Auto Stop If Signal failure

Ask any questions about using PicoScope 6 Automotive software here.
gbgunn-st
Newbie
Posts: 6
Joined: Wed May 23, 2018 9:17 am

Auto Stop If Signal failure

Post by gbgunn-st »

Morning All,

Picoscope 4425

Hope Im not repeating a earlier post, I am after some help / advise.

I have a new excavator that will randomly stop running roughly once a week, we have monitored the relevant sensors and changed with new. We believe it is a ECU failure but the OEM will not authorise with out more data.

My question is does pico scope have a way of monitoring the input signals and if one signal were to stop signalling automatically stop the recording saving a engineer sitting with the machine for days on end,

Or even if I could wire a channel in to the engine run signal from the alternator.

Thank in Advance Simon

ben.martins
Pico Staff Member
Pico Staff Member
Posts: 539
Joined: Tue May 16, 2017 1:02 pm

Re: Auto Stop If Signal failure

Post by ben.martins »

Hello Simon,

This is indeed something we can set up. We tend to use Mask's and Alarms which are some of the more advanced features of PicoScope but can be vital when trying to capture glitches like the ones you are seeing.

Initally we would have to set up a mask around a channel that is fixed while the machine is running. This could be something like Crankshaft Position Sensor, an injector signal (current or voltage) or even battery voltage. Battery voltage might be good in this case as things like CKP sensor might be tricky to find but if the engine stops so will the charging therefore the battery voltage will drop to around 24V or 12V from approx 28V or 14.4V respectively.

To set up a mask we need to have a capture to mask around. If battery voltage is likely to be used then we need to ensure the battery is fully charged and then start the machine and allow for the charging to stabilise. Connect a channel to the battery Positive and Negative and start the scope. Allow the scope to capture a whole buffer and once we are happy that we have this on the screen we can stop the scope with either the space bar or the STOP button in the lower left hand corner. To add a mask to the channel click Tools > Masks and Add Masks and click Generate.

With the dialog box open we need to instruct the software as to how the mask will work. Entering figures into the X and Y Offset will tell the software which area to look around the screen to mask and how far above and below the trace to draw the boundary. This might be easier explained in this video from Mr Steve Smith.
CAN TRIGGERING.mp4
(7 MiB) Downloaded 573 times
If it were myself setting this up I think I would have at least 1s/div across the screen, maybe even 5s/div and 1 Million Samples. In the Mask generator you can input the X Offset for the entire screen which is indicated on the far right side of the X axis. eg, 1s/div time base + 10 divisions on the screen = 10s. In terms of Y offset, if we are sticking the battery voltage then we can see some fluctuation so maybe 3V would give us enough height to only enter the mask when charging has stopped and battery voltage has dropped significantly to indicate this. You can amend the X and Y Offset of your mask to allow for an increase or decrease in error detection. Once happy with the numbers click Apply and OK and you should now have a mask drawn either side of your battery voltage allowing for a positive or negative voltage error of 3 V either side of your battery voltage signal across the entire width of the screen. I'd recommend trying this first to ensure it will 'fail' the mask when the machine switches off so start the machine with the scope and the mask running and then switch off. You should hopefully see that the battery voltage drops into the mask creating a fail. If this is not the case you may need to adjust your Y offset by decreasing it to 2V.

Once we are happy that the mask is working correctly we can then instruct the software to perform a number of tasks once the mask has been failed. The current options we have are Beep, Play Sound, Stop Capture, Restart Capture, Run Executable, Save Current Buffer, Save All Buffers and Trigger Signal Generator. The only one we can't use with our Automotive Scopes is the Signal Generator but all the others will work as expected. One that might be useful for the work you are looking to do would be the Save Current Buffer or Save All Buffers. This will store the capture every time the mask detects a failure whether that is the buffer that the mask failed on or all of them. You do need to be aware though that if the scope is left to run and the mask failure is high, you will get a lot of files and will slowly fill up your computer! To activate Alarms click Tools > Alarms.
Alarms.png
Now you would need to click the drop down arrow to highlight and tick the box for Mask(s) Fail.
Alarm Mask Failures.png
Now click Apply. From here we now need to tell the software what we would like it to do if there is a mask failure. If you now click Add this will bring up the options I spoke of earlier.
Alarm options.png
For the type of fault you are experiencing I would suggest using the save all buffers. This will mean as soon as the mask is failed the software will save all the current buffers to a file on your computer. You will need to setup a location to save the captures and then add a name for the file. We can also add multiple events here to prevent the software from continuing to capture when the machine has stopped. If left the software would continue to save all buffers after the machine has stopped as the mask is constantly being failed. To prevent this from happening we need to add a Stop Capture event to the alarm section. Once you have made your selection click OK and on the list make sure the box is ticked and then click Apply and then OK. You are now ready to start capturing. Start the machine and allow charging to stabilise. Run the scope by using the start button in the lower left had corner and let it run. Hopefully when the machine stops and you have setup your other 3 channels to monitor the possible areas of concerns it will capture what is happening at the time the fault occurs.

I hope this helps and there are number of helpful files on the forum starting with this one Triggering on CAN with Masks topic12871.html and also here - topic14691-10.html. You will also find some more information on setting up Masks to trap a fault within the Mazda 3 case study which you can find here https://www.picoauto.com/library/case-s ... ions-fault where we demonstrate the techniques with Masks and Alarms to diagnose the fault.

There’s a lot of information here but if you need any further assistance please don’t hesitate to get in touch. It would be great to see how this one works out Simon and I'm keen to see what you find. Happy to look at this with you in more depth if you need any assistance.

Kind regards

Ben

User avatar
Autonerdz
TwoWaves
TwoWaves
Posts: 387
Joined: Wed Nov 20, 2002 4:19 pm
Location: Washington State USA

Re: Auto Stop If Signal failure

Post by Autonerdz »

Hi Simon,

Good information from Ben there.
I have a new excavator that will randomly stop running roughly once a week
Considering this statement, the solution could be quite a bit simpler.

1. Use a simple repeat trigger on any signal that will stop when the engine does.
2. Set up a long capture time and adjust samples so that you show a one microsecond sample interval in properties.
3. Verify that your settings result in at least two buffers captured.
4. Disable any PC sleep functions.

That's it. We call this capturing an unattended stall.

When the engine stalls, The scope will stop triggering and your result will be there waiting for you in screen one of two unless you are profoundly unlucky and it happened right between buffers.

hillp
OneWave
OneWave
Posts: 43
Joined: Wed Feb 11, 2004 11:04 pm

Re: Auto Stop If Signal failure

Post by hillp »

Why do we need at least two buffers captured? Isn't one enough? Using a single trigger with some pretrigger capturing to catch it in the act?

liviu2004
TwoWaves
TwoWaves
Posts: 472
Joined: Sat Mar 19, 2016 8:23 am
Location: Netherlands

Re: Auto Stop If Signal failure

Post by liviu2004 »

You didn’t understood.

hillp
OneWave
OneWave
Posts: 43
Joined: Wed Feb 11, 2004 11:04 pm

Re: Auto Stop If Signal failure

Post by hillp »

You didn’t understood
Exactly. That's why I'm asking.

liviu2004
TwoWaves
TwoWaves
Posts: 472
Joined: Sat Mar 19, 2016 8:23 am
Location: Netherlands

Re: Auto Stop If Signal failure

Post by liviu2004 »

Two buffers like doing two tests each capturing one buffer.

User avatar
Autonerdz
TwoWaves
TwoWaves
Posts: 387
Joined: Wed Nov 20, 2002 4:19 pm
Location: Washington State USA

Re: Auto Stop If Signal failure

Post by Autonerdz »

The reason you need two buffers using the method I described is that you need one to waste.

If you have settings so that only one buffer is able to be captured, then when the signal fails, the scope will move on to the next capture (buffer) and wait for trigger. Your original capture will be lost and you will have only flat lines waiting for you in the current buffer. You need to set up to capture at least two buffers so that when you return to the dead vehicle, you can stop the scope and go to the previous buffer to see what happened..

muttnjeff
TwoWaves
TwoWaves
Posts: 72
Joined: Tue Jul 21, 2020 7:19 pm

Re: Auto Stop If Signal failure

Post by muttnjeff »

Anyone had any success using the mask or two trigger setups?

Would like to explore this for use on sprayers as we run into many intermittent faults.

hillp
OneWave
OneWave
Posts: 43
Joined: Wed Feb 11, 2004 11:04 pm

Re: Auto Stop If Signal failure

Post by hillp »

Alternatively, would not a single trigger, with, for instance, 50% pretrigger work? Hit the start button, and the 50% pretrigger starts collecting the first half of the trace (say a couple of seconds), waiting for the the trigger to be actuated by signal droput. Then when the trigger is tripped by the signal dropping out, the 2nd half of the trace is collected. The computer sits there with the entire trace collected, waiting for me to come back and save it to disk. (Of course if the computer powers down in the interim [battery dies, if not plugged in], the trace is gone.)

How would the results or the risks be different in this way vs the 2 buffer way?

I have a thread going on a similar topic; I didn't find this one before I started that tread. Been having problems with reliable triggering using the "level dropout" trigger. topic22306.html?&p=100815#p100815

Post Reply