[zxc]

Ok, I have a pretty much standard pimeta with:
2x637 and 1x627 OpAmps
2xBuffers on L/R, single buffer on Ground.
Gain 11 (same R values on Tangent's site)
2xWima 6.8uF
4xElna Cerafines (470uF, 35V)
100 Ohm resistors on outputs (for getting rid of background hiss)
Driven with 18 Watt 24V Elpac (regulated but not linear)
Used with Senn HD600

Altough the Pimeta got really loud on my Sony CD changer with volume knob
on 9 o'clock, I needed to turn it way up for descent sound on my computer
with M-Audio Audiophile sound card.

I was watching a movie on my PC with volume allmost maxed and after an hour
suddenly the movie soundtrack was replaced with an ugly 'tapping sound' which
is in sync with the flickering led.

Something might went hot inside but not totally sure. I pulled the opamps
out and when powered on, the led still flickers. No visible signs of
burned or deformed stuff inside... I checked the Power Supply voltages,
Elpac seems fine...

I figure something like this had to happen to someone else at some
step... Any idea what the cause might be, and how to debug this??? Please
tell me what to do??? I really liked the Pimeta...

 

[Edwood]

Short?

DO you hear loud popping noises when plugging in headphones?

That usually means a short in the ground.

-Ed

 

[zxc]

Thanks Ed! I did a continuity check on the outputs of Elna Caps, as each should be connected to either V- and Ground, or V+ and Ground. And I get a sound on all!!! So I guess V+, V- and ground are all shorted!!! (Am I wrong?)

What is likely the reason for this? Is it likely that the TLE-2426 might have blown up because I used it at max volume for an hour? Or is it the caps? How can I check? Any suggestions?

 

[zxc]

I did a search on "pimeta" and "problems" but none of them fits in this case. Though I got the impression that TLE is the vulnerable element in these type of situations... Any ideas or suggestions (before I go ahead and rip that TLE off of its place... yes, me too... No desoldering skills what so ever...)?

 

[dsavitsk]

Quote:

Originally Posted by zxc Any ideas or suggestions (before I go ahead and rip that TLE off of its place... ?

Only that you get some desoldering braid as you are liable to destroy the board if you rip it off.

 

[zxc]

Well, I have both the desoldering braid and the solder sucking tool. About a month ago, I just couldn't desolder the JFETs with those. But carefully pulling them by twisting them did the job, of-course, destroying the Jfets in the process... I had to do the same to a Panasonic Pot; and seeing the innards of the Panasonic pot, it is a really good pot. So it seems that Pimeta board is one though PCB

But before I do that; I will wait for some advice from the wise people who read this thread.

 

[Pars]

I would doubt that it is the caps. Your "continuity" checker is only beeping because when you ohm out a cap, it initially shows as a short and then charges up, winding up (ideally) as infinite resistance.

I would say the TLE is the most likely suspect at this time. Can you take voltage readings of V+ and V- in relation to ground (never built a Pimeta, so I'm not sure whether you would use input or output as your ground for this). If so, and you get appropriate voltages, are they stable? The LED flickering would seem to indicate that they have alot of ripple on them.

Beyond the TLE, perhaps buffer(s)?

 

[zxc]

Only my opamps were socketed; so I removed the OpAmps but had to keep the buffers in place.

I measured the voltage between all three legs of the TLE.

(V+)-(Ground) and (V-)-(Ground) is oscillating between 0-0.45V and 0-0.6V respectively.

I tried the continuity check on the Solen 10uF Caps that I had lying around and there was no sound. I also went ahead and measured the resistance between the inputs of the Elna Cerafine's in the amp, and they are 5 ohms on one and 12 ohms on the other. The values seemed steady.

 

[Pars]

Good reason to socket the buffers also. Well, it sounds like you have something either shorting the rails (blown buffer, bad cap), or the TLE is dead (that would be my guess still).

First thing I would do is pull the TLE (not literally!) and then recheck the rail to ground resistance. If you need to, add some solder to whatever you are going to pull, then use the solder wick. Or, just heat the leads in rotation, and try to work the component so you get one lead out, then the next, etc. Then use the solder wick to clean up the holes. You can pull the Cerafines if you like also, but the first thing I would yank is the TLE, and check the readings first before possibly ruining yet another component.

just my 2 cents,

Chris

 

[dsavitsk]

Quote:

Originally Posted by Pars Or, just heat the leads in rotation, and try to work the component so you get one lead out, then the next, etc.

I usually grab the lead with a needle nose pliar, heat the joint and pull and worry about cleaning solder up later. That is, you don't need to separate the lead from the board by cleaning the solder. ... I think that makes some sense?

-d

 

[dip16amp]

The left channel has two buffers and the right channel has two buffers so the return current path of those four buffers end up going through the one ground channel buffer. Seems like you should have four buffers on the ground channel to handle all that current.

 

[zxc]

Thanks Pars! I will definitely socket the buffers on my next pimeta. I took out the TLE easily by heating its three outputs simultaneously with the thick tip of the soldering iron. Shorting is gone; and LED does not flicker anymore when powered. So TLE was definitely faulty. I don't have any replacement TLE's so I can't try a new one and see if it was the only problem at the moment. I just hope the buffers are fine.

I also got two of the Cerafine's out and both seem fine. The resistance builds up and goes to infinity.

Quote:

The left channel has two buffers and the right channel has two buffers so the return current path of those four buffers end up going through the one ground channel buffer. Seems like you should have four buffers on the ground channel to handle all that current.

Tangent says (on his website) that he didn't notice any difference in sound by stacking buffers on ground channel; that's why I didn't stack them on Ground. But should I do it just for the sake of reliability? Any other ideas on how to avoid this happening again? Why did it happen in the first place?

 

[tangent]

Quote:

Originally Posted by dip16amp Seems like you should have four buffers on the ground channel to handle all that current.

All of what current?

Let's say zxc is trying to make himself deaf rightquick, and is putting 2V through Grados. That'll give you well over the 90dB maximum you should listen to, but it only causes 63mA of current to go through each channel. And this only happens when listening to sine waves, not music! A single buffer can easily handle 126mA.

The idea that the ground channel has to have some number of buffers (whether equal to the number in each audio channel, or twice, as some would say) is fallacious.

 

[motherone]

Did you try pulling your opamps and buffers, then powering your Pimeta up and checking all the voltages? This would at least let you check everything on the board without frying any (or anymore) silicon in the process. You may also want to use a resistor to load-test your elpac to make sure it's still working correctly. Usually if it's a cap problem, they ooze or explode (as I had a Panasonic FM 16v/2200uF do in my face the other night). Also, did you check to make sure that any of your internal wiring didn't break loose?

I know all-to-well about pulling stuff off the pimeta. I made a small wiring mistake and thought I had a bad jfet -- pulled them all only to discover the wiring mistake, then had to solder new ones in. I have the braid, a rat-shack desoldering iron, and a solder sucker -- it's almost impossible to get it out with any of these since the plated-through holes are so dang tight to the component pins. I always wind up having to refill the holes with solder so I can get a good flow, then suck all solder out in one big melted pile to get the hole cleared. Then a quick runover with the wick gets any left overs on the top or bottom.

 

[amb]

Quote:

Originally Posted by tangent ... The idea that the ground channel has to have some number of buffers (whether equal to the number in each audio channel, or twice, as some would say) is fallacious.

Given your line of reasoning, then there shouldn't be any benefit to stacking the buffers on the L and R channels either. Afterall, the ground channel is sourcing and sinking the sum of the return currents of the two channels. No?