LESU - model 28 Electrical Service Unit located in back of the motor in RO, KSR, and ASR cabinets. Could have line relay, autostart relay, internal loop supply, selector magnet driver, etc. In NON-TEMPEST units the external high-level (current loop) signals went through the LESU and then to selector magnets and keyboard contacts. In TEMPEST units, the LESU handled AC circuits but not any signals.

ESA - Electrical Service Assembly was a double shielded gold-colored box used in TEMPEST units. Contained cards to convert external low-level (+/-6v) signals. Included a selector magnet driver (SMD) connected via shielded cable to shielded selector magnets. For KSR and ASR it included a low-level keyer (LLK) to convert very low current keyboard signals (via shielded cable) to +/-6v signals. The LLK connected to the outside world and also to a second SMD input so that the printer would echo the keyboard. This low-level link was wired inside the ESA box. There was also a separate clutch magnet driver (CMD) ESA for driving the keyboard and TD clutches to sync with crypto equip pulses if necessary.

On Sun, Apr 9, 2023 at 11:32 AM Duncan Brown <duncanancy@earthlink.net> wrote:

On Sat, Apr 8, 2023 at 7:43 PM Duncan Brown <duncanancy@earthlink.net> wrote:

>Low-level signalling! There is another variable and reason you can't put
>TX & RX all on one loop.

On 08-Apr-23 21:04, Nick England wrote:

Normal low-level connection was with the keyboard feeding one of two inputs to the printer.

At first, I didn't understand Nick's statement.
But then I realized it is another Army/Navy, Kleinschmidt/Teletype Corp. difference.

A M28 KSR/ASR always had a an ESU with a local loop and a line relay for isolation from the line. So you could have a low level "loop" feeding the M28 and then the ESU converted the low level to high level to feed the selector magnets.

Kleinschmidt machines never included a isolating line relay. Except at the large comm centers, they normally connected directly to a wire line modem or FSK converter. A low-level TX signalling connection was made directly to to the crypto equipment which would typically be only a few away. There was no low-level signalling on the RX side from the crypto gear. From the 1950s through the 1980s, crypto gear was KW-9 (until mid 1960s), KW-7, & KW-26.

Duncan
K2OEQ