Yamaha A-500 and A-520 Thread

I finally got around to measuring the old caps removed from A-520 #1.

1000µF 16v Elna = 764µF, 752µF
1000µF 6.3v Rubycon = 844µF, 845µF
330µF Rubycon = 303µF, 304µF
100µF 16v Elna = 89µF, 89µF, 89µF, 91µF, 92µF, 90µF, 88µF, 91µF
22µF 16v Elna = 20µF
1µF 50v Rubycon = 0.89µF
0.22µF Elna = 0.24µF, 0.23µF

Again, the larger values have drifted a lot more than the smaller ones. I'm hoping that next month I can squeeze new filter caps into my budget.

Cheers,
James
 
I finally got around to measuring the old caps removed from A-520 #1.

1000µF 16v Elna = 764µF, 752µF
1000µF 6.3v Rubycon = 844µF, 845µF
330µF Rubycon = 303µF, 304µF
100µF 16v Elna = 89µF, 89µF, 89µF, 91µF, 92µF, 90µF, 88µF, 91µF
22µF 16v Elna = 20µF
1µF 50v Rubycon = 0.89µF
0.22µF Elna = 0.24µF, 0.23µF

Again, the larger values have drifted a lot more than the smaller ones. I'm hoping that next month I can squeeze new filter caps into my budget.

Cheers,
James

Cool

thanks for documenting the decay of the capacitors
 
Hi Karl & Mark,
Haven't seen you guys around for a while but some news if you are listening.

I've been trying to keep learning from the different people here and slowly started to understand that there are better choices of caps than simply shotgunning Nichicon FG's into an amp. So I slowly stocked up on some Panasonic and Wima foil caps, and a variety of Panasonic FC & UnitedChemiCon KYB elcos. And 2 more A-520s. One that is being used for WIFE's PC setup with a pair of Infinity Alpha 5s and the 2nd for a friend at work that will be used with a pair of Canton GL 260 speakers.

I just finished the 520 that will go to my friend and here are the results:

R204, C177, D109 & Jumper124 removed & area cleaned to remove old Sony glue.
Jumper124 replaced with new jumper.
R204 (3.3k carbon film)(left channel idle current/power amp) replaced with Vishay-Dale ¼w 1% metal film resistor (PN: CMF553K000FKBF). R203 right channel 3.3k resistor replaced to match left channel.
C177 & D109 cleaned & re-installed.

C175 & C176 (12000µF/56v/85°c)(Power Supply) removed/cleaned old Sony glue from the area. Caps measured (11490µF & 11430µF). Replaced with UnitedChemiCon KMH 15000µF/63v/105°c snap in caps (PN: EKMH630VSN153MA63T). Jumpers under the caps were protected from short circuit with 0.2mm high dialectric silicon foil. New mounting holes for the caps (because of 10mm lead spacing vs 12mm on the original caps) drilled with 1/16” bit (0.62”/1.6mm).
C173 & C174 (330µF/25v/85°c Rubycon)(Power Supply) replaced with United ChemiCon KYB 330µF/25v/105°c (PN: EKYB250ELL331MHB5D).

C171 (22µF/16v/85°c Matsushita)(Protection) replaced with Nichicon FG 22µF/25v/85°c (PN: UFG1V220MEM1TD).
C172 (100µF/16v/85°c Elna)(Protection) replaced with Panasonic FC 100µF/25v/105°c (FC1530022-360).
Protection relay replaced with Finder 24v relay (PN: 40.52).

C170 (100µF/16v/85°c Elna)(LED drive) replaced with Panasonic FC 100µF/25v/105°c (FC1530022-360).

C123 & C124 (1000µF/16v/85°c Elna)(EQ Amp) replaced with Nichicon FGFine Gold 1000µF/16v/85°c (PN: UFG1C102MHM1TO)(can also use Nichicon Muse KZ PN: UKZ1E102MHM).
C113, C114, C115 & C116 (100µF/16v/85°c Elna)(EQ AMP) replaced with Nichicon FG Fine Gold 100µF/25v/85°c (UFG1E101MPM).

C143 & C144 (1000µF/6.3v/85°c Rubycon)(Power Amp) replaced with Nichicon FG Fine Gold 1000µF/6.3v/85°c (PN: UFG0J102MPM1TD).
C157 & C158 (100µF/16v/85°c Elna)(Power Amp) replaced with Nichicon FG Fine Gold 100µF/25v/85°c (UFG1E101MPM).
C169 (1µF/50v/85°c )(Power Amp) replaced with Wima MKS 1µF/50v poly film cap (MKS-2-5-1).
1k Idle current trimpots replaced with Bournes sealed 1k trimpots (PN: 3386G-1-102LF).

C125 &C126 (0.22µF/50v Elna [missidentified on schematic as ceramic caps])(Loudness) replaced with Wima MKS 0.22µF/50v poly film caps (MKS-2-5 220N).

So you can see that I took your advice and bought 2 of those UCC 15000µF/63v/105°c filter caps from IT-tronics. They were somewhat tricky to install, there isn't much room to drill new mounting holes and stay inside the traces and still clear everything around them. And with the smaller lead spacing I wasn't sure if the leads would clear the jumpers underneath the caps so I cut some disks of silpad and added them under the caps as security against a possible short. I would have liked to also use a Pany FC for C171 but got dumb and forgot to order that one. But I at least used a higher voltage rated FG there.

DSC02147.jpg DSC02148.jpg DSC02154.jpg DSC02156.jpg DSC02159.jpg DSC02163.jpg DSC02165.jpg DSC02167.jpg DSC02197.jpg

The results are most satisfying. I just listened to the first CD through it and it sounds fantastic. I'd say better than the first one that I did, enough so that I'll probably retro-fit it with these same caps. I'm also in love with these amazing little Canton GL 260 speakers. They are really really great sounding little things with great imaging.

Cheers,
James
 
Wow! What a fantastic work! Very accurate selection of parts. Congrats, James :)

And hey, the A-520 / GL-260 combo looks very nice indeed!:music:
 
James,

I trust that you and your family had a good 2015, and an even better one to look forward to in 2016 with better sounds.

2015 has been a challenge for me, I am still cleaning up the mess......... life goes on.

Thanks for the update, you are where I would like to be on my next project.

I did update my LaScalas last year, and started a project using JBL horns and drivers for two ways with the LaScala bass bins, the quest for the last 1% improvement.........there are lots of 1-2% improvements that add up.

https://community.klipsch.com/index.php?/topic/159419-lascala-update-upgrade/

It appears that you have the best of the best hardware now, if it were me, the next mountain to climb would be performance measurement for THD, SN and linearity all of which measure one sine wave, multi-spectral analysis is I think a bigger, and less friendly, mountain to climb and is perhaps the the reason that some amps have great specs but don't sound as good as other ones. How things sound on paper is not important.

El Paso does great measurement videos on his tube amp rebuilds and can serve as a template for setting up a test bench.
https://www.youtube.com/user/essoterric

Load testing at max output is good for the spec sheet and sales, but is of little importance for daily or normal listening volumes in a home. Depending on the efficiency of the speakers, the first half watt or less is everything. My LaScalas are 104db 1w1m sensitivity; checking the meters on my amp, I frequently listen at 1/200th of 1 watt. 2 watts is ear splitting.

I was involved in many high-tech leading-edge product developments in the telecommunications industry, there is no substitute for prototypes on the bench and tweaking the hardware and the software in various configurations to "characterize performance". As you characterize the performance, you get a feel for what the design and it's variables will and won't do, there is no substitute for trial and observation.

Perhaps someday I will acquire the skills and software to load the designs into a simulator like Spice, hope I got that right, to verify the math of the Yamaha and Macintosh designers, but I suspect that they are very good with their numbers and feel for systems. On paper, this would allow for exercises like using wire wound resistors instead of carbon or film resistors to see if it changes the calculations for noise and distortion on the output, then there is the question of "how inaudible does something have to be before we can't hear it"?

Engineers are trained to optimize cost even when it's completely irrelevant, like adding 2 cents to a $3000 system. This was a very hard habit to break in Engineers that worked on my products. For a few dollars more, we did use the best components available for my products. Sometimes the total cost difference was $20-30 on a $20K+ system. It also saved months of searching for the cheapest substitute that would still meet spec.

Like wine tasting, you may be at the point where you could do "bake offs" using different Op-amps which may be the single biggest variable in sound quality once the systems are to spec. I would have to go back and look to see if you are mounting sockets for the op-amps.

Keep up the good work, I do enjoy watching your projects!

With kind regards,

Mark
 
Hi Mark,
Thanks, my family and I are doing well, very good in fact. I'm very sorry to hear that you had problems last year, may this one be a better one.

I enjoyed reading through your LaScalas thread, thanks for sharing that! I've never had the chance to hear any such highly efficient horn type speakers but that experience is definitely on the list. Very nice pics of your listening (living) room also.

I really like the idea of getting an oscilloscope and learning to use it and have been thinking of buying one of these little "DSO062 Digital Oszilloskope Analog Bandwith 1MHz 20MSa/s Oscilloscope DIY" kits on ebay. I know that they are probably a ridiculously limited little thing but at under €25 it would be something that I can afford and use to start learning on. That little LCR checker that Karl steered me onto has proven to be very useful, perhaps one of these kits will also.

Unless I'm looking at the schematic wrong (always a strong possibility with me) the op-amps in the A-520 are solely for the phono section and neither I or my friend will be hooking up a TT to our amps. So it just hasn't seemed worthwhile to play with yet, at least with the 520. However, last summer I bought a pair of Denon DCD-1290 CDPs and did start experimenting with op-amps in them.

http://audiokarma.org/forums/index.php?threads/dennon-dcd-1290-mod-thread.688774/

I had hoped to be already starting another round of mods with them by now but in my last order with Mouser somehow got stupid and ordered 1/8w metal film resistors instead of 1/4w ones so I will have to wait a couple more weeks for that (after I make a new Mouser order). Since then however I picked up a pair of Denon DCD-890 DCPs and will very shortly be starting a round of mods with them. The results so far with the 2x 1290s have been very satisfying. I use one for CDs that are too bright or harsh sounding (the one with the Burr Brown OPA2134PA op-amps) and the other for everything else (the one with the LME49720NA/NOPB op-amps). It's interesting to use 2 identical CD players with different op-amps playing the same CD simultaneously and AB them. I sit people down and don't tell them which is which and get their opinions.

Other than that I took a risk and bought a couple of damaged (but opticly good looking) Yamaha items, a CX-600 pre (Possibly/probably bad transformer) and a M-70 power amp (unknown damage to one channel). Low risk because I didn't pay very much for them (at least in terms of what such units in good condition are going for). I haven't touched them yet because of too many other projects in the way and it will most likely be at least another couple of months before I do. But if I can get them running they will be the core of my 2-channel system in the future. Last fall I found a Yamaha RX-V3900 AV amp and set it up with all Canton speakers for home theatre use. Lots of fun there.

Cheers,
James
 
So after doing that last 520 I thought a bit about replacing those big filter caps, and realized how lucky I was that I was able to drill the new holes for the leads and actually have it work. So I thought about making some patterns to make it easier to do this consistently.

I took some of the clear plastic used to hot laminate ID cards etc and used a double thickness of it to get something that would be easy to handle. Used a ruler and compass to mark it to the right dia etc and cut a pair of templates to use to mark the pcb for drilling. They didn't turn out perfectly round (heh heh) after my hand cutting with scissors, but good enough for what we want.

DSC03068.jpg DSC03070.jpg

Cheers,
James
 
To use these things I backlit the board from underneath so that I could see the traces and their relationship to the existing mounting slots and the jumpers on the board.

DSC03071.jpg DSC03073.jpg DSC03075.jpg DSC03076.jpg

This was reasonably sucessfull. But the fact is that there is very very little room to work with here. Perhaps a total of 2mm to play with. And there a couple of problems with backlighting the boards. First, the light bleeding through the existing holes blinded me when I tried to view the board fro directly above, so I had to work from an angle. And second, due to the thickness of the pcb the light diffused and I couldn't get a perfect idea of exactly where the traces positions were. So this method of marking the board for drilling was only partly successful. In the end my hole locations weren't as good as the first time around and I had to elongate 2 of the l holes with a jeweller’s file.

Which makes me realize just how lucky I was to have gotten it right the first time doing nothing more than using the Mk-1 Eyeball. I don't think that I will do this again. The next amp I'll make a pair of standoffs with 5mm pvc and leg extensions on the caps. When I get that far I'll post.

Cheers,
James
 
On the other hand, one thing that was successful was an idea about using Deoxit. In order to get into some switches with their very small hole on the top as the only access point, I got creative. Working in a hospital has it's advantages. I was able to get a hold of some needles used with infusion bags and taking blood samples. I used a wet stone to file the end of the needle to a cone shape. Inside the other end of the needle where it plugs into bottles etc the end of the needle is just large enough to pop into the end of the normal spray tube on a Dexit can. I used a hot glue gun to seal them together.

Dang! This works great! The needle end just fits into the tiny hole on top of ALPS switches such as the speaker, subsonic, mode and phono selector. It's a tight enough fit than you give a blast of Deoxit it comes boiling out of every crack and crevice of the switch. The switch is really hit with a pressurized blast rather than seeing most of the Deoxit going wastefully sideways as has been my experience up till now.

DSC03077.jpg DSC03078.jpg DSC03079.jpg

Cheers,
James
 
Very cool idea.
You could even use some tissue around the switch to catch the outflow while purging if you didn't want any cleanup afterwards.

The backlit board makes for some nice pics too!
 
Thanks Zaibatsu. I usually wrap a piece of rag around switches and potis when deoxing them.

This latest A-520 that I did had glue damage in a new spot. A 4.7 ohm fusible resistor and a 47k ohm resistor had corrosion from the glue.

For the 4.7 ohm fusible...
http://www.mouser.de/ProductDetail/...WamRRHPrfiGdDXz5OeJ5yWuRhnmzyXmYI/nNK5PfdXrY=

I couldn't find a 47k ohm Vishay metal film in 1/4 watt but found a 47.5k and took the chance that the extra 500 ohms wouldn't matter.
http://www.mouser.de/ProductDetail/Vishay/CCF5547K5FKE36/?qs=rMX6kfwPG0pDC%2bWAbQ2oFA==

When I pulled the old resistors they measured 48.1k and 48.6k so I didn't feel at all bad about using the 47.5k resistor. And the amp sounds great.

Cheers,
James
 
OK. So I've finally gotten around to replacing the psu filter caps on the very first 520 that I bought. It has proven to be a nice clean dependable unit, first in my PC system and then in Wife's PC system after I bought my AX-500.

When I replaced the filter caps in the AX-500 I tried the leg extension + stand-off method for the first time and it worked very well. And since I decided to never again drill the pcb to fit snap-in caps I used that method here.

I took one of those clear template disks and marked where the leg extensions need to be positioned. They only need 4mm of re-positioning. The neg goes 4mm to the left and the pos goes 4mm up or down. I used 1.5mm² solid wire, wrapped it tightly a full turn around the terminal for a good mechanical connection, soldered, and bent. For stand-offs I used rubber o-rings, 35mm od and 5mm thick. The o-rings were glued to the bottom of the cap with simple general purpose rubber cement.

DSC04189.jpg DSC04190.jpg DSC04191.jpg DSC04195.jpg DSC04198.jpg

Btw, that's not a + & - marking on the bottom of the left cap pictured but rather a cross to mark where the leads needed to be bent.

Cheers,
James
 
Last edited:
Since I have 2 more of these amps in the line-up to be recapped I went ahead and prepared all 6 caps now to save time.

DSC04200.jpg DSC04201.jpg DSC04204.jpg DSC04207.jpg DSC04210.jpg

Cheers,
James
 
In order to fit the lead extensions in the pcb slots without modding the pcb I used some flat faced duckbill pliers to squeeze and flatten the wire just enough to pass through the slot snugly. The o-rings got a bead of rubber cement before inserting the caps into the pcb. Then the leads got trimmed to a reasonable length and soldered. It turned out nice and clean.

DSC04212.jpg DSC04215.jpg DSC04216.jpg DSC04217.jpg

Cheers,
James
 
Last edited:
The finished results are most pleasing. This went very smooth and easy, no fuss and no changes to the pcb. The whole thing took less than an hour including removing the old caps and cleaning the board. I'm very happy with how it turned out, this will now be SOP for me when replacing filter caps that don't have the OEM leads/spacing. Btw, the old caps measured about 10700µF and 10900µF.

DSC04223.jpg DSC04226.jpg DSC04227.jpg DSC04231.jpg

Cheers,
James

edit: One final thought regarding doing this. I've read in a number of threads that long lead extensions on caps is a bad practice because of inductive and RFI interference. But in the end the total length of extra lead is no more than 10mm. So I feel pretty comfortable in doing this instead of permanently modding the PC board and taking a chance on damaging it.
 
Last edited:
Hello Bratwurst, impressive work on those filtercaps! I recently bought a non-functioning A-520, and I believe the transformer to be kaput. However, I'm having a hard time getting the exact specifications for this transfo, as the original part-number from the service manual (GA68050) doesn't turn up any matches on Google. Do you happen to know the secondary voltages?
 
Interesting thread. I have one sitting unused at the moment and in need of care. Not sure whether to try a rescue or not but your pics and comments at least have me thinking.
 
Back
Top Bottom