Oscilloscope for Engine/DME Troubleshooting
 

Hi Everyone. I've not been very good at posting things, just receiving help from all you experts. I'll try to do better in the future.

Anyway, had some issues with my recent rebuild of my 3.2 and wanted to see those sensor pulses and coil output from the DME. I found this handy dandy scope online, called a DSO 138 and for as little as $20. I paid $30 and it was already all soldered up on the board and all I had to do was build the case (supplied). Be careful if you buy the kit as it's tricky to solder those surface mount components; I'd pay the extra and get one already built.

But just amazing that a color, digital, storage scope can be had for this much money. Small screen I know (2.4") but still. I just saw one on a popular bidding site for less than $4.

Sorry about the picture quality

Mikehttp://forums.pelicanparts.com/uploa...1512169291.JPG

That is very cool!

That's what I thought. Now I'm looking for a $10 spectrum analyzer :)

Nice !

Tutorial........
 

Mike,

Would you be kind to help us electronic challenged individuals to get started using an oscilloscope? I have 2 oscilloscopes (‘70 vintage solid state professional grade and a brand new portable electronic oscilloscope) and I have no idea how to use them. I just need to learn how to measure the square waves and will be satisfied. Any help or recommendation is highly appreciated. Thanks.

Tony

Awesome info.

OK, I can see this getting into a diagnostic thread. Which is great IMO.

So the question is, how are you measuring the ignition trigger signal. I thought the ground was interrupted in the ECU to create the pulse. What connections to the ECU/coil wires did you make?

Tony, the Velleman handheld autoscaling oscope will dial in to what you need. It's pretty intuitive and only like $120

Tony, et al, thanks for the reply. Scope use for troubleshooting our cars is extremely useful IMO. For example, looking at the pulses from the reference and speed sensors. You can get an idea from a DVM that's there's something coming in to the DME but the DVM is not fast enough and also tends to RMS (root mean square) values, not peak envelope like a scope. So, scopes are basically Y axis voltage and X axis time, which is settable in milliVolts or Volts per division (Y) or micro, milli or Seconds per horizontal division. So for the reference sensor, which is supposed to give off about 2 volts, once cycle per revolution, you could set the Y (vert) to 0.5 V/div and the horizontal timebase to about 100 mS/div (1 Sec over the screen width, for 10 divisions) you should see about 5 pulses and they should be at least 2 Volts for the ref sensor (Bentley). Not sure if that's 2V peak-to-peak or RMS but should be a healthy signal. If you look at the Speed sensor, you should see a lot more pulses (6 X) at a slightly higher voltage (2.5 V).

The scope has a high-impedance input, meaning it shouldn't add very much load to the circuits on the DME. The better the quality scope and probes, the higher the input impedance. That DSO 138 cheapy scope has 1 Meg ohm (10^6) so it's not the best but seems to work for me on my DME. The scopes you mention are probably much better than the DSO 138.

As for where to put the leads, for the ref and speed sensor, attach the scope ground lead to the chassis and with the DME disconnected, at the plug, you can probe either of the two outputs wires from the sensors but the scope should be set to AC coupling. Do not put the ground of the scope on one wire and the probe on the other. The output from the sensors is a balanced 2 wire with a shield, so neither of the 2 wires is grounded. Each wire should read about the same except they will be out of phase with eachother.

Tippy, you had asked about the coil output. As far as I understand from reading this forum, the output is high (12V through the primary of the coil) which is dragged to ground for firing the coil. It's that rapid change from 12V to ground that is really an AC pulse that is transferred to the secondary and multiplied many times because of the prim to sec turns ratio. I just scoped the pin 1 output on the DME and also on the primary of the coil, looking for those negative going pulses. need to set the vertical sensitivity to 2 V/div but seems to work quite well. That Velleman scope you mention for $120, seems like a very good deal also. Would be much more flexible than this DSO jobby.

Triggering a scope to give you a nice waveform from our cars is difficult but with a storage scope like this DSO you can capture then look at a later.

I'm pretty new to this, so would appreciate any inputs from anyone with much more experience in this area than me.

Mike

Good basic scope tutorial

https://www.youtube.com/watch?v=u4zyptPLlJI

Just wanted to correct something I said above, that the speed sensor is 6X the frequency of the ref sensor. That was a wrong statement. From the ref sensor I figured out the starter motor rpm is about 180 rpm or 3 per second. The output from the speed sensor is shown below, at 1 cycle every 4mS or 250 cycles/sec. This is pretty crude math but that would say the speed sensor is giving a pulse every 84th of a revolution and I looked closer at Bentley and seems it's providing a pulse every tooth.

Mike

http://forums.pelicanparts.com/uploa...1512328513.JPG

Great info. What does the coil trigger signal look like?

Apologies for the delay in replying. I've had some intermittent issues with my DME. Here is the trigger signal to the coil, measured right on Pin 1. Looks like almost 80mS between ground pulses.

Mike

http://forums.pelicanparts.com/uploa...1512516357.JPG
http://forums.pelicanparts.com/uploa...1512516357.JPG

Make that every 50mS, which is about right considering 6 firing per revolution at 180 rpm

Believe it or not Porsche is still pro-scope after all these years. I went to Classic Fuel and Ignition training earlier this year and there was much to learn. They (Porsche) want you to be able to look at all six cylinder traces at once, “the parade”, to check for misfiring, low compression etc. It’s quicker than almost anything when hard data is needed.

Interesting. So no scanning for trigger signals?

No scanning available on pre-OBD cars but you can look at crank and reference signals. Anything with a duty cycle, etc.

Well, my brain is still stuck on ignition points and carburetors. :D

That said I had to recently trouble shoot the wife's car's ignition. No spark on two cylinders. It is a four cylinder car with wasted spark ignition controlled by a ECM. I determined a bad coil trigger from the ECM. Car to remain anonymous.

So my question is couldn't one read the trigger signal from a Carrera or newer air cooled car? Isn't that what Mike is doing in post #12?

Turns out I could use a trouble light to figure it out.

Yep, trouble light is good for on off type signals.

Much more info available watching ignition wave forms.