Power Supply Restore and Test
Wang 600 Calculator Chassis
The process of cleaning was laborious but straightforward and gave a very good result. Removing the ROM, logic boards, printer and tape unit left the chassis with power supply components mounted on the large PCB motherboard. The power supply architecture and construction was similar to the 700 and indeed the old 300 series machines, leading to a similar strategy for restoration:
- rework the primary side of the power transformer so as to remove all mains connections from the low voltage PCB and provide more comprehensive insulation across the mains wiring. This departure from originality is regarded as essential for safety.
- remove and test all electrolytic capacitors for value and leakage, reforming and replacing if necessary
- “recapping” is not DoPECC policy but past experience with Wang filter capacitors recommends comprehensive testing in this case
- bring up the supply against dummy loads and test for stability and ripple
- where series pass transistors are the regulating element, consider if short-circuit failure may cause catastrophic damage and if so, consider fitting protective voltage clamp(s)
Primary Side Rework
This was fairly straightforward, there was (just) enough room to fit an IEC receptacle with integrated fuse and switch. While not original the change is barely visible and it is very convenient to be able to detach the power cable.
All mains wiring was removed from the low voltage motherboard, the correct primary taps and blower supply were assembled and the resulting mains wiring loom was insulated and then secured with physical separation from the low voltage elements.
Filter Capacitors
Past experience with the Sprague electrolytics in Wang 300 and 700 machines has revealed a failure rate of probably 30%. With this in mind it seems reasonable to restuff the cans with high-quality modern equivalents, these fit easily into the original cans.
The electrolytics were all removed and subjected to 30 mins of controlled current reforming at rated voltage before testing (after discharging!) on a DER-5000 at 100Hz. All units had capacity within specification but around a third of them displayed lower Q and higher dissipation than would be expected. This helped to justify replacing all of the electrolytics by restuffing the cans.
Rectifier Diodes
After replacing electrolytics as above the bare chassis and motherboard was powered up - and immediately blew the fuse on the primary side of the transformer.
With only the transformer, rectifiers and filters present it did not take long with the multimeter to find that D3, one of the main rectifiers in the 5v supply, was shorted. On removing D3 it could be seen that the PCB under the diode was heat damaged, no doubt it had been leaky for some time. Both 5v rectifiers were replaced with 105v/5A devices.
The low voltage supplies now came up without blowing fuses but something was still not right - with dummy loads in place the -12v rail had ripple of about 40mV whereas the +12v rail had over 200mV. This was peculiar because the (dummy) loading of each rail was the same….
and a Design Error
Putting the low voltage ripple aside, the HV secondary was connected to test the HV supply, and the mains fuse to the primary blew again! Testing for shorts with the multimeter showed nothing suspicious. This was perplexing, so components were lifted/removed from the HV side and the short disappeared when the 30uF filter was removed.
This component was a restuffed version of an unusual vintage Sprague component, 30uF/250v and 3000uF/20v in the same can with common negative terminal. Careful checking showed that the restuffing was correct but that the PCB motherboard did not match the schematic - the 30uF was connected across the +17v secondary and the 3000uF was connected to the +250v secondary. This error was made in the original PCB markings and has been present in the original construction. It seems that the original Sprague components could withstand this error but not the modern replacements.
Reversing the can to restore 30uF/350v to the HV supply and 3000uF/50v to the +17 supply fixed both the HV short and the assymetric ripple on the low voltage rails.