I have never been able to get a sensible trace using the ignition pickup lead. After buying the pickup I put the problem down to the unusual input characteristics of my Nicolet bench scopes.
The lead has rested undisturbed until a few days ago when I tried to use it again. Once more the trace is impossible to interpret, even taking into account:
1. a possible strange fault in the ignition system
2. some sort of mismatch into the scope.
Could you tell me what is actually in the ignition pickup, or at least its equivalent circuit so that I could try to guess what it is trying to tell me?
It has 1 meg sampling rate, 2-channel (each differential) 1 megohm inputs.
I know that it does not match with normal x10 probes because its input capacitance is too high. I think it's more than 60pF, but I can't find the actual spec just now, but I know that a x10 probe cannot be matched, so I thought my problem with the HT probe was similar.
Just now, I cannot retrieve my old records from this scope, but I have just made some more with a cheap PC scope. This one does match to x10 probes but still shows strange results with the HT probe. [Attached]
I would really appreciate if you can help me with understanding these traces. I have very limited experience with the noisy and dirty automotive environment.
Michael, I would really appreciate it if you could find a reply to my problem.
If I still have not posted enough info, or if my attachment is not clear enough, please let me know and I'll try to do better.
I would also like to have permission to view attachments. Do I need any other qualifications other than being a Pico customer and a registered user?
Best regards,
Ed
Michael,
I have done some simple tests just on the MI074 HT probe.
I measure a dead short (less than 0.1 ohm) from earth-clip to BNC's screen, as you would expect.
I also measure a dead short from the HT-clip to the BNC's signal pin. This is a bit more unexpected, but you haven't revealed the construction of the probe, so maybe it is OK?
There is no connection between HT-clip and screen (it is more than 200Mohms, the limit of my DVM). There is a capacitance of 43nF between HT-clip and screen. Just to emphasise: that's nF, nano-Farads.
As I mentioned, a square-wave calibration signal with a sub-microsecond rise time is degraded to about 70 microseconds by connecting the probe. Of course the source impedance is probably quite high and so it gets pulled down by the capacitance, where a lower impedance source would not be degraded so much.
This capacitance seems to me surprisingly high.
Perhaps you could compare this data with a known good probe.
Your resistance measurements seem ok to me so I don't think you have a problem with the pickup.
Do you know how much memory your scope has? From the traces you have posted, it does not look like its collecting enough samples to give you the clarity and detail that a PicoScope would give you.
The first capture looks good to me apart from the lack of samples, and obviously the signal has not been inverted. To give you an idea, when we capture a single ignition event over say 10ms we normally collect something like 30Ks (3MS/s).
Sorry, these are the best I could get. Also because I'm new to this I didn't realise that it's conventional to invert secondary traces. Lesson #1 learned.
Because of the "wandering" baseline voltage I had a lot of trouble with triggering, so that even with one of my good Nicolet (12 bit) scopes the displays weren't much better. In other words, each time I tried to trigger on just one ignition event I had a false trigger, giving rubbish.
On this Velleman scope the sample rates used for each recording are fixed by the unzoomed timebase mentioned in the text on each record. The zoomed timebase can be seen in the top right of each screenshot.
The rates are:
50ms 2,500 samps/sec
10ms 12,500 samps/sec.
The sample buffer is only 4K.
I hope this has given some extra detail concerning the problem and I am grateful for any more advice.
EDIT: I can do a little bit better with the old captures. (Attached -- but can most people see them???).
These (slightly) better resolution show the ignition event a bit better, but lose the real problem that I'm having.
I've put a few screenshots with some commentary onto the above website.
They are just bitmap (.BMP) files, so should display straight into your browser.
I'd appreciate any explanation about these traces, and especially as to why the secondary voltage wanders away from the zero line when the misfire happens.
EDIT: also some PDF files which show the traces, several to a sheet.