PDP-8/L restoration - PSU problems.

[MattisLind]

I have now tested the idea changing the zener value in the real world (well almost - with dummy loads).

I replaced the zener with a 1.3W 4.3V zener BZX85C4V3 and the input network for the matched pair was changed into a voltage divider consisting of a 560 ohm to the 5V and a 2.4k to ground. I tested with dummy loads loading the PSU with 2.5A, 3.3A and 5 A. I got no overshot in the output voltage when ramping up the input voltage. At 2.5A the output was at 5.17V and a 5A it was at 5.08V. The crowbar triggered at around 5.7V. No false triggering occurred.

I am happy with this solution and will put it in use in my other PDP-8/L and BA08. Possibly even the PDP-12.

 

[m_thompson]

Good idea!

 

[Bitly]

I have now tested the idea changing the zener value in the real world (well almost - with dummy loads).

I replaced the zener with a 1.3W 4.3V zener BZX85C4V3 and the input network for the matched pair was changed into a voltage divider consisting of a 560 ohm to the 5V and a 2.4k to ground. I tested with dummy loads loading the PSU with 2.5A, 3.3A and 5 A. I got no overshot in the output voltage when ramping up the input voltage. At 2.5A the output was at 5.17V and a 5A it was at 5.08V. The crowbar triggered at around 5.7V. No false triggering occurred.

I am happy with this solution and will put it in use in my other PDP-8/L and BA08. Possibly even the PDP-12.

It would be interesting to see the waveforms when powering up the actual hardware. It is possible the overshoot doesn't occur in reality due to inrush current. It's possible your change will slow the slew rate of the supply when powering up the computer, but I can't think of any reason this would be bad.You might see higher currents due to the long transition through the undefined state.

CW