Crystal mounting position on PCB.

[G6ONEDave]

On a frequency counter that I have just repaired, the internal 10Mhz crystal was fitted to the back side of the pcb and not on the component side. There is a silk screened position on the component side and room to fit the crystal laid flat with it's legs at 90 degrees, as is often found on many pcbs. I just wondered if there is any technical reason for mounting the xtal on the non component side of the pcb in free air. The xtal is mounted with full length wires standing straight. this seems to be done by the manufacturers, as other counters of the same make and model have the same way of mounting this xtal. The pcb is double sided print with plated through holes also the xtal case is isolated (no earth wire to it). Or maybe this is just a foible of the manufacturer.

[unitelex]

Hi,
It may have been put on the underside to get better stability.
If for example the underside was better shielded from noise from digital circuits etc
Or could be to minimise thermal tempco effects assuming it remains cooler under the PCB

Best Regards
Chris

[Jon_G4MDC]

Another bit of crystal trivia. I was told never cut the lead wires or much worse the pins (of HC6 or HC25) with side cutters. The type that work by compression.
The eventual fracturing of the lead or pin creates a shock wave that travels up inside to the "works" and it can shatter it or shift it in frequency irreversibly.

Cutters that work in shear mode and leaving long leads are advised.

[Radio Wrangler]

That is good advice, Jon. The shock wave up the lead can damage the hermetic seal (glass beads) on the entry to the crystal's can and can crack the quartz disc at the points where connection is made.

Shock waves up transistor leads are another no-no if you want your products to be reliable.

We had a programme to reduce instrument failures by a factor of ten in a ten year period. The investigations taught us a lot, and in the end we did appreciably better than a tenfold improvement. Careful handling of all sorts of components was a big factor. Even small resistors proved susceptible to ESD.

David

[lesmw0sec]

I can't see a problem with triming the leads - provided - that the component leads have been soldered into place first, with just the excess removed afterwards.

provided - that the component leads have been soldered into place first, with just the excess removed afterwards.

This is against the Philips soldering guide of 1980 ish, cutting after soldering can fracture the soldered joint, so they say.

[Jon_G4MDC]

As David said, not only the joint but the hermetic seals and crystal blank itself.
Cutters that work with a scissor action should be used.

[G6ONEDave]

Chris the pcb is vertically positioned behind the front exterior panel, so the crystal is horizontally sticking out at the back with one lead vertically above the other and not side by side. There is also a note in the service manual stating that a cold welded case type should be used as opposed to a soldered case type. So I take it that the case requirement is also to do with the possible shock damage that might affect the xtal's accurracy.
With reguard to this mechanical / heat shock effect that several of you guys have mentioned, do I take it that that is why a lot of early solid state circuits had their transistors mounted with full leads, so that the device was in effect standing to attention? Maybe this goes someway to explain why some of thearlier electronic equipment works perfectly well, while newer gear seems to fail earlier than expected.

[unitelex]

Germanium transistors are more susceptible to heat from soldering, I have always assumed they left the leads of those transistors long (especially when PCB mounted) for that reason, to reduce heat flow up the leads.

Best Regards
Chris

[Jon_G4MDC]

Cold welded rather than soldered is specified for ageing reasons.

Soldered crystals contain residual flux which over time settles on the quartz plate and lowers the resonant frequency.

Cold weld having fewer contaminants in the can shows much less of this effect.

[G6ONEDave]

I had heard of ageing of crystals and indeed other components but did not know that the type of enclosure seal had any bearing to it.
It does mention in the manual that the unit should be left switched on for 15 minutes before attempting any calibration adjustment, to let the equipment get to a stable temperature. This does sound odd given that the entire counter is of solid state design and does not use any thermionic devices, although this does seem to be quite a common requisite for adjustments.

[Jon_G4MDC]

Oscillation causes some power dissipation in the crystal. Temperature will rise a bit after switch on never mind external influences so it's not that strange.