Whilst I've got absolutely nowhere so far with the tacho, I have on my travels so far managed to fix the oil pressure light which also wasn't working. There is a 6 wire connector behind the headlight that feeds the tacho and oil pressure guage. The pin for connecting the yellow wire had been pushed in and was no longer making contact. Id deffo check the pins on that connector.
Evening all A bit ahead of schedule. The Scope turned up early and I had a spare hour so i went to have a play. The Outcomes are as follows: - Oscilloscope connected to Grey/Green wire at the 3 pin tacho connector, and earth to battery - Square Wave form detected (this is what it should be), with an amplitude of 1.2v. At idle the signal was running at 36Hz (which from what ive read this is correct) - Oscilloscope connected to Pin 24 at ECU and earth to battery - Identical. Square wave form with an amplitude of 1.2v at 36Hz All measurements taken with the bike at Idle. Ive taken a pic of the Scope screen incase anyone with more knowledge see's that ive made an error somewhere. Given the read outs are identical at both the tacho connector and ECU, i'm now confident that the issue isn't with the grey/green wire that runs front to back on the bike - otherwise my read outs would be different. It would however appear that the tacho signal being given out by the ECU is correct but 'weak' - I believe (and please if you know better let me know) that the signal that runs down the grey/green wire should be 5 Volts, not 1.2 Volts. I need to find this out for sure as at the minute i'm comparing my figures to rough guesswork based on various forum posts. My new oscilloscope also has an inbuilt signal generator, however it only has the ability to push 3v. I tried to test the tacho(s) with the signal generator, and whilst neither of them worked fully, both were twitching which makes me think that with the correct signal, they will work. I am going to look into bench testing them using a more powerful signal generator This leaves me with - why would the ECU being giving out a weak signal, and what can i do about it? The bike runs an absolute dream so i can't see it being a terminal fault anywhere. Any thoughts or is it time to start looking for an ECU? As always, many thanks for all your input to date. I really do hope that this can help someone somewhere along the line.
Thanks That adds some more info into the mix. My reading of that is: - 1.2v at Idle may actually be normal (or could even be an over voltage) - I should see a 10-12v square pattern at high RPM - I need to check the ECU earth on Pins 18 and 19 Anything im missing?
My learning curve continues. I had the scope set wrong - so in actual fact the voltage that is being received at the tacho connector was 12.7v not 1.2v This also the voltage that the ECU is giving out at pin 24 This is pretty much a constant. No matter how may revs I give the bike, i'm getting 12. something volts, in a square wave pattern, at the appropriate variable frequency. My understanding at present is that this voltage is too high and outside of what the tacho will understand. It may also have damaged the tacho? Next i'm going to bench test the tacho(s) and confirm if they are working correctly, but I think i'm heading towards a conclusion that this is an internal Fault with the ECU that's resulted in the signal being a pulsed 12 volts rather than a pulse that increased between 1 and 5v with higher RPM.
I think you will find that the tacho works on frequency changes ( being switched on and off therefore like a Schmitt trigger square wave) rather than variating voltage. It could even be neutrally switched at the ECU? I could not say for sure! By the way pin1 and pin19 are neutral switching for the coils on a 996 Bip. 996 SPS they go to earth.
If it is an output problem then a possible solution would be to use a voltage divider. Basically 2 resistors. One on the signal wire and the other on the ground wire. If you Google ‘tachometer voltage divider 12v to 5v’ you should find the resistors values you require.
memory rusty, but oscilloscope reading in pic seems about right frequency for 1200 rpm if same configutation as other ducati tachos 1.e. - 33.33Hz per 1000 rpm.
Thanks all I'm confident that the frequency and wave output is all good. It's the voltage that seems high given the information I have available. I will potentially look into something to modify the voltage as it seems doable, but I think I need to know beyond all doubt that the tacho(s) are working. I'm a painter decorator, not an electronics engineer (or even hobbiest), so all the information I'm feeding back could of course be utter nonsense! I'm doing my best but please do feel free to forward away with anything and everything. Anyways, will be next week now. Busy weekend of family chores ahead
My polymath-engineer brain can't see why the tacho would expect to see anything other than a constant-voltage, varying-frequency (squarewave) signal Surely it's just a simple frequency-measurement device (albeit with a bit of damping).
https://chatgpt.com/share/69178beb-c58c-8013-a78a-eb1a73dc5e13 I was approaching this thinking the tacho signal behaved like the signal for the TPS. The post above was very helpful.
This may help in testing the tachos https://www.aliexpress.com/item/100...MIobG6jcDykAMVt5pQBh0VgzvKEAAYASAAEgJHOfD_BwE AI is useful for collating information. It helps to have some knowledge as it will respond and learn to corrections. A lot of the information is actually sourced from this forum. It is a tool to be used, not always believed.
Here is my 2 pence worth. The voltage on the signal wire, when measured with an oscilloscopes, does not fluctuate. Weather it's meant to be 5V or 12V in this particular case, I don't know. You do get a rising voltage reading when you measure the signal with a standard multimeter. All multimeter have some lag so it stands to reason that a low pulse frequency with a short pulse duration does not give enough time for the meter to measure correctly. With increased pulse frequency the meter reads a higher voltage but all these readings are false up to the point where the frequency is high enough that the meter reads it as a constant voltage. Measuring this kind of signal with a multimeter is at best a crude indicator whether a signal is there or not. I don't think measuring on an AC setting is going to work either as it's only a 0V, 12V square wave instead of alternating current. The OPs measurement of the signal with his scope looks to be correct to me and that would suggest the fault lies with the tacho and not the bike. Chat GPT cannot find any reference to the Marelli 16M ECU having a 5V logic output but finds lots of evidence that it's indeed a 12V signal. I can think of a few ways forward. Give the Tacho to someone with a working bike and plug it in. Disassemble the large connector on the ECU and tap directly into the signal wire with a test lead and power the Tacho directly from the battery or an external power supply bypassing all of the bike wiring. Find a signal generator that provides a 12V square wave and test the Tacho with that. I have a coil tester that may well be suitable. If you're not getting anywhere I'll have a play with it and see what kind of signal I can get out of that.
Therefore AI collects a lot of old b****cks and every body believes it ? Plenty of AI crap on youtube by the way. Yet people post comments as if it's gospel truth.
I couldn't disagree more. I find AI a highly useful tool. Like all tools it needs to be used in the right way. There are methods to ask in such a way that filters out the bull. It does a lot of the filter work for you which you would have to do yourself if you use Google or similar. It just saves time and gives better answers.
AI is a tool. It does save time by pulling everything together but you need to extract what is applicable. One area I use it is creating complex Excel formulas. I used it recently when I was working on the algorithm used to calculate the checksum used on dash mileages. Saved me hours of Excel work. It also got me into recording Macros in Excel.
