[Collins] More 75A-4 Recap

[Dr.Gerald Johnson]

I'm sure I didn't advise reading Rippel's Deadly article. I don't recall ever 
seeing it, though I read it this afternoon.

My experience has been that few micas have failed. Few disc ceramics 
have failed. That is not to say they can't just I've not found a large 
enough number failed to justify shotgun replacement.

Certainly coupling capacitors going to grids with high impedance 
ground returns like the AVC line are ultra sensitive to leakage. Mica is a 
natural substance and some can have inclusions.

For a 4 pf coupling capacitor, mica will have better temperature 
stability than all but the finest of ceramic. A cheap easily found ceramic 
might loose a significant fraction (a quarter to 3/4) of its capacitance 
when up to tube base temperature causing a loss of gain. NPO 
temperature coefficient has nearly the stability of silver mica.

Dipped mica capacitors are imperfect, the conducting layers being 
silver paint on the mica. Most molded micas (unless in a red case) were 
made up of stacks of foil and mica, which seemed to be more reliable, 
though the silver micas have better stability because of the exclusion of 
air between the silver and the mica. When used for frequency 
determining oscillators, a few silver micas will have random jumps in 
value that seem to be caused by air bubbles in the dip. The pressure of 
the molded mica would have minimized or eliminate those bubbles.

I don't favor replacing .01 disc ceramics with orange drops, even at 
audio. I do recommend the most stabile of disc ceramics which will be 
relatively large. If the replacement is smaller in diameter than the 
original, its sure that its temperature coefficient will be larger and that's 
not a good replacement. Especially in the receiver IF and front end, the 
disc ceramic's lower inductance will lead to greater gain than the same 
sized orange drop. In the old days orange drops were not the extended 
foil construction they are today making the inductance much greater. In 
vintage receivers using paper capacitors for screen and cathode 
bypasses, some have found significantly improved 21 and 28 MHz 
performance when those paper capacitors have been replaced by 
disc ceramics to reduce the inductance of the bypass capacitors.

I have downloaded the parts list and illustrations sections for the 75A4 
from the CCA web page. I had to copy the parts list to a faster 
computer because this one is too slow. I have examined the listing for all 
the questioned parts.

As for C124, that "0.5mmf" PAPER capacitor has to be a 1/2 microfarad 
paper. Paper capacitors have never been made much smaller than 
1000 mmf (pf) ( 1 nf) and a 0.5 pf capacitor would be a poor bypass for 
a B+ line at any frequency under 10 GHz where it would be marginally 
small. One can make a 0.5 pf capacitor (not with necessarily great 
quality or stability) by twisting together two #22 solid PVC insulated 
copper wires. Takes about a half inch to make 1/2 pf. Not a benefit in 
an HF radio to add a 1/2 pf capacitor. Less than the stray capacitance 
on the B+ wiring in the cable harness.

That 1/2 mfd paper capacitor probably had a tolerance of +/- 10%. 
Today the nearest standard value is 0.47. The next higher is 0.56. As for 
voltage on paper, mica, and ceramic using a higher voltage capacitor 
than the original may be of benefit in keeping leakage down but will 
always cost in the physical size of the capacitor being large. A 1/2 mfd 
200 volt paper is fairly large, it has to be nearly twice as big (thicker 
insulation in the roll) for 400 volts and with the same type of insulation 
three times the volume for 600 volts. So besides the added cost it can 
be hard to fit in the place and will have a greater inductance than the 
original. When asking for RF bypassing, added inductance is not a good 
thing. And the value of a bypass capacitor is generally uncritical. Either 
0.47 or 0.56 would work.

In the old days before about 1960 one needed to replace electrolytics 
with the voltage rating as near as possible but above the working 
voltage. Operating a 450 volt capacitor in a 300 volt circuit tended 
make that high voltage capacitor change value, and probably get a 
little more leaky because the electrolyte would eat away at the 
previously formed oxide making it thinner (and probably raising the 
capacitance). Again the higher voltage part was often harder to fit 
because it was always larger. Long about 1960, GE came out with 
replacement electrolytic capacitors that claimed to have cured any 
faults with using the higher voltage electrolytic on lower voltage so one 
could stock only 450 volt capacitors and use them for 250, 350, and 450 
volt applications. Maybe even for 150 volt applications. And they quit 
selling the lower voltage parts.

The electrolyte eats away at the oxide formed when the capacitor is 
made when there is no voltage applied. New Old Stock electrolytics are 
not of value in my experience. They will be more leaky than used 
electrolytics of the same age and for me they fail violently.

Rather than shell out $25 for an Allied minimum order, I'd go with a 3.9 pf 
NPO disc ceramic. For low leakage, perhaps a 1 KV part like Mouser's 
75-10TCCV39 for 28 cents. Or I'd lower the leakage of the original by 
removing the accumulated dirt of the past 45 years.

73, Jerry, K0CQ, Technical Advisor to the CRA

-- 
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.