Can I swap a cer. resonator for an xtal?

[DaveC]

This question has evolved out of a related question posted elsewhere in this
NG, but I consider that it is now "buried" in the many excellent responses to
that question. So...

Can I use a crystal in place of a ceramic resonator? This circuit is the
clock for a National Semi proprietary (and hence, no pin data available)
microcontroller in a car alarm. Two 33uF caps tie the 3.58 MHz res. to gnd in
a "PI" config. One end of the res. is tied directly to the Clock + (I
presume) pin of the uP, and a 1M-ohm resistor ties the other end to the Clock
- pin.

I can get a resonator mail-order, but I can find only crystals locally. I
will order the resonator if that is what is called for, but if I can
substitute, I can buy that today.

Thanks,

 

[Dave VanHorn]

Can I use a crystal in place of a ceramic resonator?

Almost certainly.
Without a spec sheet for the chip, the best advice I can offer is "try it
and see".

 

[Dave Platt]

Can I use a crystal in place of a ceramic resonator? This circuit is the
clock for a National Semi proprietary (and hence, no pin data available)
microcontroller in a car alarm. Two 33uF caps tie the 3.58 MHz res. to gnd in
a "PI" config. One end of the res. is tied directly to the Clock + (I
presume) pin of the uP, and a 1M-ohm resistor ties the other end to the Clock
- pin.

I can get a resonator mail-order, but I can find only crystals locally. I
will order the resonator if that is what is called for, but if I can
substitute, I can buy that today.

I think the answer is "Maybe".

It'll depend on at least a couple of things. Different types of
crystal vary in the amount of drive power that they expect/require
from the uP, and in the amount of shunt capacitance required to ground.

Looking through a PIC microcontroller's datasheet, I note that LP (low
power) crystals seem to be used for 32 kHz operation... not what
you're looking for, so this (low-power) mode doesn't seem to apply.

The PIC has a single mode for standard parallel-cut crystals and for
resonators. The two differ in the size of the shunt capacitors... it
specifies 30 pF for resonators (and I suspect that the "33 uF" caps
you've indicated are actually 33 pF), and anywhere from 15 pF (1 MHz -
4 MHz) up to 47-68 pF (200 kHz) for crystals. "Values are for design
guidance only".

So, my guess is that you could probably substitute a standard 3.58 MHz
parallel-cut crystal for your resonator, but might have to pull out
those 33 pF caps and replace them with caps of about half that size.

Couldn't hurt to try - if the drive level is too low for the crystal,
or the shunt capacitance is wrong, the oscillator just wouldn't start
up.

 

[DaveC]

Have you checked the voltages in-circuit with 12V applied to
all the correct places on the alarm's wiring harness and ground applied
to the other correct places?

While it's not impossible that the crystal/resonator/chip has died, it
seems far more likely to me that there's either a wiring problem or a blown
regulator/fuse.

The fuse on the PCB may not look like a fuse - it might look like a resistor
or even a capacitor. Any "F1" designations?

On-board v-reg. is putting out the specified 5 volts to all ICs. System
ground was solid, but I moved the harness ground to the vehicles main ground
point behind the instrument panel, just to be sure.

Symptom is that the microcontroller (that this clock drives) is driving the
output relays to chattering (about 10 Hz).

I think the clock is gone wonky due to an old resonator. I'm going to look at
the resonator's output again tomorrow. I've been having problems viewing it
with my 'scope; the probe loads it down to where it quits altogether. I'll
try some isolation techniques suggested by others in this NG, and report what
I see.

Thanks,

 

[Tim Shoppa]

On-board v-reg. is putting out the specified 5 volts to all ICs. System
ground was solid, but I moved the harness ground to the vehicles main ground
point behind the instrument panel, just to be sure.

Symptom is that the microcontroller (that this clock drives) is driving the
output relays to chattering (about 10 Hz).

10 Hz is a pretty typical frequency for motorboating. The regulator might
be putting out 5VDC, but is it really flat 5VDC? I know you checked the
ground, that's a very good thing to check in case of motorboating, but you
also should check for a good clean low-impedance connection on the 12V side
too.

What may be happening is that the microprocessor starts up, it commands
one of the output drivers, the output driver draws so much current from a
flaky 12V supply that the 12V on the PC board drops low, the 5V follows it
low, then the whole cycle repeats. That seems pretty likely to give you
the motorboating that you see. It could be either a high-impedance 12V
connection or a short in the output harness.

I think the clock is gone wonky due to an old resonator. I'm going to look at
the resonator's output again tomorrow. I've been having problems viewing it
with my 'scope; the probe loads it down to where it quits altogether.

What do the 5V and 12V look like on your scope?

I'd start looking at electrolytics on the PC board before I suspected
the resonator...

Tim.

 

[DaveC]

10 Hz is a pretty typical frequency for motorboating. The regulator might
be putting out 5VDC, but is it really flat 5VDC? I know you checked the
ground, that's a very good thing to check in case of motorboating, but you
also should check for a good clean low-impedance connection on the 12V side
too.

Removed all caps & tested with cap tester. All within 5 percent. Voltages at
the chips are solid.

Thanks,

 

[DaveC]

http://home.covad.net/~peninsula/Res1.jpg

Any ideas who's logo that is on the resonator?

 

[Jim Thompson]

Any ideas who's logo that is on the resonator?

Murata?

...Jim Thompson

 

[DaveC]

So, my guess is that you could probably substitute a standard 3.58 MHz
parallel-cut crystal for your resonator, but might have to pull out
those 33 pF caps and replace them with caps of about half that size.

Couldn't hurt to try - if the drive level is too low for the crystal,
or the shunt capacitance is wrong, the oscillator just wouldn't start
up.

So, it sounds like my steps should be:
1. replace with crystal
2. see if the uP runs. If not,
3. Change out the 33pF caps for 15pF (or so)
4. See if the uP runs. If not, back to the resonator...

Sound right?

Thanks,

 

[Ricardo Matos Abreu]

Old Samsung logo? (Three squares with a star inside each square)

 

[Spehro Pefhany]

Any ideas who's logo that is on the resonator?

Looks like Kyocera. There are a lot of companies that make resonators,
including some in China.

Best regards,
Spehro Pefhany

 

[Michael Black]

So, it sounds like my steps should be:
1. replace with crystal
2. see if the uP runs. If not,
3. Change out the 33pF caps for 15pF (or so)
4. See if the uP runs. If not, back to the resonator...

Sound right?

Thanks,

Desolder the resonator. Breadboard an oscillator and try the
resonator in it. You can put a gate between the oscillator and
the scope probe, isolating it more than you can in-circuit. If
it works there, then obviously something else is wrong.

Michael

 

[Dave Platt]

So, it sounds like my steps should be:
1. replace with crystal
2. see if the uP runs. If not,
3. Change out the 33pF caps for 15pF (or so)
4. See if the uP runs. If not, back to the resonator...

Sound right?

That's probably the way I'd do that sort of substitution...

 

[Jason D.]

Looks like Kyocera. There are a lot of companies that make resonators,
including some in China.

Yes Kyocera. Very sloppy stamp! I seen this on ICs in old days,
3.58 sounds like common MHz, grab one from recent junk chassis that
still use this type resonator for color burst. Heck, try 4MHz or
3.5MHz or in between, did this before on few stuff, some stuff that
doesn't care about freq worked rather well.

Wizard

 

[Gary P. Fiber]

Any ideas who's logo that is on the resonator?

Nice photo, what manufacturer and model of camera did you use?


Gary K8IZ
Washington State Resident
Registered Linux User # 312991

 

[Jim Thompson]

Yes Kyocera. Very sloppy stamp! I seen this on ICs in old days,
3.58 sounds like common MHz, grab one from recent junk chassis that
still use this type resonator for color burst. Heck, try 4MHz or
3.5MHz or in between, did this before on few stuff, some stuff that
doesn't care about freq worked rather well.

Wizard

http://www.actcrystals.com/PDF Files/SM Master Resonators.pdf

I'm using type ZTT at 3.58MHz (standard part).

...Jim Thompson

 

[Dave VanHorn]

For cap values, take the xtal's loading spec, double it, and subtract 5pF or
so.
22pF loading spec, try two 39pF.

Frequently done wrong, and by those who should know better.

 

[Dave VanHorn]

<http://www.jameco.com/cgi-
bin/ncommerce3/ExecMacro/jameco/searchResult.d2w/report?sort=BPA&search=1378
16
which states 17pF, the two 30pF caps currently in the circuit would be

fine?

Sounds about right.

There are two main types of crystals. Parallel resonant, and series
resonant.
Series resonant crystals don't specify a loading capacitance.
You'll never get a series crystal to run at exactly the right speed in a
parallel circuit.
In many applications, the difference isn't important, so people aren't aware
of the problem, but when they do this, and need precice timing, then they
get bit.

 

[DaveC]

http://www.actcrystals.com/PDF Files/SM Master Resonators.pdf

I'm using type ZTT at 3.58MHz (standard part).

I presume you're in the States. Where did you get this resonator?

Thanks,

 

[Fred Abse]

You'll never
get a series crystal to run at exactly the right speed in a parallel
circuit.

If you get it to run at all.