[Takaji]

So after months of slacking off/school, I finally decided to finish up my PIMETA. To my disappointment, I turned it on for the first time today, did a few checks, then plugged in the source and headphones. And.... I can't hear anything.

To top it off, BUFG is getting extremely hot, so I immediately turned it off and tried the board without the buffers. Still, no sound.

My PIMETA includes the trickle charger circuit, and the crossfeed circuit. I tried the PIMETA without the crossfeed circuit and checked my wiring for the trickle charger, so I would assume that those circuits are both OK.

The LED turns on, however, so it's not completely dead. I noticed that the AD8620 OPAL/R chips were only getting very slightly warm - is that normal or should they be completely cool?

If anybody needs a specific photo, I can upload one.
Thanks very much!

 

[joneeboi]

A photo would be a good step.

Try to get an aerial view of the entire board, then a close-up of the ground section as well as the underside.

Have you run through Tangent's troubleshooting steps yet?

If you remove the buffers, you wouldn't get any sound anyway, unless you tapped the signal from the sockets. You would need to jumper all the pins 3 to pins 6 which would turn it into a 3-channel CMoy. I would say that the opamps don't get all that warm to the touch, even after running a couple hours. They wouldn't be cold, but they wouldn't be burning up, just slightly warm.

 

[synaesthetic]

My dead PIMETA is doing about the same thing. Powers on, LED comes on, but no sound comes out at all, and there's about 58mV of DC offset on the right channel output (left channel has 0mV offset).

I'm pretty sure I somehow cooked my little discrete buffers and that's the problem, either when soldering them together or by accidentally powering it on without 'phones connected. Maybe when I can find a few BUF634Ps I'll mess with it again. For now it's shelved, though.

 

[Takaji]

Quote:

Originally Posted by joneeboi /img/forum/go_quote.gif A photo would be a good step. Try to get an aerial view of the entire board, then a close-up of the ground section as well as the underside. Have you run through Tangent's troubleshooting steps yet? If you remove the buffers, you wouldn't get any sound anyway, unless you tapped the signal from the sockets. You would need to jumper all the pins 3 to pins 6 which would turn it into a 3-channel CMoy. I would say that the opamps don't get all that warm to the touch, even after running a couple hours. They wouldn't be cold, but they wouldn't be burning up, just slightly warm.

Photos available here: PIMETA Pics
I apologise for some shaky photos, please let me know if you guys need better ones.

I have not tried Tangent's steps yet, because I'm afraid to leave the amp on since that buffer is getting so hot.

I'm also wondering if I made the right choice with R11G; I made it 220 Ohms since I'm only using one buffer (Tangent's site says "If you're stacking your buffers, R11's value needs to be halved every time you double the number of buffers in order to maintain the same bandwidth setting. This means that the minimum R11 value should be 100 Ω with two stacked buffers, not 200 Ω.")

 

[joneeboi]

I must say, I'm impressed with your build quality. But first things first, why don't you clean up some of that flux underneath the board? Though it's not always necessary, unkempt flux can play tricks with your board. It seems quite dirty around the area where you soldered OPAG, so I suggest giving it a good cleaning.

My other guess is that perhaps you fried the AD8610 when you soldered it onto the board. Did you just run out of SOIC-to-DIP adapters? You put in the OPAG socket, but then soldered the chip onto the board.

Regarding R11G, you have the "correct" value at 220 Ohm. There's a range of accepted values here which the datasheet will specify. If you stacked your buffers, you would need 100 Ohms, but since you only used one, you're okay sitting at 220 Ohms. What kind of resistor did you use there? I notice that it's quite different from the V-Ds.

Overall, everything looks really good on your build, except for the messy flux. You may also want to ground your pot via the Philips screw on the back of the Alps Blue, unless you have another method of grounding it. That's a minor detail probably unrelated to the soundlessness of your amp, but keep it in mind.

To sum up, clean up the flux, and maybe you fried the AD8610.

 

[JamesL]

I find it really hard to fry opamps. In my first build, I put both my soic opamps with pin 1 facing the other way, and of course, I didn't have any braid, pump, flux, so I ended up having to heat each side a couple dozen times with my radioshack iron, picking away at it with a toothpick.
I'd say, if you haven't already, go back and check the values and polarity of all the components.. The build quality looks great, but it wouldn't hurt to double check the joints. the most top-right joint on the last picture, for one of the dip packages looks kind of odd to me.. maybe its just the picture.

 

[Pars]

What does the DC offset look like, as well as the IG to OG offset? I would agree, the symptoms would make me suspect a blown ground opamp. And the AD8620 is one of the very few opamps that I have blown up (I know you are using 8610).

 

[Takaji]

I will check all this, thank you everyone for your comments...
I have only one question before I begin: what's the best method of cleaning up flux?

Thanks!

 

[kurt6652]

I have used a toothbrush and rubbing alcohol, it seemed to work good.

 

[joneeboi]

Grab a toothbrush and the purest rubbing alcohol you can get. Submerge the amp in the alcohol and just brush like mad. It's a small board with a small amount of flux, so you only have to wash, dry and repeat a couple times. The guide I used can be found here: How to Clean Circuit Boards. There are other guides, that one's a great start.

 

[bhjazz]

I'll throw in an X2 for following Tangent's Troubleshooting Tutorial (TTT!). I think the actual link is here: Basic Troubleshooting for Headphone Amplifiers. Get through all of these because one will likely be the fix!

I didn't encounter any problems with my Pimeta build, but I walked through them anyway to make sure I wasn't missing something along the way. The overheated buffer is a pretty clear signal of something upstream that is obviously not quite right.

 

[tangent]

I believe you've soldered the op-amps on the adapters backwards. The beveled edge indicates the side with pin 1, which appears to be "down" in your pictures, not up toward pin 1 on the adapter. If you desolder the chips you will see a little yellow circle indicating pin 1 on the adapter. If you look on the beveled edge of the chip, you should have a secondary indicator -- a little divot or notch, usually -- nearest pin 1.

If this is what happened, OPALR is almost certainly dead. OPAG might have survived, since mirroring a single-channel op-amp doesn't swap the power pins. Whether a single-channel op-amp survives this depends on what it has connected to pin 8, and how long it is connected to reversed power.

 

[Takaji]

Quote:

Originally Posted by tangent /img/forum/go_quote.gif I believe you've soldered the op-amps on the adapters backwards. The beveled edge indicates the side with pin 1, which appears to be "down" in your pictures, not up toward pin 1 on the adapter. If you desolder the chips you will see a little yellow circle indicating pin 1 on the adapter. If you look on the beveled edge of the chip, you should have a secondary indicator -- a little divot or notch, usually -- nearest pin 1. If this is what happened, OPALR is almost certainly dead. OPAG might have survived, since mirroring a single-channel op-amp doesn't swap the power pins. Whether a single-channel op-amp survives this depends on what it has connected to pin 8, and how long it is connected to reversed power.

I don't think I've soldered them in wrong.. I checked where the beveled edge was on both sides of the adapter. For each chip I can see the beveled edge and the notch, and they are lying directly under the little yellow dot indicating pin 1.

As for using alcohol to remove the flux, I handed the board over to my aunt who works in a lab, and she was able to scrub the back gently with a plastic brush and 95% pure methanol-based alcohol. However, when I got it back from her the entire back of the board seems to be sticky! I'm not sure what caused this (it may be because she laid the board on top of a bag?) but I'm gonna ask her to clean it again with a toothbrush.

I'm wondering if I need to replace the OPAG chip as well...

Just did a quick check, I'm getting 14.1V through the OPAG, and 13.28V through the OPAL/R. No heating up on BUFG yet. I have determined however, that when I plug in a source and cheap ipod headphones that BUFG starts to heat up.