Fairchild 670 PSU for tube preamp

[Cqwet Dbdfte]

Not really, the FET is controlled thru its source, not the gate. The topology is called "grounded/common grid/base/gate".
Works good. Idea is remove any AC signal from the gate and let the source be the input. A slow RC constant can additionally be used to control startup risetime.

 

[Walter66]

High voltage supplies based on transistors, ic's etc, often need some protective measures added..The function of D4 , C7, R5 and M2 , is to limit the voltage across the LT 317, during start-up and turn-off. Best regards

By using tube rectification (GZ34) instead of a sand rectifier, this means those parts will be obsolete?

 

[Walter66]

In fact, introducing the C7 will be somewhat detrimental to the operation of this part of the circuitry. The mosfet is in there to handle the steep voltage-rises and drops at turn-on and turn-off. Attaining close to zero ripple, before voltage-regulation, should be fairly easy with the compact electrolytics available now-a-days. If that's a concern, the easiest way to get less ripple in the proposed "Maida" schematic, before the regulator, would be to use a normal bridge-rectifier ( with a different, higher voltage transformer ), instead of the voltage-dobbler arrangement, which was probaly only used because a relevant transformer was not at hand that day Mr. Maida made the thing ( ! ) Voltage-dobbler psu's will require ridiculously more Electrolytics/capacitance than the most commonly used bridge-rectifier psu's, to attain the same ripple-levels. Best Regards

It wouldn't be a problem to reduce the input ripple of LM317 by using a simple BJT (BUX85G) in a cap multiplier circuit combined with a cap-choke network at the input.
And by using GZ34 instead of a sand rectifier, there shouldn't be a need for slow electronic ramping of the voltage, isn't it?

 

[Walter66]

I ran a sim on the regulator above, it showed a 100dB rejection of 120 Hz ripple voltage, using a ~10uF cap on C7 improved about 20 dB. No C8.
A cap multiplier is a simple device, whereas an amplifier w/reference has much more effect, LT317 has over 80dB rejection, gradually dropping to 70dB at 10kHz.

The 6.2V zener may not be enough for using a HV FET depending on its threshold voltage. The 317 needs to dropout min ~2.5V , 5V better, plus the threshold of the FET, commonly around 3V, so a 10V or so zener would work.

A cap multiplier is a simple bandaid we can leave out in a discussion about robust voltage regulators.

Many thanks for the simulation.
So you think the FET input (original Maida schemo) would be preferable in regard of achievable SNR vs. my cap input circuit?
I'd like to put a C-L network in front of both to further improve the performance.

 

[Cqwet Dbdfte]

Hand Drawn schematic:
You can't use 200V 1N4148 for this.
The 317 will die instantly, max 37V.
D3 is necessary.
No no point to use choke and vacuum rectifier. A proper HV regulator will make them look silly.

Load separation thru 1k's and several 820uF addresses a non-existent problem.

The schematic above it should be OK.

 

[Walter66]

Thanks, will give a try on the original.
Whats this 4.7 Ohm in series with the 470nF all about, a tuned Network? Can it be replaced with a 10uF For example?

 

[Cqwet Dbdfte]

That 0.46uF and 4.7R would be to stabilize the gain in the regulator, I never used it, reg would be stable without it. 10uF should work.

 

[Wilhelm]

By using tube rectification (GZ34) instead of a sand rectifier, this means those parts will be obsolete?

No, really, i would not recommend doing that,( the designer may have had other good reasons for including it) . If there's a centertapped power-transformer at hand, i would rather choose to simply use 2 standard high voltage diodes for rectification, and perhaps 50/100 % added capacitance before and after the choke. But i would NOT swap the EL 34 ( V302) adjustable series-regulater for any active semiconducter circuits ( as i've had to replace/repair a few of such in other equipment ) In place of the 5651 voltage reference-tube, i would, for improved voltage-stability, absolutely use a combination of series connected medium-power zener's to get same voltage reading. (the adjoining resistors will of course need to be adjusted for this to function) - On the subject of using a tube rectifier, like the GZ34 ( if not for historical reasons/exact copying) : it will only, ( in my opinion) make sense in guitar-amps, where it's series resistance, under the influence of somewhat shifting temperatures/current-levels, when played, can give such amps an added "natural"/"organic" compression, - especially when driven to their limits. Best Regards

 

[Cqwet Dbdfte]

GZ34's are getting very expensive and rare, where a silicon rectifier and regulator will do a better job than all the chokes and junk thrown in.
Example of a simple regulator with "retro" dual 22uF cap input filter outputting 18V pp AC.
PSU Ripple rejection over 140dB. Should be better than the geezer solution.
Output is set to 380V. Startup ramp is 15 sec with those values.
LT1085 model has ~20dB lower ripple than TI's LM317.
Use the LM317 to set up and test the circuit then replace with the $5 part.
I'll make a PCB if anyone is interested.

 

[Wilhelm]

GZ34's are getting very expensive and rare, where a silicon rectifier and regulator will do a better job than all the chokes and junk thrown in.
Example of a simple regulator with "retro" dual 22uF cap input filter outputting 18V pp AC.
PSU Ripple rejection over 140dB. Should be better than the geezer solution.
Output is set to 380V. Startup ramp is 15 sec with those values.
LT1085 model has ~20dB lower ripple than TI's LM317.
Use the LM317 to set up and test the circuit then replace with the $5 part.
I'll make a PCB if anyone is interested.
View attachment 123193


View attachment 123192

There's really no lack of ideas on how to do "a Fairchild 670 - with some updates" . Either way - all psu's have both pro's and con's. Some will look good on paper - and in software simulations - and others will just function problem-free for generations. Naturally, most of us will prefer something that's lightweight/portable, when the equipment is for live-jobs. Of course you're right ; all the "old junk , like chokes and tubes and what have we" could be replaced with a few ic's, and whatever power-supply needed, could be done in switch-mode style - so nobody would have anything to complain about. (That's what can be had if one buy's a compressor-pedal ) Best Regards

 

[Cqwet Dbdfte]

I REALLY like tubes for their properties as amplifiers, but as rectifiers they are a compromise.
IC's are great for power supplies, where the output is a big fat zero.

 

[Wilhelm]

I REALLY like tubes for their properties as amplifiers, but as rectifiers they are a compromise.
IC's are great for power supplies, where the output is a big fat zero.

It's obvious that we agree on all this. When recreating / reissuing vintage gear though, we may consider keeping a selection of design-specifics identical to the original - unless they have shown to cause trouble. ( In some cases, just for psychological reasons/ resale value ) Best Regards

 

[Cqwet Dbdfte]

The original designers did not use big chokes, vacuum rectifiers, and small electrolytic caps because they wanted to. It was because they had to.
There is now very little reason to settle for compromises done in the 1940's in an area that has no negative effect on the output, and has better performance, reliability and cost. I'm sure they would kill to have present days parts.

All old parts had some degree of "trouble", chokes/reactors are heavy, expensive, and needs shields for stray fields, tube rectifiers has low current limits, generates substantial heat, increases transformer core loading, expensive if available, large voltage dropouts. Electrolytic caps have likely improved, and with less heat from rectifier tubes they will have reduced thermal stress.
Let's not forget 10% carbon comp resistors and paper/foil caps.
Most of the magic happens behind the microphone.

 

[Wilhelm]

And back to drilling, welding, soldering, and playing ( Have Fun !

 

[Walter66]

GZ34's are getting very expensive and rare, where a silicon rectifier and regulator will do a better job than all the chokes and junk thrown in.
Example of a simple regulator with "retro" dual 22uF cap input filter outputting 18V pp AC.
PSU Ripple rejection over 140dB. Should be better than the geezer solution.
Output is set to 380V. Startup ramp is 15 sec with those values.
LT1085 model has ~20dB lower ripple than TI's LM317.
Use the LM317 to set up and test the circuit then replace with the $5 part.
I'll make a PCB if anyone is interested.
View attachment 123193


View attachment 123192

Thanks for simulating this! The Infinion FET is nearly unobtainable, any ideas on a spare type?

 

[Tubetec]

I cant fault anything in the forgoing discussion ,
but I have taken the old geeser approach with x10 smoothing caps which gives me a ramp-up lasting around as much time as it takes a tube circuit takes to reach thermal equlibrium ,
you still have to be really careful about how you arrange the passive components for maximum screening ,
In one pre psu I did , theres a choke unused ,disconnected ,yet it provides electromagnetic screening ,so it remains in place , its not a boast ,but if I move it the hum level in my audio circuit will almost certainly rise from non existant .

I did a moderniseation of the classic HP200CD oscillator , swapped out the vintage electrolytics for x10 value or more modern components in a similar package size ,
I adjusted the value of the first filter cap after the rectifier , which I adjusted on test a metal foil motor start cap to hit the target HT volts ,

I remember eliciting a comment or two from our former member PRR on the subject,
and he said the HP200 always displays an instabillity at the fundamental the mains arrives in on ,
after buffering up the supply with x10 cap values, she's rock solid at mains funadamental .

Cqwet Dbdfte,
I have to admit having trouble pronouncing that ,
Pint of Genius for your contributions sir!

 

[Cqwet Dbdfte]

Thanks for simulating this! The Infinion FET is nearly unobtainable, any ideas on a spare type?

https://www.digikey.com/en/products/detail/infineon-technologies/IPA95R1K2P7XKSA1/9491055 looks like 500+ in stock.

I'd use anything 600V or higher, super high current not needed, 5 - 6 A or so OK, which means lower gate charge and gate capacitance. The model used in the simulation is similar, but now obsolete.

I made a search on Digikey, and got ~ 175 suitable candidates. https://www.digikey.com/short/fp8qvz25
I'd choose fully insulated case to have one less thing to mess with.
The "317" normally does not need a heatsink, but I'd place on a chassis (with insulation kit) anyway, in case of an oops.
A steel chassis can be doubled up with a good piece of 1.5mm, or so, aluminum sheet, to improve heat transfer, and avoid added objects sticking out, but a personal taste thing. Finned heatsinks inside a box not so effective.
The circuit in the sim above lacks a current limiter, but as I found out, my chassis mounted PSU resistors, FET and 317, handled the a short nicely. I'm sure a rectifier tube would have fried.

 

[Cqwet Dbdfte]

A friend gave me a dirty late 1950's HP200ABR, a rack-mounted version, over 30 years ago, and it has been sitting in my garage since. I took it out recently cleaned it up and lubricated the gears. It was grimy as hell, some broken tubes, leaky coupling caps, shorted electrolytics. I found a substitute pentode in my collection, and got it working. Replaced all old coupling caps with Mylar or PP types, rectifier tube taken out replaced with 4007's, 360R chassis resistors in series with the 4007's. Substituted the old 22uF's with 220uF, and added a the regulator circuit above, after the rectifier.
Voltage was not very high so I used a TIP50 NPN that I had a bag of, may switch to FET when I get some time.
The update also included rewiring the pilot light and using a voltage divider on the heater voltage for ground reference, along with the changes outlined in the HP manual.
After a couple of hours of warmup the frequency and amplitude are pretty stable considering the 1940's tech.

 

[Walter66]

https://www.digikey.com/en/products/detail/infineon-technologies/IPA95R1K2P7XKSA1/9491055 looks like 500+ in stock.

I'd use anything 600V or higher, super high current not needed, 5 - 6 A or so OK, which means lower gate charge and gate capacitance. The model used in the simulation is similar, but now obsolete.

I made a search on Digikey, and got ~ 175 suitable candidates. https://www.digikey.com/short/fp8qvz25
I'd choose fully insulated case to have one less thing to mess with.
The "317" normally does not need a heatsink, but I'd place on a chassis (with insulation kit) anyway, in case of an oops.
A steel chassis can be doubled up with a good piece of 1.5mm, or so, aluminum sheet, to improve heat transfer, and avoid added objects sticking out, but a personal taste thing. Finned heatsinks inside a box not so effective.
The circuit in the sim above lacks a current limiter, but as I found out, my chassis mounted PSU resistors, FET and 317, handled the a short nicely. I'm sure a rectifier tube would have fried.

Many thanks for teaching the forum about those circuits and giving away some of your knowledge. I'll have to order some additional parts for a build and test, thanks for sourcing the FET.
In general, tubes are very robust. Don't have tried shorts yet, I always test with monitoring the current and have a regulated soft start from 0V to max AC when finished a new circuit. Just an isolated and regulated AC transformer in front of the circuit.