I somehow acquired a very early production 200CD audio oscillator - serial number 739. These early units had a number of differences in the circuitry from the later production models, and manuals for them are hard to come by. The common ones on the net start at serial number 8739 and above. While it's possible to apply all of the backdating changes from the newer manuals to get the correct circuit for my unit, it's a bit of a pain to keep track of the changes.
Dave of Artek Manuals dug up a manual that covers serial number 2251 and above, along with a number of service bulletins describing the reason for the circuit changes and including a schematic covering my serial number. Thanks, Dave!
I had pulled the dial, vernier, and the knobs quite a while back, and cleaned up the front panel. The rest of the case was pretty grungy with some flaking paint. I ended up stripping it, but there are some stubborn bits that I'll have to scrape off with dental picks or something, so it isn't ready for painting yet.
The front panel is missing a good deal of paint, too, but I decided that I'm just going to touch it up rather than strip and repaint it. It's old enough - it should look well-used.
Yep, serial number 739.
It has a couple of bumble-bee caps in there that have to go, and three electrolytic cans that need to be restuffed.
The tuning capacitor needed a good cleaning (as did the rest of the innards - it had been used in a lab where there were smokers), DeoxIT on the contact fingers, a dab of grease on the ball bearings and the gears, and some oil in the oil-holes of the bearings for the gear mechanism.
Someone had been in there to replace the non-polar electrolytic. They cut the B+ lead to get access and then resoldered it and wrapped it with electrical tape.
The mounting of the bumble bee caps and the associated resistor network is a bit surprising. While the bumble bees are (or were) held in place by some plastic cable clamps, there are a lot of flying joints. This is a very high impedance part of the feedback network, though, so I guess the engineers didn't want to risk leakage between terminals on a terminal strip. You sometimes see this same flying-lead construction in modern gear, if it is a very high impedance node or the circuit requires very high precision.
I checked over all the resistors. In general they were "close enough". I may do a full calibration sometime in the future, but for now I don't want to mess with the range switch.
The old-style molded mica caps checked out fine on the Capacohmeter megger. Some of the high wattage power resistors were a bit off, which is quite surprising for a wirewound resistor. Maybe my measurement technique was off, and I didn't notice a paralleled path.
Getting the electrolytic cans out to restuff them was a pain, as the wiring is tucked in under the power transformer.
A Hammond 1490-BB cast aluminum case turns out to be just the right height to prop up the power transformer when you turn the 200CD upside-down.
After re-capping, the unit was finicky and didn't want to sustain oscillation at low frequencies. The service bulletins mentioned that the 6AU5's tended to get gassy and start drawing grid current, which would be consistent with trouble on the low range. I replaced the 6AU5's and it came to life with a nice stable voltage output across the whole range.
The dial was badly water-stained. I buffed it with rouge on a wheel on my bench grinder, which took care of the water staining but also removed the satin finish. I decided to restore the finish using a lye bath as described in an old ARRL Handbook. I mixed up the lye solution according to the handbook description and put the dial in. I had some chemical gloves (from Fair Radio quite a few years back) that I used to go in and wipe off the smut that formed on the aluminum, so that it got an even finish.
Unfortunately, the handbook must have missed a dilution step, or else replaced "minutes" with "hours" - it suggested leaving aluminum in the lye bath for an hour or two! I didn't let it go that long, but I did over-do it and the dial was pretty deeply etched.
I used Lacquer-Stik that I had bought many years ago from Antique Electronic Supply to fill in the engraved dial markings, but the deep etching also picked up some of the paint, so the dial is still stained, except that this time it is a fairly even black instead of water-spots.
So the dial doesn't look very good. Worse, the etching thinned it enough that the pressure plate doesn't grab the dial. Even with the plate fully clamped, the dial spins freely on the hub.
I found some 0.010" aluminum shim washers at McMaster-Carr.
I also bought some machine screws in the same order, but the shipping was still almost as much as the merchandise.
I was lucky that the diameter of the hub was exactly 0.75" ...
... so the washer fit perfectly. It only took one washer to shim it enough for the pressure plate to grab the dial, so I have nine left. Anyone else need one?
The oscillator is still mostly within its ±2% frequency spec, except I think the X1K range might be a little low. I'll put off calibrating it for a while since I've got a frequency counter handy here in the lab.
(To do: put a photo of the finished restoration here.)