So I just got to thinking about how one would go about producing a regulated voltage output (say just a linear DC supply) without using a regulator chip and, of course, the first things that come to mind are using zener diodes or some such as a voltage reference. But that got me to wondering how they made well-regulated supplies back in the days before semiconductors. How did you get a decent voltage reference?

Anyone know?

 

I have a few of the old ones built by GFC Hammond and the one I have in front of me now looks very simple, a MC1723 and a 2N6576 pass transistor.
I do have some models that also have the Crowbar O.V. shut down.
Practically all also have output terminals for remote sense regulation.
If you are going back before semiconductor types then they would most likely had tube regulation and the equivalent to the zener back then was the neon tube and thyratrons, the equivalent to the SCR was Thyratron and the Mercury Ignitron.
Max.

 

I had heard that before, but completely forgot about it. Now reminded, I immediately found this

http://www.edn.com/design/analog/4363748/Voltage-references-hold-steady

which was a quick, interesting read.

 

In those days Mullard was still a manufacturing force to be reckoned with and they produced a book

Transistorised Power Supplies, attributed to JR Nowicki.

This has a whole section entitled High-stability reference sources.

I also remember a large book on the subject, when I was a college, produced by Texas Instruments, though I could never afford a copy.

 

In those days Mullard was still a manufacturing force to be reckoned with and they produced a book

I still have my copy of Radio Valve Data with all the characteristics for Mullard, Brimar, Cossor, Emitron, Ferranti, Marconi etc ..etc.

I have worked with both Thyratrons and Ignitrons also a neon counting tube called a Dekatron.

The Ignitron is quite the device, a large steel encased tubular glass chamber that has a main Cathode which has a pool of mercury over it, the phase angle controlled ignitor rests in the pool of mercury and when low current is passed through the mercury pool via the ignitor , it ionizes the gas and conducts to the large anode and switch off as the AC sweeps through the half cycle, for full wave control you need two, back to back.
Max.

 

My mind quickly goes to using a vacuum tube as a constant current source.
ΔIp/ΔB+ is very small when Vgk is fixed.

 

Here is an extract from the Mullard Book.

The give considerable detail of analysis and experimental results as compared to primary sources (Weston Cell)

 

Before semi's.

Max.

 

If you wanted to entertain counting the Dekatron runs on a principle that is parallel to the Stepper motor.
http://www.youtube.com/watch?v=PS9_Ognxiys
Max.

 

The question was

How did you get a decent voltage reference?

Not how do you regulate?

So how decent a reference was V3 and V4?

 

Wasn't the question regarding regulators before semiconductors, i.e., no transistors and no diodes?

 

Well perhaps I misunderstood post#1.
WBahn talked about what would happen without regulator chips.

But its true the title says before semiconductors, which takes us back a very long way if you take that strictly. In fact I am just trying to fix my microwave which gets its regulation from and incredibly massive transformer. So non semiconductors are still in use.

 

heres my valve portable psu and footwarmer schematic and studiot whats wrong with your mw?

 

As I said before it's 28 years old and needs a new magnetron. (it overheats and the cutouts shut it down after a couple of minutes of operation). The HV capacitor, fan and diode test out just fine.

I have looked at new microwaves, but they do not offer the facilities I originally picked this one out for.

 

I have looked at new microwaves, but they do not offer the facilities I originally picked this one out for.

What? Using microprocessors doesn't do what you want? How could that happen?

ps, there's a person (just al) in Electronics Resources forum right now with the same overheat problem.

 

The question was

Not how do you regulate?

So how decent a reference was V3 and V4?

Neon regulators were pretty good IIR, the output was considered at the time to be 'Fairly Reasonable' ?
I don't have numbers, unfortunately.

But its true the title says before semiconductors, which takes us back a very long way if you take that strictly. In fact I am just trying to fix my microwave which gets its regulation from and incredibly massive transformer. .

The method used is also an old one, part saturation by magnetic shunt.

Max.