Fisher 400 questions

[pageboy]

Thermal runaway, as someone else called it, is a better description of what is going on. The plate glows bright red, a loud hum comes from one speaker, and the mains fuse lets go.

The AC filament voltage is fine. Moving one filament lead down the ladder is a good idea anyway. I understand what you're saying there. I'll be doing that one. Yes, I guess even 35 volts high on the HT is within 10%, barely, so I'll let that one go.

I can install some 10 ohm cathode resistors, but how do I know what the current should be? My schematic only gives voltages.

The rectifier is a silver box with four tabs that says Siemens on the side of it, held against the side of the chassis with a pair of screws. I'll be getting that out of there. What about the pair of diodes on the HT? Should those be replaced also?

I haven't had a chance to check out the coupling caps yet. It sounds like I should replace those just out of principle. I don't understand something though. If some DC is leaking through from the previous stage, why does the bias voltage on the power tubes read perfectly normal and all four exactly the same? Shouldn't some DC leaking through affect the bias voltage? Not enough negative bias voltage is what I expected to see when I first opened this chassis and did some voltage checks. But bias voltage measures normal.

[kevinkr]

I have seen instances where leakage in old coupling caps increased as the chassis heat soaked over time (an hour or so) in cases where the cap did not appear immediately to be bad.

I would replace the coupling caps to the output tubes to start.

Check all of the grid circuit components associated with that tube and make sure they have not drifted in value.

You have a Siemens selenium rectifier. Replace with modern silicon bridge, add a small resistance in series with the output of the rectifier to trim to something close to the recommended values, probably trim to the high side to reflect your line voltage relative to the design value. (Probable 117V or so)

I believe the high voltage rectifiers in your receiver are already silicon, at the voltage required selenium would be very ungainly.

[TubeHead Johnny]

Sounds like one of the 0.047 uF coupling caps between the phase inverter and the power tube grids is leaky, which in turn drives that grid of the output tube positive, then goes into thermal runaway. To check this, disconnect that cap (the end connected to the output grid) and measure voltage of the disconnected lead to ground. If there is any measurable DC voltage there, that capacitor is bad. Make sure you keep the bias supply wire connected to the grid.

[Eli Duttman]

Pageboy,

Are you using vintage 7868s or current production EH tubes in your 400? Are your 7868s a closely matched quad?

In any event, replace all 4 coupling caps. with 100 nF./630 WVDC ECQ-P(U)s.

Yes, the OEM B+ diodes are Silicon. However, they are noisy. Replace the B+ diodes with a pair of 1000 PIV/3 A. low noise UF5408s.

[pageboy]

I have old power tubes. I'm running what came with the unit, plus I found a few extras in a tube caddy buried in a corner of an old organ repair shop. I don't know whether they've been used or not, but they seem to be in reasonable shape. I'm sure that what's in the amp now is not even close to being matched, especially since they all show evidence of running too hot - not just the one that burns up. I've been out of the loop on tubes for a few years, and I didn't realize that EH was making new ones.

How much bias current should I be looking for on the power tubes if I install some 10 ohm cathode resistors?

Our weather here is calling for rain tomorrow. Maybe I'll get some time to check into this thing again. I appreciate all the replies. I've learned a good bit right here.

[Eli Duttman]

You have mail.

Fisher used truly fixed bias. Without modification, adding 10 Ohm cathode resistors buys little. That fixed bias mandates that the 4 7868 O/P tubes be a CLOSELY matched quad. Good matching of old production tubes is, at best, problematic.

If you make the decision to switch to current production ElectroHarmonix 7868s, there are 2 pitfalls you must be aware of.

Old production tubes have Novar bases. Current production tubes have Magnoval bases. The pin circles are identical. However, the Magnoval pin diameter is slightly larger. The sockets get stretched and old production tubes can no longer be employed.

The 2nd issue is the 330 KOhm grid leak resistor value Fisher employed. That is outside the published limit of 300 KOhms for the 7868. The tubes of yore tolerated that liberty. Current production stuff has been known to run away, when overly large grid leak resistors are in place. The "best" solution for the problem is to change from 12AX7s to 12DW7s in the driver/splitter position. The 12DW7 is asymmetrical, with 1 section like a 'X7 and the 2nd section like a 'U7. The 'U7 section makes a much better "concertina" phase splitter than a 'X7 section. Some parts values change, but topology doesn't. Switching to the 12DW7 in the splitter driver positions allows the 7868 grid leak resistors to be reduced to an in spec. 270 KOhms.

[pageboy]

The purpose of the 10 ohm resistors wasn't to change the bias operation. Someone mentioned a few posts up that I could use them to measure the bias current, as that was more important than the voltage. But I don't know what current level to look for them to be.

Nuisance that they can't make the repro tubes exactly the same as the originals. Thanks for the info on those.

[Tom Bavis]

New or old, I wouldn't run 7868s over 15W dissipation... at the 19W rating, they won't last long. 40V @ 35 mA is 14W... a good spot to shoot for.

[Eli Duttman]

The purpose of the 10 ohm resistors wasn't to change the bias operation. Someone mentioned a few posts up that I could use them to measure the bias current, as that was more important than the voltage.

Being able to measure the "idle" current is of value only if it can be adjusted. Fisher's true fixed bias scheme doesn't provide any adjustments.

Nuisance that they can't make the repro tubes exactly the same as the originals. Thanks for the info on those.

Current production ElectroHarmonix 7868s fully comport with all the published electrical spec's, but they don't allow for liberties being taken. I agree Novar bases would have been nice. Unfortunately, Novar bases are "unobtainium". OTOH, Magnoval bases are available, as they are used in other current production types, including the 6d22s damper diode. New Sensor (the folks behind the EH label) did both the guitar and HIFI people a huge favor in returning the 7868 to production. An awful lot of quality equipment was in danger of becoming inoperable. Thankfully, that is no longer the situation.

[Ciscokid]

Hi,

I was thinking the 10 ohm (1/4 watt) resistor was to be added as a safety feature. If a tube goes south, then the resistor would pop like a fuse and save the OPT. At least that was the thought and it allowed me to leave the room with a bit of comfort while the amp was running.

Regards,
ck