Y-register decimal point fault
The calculators are exercised from time to time, the Wang 700s have a Wng-supplied memory test that is useful for this purpose. The memory test executed without error (which was reassuring) but the Y-register display looked a bit ‘off’. Closer inspection showed that decimal point display was incorrect for the Y-register only:
Wang 700 DP faults Y display, missing DP
Y display
- After PRIME there is no left-hand decimal point
- there are also two areas of odd glow around the decimal points for digits 3 and 4
- Entering a floating point number to X and copying to Y showed correct digits but no floating point decimal in Y
- digits 3 and 4 in Y also showed some ghost glow around nixie digits that should have been dark
- Entering an exponential number to X and copying to Y showed correct digits
- the Y-register decimal was correctly lit but was over-bright
Y display
Y display
Analyis
The 700 machines can display in fixed or floating decimal format and it’s interesting to find that the two display modes are entirely hardware-based. The memory representation of the X and Y registers is the same regardless of display mode. The firmware sets a bit for each register to say if fixed or floating display is required and the display driver hardware takes care of:
- placing a fixed decimal point at the left end of the display
- showing a decimal point within the mantissa region
- showing or blanking the exponent sign and digits
With the above in mind it’s clear that a purely hardware explanation is possible:
- fixed point mode shows a decimal point as expected
- floating point never shows a decimal point
Most of the hardware for display representation is on the Nixie control boards, the Y register board was set up for testing.
… and autofixed
After setting up the Nixie control board in extenders, the logic signals for fixed and floating point mode were checked with a basic logic probe. Both were pulsing with good levels, which seemed to be expected behavior.
…. and then the Y display was seen to be good. Proper operation restored with board removal and plugging, so perhaps a bad edge connector contact (perhaps related to Wang’s well known indifference to gold plating….)
This will be monitored going forward.