I should have been clearer when I wrote, "I have some 1960s power supplies using glow tubes and triodes which would benefit from upgrading to Zeners, since I don't need the magical glow.".
My point was replacing the VR tube with the Zener, still driving the shunt triode, instead of using the VR tube as the voltage standard. So the Zener is running at low current, below 40 mA which is the VR tube's limit, just like the VR tube. A low-noise Zener may be used for this purpose which, being low current, is not expensive, and can be stacked for higher voltages while maintaining low noise.
Regulation is always an improvement.
But, again, the question is: how does the noise of the VR tube compare to a (a) standard (relatively noisy) Zener and (b) a low-noise Zener. All have statistical properties, and thus noise, but the VR tube is going to be a lot worse.
A fascinating debate here.
Well, I am not injecting noise into the AF amp, that is pretty much immune to it, once decoupled.
All I am after, is rock steady voltages no matter what giant loads are being swung around upstream.
Also using a cathode follower (de rigeur for regulators) is going to reduce what little noise is there.
When you look at the solution involving decoupling different noise sources through the valve rectifiers, chokes, then injecting it into a cathode follower, you end up with HT which has no 50hz present, no voltage variations or ripple, and has no reverse feedback reactions with varying loads in the ouput stage.
Reminds me of Lotus cars philosophy.
Make the chassis rock solid with no torsional distortion, halve the weight, and add performance.
Nothing much has changed in either successful sphere since the early 60s.
Suprising really.
Nonlinearity does not have much of a filtering effect.6vheater said:
I never said that gas regs are noisier than zeners. I implied that gas regs may be noisier than smoothing capacitors. Do you agree? Gas regs can suffer from low frequency variation, as the discharge moves around the electrodes.
Most older radio amateurs are aware of that, as the same effect occurs in tetrode PAs.
You mean that the circuit design was such that the 'correct bias' window was so narrow that regulated supplies were needed? A redesign might have been less work than providing a regulated supply. In my opinion about the only audio stage which needs a regulated supply is a phono preamp, in order to stop mains voltage variations from causing woofer flapping.
A CF merely passes on whatever noise is given to it.6vheater said:
"No" is of course an exaggeration, but we all knew that.
Wrong.
An amp being driven heavily in AB2 with direct coupling to the grid, requires a very stable supply, far more than any preamp will.
In order for it also to get maximum linearity it also requires excellent stabilisation at the screen grid.
If the 3rd supply sags, (HT) that also makes the curves tip over at full load, so a perfect candidate for choke input filters to feed the anodes, because CIF handle heavy loads much better.
I notice on the Hiwatt 100, 200, 400, they used seperate windings to provide the same kind of stability even with capacitors.
Having said that, they were very well known for blowing up KT88s.
If you are going to run stuff well on the ragged limit then it has to have all that extra safety margin.
I don't see many amps running a pair of 807 doing 85W RMS, without breaking a sweat.
I guess we are way OT now.
Last edited:
How many people on here are building high power PA or AM modulators and so wish to use AB2?6vheater said:
We don't see many amps running 807 at any power level.
Last edited:
Actually Peter Millett used 807s for a well known amplifier of his, but it was the modern STC version.which are dead easy to get NOS and cheap.
You can buy the STD PP driver board PCBs for it off the shelf.
CHEAP.
Many people use the Octal TV v-scan version of 807, the 6BG6 - again dirt cheap, and some people discovered the small Noval QQV0xxx VHF transmitter valves for audio, because they are ultra high quality NOS Mullard valves costing peanuts.
Just ONE will do 12W of audio in PP, and at +30V on the g1 they can conduct more than 1/4 of an amp at 100V, but are low gain at only 3.3mA/v so need loads of low impedance voltage drive.
Look, I got these 4 for less than 20USD, they contain 2 beam tetrodes in 1 glass, and they are genuine Mullard! What's not to like?
Transmitter valves just love having the g1 driven heavily positive, and now in the age of mosfets it's dead easy.
Lots of people are totally unaware the 6V6 is an excellent transmitter valve, which works very well in quads (PPP), and on top of that, has excellent linear characteristics when driven in AB2.
There are curves for that, but you will have to hunt for them.
Instead of 8-12W you can multiply that by 2.
I have 4, 6BW6 here all NOS, which cost me less than 10USD a piece,
They are tiny. It's a 6V6 this time in a tiny glass.
They were used in V-scan circuits which are in fact merely a fixed frequency class A amp.
DC coupled they can happily drive a 3K ohm A-A load in AB2 push pull, and make 25w per channel, with very low screen grid currents, and will outperform EL84, but that requires a proper PSU.
One reason for direct coupled output valve grids from cathode followers is the excellent linearity, and low idle current demand. That's the perfect candidate for choke input filters.
Just because lots of people do the same old thing, doesn't mean you can't do it much better.
Just saying, (complimentary to the topic in hand, because you can seriously up the peak currents without overloading the mains transformer this way).
You can buy the STD PP driver board PCBs for it off the shelf.
CHEAP.
Many people use the Octal TV v-scan version of 807, the 6BG6 - again dirt cheap, and some people discovered the small Noval QQV0xxx VHF transmitter valves for audio, because they are ultra high quality NOS Mullard valves costing peanuts.
Just ONE will do 12W of audio in PP, and at +30V on the g1 they can conduct more than 1/4 of an amp at 100V, but are low gain at only 3.3mA/v so need loads of low impedance voltage drive.
Look, I got these 4 for less than 20USD, they contain 2 beam tetrodes in 1 glass, and they are genuine Mullard! What's not to like?
Transmitter valves just love having the g1 driven heavily positive, and now in the age of mosfets it's dead easy.
Lots of people are totally unaware the 6V6 is an excellent transmitter valve, which works very well in quads (PPP), and on top of that, has excellent linear characteristics when driven in AB2.
There are curves for that, but you will have to hunt for them.
Instead of 8-12W you can multiply that by 2.
I have 4, 6BW6 here all NOS, which cost me less than 10USD a piece,
They are tiny. It's a 6V6 this time in a tiny glass.
They were used in V-scan circuits which are in fact merely a fixed frequency class A amp.
DC coupled they can happily drive a 3K ohm A-A load in AB2 push pull, and make 25w per channel, with very low screen grid currents, and will outperform EL84, but that requires a proper PSU.
One reason for direct coupled output valve grids from cathode followers is the excellent linearity, and low idle current demand. That's the perfect candidate for choke input filters.
Just because lots of people do the same old thing, doesn't mean you can't do it much better.
Just saying, (complimentary to the topic in hand, because you can seriously up the peak currents without overloading the mains transformer this way).
Attachments
Last edited:
I have quite a few QQV0x/y (including one of the the smaller ones I used in a 144MHz Tx many years ago) and a box of 6BW6. I even have some 5B/245M (miniature 807) and bases for them.
Same here, you simply can't go wrong buying transmitter valves because Audiophools turn their noses up at them, so they go for next to nothing.
As I say, I bought a whole box full of 807 for about 1USD each.
All I had to do was stick them in an amp and check each one individually for bias, -scrap the dud ones.
Even some of the nice NOS chokes and 100V line transformers people say are scrap metal, little knowing 100V line distribution transformers often make superb output transformers, costing next to nothing.
The older US made chokes/transformers esp mil spec are top quality, and last forever.
The heart replacement has arrived.
Did it the long, hard way and measured the currents separately. Was a bit of a timing issue as the failed power transformer could be operated for approx. 1 minute at best before it gets too hot again.
Now the amp is back on the workbench again:
And the heart replacement has arrived, now I need to fit it in:
Regards,
Rundmaus
Did it the long, hard way and measured the currents separately. Was a bit of a timing issue as the failed power transformer could be operated for approx. 1 minute at best before it gets too hot again.
Now the amp is back on the workbench again:
And the heart replacement has arrived, now I need to fit it in:
Regards,
Rundmaus
That Toroid label shows 0.460kVA,
And 225A, 150A, and 2,000A.
I believe that is mA, not A.
Otherwise you have discovered "Perpetual Motion"
Those Toroids will sell out real quick!
And 225A, 150A, and 2,000A.
I believe that is mA, not A.
Otherwise you have discovered "Perpetual Motion"
Those Toroids will sell out real quick!
Yes, the tricky topic of decimal limiters in different countries... 😀
The transformer was made by a German company using the usual German decimal delimiter "," instead of the English decimal point.
So the label reads 0,225A 0,150A and 2,000A corresponding to 225mA 150mA and 2000mA. Unfortunately, no new physics involved here. 😴
Regards,
Rundmaus