Okay, after digging in to this unit I found that one of the bias feed resistors was not what was on the schematic (not too surprising since it might be adjusted at the factory).
My unit has the 10k resistor (R125), and the 5.6k resistor (R126), but it also has a 560 ohm resistor in parallel to this resistor. Now doing the math this gives 509.09 ohms. Now maybe someone added this or the factory did this to change the bias? It doesn't seem right to me. What I was going to do is the recommended change of changing the 5.6k resistor to a 2.7k to get the bias down around -20 volts. Has anyone seen this in an 800 that they've worked on?
Glenn

Ah...the dreaded Bias resistor. On my 800c, I had a bitch of a time. It too was not the value specified on the schematic. I was looking for a 5.6k, it was more like 7k. After all the work, I ended up closer to 3k. I had to swap the resistor several times to get to -20v. I should have just put a 5k pot in there. I will be doing it in the future at some point. I may remove the large cardboard cap on the chassis (not used anymore) and mound a small plate there with the bias control pot in it. This will allow me to check and adjust without going inside.
John

Thanks John.
3k at least sounds reasonable, 509 ohms sounds just not right. I haven't run the amp yet, but maybe the previous owner added the 560 ohm in series with the 5.6k to get the tubes to bias really cold? I may just start with the 2.7k to see where I'm at and measure it one it's running (now that I have the 10 ohm cathode resistors in place).

The trim pot sounds like a good idea, any idea as to how much current runs through the bias circuit? I typically have a bias trim pot on all my guitar amps.

Last night I just replaced the 1000/1000 bias filter cap and the selenium rectifier. So now I do have the room for a bias pot.

Another thing I found was that this amp has the 20uf/250v (C54) electrolytic that is crossed out on my parts list with the word "error" hand written in the margin. I also can't find this cap on the schematic. I have to trace where it's connected to, but it looks like it goes to the multiplex section.

Glenn

That cap sounds odd. I also had one cap on my 800c that was in the multiplex area. I left it since I was afraid it would screw up the alignment. Thankfully it is pulling in stations well. As for the bias pot, I think a half watt would be adequate. I found a multi-turn pot somewhere and used it on my Sansui for bias. Made it very easy to adjust the voltage. Small turning made slight changes in the voltage making it possible to get it very accurately dialed in. May be overkill, but it worked very well.
I have seen some who made test probe jacks on the chassis top and bias pot nearby. Then you can hook up a DVM and adjust the amp at anytime. That is my goal since removing it from the wood case, removing the bottom panel and running it on it's side is a big pain. A pot I saw at tubedepot can be adjusted from either side with a small screwdriver. Here is a link.
http://www.tubedepot.com/r-cts-l.html
John

Thanks John, that's a good idea. I've seen that on other amps as well.
Doh! I just placed an order from the Tube Depot this morning! Too late now.
They don't list a wattage for that pot though.

Here's some nice panel mount slot adjust pots just for that job:
http://www.surplussales.com/Potentiometers/PBM-SlotAdj/PBM-SlotAdj-3.html

Okay, I've rebuilt the bias supply and replaced all the electrolytics.
Replaced all the coupling caps, and made the change to the .1uf, 200k from the stock .047uf, 330k. Installed a new matched quad of JJ 7591 tubes.

I get the following readings off the power tubes:
[14.2mA, 13mA] [11.7mA, 10.2mA], this is with -21.6v bias and 372V on the plates.
I thought the target cathode current for the 7591 tubes was 30-35mA. Is the bias a little too negative at -21.6v? I can adjust this to bring the cathode current up to 30-35mA if necessary. This just seems a little cool to run the power tubes at.
Glenn

EDIT:
The JJ's are rated for 85mA cathode current and 19w plate dissipation.
At this rate these tubes will last forever :)

Okay, I've rebuilt the bias supply and replaced all the electrolytics.
Replaced all the coupling caps, and made the change to the .1uf, 200k from the stock .047uf, 330k. Installed a new matched quad of JJ 7591 tubes.

I get the following readings off the power tubes:
[14.2mA, 13mA] [11.7mA, 10.2mA], this is with -21.6v bias and 372V on the plates.
I thought the target cathode current for the 7591 tubes was 30-35mA. Is the bias a little too negative at -21.6v? I can adjust this to bring the cathode current up to 30-35mA if necessary. This just seems a little cool to run the power tubes at.
Glenn

EDIT:
The JJ's are rated for 85mA cathode current and 19w plate dissipation.
At this rate these tubes will last forever :)

First of all, good luck with those JJ's; you're a much braver man than I.

Yes, that bias level is way too low. I think around -15 to -16 would be a good starting point. But, remember, the important figure here is the actual current being drawn by those 7591S's. Right now you're running them down in the Class B range, and you'll probably want them around 32-33mA for best performance and lifespan.

The 85mA and 19w dissapation doesn't really pertain to realistic in-usage levels in a Class AB push pull configuration. Those are just the design maximums for that tube :)

. . Falcon

Okay, I've rebuilt the bias supply and replaced all the electrolytics.
Replaced all the coupling caps, and made the change to the .1uf, 200k from the stock .047uf, 330k. Installed a new matched quad of JJ 7591 tubes.

I get the following readings off the power tubes:
[14.2mA, 13mA] [11.7mA, 10.2mA], this is with -21.6v bias and 372V on the plates.
I thought the target cathode current for the 7591 tubes was 30-35mA. Is the bias a little too negative at -21.6v? I can adjust this to bring the cathode current up to 30-35mA if necessary. This just seems a little cool to run the power tubes at.
Glenn

EDIT:
The JJ's are rated for 85mA cathode current and 19w plate dissipation.
At this rate these tubes will last forever :)

10 to 14ma sounds way off for -21.6v bias. I assume you are measuring milli-volts across a 10 ohm "cathode fuse" resistor with no input? You should read 0.300 to 0.350 volts, which with ohms law would be 30-35ma. That 85ma figure is not bias, it is at full power and also includes screen and anode current. It is the power just before it goes ballistic. At 30-35ma idle bias current you are not running them at half rating and they will not last forever.

I understand that the 85mA and the 19w plate dissipation are maximum. I think class AB is supposed to be around 70% of these values. Anyway, I am measuring across the 10 ohm cathode resistors, and in millivolts. I'm thinking that the plate voltage is a little low at 372V, maybe that is contributing. I am running off a variac at 110VAC.
You also just reminded me that I shouldn't have any signal, but I was listening to a strong FM station with the volume at around 1/8 up. I'll re-measure with the selector to phono to see what the difference is if any. I'll dial the bias less negative to bring the cathode current more in-line with the 30-35mA.

I guess I'm just used of making single ended class A amplifiers where it's normal to run up around 90% of the plate dissipation :)

Thank to everyone for helping, I appreciate it!
Glenn

I am running off a variac at 110VAC.


Bring the variac up to 117VAC. Even though its only shown on a few schematics, all Fisher voltages of the era are measured from that. Then see what you get, both for B+ and bias current.

Thanks, I'll give that a try.

Just curious, is 4mA a close enough match for a supposedly "matched" quad?

Thanks, I'll give that a try.

Just curious, is 4mA a close enough match for a supposedly "matched" quad?

Yes, but it has been my experience that the so called "matched" tubes are not. Simply because they have not been burned in. Watch your ma carefully for the first hours and check them often over the first few days. If they are true to course they will shift in value. Almost always one of them will end up being dramatically different from the others. If you don't watch them you can burn up a tube.

The real solution is to add individual bias controls.

I've now got this set running nicely:)
I'm running a bias of -17.5 with 374v on the plates. This gives cathode current readings of 32.4,32.7,32.2,28.6. Why is one tube always a little different than the others? The JJ's sound pretty good to me and they fit correctly:)
I've been shooting for between 30-35mA cathode current, so this is right in there.

The current did change slightly over the 20 minutes I played the system. I'll keep monitoring the current over the next several hours to see if they settle in.

Thanks everyone.
Glenn

Hey, Glenn, glad you were able to get some help!:yes:

Clay

I have a question about the first filter capacitor on the B+ line.
This is the can on top of the chassis with the cardboard insulating sleeve that has the 200uf value (C102 in the schematic):
http://oldtech.net/Fisher/800B/800b-1.gif
I've re-stuffed the can with two 100uf 250v electrolytics in parallel, and before I re-solder it back into position I was assuming that the positive end is on the cathode of the second diode. From the symbol on the schematic it is represented that way, but I just want to make sure. I don't need a large explosion right now :)
This is not my first re-stuffing job, but I'd just like some verification.
Can someone please verify this for me?
Thanks.

I hate to rain on your parade but there is 395 volts on that capacitor. 250 volt rated capacitors will probably fail. I think you may want to find the correct value first.

Also, the positive end of the capacitor is connected to the diode. The negative, which is usually noted with a minus sign inside of an arrow pointing down the side of the capacitor toward the negative terminal.

If you are talking C101, that is a different story. That capacitor has 195 volts.

I have a question about the first filter capacitor on the B+ line.
This is the can on top of the chassis with the cardboard insulating sleeve that has the 200uf value (C102 in the schematic):
http://oldtech.net/Fisher/800B/800b-1.gif
I've re-stuffed the can with two 100uf 250v electrolytics in parallel, and before I re-solder it back into position I was assuming that the positive end is on the cathode of the second diode. From the symbol on the schematic it is represented that way, but I just want to make sure. I don't need a large explosion right now :)
This is not my first re-stuffing job, but I'd just like some verification.
Can someone please verify this for me?
Thanks.

C102 is part of the voltage doubler circuit. The cap is hot on both sides and was paper covered because the metal can would be hot. I have attached part of a later 500c schematic which shows the correct polarity. A 250v rating is fine in this location since each cap in a doubler only "sees" half voltage.

Thanks guys!
I thought that was the case. I was going by the schematic, and installed it as it was shown by the symbol (which is correct). Also, I have the parts list, so I installed the 250v caps as specified.
I actually was browsing the cap board layout that someone made for these receivers and it's wire the same way. Interesting that they also used two 100uf caps in parallel as I did.
Two more cans to rebuild and I'm done!
Who do you guys recommend for hi voltage electrolytic caps other than Mouser?
I've used Dave at justradios.com before, but it's a little slow getting caps from Canada. Anyone around the New England area?
Thanks again.
Glenn

C102 is part of the voltage doubler circuit. The cap is hot on both sides and was paper covered because the metal can would be hot. I have attached part of a later 500c schematic which shows the correct polarity. A 250v rating is fine in this location since each cap in a doubler only "sees" half voltage.


Sorry about that. I was too quick to shoot...

Well, I finally finished restuffing the rest of the filter cans this past weekend (five hours worth). I'd already rebuilt the can with the cardboard cover over it a few weeks ago.
It was bothering me a little that the remaining cans were getting SLIGHTLY warm after running for an hour, although I had NO hum in the amp. When I removed the cans I noticed that they were slightly bulged a third of the way up. Now I'm not sure if this was from manufacturing the can, or from the caps starting to fail. After removing the guts from the cans, you could see that the paper was king of crispy and discolored on the outside surface. I guess these things were on their way out.

The good news is that this thing sounds GREAT after replacing all the PS caps. It's hard to believe that the power supply contributes so much to the sound in this amp. I know they're critical in single ended amps, but I didn't think they made much of a difference in PP amps (within reason of course).

Thanks to everyone for all the help on this receiver!

Glenn

That's dedication. Congrats!!

While I know the "right way" to do things, I often continue to use MY equipment with vintage capacitors if they still test OK. But I know that is a short cut to a problem. Now if I am repairing equipment for someone else, I will insist on replacement. That can be from re-stuffing, to new clamp mount or an under the hood terminal strip replacements.

I know that it has been discussed many times, but don't ever tack a new capacitor into the circuit without electronically disconnecting the old one...

Yeah, I was running it for a while with the old cap cans, and it seemed okay, but it always nagged at me that they could be marginal.
I'm the same way, I repair antique radios for other people, and I would never let a set leave my bench without new components as needed.

Now if I can just figure-out how to add a headphone jack to this 800B :)
Maybe an impedance matching transformer connected to the speaker terminals?
I see the 500's have some sort of voltage divider driving their headphone jack.

Glenn

Have you looked at the 500c/800c schematic? I think they just use a 330 ohm resistor and switch to drive the headphones. Only question is where to add the jack. I guess you could run a wire out the back without drilling a hole in the faceplate. Or perhaps a box with resistors and jack connected to the speaker outs? John

Hi John-
Yes, I thought it was some type of resistor network (it's been a while since I looked at the schematic).
I was just planning on putting all the new components in a project box like you suggested.
I'll probably give it a try this weekend.

Thanks
Glenn

Here's how the 500 is connected:

http://webpages.charter.net/porkchop/tubeamp/500_headphones.jpg

The good news is that this thing sounds GREAT after replacing all the PS caps. It's hard to believe that the power supply contributes so much to the sound in this amp. I know they're critical in single ended amps, but I didn't think they made much of a difference in PP amps (within reason of course).


Yep, that's the same experience I had with the last 500C I worked on. Just sounded kind of "doggy" under high output.

Replaced the two doubler caps... WOW. Now, THAT'S what it's SUPPOSED to sound like! :thmbsp:

Regards,
Gordon.