Trying to catch an intermittent shutdown at idle in a 1997 Audi A8. There's lots of extremely short duration noise spikes, positive and negative, on the signals. This could trigger an unwanted data collection event, so I've set a level dropout single trigger for when the fuel pump relay output drops below 1 V for a second. I have a 50% pretrigger running. The total scan time is 5 seconds. The problem is that the 4423 scope seems to trigger at random. After I hit Go, the pretrigger will run for up to about a minute, tops, before the trigger kicks in the remaining 50 percent and stops data collection. But there is absolutely no sign of the signal (channel D below) dropping below 1 V for a whole second. The car would have shut down with the fuel pump off for that long anyway. What am I doing wrong? Does it have to do with my choice of collection time and number of samples?
The file's too big to email through my email program. However, things are getting weirder, with the scope not triggering at all now, even when I physically unplug the probe from the voltage source on Channel D. And not even when I have it plugged in and shut the car off with the key. I know there's no voltage on Channel D then because I had a VOM also hooked up. Then, when I restarted the car with the key, the scope triggered. I can't seem to attach that file to this post either, but it's small enough (2377 KB) that I was able to email it. Windows Task Manager shows a lot of CPU being used, and shows Picoscope 6 as a 32 bit program. 32 vs 64 bit?
The laptop is a 2 year old HP running Windows 10, with Celeron N4000 CPU at 1.10 GHz, with 4 GB of installed RAM. 64 bit operating system, x64-based processor.
According to the description of the trigger, "level dropout" actually trips on decrease or increase, so it seems a bit misnamed. 1 second before time zero on the waveform that actually triggered (screenshot earlier; sorry there are 3 of the same; I couldn't at the time tell they'd been successfully attached), the voltage increased from zero. So why did the scope trigger upon voltage increasing, but not when I disconnected the input cable's + probe from the signal voltage source? Or shutting down the vehicle with the key, which dropped out the relay supplying the signal? Both of those attempts were for much longer than 1 second. (These questions are in addition to the random triggering that precipitated this thread, in which no valid trip points happened, but the scope triggered anyway.)
Before we continue to look at the “Level Dropout” Trigger could you take a look at a number of options here from Ben and Autonerdz topic20161.html
The Autonerdz “unattended stall” procedure would work well for your channel D input once dropping to near zero volts at engine stall
Looking at your first image we have a 4423 capturing Primary ignition and injector voltage. The injector voltage appears over-ranged and may also exceed the max input voltage of 100 V. (This could influence the behavior of the scope)
Could we add attenuation to channel C to capture the induced voltage, so removing this “unknown” from our evaluation of the level dropout trigger?
Could you confirm the relevant attenuation is applied to Channel; B as we have what appears to be 25 V for primary ignition events? (perhaps a setting issue)
Could you confirm we are using the latest PS6 build of software?
Reducing the spike/noise level will help too.
Could you confirm we are earthed to a good clean chassis ground and our test leads are routed away from sources of ignition/injection noise?
I will try to replicate your set up given the issues you are experiencing posting your waveform
Thanks for responding. I'll look at the suggested options from Ben and Autonerdz. I'll put a 20:1 attenuator on Channel C, injector voltage. (I could alternatively look at injector current with a clamp-on, but the wiring is a little tight without cutting the wiring harness plastic sheath away.)
The Primary ignition voltage is already run through a 20:1 attenuator.
I will download and install the latest PicoScope 6 Automotive software (6.14.25). According to the startup banner, I currently have 6.13.14.4038. (different rev numbering system?)
I've earthed the underhood probes to an aluminium air conditioning line that's clamped to sheet metal in a number of locations. I could likely find something better, though the battery is in the trunk (boot) for the Audi A8 of this generation.
Away from potential ignition/injector noise? I don't think so, since we're measuring ignition and injector waveforms. Not sure it's possible.
OK, I updated the software to 6.14.25, and put a 20:1 attenuator on channel C, injector voltage. I read the thread about capturing an unattended stall (Auto Stop If Signal Failure, topic20161.html?&p=100813#p100813), with the autonerdz post giving the much simpler option (than using mask with alarm). I didn't understand the why of their advice, so I posted a question there: 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?
So I ran with my original triggering strategy, fuel pump voltage (Channel D) drops to less than 1 volt for one second, with a 50% pretrigger. I started the scope, and it collected the 50% pretrigger, constantly keeping it updated. Then I unclipped the Channel D positive lead from the fuel pump voltage. No trigger. Several seconds later, I reconnected the Channel D probe. Now it triggered. File saved, attached. Then I started the collection again, and this time an actual stall occurred. No trigger. I must not understand something about the level droput trigger. Is there a a setting for direction? I did set the hysteresis to 0%; the default is 1.5%. I think I need to study the hysteresis option. Another observation - while I was testing, at one point, with the pretrigger 50% running, the 50% signals seemed to stop constantly updating, but the trigger never tripped (as it should not have).
update - psdata file didn't successfully attach. It's 10332 kB. Is this too large? I'll email it. screenshot below, showing what seems to be a correct triggering if I had wanted it to trigger on level coming up when I reconnected Channel D.
Hello and thank you for the posts. Sorry for the late reply.
The principle of the following trigger settings allows the user to either watch for changes in stable signals (E.g. during monitoring) or, capturing intermittent events such as “cutting out”/stall when the vehicle is unattended (As the team at Autonerdz have mentioned)
The video below demonstrates how PicoScope behaves like a Chart Recorder when the trigger conditions are set outside of the typical operating parameters of the signal.
In the video we are capturing the crankshaft sensor and diesel common rail fuel pressure sensor signals. (Note the signal drop out of Channel B between 47 & 49 seconds)
We have set a Rising Edge, Repeat Trigger @ 7 V with our Pre-Trigger set to approx. 90%
Note our crankshaft sensor signal is typically 0 – 5 V and therefore the trigger conditions above should never be achieved.
Note how the fuel pressure signal (Channel B, Red) momentarily drops to 0 V indicating a fault. No doubt the vehicle also would respond with a hesitation.
If the capture is being monitored, the user has the time is takes the “drop out” of Channel B to travel from its origin (at the right-hand side of the waveform) to the left-hand side of the screen to stop the capture before the glitch is lost forever (Approx. 18 seconds)
Using the following triggers, we can use the same technique to capture such events unattended.
Test 1 Repeat trigger falling edge at 500 mV Channel C Fuel pressure signal (90% Pre-Trigger)
When the fuel pressure drops to 500 mV the car should stall and we capture the event (Hands free)
Test 1 Repeat Trigger
This capture will continue to run after the stall as we have used a repeat trigger. On return to the vehicle the scope would still be running with buffer 1 containing the stall data and buffer 2 displaying current data. (Current data may or may not be relevant)
Test 2 Single trigger falling edge at 500 mV Channel C Fuel pressure signal (90% Pre-Trigger)
When the fuel pressure drops to 500 mV the car should stall and we capture the event (Hands free)
Test 2 Single Trigger
This capture will stop immediately after the stall as we have used a single trigger. On return to the vehicle the scope would have stopped and display1 buffer (With offending data)
Test 3 Single trigger Level Drop Out Channel A Crankshaft sensor signal (90% Pre-Trigger)
When the crankshaft signal remains high or low for 100 ms the car should have stalled and we capture the event (Hands free)
Test 3 Level Drop Out Single
I hope the above help and please feedback for any clarification