Problem in connecting quartz crystal to oscillator circuit via a BNC connector

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
Hello engineers
I am designing a quartz oscillator following to bellow topology
NV_0518_Frenzel_Figure03.jpg
However, I am not strictly following to this topology. Instead of directly connecting Quartz crystal to the circuit, I put a BNC connector between quartz crystal and circuit in order to connect them together. Unfortunately, my configuration did not work.

So in my view point, I think the BNC connector disperse/ or consume power, leading to not enough energy to drive the quartz crystal oscillation.

Therefore, please help me to treat this problem. Do we need any additional circuit for BNC connector ? Or could you please give me some ideas for solving my problem ?
I really need helps from you.
Thanks in advanced.
 

dendad

Joined Feb 20, 2016
3,251
have you tried the circuit with the crystal connected directly to it first to ensure it works?
How much cable do you have between the oscillator and the crystal? It may be there is too much capacitance there.
Why do you want to connect the crystal via a BNC connector anyway?
 

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
have you tried the circuit with the crystal connected directly to it first to ensure it works?
How much cable do you have between the oscillator and the crystal? It may be there is too much capacitance there.
Why do you want to connect the crystal via a BNC connector anyway?
Hi
my oscillator worked well with direct connection.
my cable length is around 15cm.
I design my circuit for especial purpose, that''s why the crystal could not be directly wired to oscillator circuit.
Could you give me some ideas or any reference figure/material (doccuments, articles,..) related to my problem ? I think it should have a additional circuit like this: crystal connection points on oscillator circuit > connect to an input of an additional circuit > output of the additional circuit connect to BNC connector mounted on PCB.
 

SLK001

Joined Nov 29, 2011
1,543
A crystal oscillator is a very delicate creature. If you add too much stray capacitance, stray resistance or stray inductance, it will just lie there dead. What you have done is add all three "extras" and expect the xtal to work the same as without the extras. Perhaps if you told us exactly what your "special purpose" was, we could offer alternative approaches.
 

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
A crystal oscillator is a very delicate creature. If you add too much stray capacitance, stray resistance or stray inductance, it will just lie there dead. What you have done is add all three "extras" and expect the xtal to work the same as without the extras. Perhaps if you told us exactly what your "special purpose" was, we could offer alternative approaches.
Hi SLK001

Actually, I am using a quartz crystal as a sensing element. Anh this sensor can not be integrated onto the oscillator PCB board.
My sensor system is similar to the bellow figure
QCM200.jpg
So, could you recommend me any another approaches to connect a quartz crystal to an oscillator circut ?. My distance is less than 20cm.
 

Thread Starter

Ngô Đình Nhân

Joined Jan 8, 2017
45
There is enough space in the quartz crystal holder for SMD components.
Put the oscillator there.
Hi MrChip
For my applications, I can not put crystal and its oscillator in same location.

And also, I only want to use a BNC connector, I am not using a coaxial cable. Because with a BNC connector, I can plug and unplug quickly.

Do you have any ideas for my problem ?

Best regards
 

crutschow

Joined Mar 14, 2008
24,045
What is the oscillator frequency?

If y0u insist on that configuration, then you will have to measure the capacitance of the BNC connector and wire to the crystal, and then determine how to get the oscillator to work with that extra capacitance (which may or may not be possible).
A different oscillator configuration may be more tolerant of stray capacitance.
 
Last edited:

BobTPH

Joined Jun 5, 2013
2,192
Every thing I have ever seen about using quartz crystals say they should be as near as possible to the oscillator. If I recall correctly the layout guidelines for PICS suggest the crystal should be within 1/4 in of the osc pins.

Bob
 

Phil-S

Joined Dec 4, 2015
140
Give these guys at least a hint of what the "special purpose" is. Something to do with Ghostbusters?
Does it vibrate? Sit in some flux field? There might be something available that would suit your purposes without the hassle of BNC and those etched components
 

sbkenn

Joined Apr 15, 2017
21
Crystal oscillator circuits are VERY sensitive to capacitance/inductance, less so to resistance. Even on a PCB, connecting test probes to the crystal will often prevent it from starting. I worked in a volume manufacturing establishment, and had great difficulty persuading my boss that it was a real problem. A year later, there was a substantial article in EDN magazine about precisely that problem. Whatever you are trying to sense, I suggest that you find a different method.
 

BR-549

Joined Sep 22, 2013
4,938
Oscillator....and sensor is confusing. A crystal that has been cut for stability......I would think.....would make a bad sensor.

Of course it depends on the true application. A crystal is still a piezo device and has those characteristics. So it can sense, but not have anything to do with oscillation.

So......do you want to sense or oscillate?
 
This circuit will deliberately not work at your configuration. You should try 2...3-transistor circuits which have "buffered" drive and sense signals. Cannot reproduce now their correct english names, but they can be found in the net. May be it will be necessary to add phase shift compensation network to coax. cable.
It is possible that such devices have generator with parasite power inside crystal holder, it is big enough. This allows easier temperature compensation, which is needed if device works in vacuum (film thickness meter) or something like.
Also shown devices may have not coax connection but simple two separate wires, which have much less capacistance. Consider using this if your setup is not exposed to EMI.
 
This almost looks like a thickness monitor. e.g. http://www.hsdengineering.com/images/Kronos QM_300_001.pdf

The connector was similar to an SMA connector, maybe smaller. Within <2 foot there was a transformer, then the cable to the instrument was a PTFE sheathed RG174 like cable that could be quite long.

I had to buy two making connectors and two crystals of different frequencies to use as a performance verifier.

We etched off the stuff we deposited and the remaining gold to make our own re-furbished crystals.

or this: https://products.inficon.com/getattachment.axd/?attaName=de4133ef-e576-489e-b763-64c5fc55497b

The connector might be a Microdot connector.

Who knows what the OSC100 is: http://colnatec.com/?s=osc100&x=0&y=0 (Thin film Oscillator) This https://products.inficon.com/getattachment.axd/?attaName=ce6a66fe-d8c9-42bb-b92d-608e59429802#page=39 gives me an IDEA of what the OSC100 is. It (the OSC100) MIGHT be an option for you.
 
Last edited:
Top