For a possible future project, I made hi-res photo's of the amplifier board in the delay line from an Olivetti Programma 101 calculator. I never had the complete P101, the delay line is the only bit given to me long ago. So some non-destructive experimenting seems a responsable thing to do, like interfacing it to an Arduino to demonstrate the delay operation.
My Delay line is an "L.M.S - UME/2", with a four transistor amplifier board. Other delay lines appear to have a six transistor board.
According to Wikipedia, a total of 240 bytes of information were electrically stored in magnetostrictive delay-line memory, which had a cycle time of 2.2 milliseconds.
The P101 is based on discrete components like transistors, resistors and capacitors. Its main operating voltages were +20V and -10V. From *1, it became clear that 20V is used as the logic voltage. The electrolitic capacitors have 25V as maximum voltage and none is directly on the supply line. There are four transistors, all 2N708 NPN. The complete circuit is at one side of the delay line, it is probably the amplifier creating the logic levels used in the P101 logic. All components on the board seem transistors, resistors and capacitors and one diode.
*1) https://hack42.nl/wiki/Olivetti_Programma_101_Introduction_to_logic
At 20V, the amplifier draws 19mA, that is 400mW. The Zener plateaued at 9.1V.
Reverse engineering is made a bit simpler by using images of the two board sides and reversing one, bigclivedotcom style.
| Send/Receive overview |
 |
An overview of the send and receive section. The delay line wire has 11 windings, of 150 mm diameter. That makes 11 * 150 * 3.14 = 5181.00 mm, that is 5 meter! |
| Send section |  |
The send section is connected to wires going outside the delay line package, probably the driver circuit. The resistance of the coil is about 16 ohm. |
| Receive section |  |
The receive section is connected to the amplifier board inside the delay line package. The resistance of the coil is about 36 ohm. |
| Context for the delay line |
 |
From the Technical description on the Hack42 site, a context for the delay line. |
With the preliminary schematic, it feels save to apply some voltage to the board and see what happens. There are very old electrolitic capacitors there so some care is taken. Slowly increasing the voltage and measure the current should be save. At 10V, the current is 4.5 mA. The diode, expected to be a Zener is at just 7.5V. At voltages above 12V the Zener diode voltage plateaued at 8.8V, total circuit current just 8mA. For now I assume 12 V is a safe voltage to operate the amplifier.
With a simple pulse generator program on a Teensy LC, pulse trains 1 to 10 are generated of 1 us, with an interval of 1 ms. This resulted in scope images like the one below. I guess the delay line is working. The delay is 2 ms minus 7 * 20 us, 1860 us ?
Local links:
Global links:
Last update: 2026-02-05
