I think we may be missing the point here. I think the OP is concerned about the reliability of the oscillator, not its precise frequency. The load capacitors will change many characteristics, including the voltage across the crystal and the startup time, or if they are off enough, they may prevent reliable startup at all.
How about one of these probes?
https://www.ebay.com/itm/332393676082
I think we may be missing the point here. I think the OP is concerned about the reliability of the oscillator, not its precise frequency. The load capacitors will change many characteristics, including the voltage across the crystal and the startup time, or if they are off enough, they may prevent reliable startup at all. I think you would select the capacitors by starting with the datasheets and a little math, then doing final testing with various values and testing both the crystal voltage to make sure it isn't too high and then the startup characteristics over the range of temperatures you expect the device to be subjected to. If you want to measure and view those things, you need a probe such as the one I mentioned--and even that is going to have at least a small effect.
This isn't tuning a transmitter. How accurate do you thing the average internal MCU oscillator circuit is anyhow, even with a proper crystal?
EDIT: It's not differential. Thats going to be an issue because you generally can't ground either side of an XO circuit....
EDIT: It's not differential. Thats going to be an issue because you generally can't ground either side of an XO circuit....
There's no need for differential probe to measure crystal oscillator. Idea that part of the circuit must be grounded for single-ended probe is bizarre
Here's Epson tech papers regarding crystal oscillator measurements:
https://www5.epsondevice.com/en/information/technical_info/pdf/tech_notes_e201302.pdf
https://www5.epsondevice.com/en/information/technical_info/pdf/tech_notes_e_oscillator_circuit_evaluation_method_2.pdf
As far as grounding, are you proposing not using the probe ground lead? If not, where do you propose connecting it? Some XOs, like the last one I worked on, are in isolated circuits. Every circuit is different and IMO the easiest, quickest way of measuring startup and drive level characteristics would be an active differential probe right across the crystal. I'm not sure where your "bizarre" comment comes from--a single ended probe connected to a regular oscilloscope has a ground lead that when connected will ground that part of the circuit. Now it may be the case that many XOs can be measured easily with a single-end low capacitance probe, but I assure you that some can't.
One thing that can be done is to have the xtal manufacturer verify the design will work over a wide range of evironmental conditions. Need to supply the relevant part of the schematic and samples of the circuit.
As far as grounding, are you proposing not using the probe ground lead? If not, where do you propose connecting it? Some XOs, like the last one I worked on, are in isolated circuits. Every circuit is different and IMO the easiest, quickest way of measuring startup and drive level characteristics would be an active differential probe right across the crystal. I'm not sure where your "bizarre" comment comes from--a single ended probe connected to a regular oscilloscope has a ground lead that when connected will ground that part of the circuit. Now it may be the case that many XOs can be measured easily with a single-end low capacitance probe, but I assure you that some can't.
Any XO circuit needs power, it has positive supply rail and negative supply rail which usually is considered as ground in unipolar circuits. So any XO already has ground where it's signals are referenced-to. Connect ground clip of your probe there. Saying "you generally can't ground either side of an XO circuit" does not even makes sense or we got some strange language barrier problem here. [edit] By grounding any side of XO I understand - short it's output or input to ground. Obviously such suggestion is bizarre to me.
Most of the oscillators are either Colpitts or cmos inverter topology, boths has grounds, both can be measured using fet probe. No need for fancy FET differential probes I am not sure even exist. Could you show pointers to differential probe you are talking about?
They exist.
https://www.tek.com/datasheet/differential-probes-2
I've been stalking them on eBay, but they still go for big bucks unless you buy them as-is, which means someone probably burned it out--they're quite sensitive.
They exist.
https://www.tek.com/datasheet/differential-probes-2
I've been stalking them on eBay, but they still go for big bucks unless you buy them as-is, which means someone probably burned it out--they're quite sensitive.
It is widely known that differential probes exist. I did ask you differential FET probe. Hint: one with FET transistor input. TEK differential probes (P624*) are *not* FET probes because they have differential 200k input resistance and 100k common mode resistance (attach) - as soon as you connect such to your oscillator, it (oscillator) will stop operation.
200K is still much higher than the impedance of the small input capacitance at most XO frequencies
200K is still much higher than the impedance of the small input capacitance at most XO frequencies
Before we continue this debate, you better show ANY mention of using differential probes for XO measurements in papers from reputable sources, preferably crystal/oscillator manufacturers. I did show Epson papers suggesting that only single FET probe and little brain cells needed to check XO.
If I happen to acquire a suitable probe, I will try it and report back.
If I happen to acquire a suitable probe, I will try it and report back.
Good luck with that. Before you invest into probe - solder some 100k load resistors to various crystals and see how it goes. Start with 32KHz tuning fork crystal.
I think we may be missing the point here. I think the OP is concerned about the reliability of the oscillator, not its precise frequency. The load capacitors will change many characteristics, including the voltage across the crystal and the startup time, or if they are off enough, they may prevent reliable startup at all. I think you would select the capacitors by starting with the datasheets and a little math, then doing final testing with various values and testing both the crystal voltage to make sure it isn't too high and then the startup characteristics over the range of temperatures you expect the device to be subjected to. If you want to measure and view those things, you need a probe such as the one I mentioned--and even that is going to have at least a small effect.
This isn't tuning a transmitter. How accurate do you thing the average internal MCU oscillator circuit is anyhow, even with a proper crystal?
the best way to test it is to write a little program to toggle a pin with the maximum frequency possible and measure there ...
How did we go from discussing a 12MHz MCU crystal to a 32kHz low-power unit with an ESR that is probably 1000X higher?
How did we go from discussing a 12MHz MCU crystal to a 32kHz low-power unit with an ESR that is probably 1000X higher?
We discussed crystal oscillator measurements, suggested methods that work for most of the crystals. Fact that you push towards HF 12MHz crystal to make your invention of using differential probe work, is your problem. I already said that - good luck with that.
OP needs just one thing - ensure that load capacitors have right capacitance. Easy way is to measure frequency of XO w/o loading it much with probe. If you have option to output 1:1 or divided frequency to dedicated pin - then you can use 1:10 generic scope probe and call it a day, considering that your scope have good enough frequency counter. Otherwise you need to use dedicated counter. If there is no pin to output MCU clock - you better use FET probe to measure XO output. You need to check couple of devices/crystals and ensure that they fall into specified frequency tolerance, their "ppm specs".
Drive strength measurements are needed when you need to find lowest possible, still stable drive strength in low power application. Obviously it would be every real-time clock that uses low frequency crystal (32KHz). When you use HF crystals and saving 1mW is not important - just drive your crystal at max as whole industry do, that's it.
The OP stated 12MHz and an MCU, IIRC. My ramblings henceforward reflected this, that is something we call context.
If you are stuck with a particular MCU and production changes, you may need to change the circuit or the crystal to get reliable operation, and if you can't get reliable operation with a particular MCU under test conditions, you may need to use an external oscillator. These are the things the OP needs to know, since he is clearly concerned with reliability beyond the "hey, it works OK" level.
What exactly is method of measuring drive strength with differential probe? Please describe steps in enough details.