I have two 455 kHz crystals that were used in the phasing circuit of mid-century communications rcvrs like the Halli SX-28. Is there a simple means of testing them out of circuit?

Thanks,

Tom

You can just put it in series with a signal generator and scope. Look for the peak at it’s resonant frequency.

Hm. That's what I thought, and have already tried it. Neither crystal seemed to respond. I took both apart and found what appeared to be clean, unbroken crystals. I cleaned them carefully with isopropyl alcohol, reassembled them and retested. Still no luck. It just seems odd that the crystal should just die.

Tom

I wonder if you're feeding it enough of a signal. Touchy business, because if you hit it hard enough, it could shatter I suppose. Like the legendary flashbulb tester of olde.....

You need to raise the signal strength.

Thus far, when checking crystals using this method, I've never had to go beyond several volts peak-to-peak from the signal generator to make things happen.

So far, no incidents like what happened on the 'Silicon Avatar' episode of Star Trek: The Next Generation.

I assume these must be in FT-243 or similar holders.

You may have shifted the frequency slightly in your process.

Rich

Filter networks are a world unto themselves. In the old communications receivers they were trying to narrow the selectivity of the IF passband to one kHz or so by using a crystal to sharpen the peak or 'Q' of the IF system between the first and second IF amplifiers way beyond what could be accomplished with LC networks. Pre-packaged crystal filters with resistive terminations did not exist in the 1940s when the SX-28 and similar receivers were designed, so receiver manufacturers designed the networks themselves. They consisted of fixed and variable capacitors in series and parallel with the crystal, which is itself in parallel with part of the secondary of the first IF transformer, and some shunting resistances. The filter crystals were unique to the circuits they were used in. You can't even swap the crystal from an SX-28 with a different contemporary receiver like the SX-25 and get the same results. Without the rest of the radio it goes in, you will not be able to get the same response curve at the same frequency from a loose filter crystal. But if the crystal is good, you should be able to see spikes or peaks in the response curve (there is usually more than one if the crystal is out of circuit) as it goes through its natural resonance points. A sweep generator and scope, or a device like the NanoVNA should be able to find them.

You should not need very high RF voltages--I'd start off with a volt or so--but the signal generator has to be capable of driving a low impedance load. Any generator meant to work into a 50-ohm load should be fine. As for the scope, keep in mind that a 30-pF scope input looks like 12-k ohms at 455 kHz. This may load the crystal so much as to throw it out of resonance, so an active (FET) probe would be a good thing to have, but if not available at least use a 10:1 probe.

The Q of these crystals is *very* high. It's possible you are missing the peak if you scan the signal generator rapidly, or if a digital type, increasing the
frequency in steps that are too coarse. An example of the one in an SX-11, full width at half max is around 100 hz or so.

You can easily build a simple circuit to test if a crystal will oscillate.
There are circuits that will light an LED if the crystal is good. There are also circuits you can connect to a frequency counter and see the crystal frequency. I have built similar to the latter when checking some FT 243 crystals.

A friend had a Allied Knight G-30 Grid Dip Meter which the coils fit a crystal socket. He would always test crystals with it and listen to a receiver to see if it was oscillating.

-Steve

Thanks for the responses. I like the ccts. that Dale posted and I'm going to build the second one (the first one uses transistors I don't have).

Tom

2N2222's or PN2222's or 2N3904's should work fine in place of the BC548's.

I bought 2 of these. It will give you the fundamental freq. Not so good as a counter. Too much input level required.

https://www.ebay.com/itm/392053655272?h ... SwkcFbDKQF

Russ