I may be overthinking this but need some guidance. Working on an automotive ECM and have a Hitachi HS8/2623 microcontroller connected to an ECSL200E which I cannot find a datasheet for. When I power up the board, I have correct input voltages at all Vcc areas on micro but there doesn't seem to be any oscillation from the crystal. It looks like a Pierce type setup, I have roughly 1.5VDC on either side of the crystal and the scope when connected to ground and I probe the crystal, I get no signal on either side. Both caps are not shorted and pathway to and from the micro is complete. What am I missing? I can provide picture if necessary.
https://www.renesas.com/us/en/docum...p-h8s2626f-ztat-h8s2623f-ztat-hardware-manual

 

It is not always possible to probe the pins of the crystal with an oscilloscope. The capacitance of the probe could cause the crystal to stop oscillating.

Also since the crystal is controlled by the MCU, it is possible that the initialization is not properly starting the crystal oscillator. Many MCUs have an internal RC oscillator that can be used as the default clock. After startup, the MCU can switch to the configured clock source. I don't know if your MCU works in this manner.

 

I wondered about the probe capacitance which makes sense. Is there a way to see if this XTAL is actually working in circuit then?

 

Usually it is the crystal oscillator that provides the processor clock for the processor. So if the processor is functioning as expected then the clock is clocking. Possibly not at the right frequency, but working.

 

Okay. I was hoping there was a way to verify by measuring the frequency right now or visually verifying that it is working? I am going to put it back on the machine tomorrow to see what happens. Pretty sure the processor is okay.

 

Did you try s 10X probe? I can measure the crystal frequency on my PIC boards with s 10X but not a 1X probe.

 

Did you try s 10X probe? I can measure the crystal frequency on my PIC boards with s 10X but not a 1X probe.

And connect it to the crystal by a 1pF capacitor which will very much limit any disturbance of the crystal circuit.

 

Not sure about that micro but all PICs I know have an output that if enabled for that would give a pulsed output with a certain submultiple? of the clock. I rarely used it having implemented, very early, a simple / robust clock based on high priority interrupts that I could "peek" through a certain OUT pin.

This last served as a reliable reference for my scope to decode repetitive signals where timing was higly relevant.

 

I may be overthinking this but need some guidance. Working on an automotive ECM and have a Hitachi HS8/2623 microcontroller connected to an ECSL200E which I cannot find a datasheet for. When I power up the board, I have correct input voltages at all Vcc areas on micro but there doesn't seem to be any oscillation from the crystal. It looks like a Pierce type setup, I have roughly 1.5VDC on either side of the crystal and the scope when connected to ground and I probe the crystal, I get no signal on either side. Both caps are not shorted and pathway to and from the micro is complete. What am I missing? I can provide picture if necessary.
https://www.renesas.com/us/en/docum...p-h8s2626f-ztat-h8s2623f-ztat-hardware-manual

Add a 100K-ohm resistor between your probe and the leg of the oscillator. You're not looking for a perfect shape, you're looking for a regular change, and the oscillator causes the probe to draw less current, which might let the oscillator keep going. Not perfect, but in a pinch, worth a shot.

 

Add a 100K-ohm resistor between your probe and the leg of the oscillator. You're not looking for a perfect shape, you're looking for a regular change, and the oscillator causes the probe to draw less current, which might let the oscillator keep going. Not perfect, but in a pinch, worth a shot.

That’s not going to change much when there is already 10 Meg on a 10X probe.

 

What is the symptom that makes you suspect that the oscillator is not working?? They seldom just fail for no reason.

If the case of the crystal is not tied to the circuit board ground buss, probing the case should allow you to verify that it is working. Crystals are sort of delicate and they are subject to damage if bumped too violently.

 

Thank you everyone for all of your responses so far. I will give them a try and see what I can find. I believe this MCU has internal and external XTAL and I am trying to measure the external crystal right now with no luck. I am doing it for two reasons. A) Because I want to see that the clock is working and B) for future diagnostics involving crystals. Up until now, most of the automotive ECUs I have come across, you can just read it with the scope with no problem. This one showed no pattern so I was curious. I had read that the impedance of my scope may cause this so I decided the fastest route was to ask the experts, hence, here I am.

 

What is the impedance of your oscilloscope probe?

 

What is the impedance of your oscilloscope probe?

The scope connection certainly can have a huge effect on the crystal oscillator operation, including stopping it completely. The data sheet is very obvious in the amount of information NOT provided about the crystal oscillator circuit. But given the very small capacitance values mentioned it seems that it is a low power oscillator circuit.
AND, Given that the TS has made no response to any questions about the scope connection, I guess that it is a clip lead on the end of three feet of RG58 cable. That will stop almost any oscillator.

 

The TS has not made any response because he was a little busy. My apologies. Here is a link to the scope probes being used:
https://www.circuittest.com/op-60a-oscilloscope-probe-60mhz.html

 

OK, and even in the "10X" mode the input capacitance is 10 pF. That could be part of the problem.

 

Okay it seems like the common theme is the scope probes so I will take that as probably my issue. Thank you everyone for your help.

 

The fact is that in some cases it is very challenging to measure without disrupting. (Also Known as "The Heisenberg Uncertainty Principle" ) That states that the energy removed by measuring a variable changes that variable. Which is why we use multimeters with higher impedance inputs.

 

What was the attenuation setting on your oscilloscope probe, x1 or x10?
We are assuming that you had it set to x10. Generally, we always set it to x10.
What is the frequency marking on the crystal?

Here is something to try. Put 5pF to 10pF capacitor in series with the tip of the oscilloscope probe. Measure with the free leg of the capacitor. You can try setting the channel input to AC and VOLTS/DIV to 500mV after the oscilloscope input has been adjusted for x10 probe, otherwise you would set it to 50mV with x10 probe, the result being 500mV/DIV measured.

 

The attenuation setting on the probes was x1.
The only marking on the crystal is ECSL200E and I could not find any datasheets on it.
I have the ECM coming back for final sealing so I will try the 5-10pF capacitor in series with the probe. The MCU is working as the machine fired up and controlled the engine as it should. My problem is fixed but still curious on crystals and MCUs and how they interact and diagnostics. You guys have provided much food for thought.