Fairchild 6200B Curve Tracer

[tridac]

This was an Ebay find years ago. Initlal fault EHT supply fixed, and added a small fan around that. Used for years but recently, one axis dead. A whole set of plugin cards and not obvious what does what. 200volt regulator is labelled and fixed that, but overvoltage must have taken out something else. Nothing from a web search, so any ideas for a schematic ?, or would pay for a copy of the whole manual. A bit bulky, but very usefull thing around the lab...

[wasedadoc]

Which axis is dead?  If the voltage one (horizontal?) first check the output at the DUT terminals.  If nothing there, then the generator side is faulty.  If there is then the amp and deflection circuits are suspect.

Does the position control for the dead axis move the trace?

[tridac]

Thanks. Been put aside for about a year, after not finding a schematic, but powered it up this morning. The psu rails look ok and are adjustable at +225 and +/- 45 volts. Some of the pcbs are marked 225v, so was able to fix that rail last time, where the series pass transistor had gone short. Testing with a transistor, the fault now is just one, rather than a series of steps displayed, so looks like the stair step generator is at fault. X and Y voltage levels and pos controls look ok. Amazing number of  precision metal film and wirewound resistors on the boards, but perhaps they originally built it for their own in house work, since Fairchild wrote the book on planar technology. Date codes on parts are 1967 / 1968, all silicon 2n... devices throughout afaics...

[tridac]

The gods must be smiling on me this week, as the manual was uploaded to bitsavers on the 23rd August. Checked there in the past, nothing, but the manual has theory of op, complete schematics and parts lists. Great scans too, so whoever did that, a big thanks. Had started to analyse the circuit and one of the boards looked like a 4 bit binary counter, with scaled resistors on the output of each stage, so pretty sure that was the stairstep generator. Now confirmed. Manual dated 1967, so the design must date back to '65/66 or so, when most test gear was still tube based. Amazing company...

[tridac]

After some kind soul uploaded the manual to bitsavers, spent a day or so debugging the step generator fault and a few other items. It uses  4 bit binary counter, with scaled resistor D-A values on the outputs to generate the step waveform, driven by clipped ac from the 45 volt regulator transformer winding. Everything looked ok on the counter stages, but not  counting. In circuit be and cb junctions looked ok, but deceptive. Replaced both transistors on the first stage flip flop, and it all sprang into life. Plastic transistors used throughout the board, and supply rail is 45v, with 40 v swing on the collectors, so replaced every one with 2n2222a metal can. Also changed the 68k 2w composition resistors to metal oxide, as they run hot to touch. The 225 volt regulator series device, the original fault, ran too hot to touch, around 110 C, so fitted a small fan on card cage to keep things cool. Temp now is around 40 C, which is fine.

It's not a fault of the original design. This unit was switchable 115/230 volts, whereas line voltage here in the uk is usually 245 volts, a nominal 240. This puts added strain on series regulators and probably accounts for the initial failure. Have seen this on some other US kit, the HP5370 counter and the Guildline 9340 thermometer, for example. On the latter, used a heat sink resistor bolted to the chasssis to lose some volts on the input, but it can be a real problem for older US kit run here in the uk. More modern switch mode psu's are no problem.

Just need to finish calibration and should be good for anther 10 or 20 years...