Heh, heh, I think that my burn-in method will be to play music through it for a few hours every day.
OK. Down to business.
The last of the old caps.
C141: 22µF/25v = 23.04µF, 0.90 ohms, 1.1% vLoss. Replaced with FM 50v.
C142: 22µF/25v = 22.67µF, 0.92 ohms, 1.1%.
C154: 10µF/25v = 10.32µF, 1.8 ohms, 1.7%. Replaced with MKS2 50v.
C157: 10µF/25v = 10.02µF, 1.8 ohms, 1.6%.
C103: 10µF/25v = 10.22µF, 2.0 ohms, 2.5%.
C106: 10µF/25v = 10.07µF, 2.1 ohms, 2.6%.
C109: 10µF/25v = 10.18µF, 2.0 ohms, 2.6%.
C112: 10µF/25v = 10.19µF, 2.1 ohms, 2.6%.
C113: 10µF/25v = 10.15µF, 2.2 ohms, 2.7%.
C116: 10µF/25v = 10.06µF, 2.8 ohms, 2.8%.
C119: 10µF/25v = 10.04µF, 2.1 ohms, 2.7%.
C122: 10µF/25v = 10.24µF, 2.1 ohms, 2.8%.
C125: 10µF/25v = 10.11µF, 2.1 ohms, 2.6%.
C128: 10µF/25v = 10.30µF, 2.1 ohms, 2.4%.
C247: 1µF/50v = 0.952µF, 3.2 ohms, 0.9%.
C248: 1µF/50v = 0.961µF, 3.3 ohms, 1.0%.
C258: 1µF/50v = 0.975µF, 3.5 ohms, 1.1%.
C259: 1µF/50v = 0.965µF, 3.3 ohms, 1.0%.
C151: 1µF/50v = 0.965µF, 4.4 ohms, 1.7%.
C152: 1µF/50v = 0.951µF, 4.4 ohms, 1.6%.
C153: 1µF/50v = 0.933µF, 6.5 ohms, 3.0%.
C154: 1µF/50v = 0.962µF, 4.0 ohms, 1.5%.
That last cap, C154 is the 2nd cap that I found that the silk screen is backwards to the way the cap was OEM installed. The board layout drawing shows the minus the way the cap was installed.
So all of this listing the old cap values was probably completely boring to anyone but Karl and I but I found it interesting to see how the old caps had held up. And most of them had held up well. The Elna Duorex were all pretty good. The Elna brown and Nichicon Muse were a crapshoot, mostly good but some sizes had dropped as much as 20%. The Nichicon 1µF BP all had pretty high ESR. It looks like C153 was on it's way towards failing.
This is the first time that I ever thought of using the rather obvious tactic of using the cap date codes to pin the age of an amp. The newest cap was produced in week 18 of 1993. And we can see that Yamaha had no problem using 2 year old caps because a fair number of the Nichicons were from early '91. Most of the Elna's were made in '92, a few in '93.
As I was working on this board I freaked out when I noticed a scratched through trace off of one end of R222 in the Phono section. I thought at first that I had somehow done it and was ready to hit myself. But I realized that it was in fact a deliberate cut in the trace about 1mm wide. After checking the board layout one can see that the trace is not supposed to be joined at that spot. So this was a mistake in the etching photo-tool that was corrected by scratching the trace. There is a second small cut trace up near R232 or so (I didn't pin that one down exactly).
This stuff leads me to think that my pre-amp is a very early production model. I did professional production etching for 12 years (photo-chemical machining). The normal way to correct such a photo-tool error is to scratch out (or touch up) the feature on the photo-tool ink as a temp measure and then make a corrected one for mass production. This mistake wasn't caught until after a fair number of boards had been produced. Add to that how many caps and resistors have been tacked onto the trace side of the board, as well as a number of by-pass film caps added in parallel to an elco by opening up the through-holes and shoving them through together.
It would be interesting for 808_state and devnull etc to post some pics and cap date codes if you ever do a recap.
So, to do a final re-cap of the re-capping, I counted 98 elcos in this pre-amp. I was able to replace 43 of them with Wima MKS2 film caps. That's something over 40% I think and that makes me happy. There were another 4 or so 10µF that I would have liked to swap with film but there was just no way that they would fit because of space. And when you see the photos one can see that I barely got a lot of the 10µF Wima's in there.
I made a number of pragmatic choices for the new caps based on how the new ones measured. Some of my early choices got set aside because I didn't like the way that they measured. And ALL of the 470µF that I bought measured low. FC and FM in 25v and 35v, KL in 25v and 35v, KZ in 25v and 50v. I don't know what it is about that value but the highest new cap measured around 460µF and lowest around 440µF, average about 450µF.
Oh, btw, the 2x Mundorf's measured 6612µF and 6656µF. So, well on the low side of nominal but still way above the old Nichicons.
OK. enough babble. Photos in the next post...
Cheers,
James