Space race PSUs

Yup!

High level overview of one system here:

Enjoy this series -- they test the power supply in this one, but it's well worth watching all of them.

Going back further, no idea, you'd have to find documentation on it, if it's out there. Best case, someone's done a video series about it and there you go; or it's on archive.org or something like that. Worst case, file a FOIA regarding a particular mission and see what they have (or ask the e.g. Library of Congress? where the docs are, if they're already public).

Tim

Great, thanks!

In the V rockets the Germans used magnetic amplifiers and hence everything was based on AC -- but long-term operation wasn't exactly a goal. ;-)

Best regards, Piotr

Here is the Apollo Guidance computer 28V to 4V buck switcher:

I would love to know more about the earlier Gemini (IBM all descrete) computer but there is much less info on the web than for Apollo. From what little seen I suspect that had DC-DC transformer converter for the logic supplies so kind-of switch mode too?

piglet

N.B. Here is the AGC power supply for the DSKY electroluminescent display, HV AC square wave. Note the mag-amp:

piglet

The AGC and hence man's trip to the moon was made possible by decent-quality PNP transistors for current sources.

Much of that hysteric buck converter seems designed around having to use an NPN for second gain stage amp Q2 in a way that would be considered awkward today, probably because without it the diff amp wouldn't have enough gain to perform well at the desired switching frequency

Or more accurately, to make Sziklai pairs from

All the PNPs in that circuit are saturated switches.

They can't have been that great - needed three in parallel Q9 Q10 Q11 to drive the NPN power switch.

piglet

Not sure the selection was as limited as you suggest. Probably the earliest that design was under consideration was about 1963. We had lots of good parts by then. E.g., here's a story about Fairchild's 2N1613, silicon planar NPN in TO-5 package, introduced in 1960 at Wescon. By 1963 there were multiple second sources.

Ya I had thought Q4 was acting as some sort of load for Q2, it's not I neglected to notice C6 it's for sync-signal-injection of some kind.

I couldn't see initially how the diff amp works at all with Q2 connected that way from a DC-bias perspective but the AC voltage swing at that node will be small so I guess there's enough headroom.

I was mistaken piglet is correct there aren't any PNP current sources in that circuit.

I didn't start learning about analog design until ah, more recently so it's an odd-looking circuit to my eyes, when I see solid state designs from that era I guess my go-to is "I hear the PNPs sucked" like you have to do weird stuff to shove an NPN in Q2s position, if you were designing a simple high-gain diff amp today you'd likely use a current-source loaded PNP and be done with it.

Thank you very much, piglet!

Best regards, Piotr

The more I read it, the more amazed I am. E.g. the three phase AC synthesis using 8 steering waveforms and a bunch of windings connected in a crazy way. Not very unusual today, but a top notch engineering back then. Thanks again!

Best regars, Piotr

Interesting design which cancels out low harmonics and high harmonics are then easy to filter out.

Each unit delivers only 1.25 kW but weights 20 kg and are quite bulky.Then there are three such units. Apparently there are some critical AC loads that would justify the use AC aboard Apollo.

The 115/200 V 400 Hz is a standard aviation voltage, so apparently some existing AC devices were used.