Information on the Wallis VCS 103/3 -- Very high current on indicator LEDs?

[Wim_L]

I've got a nonfunctional Wallis VCS 103/3. This is a 10 kV 3 mA power supply. I hesitate to call it broken, it may be defective, but the more likely cause of it failing to operate would be that the link plug, which should be plugged into the remote control connector to bridge a few control lines when remote control is not active, is missing.

I have managed to find a service manual with schematics for the Wallis R603-06P, a different model of high voltage power supply but with a few similarities. Plenty of differences too, and I have not found a manual specific to the Wallis VCS, so I'm reverse engineering its internals to find out which wires need to be connected at the link plug.

One oddity which is becoming obvious is that they seem to be driving their front panel indicators at unusually high currents. The indicators are 5mm red LEDs. The service manual schematics for the R603-06P shows some are being driven from a 26 V source over a 680 ohm resistor. Assuming a forward voltage drop of about 2 V, that still puts about 35 mA through that LED. That seems unusually high -- most modern 5mm indicator LEDs specify maximum currents of 20 to 25 mA.

Was it common back then (not sure how old it is, but it seems like a relatively old unit, some of the cable ties are crumbling) to drive LEDs at such high current, perhaps to compensate for fairly low efficiency?

My preliminary analysis of the circuitry inside the VCS 103/3 indicates even higher currents are to be expected here, perhaps up to 45 mA. This, however, may not be by design. It's obvious some not so clean modifications were made to this unit, including to the circuitry driving the front panel indicators. that part is shown in the pictures I'll include. They also made the choice of linking the main positive rail of the power supply -- a bit over 30V -- to the remote control link connector. Fused, but no other current limits. A closer look at the PCB suggests that a resistor has been removed from this path, though with no indication of the expected value, and replaced by an improvised arrangement of diodes and resistors on the return path of this positive rail.

[SeanB]

Looks like it comes from the 1970's. LED units then were rather dim and low efficiency, so running them at 30mA plus was rather common to get the light output up to something that could compete with room lighting as an indicator. If you replace the LED units use a resistor to get a current of 3mA through the new LED's, they will still be an order or two magnitude brighter than the original MAN100 led's if they are in the original dark red package.

As to the resistor value, looking at the rest of them you will probably be fine with a 100R 5W resistor as a replacement for the wire link. Try to get a matching brown coated resistor, though that will likely be a 3W unit.

[Wim_L]

Thanks, that confirms my suspicions. I'll leave the LEDs as they are, it looks like they are the original ones. I suppose the high current might limit their useful lifetime, but that will be something to be dealt with when it happens. I think I'll change transformer taps though. The mains now connects to the 225 V tap, which is fine as long as I'm using it with an isolation autotransformer that can be set to this value. Local line voltage here seems to be 238 V though, so a switch to the 245 V tap would make sense.

It seems like the main PCB has useful hints on the silkscreen about which wires should be connected on the silkscreen. Most pins have unique numbers, but some seem to be matched (e.g. a pin 4 and 4A - there are several more like that, and at least two of those pairs look like they should definitely be connected. A bit more work to be done before I'm certain about the others).

I do still need to find a matching plug for that remote control connector (see picture) if I want to connect the wires through an external plug. Poking a few wires in there might do the trick for a quick test but is unlikely to be a stable long-term solution. Or, alternatively, just forget about remote control and solder the necessary links directly to the board.

[iamdarkyoshi]

I have an old RGBLED sign that I desoldered the leds from. When testing the leds to obtain their specs, the blue LEDs can handle 5v unballasted and get rather... warm, but after 10 minutes it still was working so I would agree that the current drive is normal.

[SeanB]

You need some 0.15in spacing board interconnects to make that plug, and then simply remove the polarising pin. If you make the cable version solder the wires to the pins then use waxed card stock ( paper milk or juice box) to make a small rectangular sleeve and fill with epoxy resin to seal the back. For the local control do the same and fill it to make a block of resin.

[Wim_L]

Successful repair, with a bit more improvisation required than first expected. 100 mil pin pitch headers were far too small, but 150 mil were just a little bit too big. In the end, I cut loose pins from a 100 mil pin header and soldered those into a quick and dirty little PCB I made with a Dremel. Not the most elegant looking thing, but it does the job. If I ever need the remote control functionality, I'll probably remove the entire connector and replace it with something more common like a 15-pin D-sub.

(The connections do need to be fairly reliable. A loose connection on the remote will set the voltage output setpoint to about 11 kV, which could get interesting depending on what's connected to it.)