch the probes (red on minus battery and black in the socket) i get 25V as on the battery tests above. And measuring between BT1 minus and pin 8 of U4 shows 25V and not 11.8.
My battery tests also showed 25V. I have a really cheapish DMM so I was thinking maybe it can't measure lower than 25V and no minus voltages?
And by the way, I've cut the ground traces at J2 to prepare for installing the ODAC. Maybe this is why? I don't know...
@Rullknufs - It is unlikely that the fault is due to DMM or you cutting the ground traces. The 7812 and 7912 are normally robust enough to take some beating unless you have switched their positions (a mistake I did on my first build). My current theory is that the MOSFETs might be the issue. These creatures are very delicate and need to be handled very very carefully.
The problem was that I had my DMM on AC instead of DC. But I also borrowed a high-end DMM from a friend and measured everything and everything was as it should be. Thank god!
So, now I've mounted it in my case (no pins touching the case, triple-checked that!) but I haven't mounted that ground wire I've seen some people do. And since I've cut the ground traces, is it ok to use it even though I haven't installed my ODAC yet?
Can I run it as O2 only until I mount the ODAC?
So, now I've mounted it in my case (no pins touching the case, triple-checked that!) but I haven't mounted that ground wire I've seen some people do. And since I've cut the ground traces, is it ok to use it even though I haven't installed my ODAC yet?
Can I run it as O2 only until I mount the ODAC?
RullKnufs : You should connect the ground wire to the chassis, but not connecting the ODAC and the Amp should be ok
Should I solder it to the chassis, does that work? I am pretty new to this.
In that case I'll try it tomorrow, using some old Steelseries Siberia headphones and not my Denon AH-D2000 for the first try
Okay, so I can just take a piece of wire, solder it to the ground pin on the board and then drag it under the whole PCB to the back of the board. Then scrape some paint of the back plate (is it ok to use the back plate or does it have to be the chassis?) and then squeeze it tight? Or how do you mean? I have maybe a millimeter clearance at either the front or the back of the board in the chassis so I could put a piece of cable there I think.
@Rullknufs : Here is what I did - Soldered a wire to the ground ran it under the board. On the other end of the wire I removed the insulation for about 6 mm divided the copper conductors into 2 like 2 braids and put it in-between the back plate and the chassis in such a way that one of the screws runs in-between them. I also made sure that the strands are spread so that the back plate doesn't show a bulge.
Hi folks,
I've been lurking for quite some time now, slowly munching the entire thread during the past months. A few days ago I received my O2 kit, and last night finally found the time to build the thing.
Notoriously goofing up, I was quite surprised that the resistance measurements turned out fine and swiftly went on with the DC voltage part as of the O2 Details: initial testing section. The supply voltage check across the battery terminals went fine, close to 24V total and an almost perfect 50/50 split across each. All sockets still unpopulated I reached for the power switch in order to retry those two checks. Lo and behold, LED D7 surprise - it lit up for like 0.1s, said "tic" and went back to black.
After a few minutes for the oh-no moment to pass, off to the O2 Details: troubleshooting section it was. Went through the power supply measurements step by step, with the following results:
power switch OFF:
D3: +23.18V
D4: -23.19V
D1: +11.96V
D5: -11.26V
BAT1: +11.95V
BAT2: -11.26V
LED7: 12.91V across
power switch ON:
D3: +23.08V
D4: -23.16V
D1: +11.96V
D5: -11.24V
BAT1: +11.79V
BAT2: -14.34V
U2.2: -12.96V
R5: +11.79V
R9: -14.29V
U2.3: +-0.005V bouncing
R6: +-0.005V bouncing
LED7: 14.35V across
With the LED in some dysfunctional state there's little wonder as to why the bipolar voltages are out of symmetry by some margin as long as the power switch is on, but I am puzzled as to the deviation on the path to the battery terminals. Apparently U6 is off, delivering a solid -11.25V where ~-11.8V should be. I did not backtrack the entire path to the power plug at the time, but the symmetry was almost perfect prior to the D7 accident.
Anyway, I took a maglite and a magnifying glass to the solder joints of U5 and U6 - nothing out of the ordinary, reflowed them anyway, to no avail. Subsequent measurements essentially gave the exact same results.
By the way, it seems U5 and U6 are a bit mixed up - in steps 6, 7 about the regulated supply, U6 is mentioned in the positive step / U5 in the negative. According to the BOM, U5 is the positive regulator, U6 the negative one, and I've been quite paranoid about those two, with good reason lately.
From what I gather, I definitely need a replacement D7 and perhaps U6 as well. Here is where my semi-educated a.k.a. half-arsed guess ends and I decide to humbly reach out for your collective knowledge and advice. Any idea as to what may have caused the LED to fail or the negative regulator to relax its output voltage? All clues highly appreciated. Not entirely sure the latter issue will go away once the former is fixed - clarification will yield eternal gratitude as well
Thank you very much for a great design, and a community to support it no less!
Best regards,
Chris
I've been lurking for quite some time now, slowly munching the entire thread during the past months. A few days ago I received my O2 kit, and last night finally found the time to build the thing.
Notoriously goofing up, I was quite surprised that the resistance measurements turned out fine and swiftly went on with the DC voltage part as of the O2 Details: initial testing section. The supply voltage check across the battery terminals went fine, close to 24V total and an almost perfect 50/50 split across each. All sockets still unpopulated I reached for the power switch in order to retry those two checks. Lo and behold, LED D7 surprise - it lit up for like 0.1s, said "tic" and went back to black.
After a few minutes for the oh-no moment to pass, off to the O2 Details: troubleshooting section it was. Went through the power supply measurements step by step, with the following results:
power switch OFF:
D3: +23.18V
D4: -23.19V
D1: +11.96V
D5: -11.26V
BAT1: +11.95V
BAT2: -11.26V
LED7: 12.91V across
power switch ON:
D3: +23.08V
D4: -23.16V
D1: +11.96V
D5: -11.24V
BAT1: +11.79V
BAT2: -14.34V
U2.2: -12.96V
R5: +11.79V
R9: -14.29V
U2.3: +-0.005V bouncing
R6: +-0.005V bouncing
LED7: 14.35V across
With the LED in some dysfunctional state there's little wonder as to why the bipolar voltages are out of symmetry by some margin as long as the power switch is on, but I am puzzled as to the deviation on the path to the battery terminals. Apparently U6 is off, delivering a solid -11.25V where ~-11.8V should be. I did not backtrack the entire path to the power plug at the time, but the symmetry was almost perfect prior to the D7 accident.
Anyway, I took a maglite and a magnifying glass to the solder joints of U5 and U6 - nothing out of the ordinary, reflowed them anyway, to no avail. Subsequent measurements essentially gave the exact same results.
By the way, it seems U5 and U6 are a bit mixed up - in steps 6, 7 about the regulated supply, U6 is mentioned in the positive step / U5 in the negative. According to the BOM, U5 is the positive regulator, U6 the negative one, and I've been quite paranoid about those two, with good reason lately.
From what I gather, I definitely need a replacement D7 and perhaps U6 as well. Here is where my semi-educated a.k.a. half-arsed guess ends and I decide to humbly reach out for your collective knowledge and advice. Any idea as to what may have caused the LED to fail or the negative regulator to relax its output voltage? All clues highly appreciated. Not entirely sure the latter issue will go away once the former is fixed - clarification will yield eternal gratitude as well
Thank you very much for a great design, and a community to support it no less!
Best regards,
Chris
Last edited:
Yours would not be the first bad LED.
See:http://www.diyaudio.com/forums/head...headphone-amp-diy-project-22.html#post2803953
and the result
http://www.diyaudio.com/forums/head...headphone-amp-diy-project-23.html#post2805992
Sounds like your LED may be open (dead). Pull U2 out of the socket and measure the resistance across R6 to see if it is around 40.2k.
I would also suggest re-measuring that full set of "power switch on" voltage readings with U2 out of the socket (and headphones not plugged in), and measure the voltage between U2 socket pins 4 and 8. If you get around 18Vdc - 24Vdc between U2 socket pins 4 and 8, and R6 is 40.2K, then the LED has to light if it is good, in the right way, and the solder joints are good with no shorts. If U2 is shorted internally that could also do the LED in, which is why the measurements are with U2 out of the socket. The LED should have around 3Vdc across it when working properly, with the rest (24Vdc - 3Vdc = 21Vdc) across R6.
Last edited:
adgr,
thanks a lot - I need to get used to the forum's search engine, trying to look up posts I have read weeks ago does not quite work out for me
Seems like the LED really is fried. R6 is 40.3K with U2 out of the socket and ~24V between its pins 4 and 8. Still I get 11.4V across the LED and 12.48V across R6 when the power switch is on. LED orientation should be alright, the package is hard to mount the wrong way and dmm probe polarity matching the pcb drawing yields a positive voltage.
Now I get roughly +12.5V vs. -11.5V (minus incremental voltage drop) across the power supply path, U2 pin 2 is at -8.5V, almost as it should be.
I did the initial testing with all sockets unpopulated, though leaving out gain resistors and U2 should not have such a devastating effect afaics. In an irritated attempt I inserted U2 and tried again... At least I had no opamps blow up on me.
One thing suddenly dawns on me - there's a double via close to C9, somewhere I've read it would be good to fill those with solder. Seems plausible considering the track width. Where there's a via, there's a way, so I did the same to some others, taking good care not to short anything out, esp. around that gain switch corner leg. Could I have wrecked anything by doing so? I was pretty careful applying heat so I doubt the pcb is toast. Double-checked for solder bridges or other unwanted contacts w/ flashlight and magnifying glass once again, found nothing.
I really wonder where the 11.4V across the LED come from. Judging by the numbers, LED voltage is close to the negative rail whereas R6 voltage is close to the positive. Except I'd rather not want to split the overall voltage in half but 7:1 as you outlined.
Could any build mistake sort of tie D7 and R6 to the corresponding voltage rails? I'd rather understand what went wrong before I mindlessly kill a couple more LEDs by trial and error.
Thanks again for pointing me in a sane direction!
thanks a lot - I need to get used to the forum's search engine, trying to look up posts I have read weeks ago does not quite work out for me
Seems like the LED really is fried. R6 is 40.3K with U2 out of the socket and ~24V between its pins 4 and 8. Still I get 11.4V across the LED and 12.48V across R6 when the power switch is on. LED orientation should be alright, the package is hard to mount the wrong way and dmm probe polarity matching the pcb drawing yields a positive voltage.
Now I get roughly +12.5V vs. -11.5V (minus incremental voltage drop) across the power supply path, U2 pin 2 is at -8.5V, almost as it should be.
I did the initial testing with all sockets unpopulated, though leaving out gain resistors and U2 should not have such a devastating effect afaics. In an irritated attempt I inserted U2 and tried again... At least I had no opamps blow up on me.
One thing suddenly dawns on me - there's a double via close to C9, somewhere I've read it would be good to fill those with solder. Seems plausible considering the track width. Where there's a via, there's a way, so I did the same to some others, taking good care not to short anything out, esp. around that gain switch corner leg. Could I have wrecked anything by doing so? I was pretty careful applying heat so I doubt the pcb is toast. Double-checked for solder bridges or other unwanted contacts w/ flashlight and magnifying glass once again, found nothing.
I really wonder where the 11.4V across the LED come from. Judging by the numbers, LED voltage is close to the negative rail whereas R6 voltage is close to the positive. Except I'd rather not want to split the overall voltage in half but 7:1 as you outlined.
Could any build mistake sort of tie D7 and R6 to the corresponding voltage rails? I'd rather understand what went wrong before I mindlessly kill a couple more LEDs by trial and error.
Thanks again for pointing me in a sane direction!
Yep, that sounds like a defective LED. It is producing 11.4Vdc across it (they can do that when bad) instead of the 3Vdc it should have. So you wind up with the 11.4Vdc (LED) + 12.48Vdc (R6) = 23.88Vdc, essentially equal to your [ (V+) - (V-) ] supply difference (24Vdc) across those parts. So all is good - except your LED.
The 24Vdc you measured from U2 4 to 8 says your ac power supply is working OK. If you pull out the AC adaptor plug, with the batteries charged, you should get somewhere around 18.8Vdc from pins 4 to 8, plus or minus a volt or so, depending upon how charged your batteries are.I would suggest getting your LED working first, since that little circuit is fairly easy to test and debug, and a bunch of parts make use of the LED's reference voltage. You may find more things start working when the LED is fixed. In those links I posted Rocket Scientist eventually figured out that the poster's other problem, of the NJM4556 exploding, was due to setting the PC board on a scrap wire end from the build which shorted some things.
Not related to the defective LED.If you are going to have to order another LED, given that shipping is >> than parts cost, I'd suggest getting a couple of LEDs, and another of each of the op amps including U2, and in your case another negative regulator chip, and even a pair of mosfets, to have on hand for further troubleshooting and/or in case something gets accidentally fried with a slipped DMM probe (happens all the time) during troubleshooting. Those parts are all fairly cheap and are the most static sensitive.
Last edited:
The battery part I have not tried yet, but the pair was in a separate charger over night. Just to confirm: they are at ~9.6V each hence 19.22V total across the battery terminals, 19.13V across U2 pins 4, 8 and again a split of ~12.5V across D7 vs. ~6.5V across R6. Greedy little voltage reference. Never knew an LED could do like an investment banker. There's a lot to learn for a newbie.Yep, that sounds like a defective LED. It is producing 11.4Vdc across it (they can do that when bad) instead of the 3Vdc it should have. So you wind up with the 11.4Vdc (LED) + 12.48Vdc (R6) = 23.88Vdc, essentially equal to your [ (V+) - (V-) ] supply difference (24Vdc) across those parts. So all is good - except your LED.
Yes you're right, I remember now. A scrap yard makes for a different kind of proving grounds entirely...I would suggest getting your LED working first, since that little circuit is fairly easy to test and debug, and a bunch of parts make use of the LED's reference voltage. You may find more things start working when the LED is fixed. In those links I posted Rocket Scientist eventually figured out that the poster's other problem, of the NJM4556 exploding, was due to setting the PC board on a scrap wire end from the build which shorted some things.
Definitely so, if not in the O2 debugging session I can use a few regulators for other projects as well. Got a plastic tube full of OPA2134s though, so the DIP8 spares will be purely O2 insurance.
Thanks again! I guess that concludes my fiddling for today. Will order a set of replacements and keep you posted how things go.
Best regards,
Chris
When LEDs go bad like that all bets are off on how they will behave, with a complete short or total open possible at the two extremes, but I'll bet yours is acting more like a 11.4V zener diode than a resistor. In other words, instead of splitting the battery supply voltage like two resistors in series would do. that bad LED may always maintain 11.4V across it like a zener, with the difference between the power supply voltage and 11.4Vdc being left across the resistor.
Just finished assembling my O2 and ODAC into the case. There is one problem though...
I get sound, nothing is smoking or burning or so. But there's something seems to be a bit loose. Was playing some music and though the volum was very low, tried the gain switch too but didn't help much. Then I was about to move the amplifier and then kaboom in my headphones and the volume was back to normal (I had it almost all the way up). Needless to say I almost **** my pants.
Turned the volume back down to normal listening levels and the volume was extremely low, could barely hear something was playing and when I touched either the USB cable, headphone cable or DC input cable the volume was like jumping from extremely low to normal levels.
What could the problem be?
I get sound, nothing is smoking or burning or so. But there's something seems to be a bit loose. Was playing some music and though the volum was very low, tried the gain switch too but didn't help much. Then I was about to move the amplifier and then kaboom in my headphones and the volume was back to normal (I had it almost all the way up). Needless to say I almost **** my pants.
Turned the volume back down to normal listening levels and the volume was extremely low, could barely hear something was playing and when I touched either the USB cable, headphone cable or DC input cable the volume was like jumping from extremely low to normal levels.
What could the problem be?
Alright, might not be a problem at all... Switched to the USB cable that followed with the ODAC and now it works fine. I had used a longer cable I found in a box so I could use the back USB ports of my computer. Now there's no problem at all, I can lightly bend all the cables and rock them a little and it sound just like normal. Seems I'll have some tough work trying to find a good USB cable...
I do have USB-ports on my Dell U2410 monitor, could I use those ports? I haven't got any USB cable yet between monitor and computer but if I got one, could I use those ports without any loss of audio quality?
I do have USB-ports on my Dell U2410 monitor, could I use those ports? I haven't got any USB cable yet between monitor and computer but if I got one, could I use those ports without any loss of audio quality?