Fisher Coupling Caps......????

I've just been into my 3rd Fisher 500.....one 'B' and two 'C' units. In all 3 I made the decision to replace the 4 coupling caps. Fisher uses .047 values in this position....but 2 are 400 volt and 2 are 650 volt. Anyone have a clue as to 'Why?' this might be? I've seen other units where the values were slightly different....perhaps a .1 and a .01......but both at 600 volts. I'm told this provides 'stability' in some way....but honestly, this is a bit beyond my current understanding level. However, this 'voltage disparity' confuses me 'again'!!! Thanks for any comments.......... Tom D.

Tom,
Take a DMM set to DC voltage and put it across the caps while playing. The answer will become immediately apparent.

The way the phase inverter circuit is designed on these puts different voltages on each cap of the PP pair. Not a bean counter thing. Some would argue it's part of the Fisher sound.

Even when I go to .1uf and change the grid bias resistors from 330k to 200k or 220k, which I do whenever I use new production EH 7591s, I stagger the voltages on each cap in the pair when possible.

Hey Don,

Thanks for the info. I guess I'm just curious.....if the voltages are 'different'....then clearly the need for suitable range is necessary. But, is there a problem with using a 650 volt cap on both sides even though it may be 'overkill' ? My only reason for asking is that I only 'stock' the 650 volt versions. It's what I've used on all 3 of these units in all 4 positions. I figured if 400 was only necessary on 2 of them, then having a cap rated for a bit more wouldn't hurt. But.....if there's a good reason to 'drop' the two back to a 400 range, I can do that.

Also, you mentioned changing resistors and going to a .1 coupling cap. Can you tell me a bit more about that? Does that increase range / response enough to make it 'audible'.....or is that more of an 'I can see it on the scope' kind of mod?

Thanks Don!

TSD

No problem using the higher voltage rated caps. The 0.1 uF is necessary when you change the grid bias resistors from 330k to 200k; the 330k resistors are fine if you just use NOS tubes, but with new tubes use the 220k or lower resistors. It was mentioned on Jim McShane's site: "There have been some cases where the new production 7591A tubes (EH and JJ) have had trouble in circuits using a very high value grid resistor - as vintage Scotts, Fishers, and others do. The best solution is to change all the grid resistors to 180K-200K (from the old value over 300K - sometimes they even exceed the old RCA tube manual
recommendation!), and change the coupling caps to .1 uf or .22 uf."

Agree with Lee on the voltage. Except for Historic accuracy, all you are doing is providing some voltage headroom to the new caps. I've got 630V caps in my 800C and did the same with my 400 (7868's) and the Sansui 1000A.

Dave Gillespie posted a very nice explanation on the .047 vs .1uf and 330K to 200K on the grid couplers. There is a timing issue, that is better served (as I understand it) with the resistors @ 220K and the cap @ between .086(IIRC) and .1uf. So many have used .1uf without problems that the mod has evolved into a 220K (resistor) and a .1uf cap. Note that this modification/adjustment works equally well with original type tubes also. If I can find the explanation from Dave, I'll note it here.

Larry

Making the cap/grid resistors change keeps the time constant correct, I believe. One offsets the other. I don't think there's much of a change in output power, volume, distortion, or tone.
If using vintage tubes, I just use .047 and leave it be. For any new production tubes, or if someone is likely to switch to new production, I always make the change.

Any adverse effects of placing some clear shrink on K40y's used as coupling caps since the outer casing is metal?

Roger

rshadley,

I always do shrinkwrap 'em. Why take a chance?

Wharfcreek,
If you've got the same voltage caps, use 'em. It won't do any harm. I've done it many times.

One more thing about this issue of changing .047 caps to .1 caps and changing the grid resistors from 330k to 200k or 220k.

One unexplored issue, much more of an issue in an integrated than a receiver, is that the higher value caps are often considerably larger.
In an X-101-C for instance, good luck getting a .1uf K40Y cap quad in that tight spot near the phase inverters.
Plus, it's an effort to move to new production EHs on an integrated with cabinet spacing considerations and all.
Just one more reason why I use vintage tubes and leave the cap values alone in a vintage unit.

You can read in multiple places about how the 330k grid resistors run the 7591s at the ragged edge and all, and it's true, but the vintage ones are tough tubes designed and built late in the tube era when a combination of events led to a period of excellence in American tube design and manufacturing.

If leaving the Fisher stock is so fatal to the tubes why can you go on ebay any day of any week in any year and see units for sale with the original tubes still there?

Just so you understand my concern in context, here's what I do:
1. Receiver using 7591s? Replace output tubes with EH, change output coupling caps to .1uf and change grid resistors from 330k to 220k - always.
2. Integrated? Use old-school small-bottle 7591s, leave coupling caps .047 and leave grid bias resistors if I plan to keep it or if I've discussed the issue with the friend I'm doing the restoration for. Otherwise, do the cap and grid resistor change -even if using the old stock small bottle tubes. It may make the tubes last longer.