I thought the capacitance of Y5V dielectric capacitors depended mainly on temperature rather than voltage?
He specified a 50V unit, so by this chart anyway the capacitance at 12V would be down about 8% for a Y5V
Fileelta-Cap-versus-Spannung-engl.png - Wikipedia, the free encyclopedia
But those MLCCs have improved so much over the last year or so that there really isn't any point in having to buy a Y5V anymore. A 2.2uF 50V X7R at Mouser (810-FK26X7R1H225) is a couple of cents cheaper at $0.42USD and thinner (which is good for this app given where the C7 is located).
A suggestion I made way back when was probably responsible for his change in C6/C7 values. The datasheet for the negative regulator suggests the 1uF or greater for oscillation prevention, Co on the first page:
http://www.onsemi.com/pub_link/Collateral/MC7900-D.PDF
Last edited:
Hi all. I built another O2 and it worked great. However, when I soldered a lead to the input pin 1 and then put the cover on (grounding to case like NWAVGUY said to do) I got nothing from the amp. I unplugged everything and looked. Everything looked fine with the ground lead to the case. I slid the board out and one of the batteries was burning hot.......almost melted. Visibly distorted. The other battery was fine. I clipped the ground lead off, took the batteries out and reassembled the board into case. Works perfectly. I checked all measurement again and all seems fine. I am not using batteries right now. What happened?
It is possibly the battery terminals on the PCB shorting against the case. That malfunction explains the battery behavior. Electrician's tape in a strategic location should solve that problem, or you'll need to resolder. Have a good look through the back of the unit before attaching the front and back panels... just the PCB. Reassemble with the ground to case after being satisfied there are no other parts in contact with the case.
One of your battery terminal (very unlikely) or a component pin is protruding too much (most likely) touching the case causing short circuit when you ground the input to case. Trim all component pins to bare minimum using a sharp nipper/cutter. Look from the sides for any visible signs. Also put some clear thick anti-static plastic(used in static sensitive electronic comp. packing) between pcb & case to prevent any shorting. Alt. you can use multiple layers of insulation tape and/or thick cardboard.
Caution-Do not use PC motherboard type anti-static packaging material. Do not use batteries until the issue is sorted out, use AC power source only.
Thanks availlyrics. I definitely want to get to the bottom of it. It is working great without batteries (I threw both of them away at work, we have a recycling company pick up batteries.) I will open her back up and see what I can find. Hopefully I will find something.
Thanks everyone for all your help. I will let you know what I find.
Thanks everyone for all your help. I will let you know what I find.
I could agree with that except the unit works fine with AC only and no batteries. That severely limits possible component leads in my view. YMMV.
No voltage on rails
I built an O2 that worked for a while. I handled the board and it stopped working, so I figured it was the MOSFETS. I ordered two of each plus replacements for all the chips, just in case. When they came in, I swapped out both, but no change. I pulled out the meter and went through the voltage checks and found I had nothing on the rails (pin4 and 8 of U4/ opposing sides of C10, but before the MOSFETS was ok. I tried the other two MOSFETS and changed out all the other chips just for good measure. Its still not passing audio. I'm not sure how to check if the MOSFETS are good, but I get the same resistance between the middle and right legs of both of each set. I still can't get voltage on the rails, but there is 9 volts on each outside leg of the big one and -9 volts on each outside leg of the small one. Only mV on the center pin of each that seems to vari, but is never over 1 volt. Any help is appreciated.
I built an O2 that worked for a while. I handled the board and it stopped working, so I figured it was the MOSFETS. I ordered two of each plus replacements for all the chips, just in case. When they came in, I swapped out both, but no change. I pulled out the meter and went through the voltage checks and found I had nothing on the rails (pin4 and 8 of U4/ opposing sides of C10, but before the MOSFETS was ok. I tried the other two MOSFETS and changed out all the other chips just for good measure. Its still not passing audio. I'm not sure how to check if the MOSFETS are good, but I get the same resistance between the middle and right legs of both of each set. I still can't get voltage on the rails, but there is 9 volts on each outside leg of the big one and -9 volts on each outside leg of the small one. Only mV on the center pin of each that seems to vari, but is never over 1 volt. Any help is appreciated.
Does the LED still light when the power switch is on? Also take a voltage reading between pins 4 of U2 and ground (the metal shell of the gain switch), and between pin 8 of U2 and ground
You have probably already done this, but double check against this photo on NwAvGuy's website that you have every chip and both mosfets back in the correct orientation, the "dot" in the corner on the chips and the direction of the metal back (plastic tab at the top on the big one) on the mosfets. Double check the number on each chip vs. the photo too.
http://lh4.ggpht.com/-LCfqQew0VFo/TnIzAIYxFKI/AAAAAAAABaw/QRdrYRjUY_M/s1600-h/o2%252520final%252520pcb%252520large%25255B4%25255D.jpg
Last edited:
The LED is on. I got -9.06V on 4 and 9.09V on 8 of U2.
I'm pretty sure I put them in correctly and it was working, but I did check diode orientation, pin 1 to the dots on the chips, and orientation for the MOSFETS and voltage regulators(which I don't think are part of the PS when using battery power). I only have U2 in now because I have troubleshot it to the PS.
I think there must be something in the circuit that is smoking these MOSFETS. I checked the two I bought in circuit with the middle leg lifted and they all read 0V on the middle leg, but 9V (Q1)and -9V (Q2) on the other two legs.
I'm pretty sure I put them in correctly and it was working, but I did check diode orientation, pin 1 to the dots on the chips, and orientation for the MOSFETS and voltage regulators(which I don't think are part of the PS when using battery power). I only have U2 in now because I have troubleshot it to the PS.
I think there must be something in the circuit that is smoking these MOSFETS. I checked the two I bought in circuit with the middle leg lifted and they all read 0V on the middle leg, but 9V (Q1)and -9V (Q2) on the other two legs.
Good about the LED. He uses that LED voltage as a reference for the U2 comparator. If the LED is off nothing else will work.
Something is wrong in your U2 comparator circuit that drives the mosfets. In the case of Q2 next to the battery the -9V on the source (furthest end from the battery) is correct, but the other end (gate) should be at +9V from the U2 circuit to turn that mosfet on. Right now U2 is never telling the mosfets to turn on.
Did you replace U2 along with the other chips? Go ahead and measure the voltage on the rest of the pins on U2 to ground (metal shell of gain switch). That should show up the problem. Other than the LED and U2 everything else in that circuit is a capacitor or resistor that would be immune to being static zapped by touching them.
Also - try reheating every solder joint for a second or two on U2 and the surrounding circuitry, even if they look perfectly good. You might have had a bad solder joint that moved when you touched it. Take a really close look at the solder joints on the mosfets too under a magnifying glass. I know that at least for me and my eyesight it is nearly impossible to see a solder bridge on the regulators or mosfets without a magnifier.
Last edited:
I did replace U2 as well as U1, U3, and U4. They are all socketed and I figured I would get spares since I already paid for shipping and they were cheap. I took measurements off of U2 and then I swapped for the original U2 and found they read the same. Here are the readings I got:
Pin 1. 8.55 V
Pin 2. -6.86 V
Pin 3. -6.13 V
Pin 4. -8.77 V
pin 5. -6.13 V
Pin 6. 8.55 V
Pin 7. -8.77 V
Pin 8. 8.77 V
I'm not sure what I am supposed to be seeing since I didn't really look at this part of the schematic. I don't know if I would know even if I did
I did re-solder U2 and all the surrounding components, but no change.
Pin 1. 8.55 V
Pin 2. -6.86 V
Pin 3. -6.13 V
Pin 4. -8.77 V
pin 5. -6.13 V
Pin 6. 8.55 V
Pin 7. -8.77 V
Pin 8. 8.77 V
I'm not sure what I am supposed to be seeing since I didn't really look at this part of the schematic. I don't know if I would know even if I did
I did re-solder U2 and all the surrounding components, but no change.
Well this is an interesting one. Your LED is bad, but in an odd way that hasn't popped up before in the thread. Instead of going out entirely or shorting, the LED is still lighting but its forward voltage is a volt higher than it should be. Your LED probably was what got static zapped when you handled the board - it is right there on the board edge. Quite a few people have posted with bad LEDs, but until now they have all been open or shorted (and sometimes loudly pop when they short, people say).
Referenced to the negative rail for the comparator, your LED is dropping [(pin 3 or pin 5) - pin4] = (-6.13Vdc) - (-8.77Vdc) = 2.64Vdc. It should normally only be dropping around 1.78Vdc. See my measurements on it here:
O2 amp CRC, diode, cap, and heatsink mods - Page 11 - diyAudio (photo at the bottom of that post)
So given that extra volt of drop across the LED, all the rest of your voltage measurements are just the circuit doing what it is being told to do.
Replace your LED and you should be back in business.
Last edited:
AC power offers at least some sort of basic short circuit protection with the help of regulators v/s batteries with high current capabilities & no protection. urshurak776's problem cropped up only when he grounded the i/p to case indicating higher chances of component pin touching case as against battery connectors having standard dimensions shorting to case.
Last edited:
Wow, your a genius! I never would have guessed the LED was dropping too much voltage and preventing U2 from turning on. I'm afraid this may have been a self-inflicted wound. This would be a caveat to others who don't follow the instructions like me. Let me explain. I used a blue LED instead of the red because all the red LEDs I have didn't light very bright, but the blue one did, so I used it.(I didn't think about it, but this is dumb since the brighter the LED, the less battery life). I think the batteries drained enough that they couldn't overcome the threshold and made it appear as I damaged something. So, let me say that I'm a dumbass and thank you for bailing me out.
This would be a caveat to others who don't follow the instructions like me. Let me explain. I used a blue LED instead of the red because all the red LEDs I have didn't light very bright, but the blue one did, so I used it.(I didn't think about it, but this is dumb since the brighter the LED, the less battery life). I think the batteries drained enough that they couldn't overcome the threshold and made it appear as I damaged something. So, let me say that I'm a dumbass and thank you for bailing me out.
