Opening up the DEC Precision Display 340

Simulating a P7 point-drawing CRT is hard; the behavior of the phosphor is subtle and distinctive, and the optics of seeing the image painted on an actual period correct CRT are awesome.

The ICM has the DEC 340 cabinetry and the CRT, but most of the logic is missing. Replacing the missing circuitry is a work-in-progress; the analog circuitry to drive the deflection coils and the beam intensity have replaced with re-implementations of the original logic.

However, the digital side has not yet been done.

The PDP-6 that the 340 is attached to is in similar shape, the cabinetry and panels are intact or replaced, but the insides are gone and have been replaced with a Pi.

Because the -6 is being simulated on a Pi, there is no need to replace the 340 I/O bus circuitry; that can be simulated on the Pi, and it can send the X/Y point coordinates and control signals out a GPIO port.

So the remaining bits that are needed for the 340 are two GPIO driven DACs generate -10 to 0 volt outputs based on +/- 1024 coordinate values, and figuring out how to wire the DAC outputs into the existing analog circuits.

The PDP-6

The Display 340

The insides of the 340

The top row contains the beam focus and bias card, two deflection pre-amplifiers (vertical and horizontal) and two unpopulated cards which I believe to be intended as Digital to Analog cards to drive the pre-amplifiers.

The second row (with the fans) are the deflection amplifiers.

Below that is the CRT, and below that the logic power supply. Below that but not visible is the high voltage supply for the CRT.

The top row cards are based on the original cards, but some unobtainium components have been replaced, and some of the circuitry has been redesigned to accommodate that. This means that the original circuit diagrams don’t always match what I am seeing on the cards, and the card connector numbering is completely unrelated.

This has made figuring out how to send coordinate data into the pre-amplifiers an uphill slog.

I spent the morning tracing out circuits on the pre-amplifier cards and am reasonably certain that I have located the correct data in line; this turns out to be likely correct as tracing the backplane wiring for the proposed line leads eventually to two pairs of wires that are tacked onto the unpopulated cards; the pairs are several feet long and are not attached to anything, suggesting that they are there to allow test data from a signal generator to be injected.

The pre-amplifier card

The card with the interesting wires

Next up, disconnect all of the high voltage bits, attach some meters to the CRT deflection coils, and apply signals to the wires to see if they really are deflection inputs.