Fisher 400 output valves current draw

[Radio Wrangler]

Full marks to Vigortronix!

They've shown polarity dots on their diagram giving wire colours! So it should be possible to get the phasing right first time.... but I'd still check anyway.

They also include thermal fuses in each section of the primary.

David

[PaulR]

Thank you for the replies.

The transformer arrived today and I wired it up as instructed. I now get 223V out which is great. The heater voltage on the output valves is 6.4v so that seems to be a great success.

That valve still seems to be running away and I switched it off at getting on for 700mV across the 10R resistor. My son is content to buy a set of NOS valves - at least he was the last time I spoke to him. The bias starts off at nearly -17v but drops as the valves warm up, particularly the suspect one.

I assume that your kind offer, Andy, is to separate the DC heater supply and the bias as well as making it adjustable. With new valves that should be ideal. The only problem would be that I cannot see anywhere to fit it as everything is so crowded. I am aware that some space will have to be found if I replace the power supply caps. It is very difficult to judge without having it to hand. Would it be very cheeky to ask you to let me have it to try to fit it in? I would return it if I couldn't find anywhere to mount it. Let me know the costs.

Thank you once again, all. The unit would be scrap without the help.

[snowman_al]

Well that is progress.
We would all be interested in what you get as anode volts (pin 9) and screen volts (pin 1/7) on the 7868s now?

The other voltage of interest is that on the V8, 9, 10 and 11 heaters at C60A.
That is how much -ve bias voltage is available 'to play with'.
Judging from the various Fisher 400 mod threads, to get 35mA through each valve, you need a range of -17 to -20 volts.

You might want to test quickly with the 'run away' 7868 removed?

Alan

[PJL]

Is it worth having a go to see if the runaway valve can be 'repaired' or at least stop it pulling grid current?
Lets first see if there is any cold resistance between grid and cathode, take the valve out and measure the resistance between grid and cathode.
Can you post a picture of the valve showing the getter. Also a view of the internal glass around the pins, we are looking for the glass to be discoloured.
Do you have a gas hob with spark ignition.

[PaulR]

The screen voltages are 359V. The anodes are 396V. The voltage at C62A is 20V.

I am attaching photos of the errant valve. I cannot get any resistance reading between the grid and cathode

Thanks

[PJL]

Gettering looks poor so most likely it is air. I would clean up the pins and underside of the valve with some servisol or similar.

To see if the valve will settle down you would need to rewire it to run in auto-bias by replacing the 10 ohm cathode resistor with say a 470 ohm 2W one then disconnecting the grid and wiring that to earth via a 1K resistor. Monitor the voltage across the 470 ohms, switch on and do not let it exceed 20V which would be 40mA. With luck as the valve warms up the gettering will absorb the air which should show up as a slow fall in anode current.

[PaulR]

Thank you for the suggestion.

I spoke to my son earlier and he is going to let me know what he thinks about replacing all the valves now it has come down to that. The getters on all of them are bad so it may be as well to think seriously about going down that route.

Ironically he says he doesn't really want it back as he has a Linn amp and pre amp in a unit designed for them downstairs and he hasn't really got room for it in his office. he uses a Rogers HG88 there all day so I may well be asking about replacing ECC807s with ECC83s at some point. I read somewhere that someone had done this and it still had plenty of gain.

It looks as though I will be using the Fisher for the foreseeable future! I might offer to go halves with him over the NOS valves

[qualityten]

That sounds a very happy outcome after all the work you've done.

[PaulR]

Yes it seems fair.

[PaulR]

Well my son announced today that he had bought the NOS valves. Just waiting for them to be delivered

[PaulR]

In view of the amount spent on this set I am pondering having a go at replacing the power supply capacitors using the Audiokarma thread as a guide. Does anyone have any views on this.

Quote "Never leave well enough alone"!

[Radio Wrangler]

Any bulges, leaks, or cracks would make it a good idea. Is here much ripple on the power rail? Unless it's bad, I'd try out the new valves first and see if they fix the bias current problems first, and then do the capacitors, in order to not change too many things at once.

David

[PaulR]

They actually look like the day they were fitted and there is no hum at all. I will test the ripple on the HT when I fit the new valves. The problem would be that everything is very tightly packed and I would probably have to disconnect a lot to fit new capacitors.

I would really rather not do it unless there is advice to go ahead.

[Radio Wrangler]

There are factors pulling in both directions: the fiddlyness and risks of replacement versus it being the sort of tightly packed amp which runs hot and gives its components a jolly good cooking.

If it doesn't need it at the moment, then it comes down to your preference.

David

[snowman_al]

I've been trawling through several of the threads devoted to the Fisher 400 (500/800) receivers on the USA web sites and there is one thing that sticks out, get the bias supply right.

Those by DC Gillespie are particularly pertinent. Summary is that the Fisher 400 runs the output valves at 85-90% of their rating. Fine when the 7868 was cheap and plentiful, but that was 40 years ago.
The general recommendation is to improve the negative bias supply to get a greater negative voltage and hence reduce the dissipation of the output valves (to 65 -70%). That greatly improves their life and runs the whole unit much cooler. General opinion is that the Fisher circuit is flawed, something there have been comments on here too.

I think the way forward is to fit a couple of the new 7868s at a time and measure the cathode current as you have been. Also checking the bias volts on them. If all is well it will give you a rough idea of possible pairs too.
Then depending on the actual current they pass at the nominal bias voltage you have to decide if it is necessary to make it adjustable.
If they are all fairly closely matched just making the whole supply adjustable is easiest.

Once you are sure the new valves are okay, stick them all in and measure voltages along the bias supply chain.
AC volts across CR1.
DC volts at pin 4 of V8/10.
DC volts across R138 (15Ω).
DC volts at the junction of R34 and R35.

It would be worth while getting 2 or 3 1500 (or 1800uF) 35 volt caps as replacements for C60A, C60B and C42. You could then try piggy backing one across the existing C60B and seeing if it improves the voltages.

Certainly one step at a time...

[PaulR]

I think I will leave the power supply caps as they are for the moment at least. I was thinking back to a Sansui 500 I did a while ago, but in that case the electrolytics started to drip electrolyte when it came back into use. These seem ok at the moment.

Andy (Diabolical Artificer) has kindly offered to make up and let me have a separately powered adjustable bias voltage supply which will supply more -ve voltage than the current set up. Presumably I would connect this directly after the supply to the valve heaters. It also depends on whether I can find somewhere to mount it.

I have already replaced C60 A and B and C42 together with the rectifier when I was first trying to get the bias more negative.

I was looking at some Fisher literature from the time it was made and it is described as for "economy minded music lovers" so presumably they did a bit of pound (or dollar) watching when designing it.

[snowman_al]

That's a nice offer from Andy, but I do not see any advantage in more complication.
The transformer you have should be able to supply enough voltage to take a suitable tap off, either at the heater point or even before it.
You need to make the measurements requested. All the Fisher threads use the existing hardware and get a suitable range..

You are correct with respect to price per dollar I'm sure, and suspect the original amp had to make 25 watt a channel for marketing reasons and they chose the bias to make that.
Alan

[Diabolical Artificer]

Pics of tfmr and adjustable bias supply, adjustable from -9v to -25v. Footprint 2 7/8" by 2 1/2". Not sure of the ripple at present, only just finished it,but it should be less than 10mV No worries if you don't need it. The bias supply as fitted allows no adjustment, which means you can't go over -13v, puts Paul in a corner if his re-placement valves run a tad hot.

Andy.

[Diabolical Artificer]

Finished your adjustable bias supply Paul, adjustable from -7v to -25v approx. At -13v @ 7.5mA I get a ripple of 4mV. I had to make the DCR a bit bigger,so instead of about 2k that the original has, it has a DCR of about 5k,should make a difference as you have 330k grid leaks resistors.

PM to follow, Andy.

[PaulR]

Thanks Andy. I am not sure whether I will need to use it yet, but it will be a good back up if necessary. I changed the grid leak resistors to 220K following advice given on this thread.