[Collins] testing 51J4 filters?

[Dr. Gerald N. Johnson]

On Wed, 2007-11-14 at 20:41 -0500, eugene at hertzmail.com wrote:
> Hello all, 
> 
> Finally found a 6.0kHz mech filter for my 51J4 (also got a 1.4 too).
> Problem is, my 51J4 is inoperative presently. I was wondering if there
> are some simple tests I could perform on these filters that might tell
> me either they work, or are likely to work? Or perhaps that they are
> dead? Test equipment available includes rf generator, scope, hp410B
> meter. Are there some simple tests I could perform?
> 
> Thanks
> Eugene w2hx
> 
Tube type mechanical filters have a fairly high impedance, enough
inductance that the transducers are typically resonated with about 100
pf. Makes them like the impedances of IF transformers.

I'd drive from a 50 ohm generator by using a 100 pf series capacitor to
tune the input in series resonance. Don't need watts of drive and don't
need or want any DC bias though they will stand a little. Failure off a
paper coupling capacitor in some Collins receivers tends to destroy the
filters with way too much DC.

Resonate the output with a 100 pf variable and hook on the scope probe.
That input cap can be adjustable too, if you wish, but won't be as
affected by the generator as by the scope probe. There can be a bit of
interaction between the end tuning and the filter ripple. Ought to be
about 10 dB insertion loss, no more as I recall. I'm sure I've connected
the input directly to a signal generator without tuning and put the
scope directly on the output without tuning and found a semblance of a
response curve. But the nuances of tuning if checking for the precise
bandpass edges will be necessary since one filter failure mods is from
coupling wires breaking their welds and so shifting the filter response.

Checking bulletin 1031 (available on some Collins sites) it says ferrite
transducer filter have 4 to 5 dB insertion loss and older nickel iron
transducers have 8 to 12 dB insertion loss. Another collection of filter
data in my computer says drive should not exceed 2 volts RMS at the
filter terminals and that the resonating C might be as much as 130 pf.
-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer