Oscilloscope EHT capacitor
Back-to-basics repair on a Telequipment D67A

see copyright notice. Page created 28-Sep-2006 updated 13-Nov-2019. Use the button groups above to navigate quickly around the site.

Since 1980, the most important instrument in my workshop has been a Telequipment™ D67A oscilloscope. Its purchase nearly broke my bank, but it was definitely worth it for all the features - dual trace, dual timebase, 25MHz bandwidth, variable everything, and a bright, sharp display thanks to the overall 10kV accelerating potential. That is, when all the capacitors in the EHT multiplier are working.

One day in 2006, I was using the scope to trace a 1kHz sinewave through a faulty cassette recorder, when I heard a faint, but marrow-chilling, "fizz" from the bowels of the instrument, accompanied by a noticeable dimming of the display. This was a re-run of events some 18 years previously, when I'd had to replace a capacitor in the EHT circuit. Sure enough, when I slowed down the timebase, I could see a tell-tale 50Hz intensity modulation of the trace.

Wearily, I took the side panel off and pushed my trusty EHT snake under the PDA connector cap. This contraption (pictured) comprises fifty 10M resistors sealed into a plastic tube, such that a 60µA meter from the cold end to ground will measure 0-30kV. What should have been 8.5kV (the cathode runs at -1.5kV so that the deflection circuit can operate at a sane potential) measured barely 2.5kV.

Gingerly, I replaced the meter with a 470k resistor (one false move and sparks start flying...) and used the scope itself to examine the signal across it. Sure enough, there on the screen was irrefutable evidence of an EHT fault, in the shape of 50% "hum" atop a seriously reduced DC voltage.

It took me a while to remember how to access the EHT multiplier. It's housed in a clear plastic box under the CRT towards the back. First you need to reach under the CRT shield with long-nosed pliers, and unscrew the two bolts supporting the box. Then turn the scope upside down and remove the base plate (six screws). Ease the box forward, then its cover can be pulled off and the board slid out.

The manual warns that there's no EHT bleeder, so you need to be careful to discharge all the capacitors before handling the board. It's also important to keep all soldered joints smooth and free from sharp edges which could cause corona discharge at these voltages.

There are four 22nF 5kV polyester paper [sic] capacitors in the multiplier. The input is 1700V AC from the mains transformer, equivalent to 4.8kV p-p. The multiplier (almost) doubles the p-p voltage, to provide +8.5kV DC. 22nF sounds very low for a 50Hz circuit, until you realise how little current is required - probably just a few microamps even with a bright trace.

A quick run round the components with a multimeter didn't immediately reveal the fault. The rectifiers looked open-circuit with the meter's 3V test potential, but tested OK on a 12V supply, and all the capacitors showed a slight "kick" on the highest resistance range, as I would expect. But on closer inspection the one marked X on the circuit diagram (C411) showed a very high (5M) resistance. Accordingly, I removed it and substituted the only remotely similar one I could lay hands on, a 4700pF 4kV ceramic disc. That restored a well-nigh perfect bright trace, so I was tempted to leave it there - but first I thought I'd dismantle the faulty component to see what had happened to it.

A few clouts on a socket spanner held against the end cap knocked the whole assembly through and out of its plastic tube. There were still no visible signs of failure, so I carefully removed the securing tape and unrolled the foils until I found a nice patch of carbon surrounding a hole in the polyester film. As it was about 3cm from the inner end, I cut about double that length off each film, leaving the metal foils slightly shorter than the polyester. To be sure to be sure, I cut two small pieces of polyester [I think] film from a food roasting bag, and folded them over the ends of the foil, then rolled the whole thing back up, stuck on a bit of insulating tape, and pushed it back into the tube.

To my surprise, the repaired capacitor measured 16nF, and to my even greater surprise it worked perfectly back in the EHT circuit. Now I'll keep my fingers crossed each time I switch the scope on. After the earlier failure I managed to get a replacement from Tektronix; I made such a careful job of fitting it that I can't tell which one it was, but I've an idea it was the same, C411. I don't expect it would be very easy to get a replacement now. Perhaps I'll try if it fails again, but it might be more fun to see what I can do from scratch with some aluminium foil and a roasting bag!

top ▲