Connecting EZ81 Together

[shiny phil]

Reading the data sheets for ez80/81,a limit of 50uF for smoothing C. An idea (stupid?) sprung to mind. Use silicon rectifiers to get HT volts. Connect, say 10ohm R in series with each anode (ht current sharing) of the EZ81. Connect the remaining ends of the 10r to the silicon recs. The cathode to the smoothing components. The ez81 diodes are connected in parallel. In series with the ht. Does this double its current rating? ,and allow a bigger smoothing C than 50uF to be used? .The ez81 will allow the ht to rise slowly as usual.

[glowinganode]

Sounds like a plan and should work.
I wouldn't bother with the 10r sharing resistors as the effective internal (anode) resistance is likely to be a lot higher than that. The two halves should be pretty well matched anyway.
This sounds like a leading question, why do you ask? Are you trying to cure the symptom of a different fault?

Rob.

[Alan Stepney]

Why would you want more than 50uF for smoothing?

[Silicon]

I intend to try this one day. I would use silicon rectifiers to charge up the reservoir capacitor (check surge voltage rating of the capacitor).
This would be connected to the anodes of the valve rectifier.
The cathode would be connected to the smoothing capacitor.

The circuit would give a slow warm up and the valve rectifier would be working at lower voltages. An arc-over between anodes and cathode in the valve rectifier would cause less damage than the conventional arrangement.
The crest factor of the current pulses through the valve rectifier would be much reduced.
I would expect the voltage drop across the valve rectifier to be reduced.

[kalee20]

Got it now - you are just using the EZ81 to pass DC, after the reservoir capacitor. Yes, voltage drop, peak current, etc will be much less than the conventional arrangement.

Originally Posted by Silicon

An arc-over between anodes and cathode in the valve rectifier would cause less damage than the conventional arrangement.

Absolutely, yes, couldn't agree more. BUT! what about breakdown in the silicon rectifiers? Personally I would trust the valve far more readily. A breakdown in the silicon is likely to be permanent. Goodbye mains transformer, reservoir capacitor! Whereas a flashover in the EZ81 is likely to clear almost immediately. I've never known this happen in receiver valve rectifiers.

[Maarten]

The most likely permanent breakdown faillure mode in valve rectifiers is a short circuit between heater and cathode. It doesn't happen often, though. Only with a few selected models of which the EZ81 isn't one.

[shiny phil]

Originally Posted by Alan Stepney

Why would you want more than 50uF for smoothing?

No particular reason. except more capacitance gives better HT smoothing. less hum.

Originally Posted by glowinganode

This sounds like a leading question, why do you ask? Are you trying to cure the symptom of a different fault?

I woke up and this idea came to mind. A light bulb moment may be?. Or more likely. During Xmas through to new year I've taken a liking to Emperador wine

[shiny phil]

Originally Posted by Alan Stepney

Why would you want more than 50uF for smoothing?

within reason: cost, size of cap . more smoothing is better than less

[Radio Wrangler]

In some ways better, but not all.

Increasing the size of a smoothing cap means its voltage droops less between charging cycles. This in turn means that there is less time left for the re-charge activity, so the recharge current has to increase... so the current seen by the rectifier becomes a progressively narrower pulse of progressively greater current as the capacitor is increased.

This increases the losses in the rectifier and all the series components feeding the rectifier.

You've usually got room to increase smoothers, but if you do it a lot, you need to check stresses on other components.

David

[shiny phil]

Originally Posted by kalee20

Absolutely, yes, couldn't agree more. BUT! what about breakdown in the silicon rectifiers? Personally I would trust the valve far more readily. A breakdown in the silicon is likely to be permanent. Goodbye mains transformer, reservoir capacitor! Whereas a flashover in the EZ81 is likely to clear almost immediately. I've never known this happen in receiver valve rectifiers.

I see your point. I looked on the internet, found VOX AC4 (EF86 EL84 EZ80) 4 watt guitar amp - 1960's. In some versions the EZ80 was replaced with solid state rectifier diodes 2 series diodes / valve rectifier diode 4 in all. Maybe VOX had reliability problems with the diode going s/c back then.

Would putting say 1000pf/1kv cap across each rectifier tends to equalise the transformer HT volts across the diode during the non conduction part of the cycle. Also reduces diode switching noise

[Herald1360]

I don't think 1000pF would have much voltage equalising effect- it's very small compared with the time period of reverse polarity on the diodes. It would reduce the reverse recovery hash though. Adding 1M resistors as well would have more equalising effect and wouldn't upset anything else.

[kalee20]

Originally Posted by Radio Wrangler

Increasing the size of a smoothing cap means its voltage droops less between charging cycles. This in turn means that there is less time left for the re-charge activity, so the recharge current has to increase... so the current seen by the rectifier becomes a progressively narrower pulse of progressively greater current as the capacitor is increased.

Correct! Although the way you have written it, it reads as though the current rises to infinity and the pulses fall to zero time, as the capacitor is increased without limit.

In practice, although the current does rise, and the charge time does reduce, the circuit series resistance puts an upper limit to the peak current. So you could increase capacitance without fear - ripple will tend to zero, but the peak current won't exceed this limit, although it will get ever closer.

What can hurt, is the duration of the surge when you switch on. So the advice to not use an excessive capacitance is still good, but for a different reason.

[G8HQP Dave]

In many PSUs the amount of hum is set by the grounding arrangements, once you get past normal sensible capacitor values. Most people will instinctively get the grounding wrong, unless they stop and think about where the charging pulses go and remember that every piece of wire has resistance. Old radios and guitar amps usually get it wrong, but back in those days some hum was regarded as normal (or, dare I say it, preferable?).

[turretslug]

Very much so, up until maybe the 'sixties there seemed to be a somewhat cavalier attitude to grounding, commoning etc. with a supposition that chassis, wiring and so on possessed zero impedance from DC to daylight. I wonder if the arrival of solid-state circuitry with its higher currents, low impedance, even DC coupling finally focussed sharper scrutiny on this area.

Certainly, attention to grounding detail can unlock lower hum performance that eludes any amount of extra microfarads. "Traditional" components, such as multi-section electrolytics with one negative terminal, aren't always helpful.

[G8HQP Dave]

Yes. The big mistake, which appears in almost all early domestic stuff (and some better things), is to wire the HT secondary CT straight to the nearest chassis point instead of to the electrolytic negative. This immediately injects charging pulses into the chassis and hum-free operation then becomes impossible.

[turretslug]

Even my Eddystone 750, a pricey set in its day, suffered this puzzling complacency- the output of the HT centre tap fuse was connected to a random chassis point via several inches of wire. It was the work of a moment to reconnect it to its appropriate, correct place- the negative tag of the reservoir capacitor, about an inch from the fuseholder! This completely eliminated an admittedly very low level but discernible warble. I can't believe that the designer(s) intended this discrepancy, I imagine it creeps in somewhere between prototyping and production floor. Another time, I silenced a friend's SRX30 in almost exactly similar circumstance.

Re. the original post, this just goes to show that not only does PSU wiring detail provide a more elegant (and many would say better) solution to hum reduction than "more uF", it's also free! (bar thinking time).

[G8HQP Dave]

I suspect that most designers were merely copying the current state of the art. Everyone randomly grounds the CT so that must be OK; a little hum is 'normal'.