Email or username:

Password:

Forgot your password?
All posts Ken's posts Post Back to profile
Ken Shirriff

The Bendix CADC is an analog computer used by fighter planes in the 1950s. It computed airspeed, Mach number, and other important parameters. I reverse-engineered how it performed these calculations with tiny gears, differentials, and cams. 1/12
11 February at 17:47 | Open on oldbytes.space
21 comments
Ken Shirriff

The basic idea is that specially-shaped cams convert values to logarithms. Differential gears add and subtract these values. Other cams can take exponentials or other functions. 2/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

The central "Mach section" computes log static pressure, log pressure ratio, impact pressure, and Mach number. This diagram shows how the calculations are performed with differentials βŠ• and cams. 3/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

Here's a closeup of the gears and differentials that perform the computations. 4/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

The outputs from the CADC are transmitted electrically to other parts of the aircraft.
Cylindrical devices called synchros convert shaft rotations to 3-wire electrical signals. This diagram shows the synchros and other components inside the CADC. 5/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

The CADC receives pressures from the aircraft's pitot tube. One problem is that the airflow over the aircraft distorts the static pressure reading. The fix is to apply a Pressure Error Correction factor. This factor is determined by the Compensator, an external unit. 6/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

The Compensator uses a 3-D cam to generate the correction factor, sent to the CADC as a synchro signal. The CADC uses a servo loop to generate a rotation corresponding to this factor, then multiplies it with a differential. 7/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

An amplifier board uses transistors and magnetic amplifiers to drive the motor that turns the shaft. They didn't use a printed circuit board, but components soldered to metal pegs. Here's the schematic I reverse-engineered. 8/12
11 February at 17:47 | Open on oldbytes.space
Ken Shirriff

The CADC is a boring cylinder from the outside. You wouldn't expect that amazing mechanical complexity would be found inside. 9/12
11 February at 17:47 | Open on oldbytes.space
Mars 1024 πŸŽ„πŸŽ

@kenshirriff : Wow! You are a legend! 😲😲😲

Thank you, kind Sir, for this knowledge!

11 February at 18:17 | Open on techhub.social
Erin

@kenshirriff I understand the idea / how a function is encoded in these cams, but how you'd manufacture them isn't clear to me and isn't coming up easily in search (I'm getting a lot of cam shaft results, which seem relatively simple compared to the cams in an analog computer). Do you know about any resources on how such cams were made?

12 February at 1:20 | Open on mastodon.social
Ken Shirriff

@ChateauErin I think they would machine the cams the same way they would make other metal parts with a complex shape. E.g. using a milling machine and a template.

17 February at 18:19 | Open on oldbytes.space
Erin

@kenshirriff thanks. I don't know why I didn't think of milling machine; part of me knew that they were plenty old. "and a template" was something I absolutely didn't consider except abstractly (I'd imagined something kind of pantograph-y) so this should hopefully be enough to send me down that rabbit hole

17 February at 21:35 | Open on mastodon.social
DrScriptt

@kenshirriff do you know if this is the same Bendix as the Bemdix G-15 (?) that Usagi Electric is restoring?

11 February at 18:59 | Open on oldbytes.space
ageofcool
Volvodadfast

@kenshirriff Thank you for sharing this. I wonder if the reason for doing things this way was for radiation hardening.

11 February at 20:37 | Open on mastodon.social
Go Up