[R-390] Capacitor Heresy

[K2CBY@aol.com]

It has long been an article of faith that you begin to rehab an R-390/R-390A by yanking out all the paper capacitors -- particularly the notorious "Brown Buggers" -- and replacing them with Orange Drops. I recently went through this process with an R-390A Teledyne IF subchassis (Contract No. 3785C-PC-63, Serial No. 6369), and the results were surprising.

After about four hours of squinting, sweating and cussing I managed to excise all of the original capacitors and replace them with modern polyester and polypropylene types of equivalent capacitance and voltage rating (more on this later). Removing the BFO to access the capacitors connected to points beneath the tuning shaft was a particular nightmare because some clever techie (or maybe the factory) has Lock-tite-ed the set screws on the bellows coupling. I persevered and managed to finish the job without singing the wiring harness more than a couple of times (love that smell of scorched plastic) and a minimum of other collateral damage. I then tested the IF strip in the receiver and noted no real change in performance –in terms of either gain or noise figure.

This left me with a pile of 19 old capacitors on the bench. As I was about to sweep them into the trash bucket I started to wonder "Just how bad are these things?" so I started a little investigation.
 
They all looked OK mechanically. There were no cracked cases, oozing gunk or peculiar bulges. Nothing smelled out of the ordinary.

Although I hadn't paid too much attention to it when I pulled the capacitors, I noted that there were two distinct types. Teledyne (and probably all the other contractors) used the brown Bakelite cased tubular units only as bypass capacitors -- B plus line, cathode to ground, or screen to ground. The plate to grid coupling capacitors were all metal cased with plastic or epoxy seals – the type Sprague used to call "Vitamin Q," although the ones in this unit were by Astron and General Instrument.

I started out by measuring the capacitors on an ESI capacitance bridge at a 1 kHz test frequency. Every last capacitor measured within 8% of rated value. Since the spec on many of these was only 20%, I thought this was pretty impressive. I then measured the dissipation factors. These ranged from a worst case of .013 for the brown tubulars to less than 0.01 for all the metal-cased units. For the 0.1 µF capacitors, the computed series resistance was < 21 ohms in all cases; and for the 0.033 µF capacitors the series resistance was less than 68 ohms. In all instances, the metal cased units had less than half the series resistance of the brown tubulars.

I then measured the insulation (shunt) resistance of the capacitors on a ZM-11 bridge. For the 0.1 µF "Brown Buggers" the values ranged from 60 Meg to 75 Meg; and for the brown 0.033 µF units, 800 Meg to 2,800 Meg. The metal-cased capacitors ranged from 8,400 Meg to 10,000 Meg (the limit of measurement). It should be noted that these measurements were taken with applied voltages (up to 500) substantially higher than the rated working voltages of the capacitors.

By way of comparison, new out of the box Orange Drops had a measured D of .004 and an insulation resistance in excess of 10,000 Meg (limit of measurement) for both the 0.1 µF and the 0.033 µF units.

I drew three conclusions from these tests. First, the Collins engineers were no dopes. They confined the brown tubular capacitors to non-critical applications and used premium-grade, metal-cased units where leakage resistance and dissipation factor really made a difference. Second, the original capacitors aged remarkably well. There wasn't one of them in my IF strip that actually needed replacement. Third, unless you are a glutton for punishment or just love to see the orange sparkle of fresh capacitors glinting from inside the radio, it probably doesn't pay to re-cap unless the receiver is showing symptoms of distress.

The B plus and screen bypass capacitors are most likely to fail, and if they start to leak the plate and screen voltages will be noticeably low. The interstage coupling capacitors are pretty safe because they are higher quality. If they commence to leak, the failure will be obvious because the grid of the following stage will be driven into conduction, resulting in zero or positive grid voltage and vastly excessive plate current. The cathode bypass capacitors are least likely to fail since they operate at a tiny fraction of their rated voltage.

The only exception is the infamous C-552 (0.01 µF 300 dcwv) that couples the plate of V501 to the mechanical filters. Because its failure will fry the filters, it should ALWAYS be replaced with a top quality new part with a voltage rating of at least 350. It's also easy to reach, and there is plenty of room to fit a replacement.

I have a couple of additional observations. 

Capacitors come in small packages these days, and by 21st Century standards Orange Drops are pretty bulky. I instead used CDE Sub Miniature Metallized Polyester  DME Types, which are about 1/4 the volume of an Orange Drop. The types are DME2P1K 0.1 µF 250 volts (Mouser part no 5989-250V.1); DME 4S33K 0.033 µF 400 volts (Mouser 5989-400V.033) and DME 6S1K 0.01 µF 600 v (Mouser 5989-600V.01) for C552.

Although the capacitors looked pretty good in my IF subchassis, the resistors were another story. I measured each of them, and almost half were more than 15% out of tolerance. Invariably, the resistance was higher than it was supposed to be – in a couple of cases, about 25% high. I also found a couple that were charred – but amazingly enough one of these still on value. I left the grid resistors and AGC bus resistors pretty much alone since those values are non-critical but changed all the others that were more than 10% off.

In conclusion, maybe it pays to keep a closer eye on resistors than to routinely replace all capacitors.