CA-810 Restoration

jheu02

AK Subscriber
Subscriber
Since it's been so long since I've given this amp any attention, I thought I'd start a new thread as it's time to (really) start working on it. I've become a little more methodical over the past several years doing recaps and some troubleshooting, so this hopefully won't be quite the stab-in-the-dark restoration that I originally posted about here.

I've finished up a couple other long term projects and now it's time for this one. From the other thread I knew that the preamp functions worked but wasn't too sure about doing other testing, i.e. removing OPTs and powering up, etc. Well tonight I decided to put it on my work table and check the outputs. The left channel 2SB555 is toast: OL on both b,e and b,c as well as all other combos except c,e where it read 1.377V.

The R side OPTs checked fine, as well as the L Channel 2SD425.

So, 1st question regards a new mouser (or other order):

Since the recommended replacements are MJ21193/4G, should I just go ahead and replace all 4 OPTs with these, or just get replacements for the L side (i.e. would there be issues if the R side was originals and the L side new)?

If I should go ahead and replace all 4 (my current thinking) do I need to order several of each to try and match, or just get 2 of each and hope for the best (or find a company that does matched pairs)?

If I should get extras for matching purposes, what other amps could use these as replacements since I don't do this for a living, and the extras could well likely go unused unless I sell them?

What testing could I do in the meantime? Pull the outputs, reconnect disconnected boards, and see if it'll come out of protection?

Thanks for you're thoughts.
 
...

a. Since the recommended replacements are MJ21193/4G, should I just go ahead and replace all 4 OPTs with these, or just get replacements for the L side (i.e. would there be issues if the R side was originals and the L side new)?

b. If I should go ahead and replace all 4 (my current thinking) do I need to order several of each to try and match, or just get 2 of each and hope for the best (or find a company that does matched pairs)?

c. If I should get extras for matching purposes, what other amps could use these as replacements since I don't do this for a living, and the extras could well likely go unused unless I sell them?

a. I'd replace all of them. The MJ21193/4G pair are nice transistors, made for audio and have higher SOA and probably are the only TO-3 still in production for BTJ's in this range.

b. I believe that if the transistors come from same batch production, they are likely to have pretty much same numbers. I ordered a bunch of TO-92L (ok, it's easier) and measured their hFE and they had all the same number (560). Since you are going to order from a major supplier, good chances that they are from same batch.

c. Quite a few vintage amps and receivers use TO-3 devices: Sansui, Kenwood, Yamaha... Some units from 80's, too.

Now, take a close look at my avatar... :D

--
 
Last edited:
Thanks for the thoughts on this Karl. I was thinking the same thing re. transistor batches, but thought I'd try to get some other opinions.

I did a little more looking around/troubleshooting and noticed a purple Matsushita cap that was a definite replacement sometime in this amp's past. The sleeve was pulled back around the top of the can. C603 IIRC, 220uF/10V. It also had what looked like smoke marks on it. So, I pulled and checked it with the capacitance function of my mulitmeter...open, no reading whatsoever. Got a Panasonic FM 220uF/25V from my stocks and soldered it in, then pulled all the OPTs again and put it on the DBT. Nice check and relay clicked in ~3 seconds. With that working I put the R channel outputs back in, checked the bias on the DBT and set it low, then reset on line power and checked to see if there was sound. Other than scratchy pots, I got nice sound.

So, tonight, I'll swap the good outputs into the L channel and repeat the process of checking setting bias on the DBT and then if all goes well, listen to the sound from that channel too. If the bulb doesn't dim properly, I'll know there's more troubleshooting to do on the left side before I get into serious restoration/recapping.

EDIT

I swapped the OPTs, bulb dimmed, checked bias...60ishmV...way too high. Dialed it back 0mV on the dim bulb and then went to line power. Brought the bias up to 19.7mV..close enough to 20 for this check. Lo and behold... :music::music:.

Just a quick check, but didn't detect distortion or any other maladies. So, I can start the recap process while placing an order for new outputs and other miscellaneous items for other projects. Now to look for the thread on replacing the meter lamps which are out.
 
Last edited:
Last edited:
Thanks for that link. While I'm getting my order together, are there any "best practice" things to replace while I'm in there...typically noisy transistors, fusistors? I've looked at a couple threads about the fusistors and I'm not positive they can just be replaced by same wattage flameproofs. It seems in some threads the wattage of certain ones is increased (CR-2020 threads). The 2SD234s in those CR2020 threads were a recommended replace due to hot running (TIP41C, though I have TIP31C on hand...good enough?), but I didn't see it mentioned for the CA-810. Other transistors on the amp boards (2SC1624, 2SA814, 2SC1124)? Also, does anyone have a version of the schematic with the audio path highlighted in Red or Bold black like some schematics do? I have an idea of its path, but wanted to crosscheck and maybe upgrade some of the components).

Merrylander, avionic, other Yamaholics...ideas?

Here's a couple "before" pics:

16440922489_90c40238a5_b.jpg


16006947673_1131895e97_b.jpg


16626050682_694618888c_b.jpg


16004551134_ed3ae83d37_b.jpg


FM cap inserted where the purple Matsushita one that was open was.
16004540254_ef89637c34_b.jpg


16600746466_1e049255d4_b.jpg
 
Here's a couple shots after recapping. I used a mixture of caps I had on hand, but wanted to try out Elna Cerafines (ROA) in some places and combine that with the warmth of Silmics (RFS) also:

Tone Board
16626729311_959f86df5b_b.jpg


Function Board (forgot to take a before pic of it)
16626729901_0095ae7802_b.jpg


16420532447_4de34fce81_b.jpg
 
FWIW, In my research on restoring the CA-810, I did not come across any recommended transistor replacemnts. Hopefully one of the pro's can confirm for you.
 
That's what I was seeing (or not seeing in this case) also. I may check some CA-800/1000 and/or CA-1010/2010 threads to see if there's common transistors there that are known to be troublesome and just easiest to replace now while things are torn apart.
 
Input caps

I can't remember if the input caps (for both L & R channels) are 1uF or 10uF. But if they are 1uF, you could use a high quality metallized polypropylene to replace the electrolytic ones.
 
I can't remember if the input caps (for both L & R channels) are 1uF or 10uF. But if they are 1uF, you could use a high quality metallized polypropylene to replace the electrolytic ones.

IF those are on the Tone Board, I did change them to 1/50 Panasonic ECQs, the other 1/50 caps on that board I changed to 1/63V Kemets. Unlike Pioneer and Kenwood, it seems Yamaha didn't choose to use special caps in certain areas, like Low Leak in audio sections, etc. Those are easy to spot in those other manufacturers, but not so easy here. I didn't really want to try just one "flavor" of cap in this amp, so I'm mixing it up. Just want to do it logically. And maybe change out a couple resistors if they're in the direct audio path to my Dale metal films.
 
Both Kemets and ECQs are very good choices for any 1uF electrolytic.

Forgot to mention, very nice work and pictures! Clean and well documented, mixed caps are fine.
 
Since the amp had issues with the L channel, I started tearing into that side of the board more and discovered that the 85mA 68ohm fusistor is open. I've been searching the threads, but not sure what wattage rating of a metal oxide I should use as the replacement.

The others also have some smoke marks on them, so I'm thinking of replacing them also. 680ohm 25mA and 390ohm 25mA. 1/2W metal oxide work for the 25mA versions? What about the 85mA...2W?
 
I had seen that second link, but Dave states he uses the same wattage and value. Unfortunately, the resistor itself doesn't indicate wattage, and the SM just lists the Yamaha Part number, which is derived from the Micron part number. HW104680.

A little deciphering: the other fusistors are labelled HW105680 (25mA 680ohm), and HW105390 (25mA 390ohm). So the 4 and 5 in the numbers are decimal places. I wonder if the 1 indicates 1W. Unfortunately, the sizes differ between the 390 ohm and 680 ohm, even though they have the same current rating and numbering scheme. They are approximately the same sizes as the 1 and 2 W flameproof Rs on the boards (the red-bodied metal oxide ones). I may just initially go with 1/2W MF and if it opens right away I'll know a need a higher value, but I can't find any info on the web about these specific fusistors.

EDIT:

I wonder if I can use ohms law here...P=I^2 R? So, P = .085^2 x 68... P = 7.225^-3 x 68... P = 0.4913W (IOW a 1/2W resistor)? Using the same formula on the 390ohm fusistors yields a 0.24375W rating (1/4W) and for the 680ohm ones a 0.425W rating (1/2W again). Hmmm? That would also explain the size difference between the 390s and 680s, and why the 680s are about the same size as the higher current rated 68ohm ones.
 
Last edited:
This is just speculation, but here goes. A friend and I were puzzling over the schematic to an A-1020 yesterday and noticed that all of the fusable resistors have a little square next to the Rxxx number shown on the schematic. The square is half shaded (diagonally). We speculated that that might be a marking to indicate 1/2 watt as none of the non-fusable resistors have this marking. It might be worth checking into, but we could also be completely wrong.

Cheers,
James
 
Yamaha's symbol in the CA-810 anyway is an open circle at each end of the fusistor depiction, so, that was probably just to indicate a special type of resistor in the circuit and make the tech think twice about what he was inserting during a fix.
 
Differently from those used in Sansui's, which go higher in impedance to prevent a disaster, the ones used in Yamaha's seem to "open", like a fuse or have a thermal fuse inside them like transformers have, too. So these are pretty complicated to find.

Yes, you could use the 1/4W or 1/2W resistors (lifted from board) as replacements for the smallest ones. Maybe flameproof ones are safer for this application?

316kw0w.jpg


or7211.jpg


--
 
Last edited:
Differently from those used in Sansui's, which go higher in impedance to prevent a disaster, the ones used in Yamaha's seem to "open", like a fuse or have a thermal fuse inside them like transformers have, too. So these are pretty complicated to find.

Yes, you could use the 1/4W or 1/2W resistors (lifted from board) as replacements for the smallest ones. Maybe flameproof ones are safer for this application?

316kw0w.jpg


or7211.jpg


--

That pic of the fusistor helps explain why they're bigger than "standard" resistors of that era. Thanks for sharing it.

I've read that the metal oxide ones are well suited for this, with metal film and wire-wound being usable too, since all are flameproof.

I've got them on order now to use as replacements. The 680s measure 700 and change and the 390s are almost spot on, so they're certainly within tolerance levels should I choose to reuse them. With known damage in that area I hesitate to, though, as who knows how close they are to going bad, if such a thing happens to them. They have undoubtedly seen some elevated temps, smoke and maybe even some sparks. No way of knowing what voltages they may have been subject to already.

I thought I may have also had a bad 2SC1124 as I was only getting one of the 6 way diode tests to read. But, I used my Dremel wire brush on it to clean up the leads from whatever oxide layer was on it, and then got good readings.

With the 68ohm resistor open though, I wonder why I got sound when I moved the outputs to that channel to test and set the bias? I would have expected some wonky readings going on and maybe no sound or at least distortion or something. Not great with circuits, but it looks to be potentially part of the load sharing prior to the outputs.
 
I was going through some Internet links I'd saved and came across one that might be of interest. As I searched the site, I came across these. They even draw less current than the standard ones and are rated for 16000 hours.

https://www.1000bulbs.com/product/97272/AIRL-8111.html

Now I just need to look at my equipment and see what other types I need.
 
Back
Top Bottom