Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

how to measure the output of a crystal oscillator?

Status
Not open for further replies.

dl09

Full Member level 4
Joined
Feb 12, 2020
Messages
226
Helped
2
Reputation
4
Reaction score
1
Trophy points
18
Activity points
1,688
should i use a multimeter to measure the output of a 32 kilohertz crystal oscillator?
 

Only if the multimeter is specified to do so, most will not work above 2 - 3 KHz.

The two most important things to measure from a crystal oscillator are the amplitude (how big the signal is) and the frequency (how many times a second it repeats). Some multimeters have a frequency measuring function which would probably work but 32KHz is too high for most to be able to measure amplitude.

Another word of caution: if your crystal oscillator is low powered, for example using a tiny cylindrical watch crystal, the very act of connecting a meter to it may stop it operating. The correct instrument is an oscilloscope rather than a multimeter.

Brian.
 

the multimeter i use has no frequency measuring function, but will it detect a voltage at all, if the crystal oscillator has a frequency of 32 kilohertz?
 

there is no way a multi-meter can give a useful result - it will load the Xtal and likely stop it working - only a scope with a 100x probe or sensitive freq counter will do ...
 

Hi,

difficult to answer.. I just don't have the manual of your multimeter by the hand.

Klaus
 

i have a 74m294, which is a frequency divider, it can divide the frequency by up to 1 million, if the output of the crystal oscillator is sent to the 74m294, then i connect the output to the multimeter, can i measure any voltage changes?
 

I've never heard of a 74m294 and searching for it turns up hair dye!

If you had something that divided the 32KHz by 1000000 without loading the original you should get 32000/1000000 = 0.032Hz out from it but what you would measure is the output of the divider, not the oscillator. You can probably measure a rise and fall in voltage but it wouldn't tell you anything useful.

Brian.
 

the actual name is sn74ls294, i think it is made by texas instruments, sorry i zoned out and typed the wrong name. if i measure a rise and fall in voltage, then i just want to make sure the crystal oscillator is producing an ac voltage at 32 kilohertz. the sn74ls294 can divide by n. n can be set from 2 to 32 i think.
 

You may be able to measure a few volts of AC on a 32kHz output ( 5V ) depending on your multimeter ...
 

Hi,

Rough idea, may or may not work (someone correct me if it's a bad idea):

Crystal oscillator > op amp voltage follower > 30kHz to 34kHz bandpass filter > op amp voltage follower > RC filter to get average voltage > op amp gain stage (gain of 2) > op amp voltage follower > DMM on DC voltage setting.

Or, but guess also needs buffering (high impedance input, low impedance output) RC filter > DMM on DC voltage setting.

- - - Updated - - -

Hi again,

Even better, maybe..., if you just want to check the output is approximately 32.768kHz, use your divider IC or a CD4060 and divide down to 2Hz or 3Hz or 4Hz (or whatever your eyes can see change on>off>on>off>etc. enough to count) and feed the output to an LED, visual indicator and extrapolate the approximate upstream frequency...
 
Last edited:

Hi,

Rough idea, may or may not work (someone correct me if it's a bad idea):

Crystal oscillator > op amp voltage follower > 30kHz to 34kHz bandpass filter > op amp voltage follower > RC filter to get average voltage > op amp gain stage (gain of 2) > op amp voltage follower > DMM on DC voltage setting.
This won't work. All you get is zero reading.
If you want to see some amplitude related value, then you need a recifier before the RC.

The OP says:
if the crystal oscillator has a frequency of 32 kilohertz?
Thus he's rather interested in frequency than in amplitude.

Thus this is a good idea:
Even better, maybe..., if you just want to check the output is approximately 32.768kHz, use your divider IC or a CD4060 and divide down to 2Hz or 3Hz or 4Hz (or whatever your eyes can see change on>off>on>off>etc. enough to count) and feed the output to an LED, visual indicator and extrapolate the approximate upstream frequency...

Klaus
 
  • Like
Reactions: d123

    d123

    Points: 2
    Helpful Answer Positive Rating
The idea is fine but I think it is going right over dl09's head. Designing and building filters and amplifiers and the rectification you didn't include is rather advanced for someone who doesn't understand the basics of how to use a multimeter.

dl09, can you tell us the exact purpose of your circuit and if possible show us a schematic of the oscillator wiring. For anyone who has an oscilloscope, what you ask is a trivial task but with the wrong test instrument it is almost impossible for anyone, regardless of their expertise. It is like you are searching for instructions to drink coffee with chop sticks.

Brian.
 
  • Like
Reactions: d123

    d123

    Points: 2
    Helpful Answer Positive Rating
32kHZ can be measured easily with modern multimeters but it will probably charge the oscillator by probe capacitance and therefore oscillator may stop.A simple buffer voltage follower will definitely help you even build up with a simple 741 OPAMP.
Best method is using a low capacitance oscilloscope probe with a frequency and amplitude readout featured oscilloscope..
 

Hi,

Thanks for the correction. Must ask: Are precision rectifiers + whatever else may be needed rather than absolute value circuits appropriate to measure the average pulsed DC voltage?
 

Hi,

With your idea:
Bandpassfilter filters out DC level, remaining pure AC.
Then following RC filter show the DC level. Since DC already is filterd out, I expect zero voltage at RC output.

AC voltage (without DC) fed to a rectifier will shift the negative part of the AC signal to the positive part.
Thus you get a pulsed, positive only signal. It now contains DC, which can be filterd (averaged) using an RC filter.
This DC voltage will give some clue about the amplitude, but it depends on waveform.

The output of a narrow bandpass filter will be a pure sine.
The DC value of a rectified pure sine is: 2 x amplitude_sine / Pi.
Mind that the output amplitude of a narrow bandpass filter will be very sensitive to a frequency mismatch.

Klaus
 
  • Like
Reactions: d123

    d123

    Points: 2
    Helpful Answer Positive Rating
Must ask: Are precision rectifiers + whatever else may be needed rather than absolute value circuits appropriate to measure the average pulsed DC voltage?
Precision rectifiers are more linear than simple diode rectifiers at low signal levels due to the forward voltage drop in the diode being offset by an equal drop in the feedback path but they are no better at distinguishing between different waveform shapes.

Brian.
 
  • Like
Reactions: d123

    d123

    Points: 2
    Helpful Answer Positive Rating
then i just want to make sure the crystal oscillator is producing an ac voltage

Although many interesting ideas are emerging, it seems that OP just wanted to detect whether the crystal is working or not; again he started a thread saving words which led to different understandings. Maybe he wanted to say 'detect' frequency instead of 'measure' frequency. Another reason to guess this is that he said 32KHz frequency without the additional digits (32,000, 32,768, ...?). This is the effect of a weakly elaborated question.

BTW, @dl09, whenever you refer to a crystal, keep in mind that it is a passive circuit, while crystal oscillator (XO) is rather an active, self-oscillating circuit that does not need any additional components to have its output oscillating, and in this particular case a simple LED properly biased after a logic gate could solve your problem; anyway if you persist want using the multimeter to do that, as said above, yes it could give some value at in the AC scale, but perhaps you should add some capacitor in series to decouple any DC part of the signal.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top