How to select correct replacement crystal?

[TheMG]

Attempting to repair a Canon SX170IS camera. It is not turning on at all and I've noticed the lid has broken off the top of the SMD quartz crystal for the MCU. Scope confirms when I press the power button voltage goes high on the crystal but no oscillation at all.

Only markings on the original crystal are as follows:

27000
K332F

What I've been able to gather is that the manufacturer is Kyocera and it is a CX3225* series crystal, at 27MHz. The 332F is date code and factory.

Where I'm stuck, looking at Digikey, these crystals are available with a load capacitance of 8, 12, or 18pF.

https://www.digikey.ca/products/en/crystals-oscillators-resonators/crystals/171?k=kyocera+27mhz&k=&pkeyword=kyocera+27mhz&sv=0&pv16=11780&pv46=13601&sf=0&FV=ffe000ab&quantity=&ColumnSort=1000008&page=1&stock=1&pageSize=25

How do I know which one I need? Buy all 3 and trial-and-error?

[ledtester]

Caveat: I'm no expert, but...

My guess is that it won't make much of a difference -- just go with the middle value to minimize how much the crystal will get pulled from its natural frequency. I would expect the deviation to be < 10 KHz.

Here's a video showing what happens when you add extra capacitance to a crystal oscillator circuit. In this particular case between 0 and 9.8pf was added through a trim cap, but note that the frequency only changed between 14.062 Mhz and 14.067 Mhz:

https://youtu.be/_OsvXQe_l2E?t=1m50s

So I would imagine a 6pf difference on a 27 Mhz crystal would yield a similar size effect.

For a radio it's important to have a precisely set frequency. For a camera which is essentially a digital computer probably not so much.

[TheMG]

I haven't messed around with crystal oscillator circuits much, so wasn't sure if maybe the wrong crystal would not oscillate or something like that. But if it's just going to pull the frequency a little, it shouldn't matter in this application.

I'll order the 12pF and see what happens. Also hopefully that's the only problem, really bizarre that an SMD crystal would be physically damaged with no signs of any other damage to the camera, not even a scuff on the outer plastic housing. Anyways, if it doesn't work I'm just out a $1.10 crystal no big deal.

[TheMG]

Well I got the crystal and soldered it in, the camera now powers up. So the crystal with the top popped off was definitely the problem.

However... as careful as I was to mask off that part of the PCB with kapton tape and keep the temperature as low as possible on the hot air station, it appears I may have damaged one of the flat flex connectors next to the crystal. Lens operation is erratic and changes depending on which way I apply pressure to the flat flex, there is also hardly any resistance to pulling the flex out of the connector.

Perhaps my mistake was soldering with the cable still installed in the connector. With the heat, that probably softened the plastic on the tensioning lever enough for the tension on the pins to deform the plastic, so now the lever no longer applies enough tension to the pins to make a reliable contact.

Now the decision... attempt to replace the connector or time to cut my losses on this one...

[2N3055]

Stick  some kapton tape on isolated side of the flex to make it thicker..?

[TheMG]

Tried that. Nice snug fit but no improvement.

It is entirely possible that there is something else wrong with the camera though. There may be more to the story than "it just quit working" (not my camera).

Lens assembly is quite complex and has a multitude of contacts and stuff so the camera knows the position of the lens mechanism. I have powered the motor with an external power supply and it operates freely over the entire range, so it's nothing to do with the actual mechanism being sticky or anything.

If only there was a service mode on the thing to be able to make sense of what's going on, beyond the lens moving in and out erratically and the camera spitting out a not-very-helpful "lens error"...