Electronics onboard Apollo 12 CSM "failed" while the IU kept on working?

[bknight]

For some reason, might be just old age, I laid in bed last night thinking about A12.  One aspect of the mission was the lightning strikes.  I don't know enough about electronics to answer a question.
Why did the electronics on the CSM "fail" and need to be restarted including the SCE to Aux?

[Peter B]

Murray and Cox dealt with that in "Apollo - The Race to the Moon", but I don't have it to hand.

https://history.nasa.gov/afj/ap12fj/a12-lightningstrike.html

This discusses the event but I scanned it only quickly so I don't know if it explains much about the why of the event - specifically why it affected the CM but not the IU.

[JayUtah]

There were two lightning strikes, about 20 seconds apart.  The first tripped a safety switch that disconnected the fuel cells from the bus, meaning that the bus fell back to the re-entry batteries.  The second changed some of the bits in digital logic that accompanied the CM's IMU.  This caused the unit to think the CM had suddenly changed direction, and it gave commands to the IMU that caused it to tumble.

Because the bus voltage had dropped quite a bit, the sensor electronics got confused.  This wasn't fixed with the voltage came back up to a sizable fraction of the normal 28 volts.  Switching to the auxiliary and then immediately back to the main signal conditioner would probably have worked too.

The IU wasn't affected first because it was already battery powered, and thus it was already more robust and didn't need the specific cutout that affected the SM.  And its computer was a lot more primitive and made more use of vacuum tubes and other components that aren't as susceptible to electrostatic discharge.

[bknight]

So basically it was because of the fuel cells dropping continuos power supply.  Then the rest cascaded in the less primative CSM.  I wasn't aware the IU has vacuum tubes.
Thanks as always.

[JayUtah]

I wasn't aware the IU has vacuum tubes.

Actually now that I think about it, I don't think that's correct.  I think it used IBM's version of integrated circuits.  In any case, they were less susceptible to electrical discharges.

[Obviousman]

https://spaceflight.nasa.gov/outreach/SignificantIncidents/assets/analysis-of-apollo-12-lightning-incident.pdf

[JayUtah]

https://spaceflight.nasa.gov/outreach/SignificantIncidents/assets/analysis-of-apollo-12-lightning-incident.pdf

That's always a great read.  However, it talks about everything except what made the signal conditioners go wonky, and what happened to fix them.  Not surprisingly, the authors spend considerable time discussing the guidance system failures and the power system failures.  The signal conditioning electronics enter the picture only because, in hindsight, we look at John Aaron's professionalism and quick response as a slice of heroism.  Popular culture has focused more on that than perhaps the technical analysis warrants.

By the way, I verified -- no vacuum tubes in the Saturn V digital computer.  Now I'm wondering how the idea got into my head that it had them.  But the actual story is even more relevant.  So the AGC hardware was built with bona fide integrated circuits made not unlike how we make them today, by layering substrates with different conductive and semiconductive paths.  IBM took a different approach at the time.  They produced compact PCBs and populated them with standard miniature components of the time -- transistors, resistors, diodes, etc.  Then they potted the whole thing into a "chip" that had standard form factors and standard pin layouts.  These could then be plugged into larger boards and backplanes to build up a more complex device.

But the traces on these boards were farther apart than on the Fairchild-style photographic processes. And the individual components were not only potted but also had their own "packaging" from their use as single components.  This means that electrostatic discharge would have less an effect on these circuits.

[Jeff Raven]

Great information as always, Jay.  Thank you!  Yet another example of the differences between the systems then and now (not to mention from supplier to supplier), and why the "they supposedly went there before, why do they have to test it again?" argument that some CTs throw out makes no sense. Assuming modern components would have the same tolerances as the ones from Apollo is almost as dumb as assuming a modern smart watch would react the same way to the Van Allen Belts as the Omega Speedsters did.

[Bryanpoprobson]

Bit of an aside here but this has reminded me of a problem I had many years ago. I used to work for BT, one of my many jobs was in the Electronic Repair Centre. I maintained some telex signal generation equipment and I had a regular output failure that required changing a Germanium Transistor. The engineers who tested the kit before sending it to me said, "The output was ok when I tested it first time, but then it would fail!"
What was happening was the test probe they used would blow the transistor when they removed it from the test port, there was 5 outputs and 5 transistors. The transistors cost £35 each and I was going through hundreds of them. So I found a Silicon equivalent which was more robust (so would never blow when testing) and only cost £0.32p each. So I sent this off to the BT hierarchy, their reply,
"Continue using the correct transistor as it would cost more to have the diagrams changed!"
I ignored them and as the various units got repaired I never saw that fault again. 
"Oh and I used to modify any units that were in for other problems!"