Hi, I'm building a test bench for testing some automotive parts but need to simulate the output signal of RPM/Crankshaft Sensor for ECU to avoid including the real flywheel in the test bench.
the own sensor has 3 wire and I think it's inductive type because : I seen 800 ohm Resistance between two pin and when I connect oscilloscope probe to this two pin, I see sine wave on the oscilloscope. although in the car, when I unplug one of this two pins the engine will turn off and also if I unplug this pins, the car will not turn on. the frequency of output signal of this sensor changes according to the throttle change, for example when I push throttle pedal to 1000, the frequency reach to about 1KHz.
Can I simulate this signal for the car ECU with a function generator or create special circuit without using the sensor?

 

Welcome to AAC!

Can I simulate this signal for the car ECU with a function generator or create special circuit without using the sensor?

Yes. What you ask is doable.
Can you tell us the signal amplitude at, say, 1000 rpm? A screen-shot of the waveform would be helpful.
What do the 3 wires connect to?

 

One of the problems to overcome is they (manufacturers) tend to use two types of sensors, the Inductive sensor which outputs a sine wave and does not require external power and a Hall Effect Sensor which may be powered using 5 or 12 VDC and outputs a square or rectangular waveform. Therefore before we can simulate a sensor output waveform we need to know which type sensor is used since the ECU will be conditioning a sine or square waveform. The engine RPM sensor usually counts teeth on a flywheel but "usually" is not always the case. So any number of pulses per a unit of time can be worked out to provide engine RPM. So for example, using your example, 1,000 RPM was reading 1 KHz or 1,000 cycles per second. This begins with the number of pulses per revolution depending on the number of teeth on a flywheel or whatever the source of the signal.

Overall, as Alec_t points out it is very doable and can be accomplished in most cases with a simple function generator capable of sine and square wave or pulse outputs.

Ron

 

Used an old VCR Capstan motor & fited the reluctor parts to it & vary the motor speed works for me. In the Toyota this fed the engine ECU. But I use it for other Ign system tests.

 

This is the wave form from the Test unit & one from a Mazda Crank shaft sensor

 

thank you all friends. I'm glad to be among you. here I attached some photos of the sensor and screenshots from oscilloscope.
in the following picture the top black connector is the sensor side and the bottom black connector is the ECU side. I guess that one pin is signal out, the second pin is Ground and the 3rd wire is Shield for noise reduction. although voltage on first two pin is about 1.8V DC and about 3.8V AC (with voltmeter).



in the following picture I connected this two connector together with external wires for connecting the oscilloscope probe.



signal measured with oscilloscope with RPM shown on Instrument Cluster about 1000 RPM.

signal measured with oscilloscope with RPM shown on Instrument Cluster about 2000 RPM.



Thanks.

 

Surprisingly clean sine wave signals! Quite different in shape and frequency from the Mazda and Toyota signals Debe shows.
Judging by the frequencies you show for the given rpms, it looks like your crankshaft has a reluctor wheel with about 54 'teeth'. Is there a 'missing tooth' that the ECU needs to sense for a 'zero angle' reference (similar to this wheel)? If so, that makes the sim a tad harder.
Do you want to simulate just the signals you show, or also signals like Debe's?
What is the maximum amplitude sine/triangle/square wave your function generator can give and what load impedance can it drive?

 

thank you for the good news Alec_T. yes it's different with signals shown by Debe. with my studies I guess that the reluctor wheel in this type of car has 58 teeth and two missing teeth (60 teeth at all). this while show the TDC to ECU 12 degree before happening.
Yes I want to simulate the signals I shown in previous picture to ECU. I think that the signals shown by debe is for a different sensor. here I attached the picture of my function generator for your reference.
Thanks.

 

a picture of the reluctor wheel

 

Interesting profile on the flywheel for the sensor. What is the vehicle its on?

 

It's on Peugeot 206

 

This is just for interest its part of a Ford EFI course I did some years ago. It was on EF model Ford falcons here in Australia. The crank shaft sensor on these used a toothed wheel on the crank pulley & has a similar wave form as yours.

 

This missing tooth has been common practice for many years. The camshaft will also have identifiers and in that way, you can watch both signals at the same time and watch for imperfect variations. By comparing cam and crank signals, you can do relative compression, watch for engine misfire, stretched timing belts etc. There is a very good book on this by Graham Stoakes that covers a lot of automotive waveforms. Debes waveform is old school where the distributor reluctor was used for ignition primary. We got away from that setup long ago because using multiple tooth crank sensors offered up degrees of rotation as well as cylinder order. The only way you ate going to simulate the missing teeth is to built a smaller setup with a motor and reluctor wheel which simulates your car. Knkw any good machine shops. Other than that, not sure how you would take out two pulses.

 

The suggested smaller setup would give the most accurate sim, but you might get away with just blanking 2 out of 60 sine cycles from the function generator. The TTL/CMOS output of the generator could be used to clock a base-60 counter and the counter could control the blanking of the 50Ω sine output of the generator. Alternatively, a MCU could be programmed to synthesize the complete waveform required.

 

thanks to debe for the book and making me sure of the sensor's working method. I had three idea to do this too, the first was the bwilliams60's idea to buy a electric motor and couple it with reluctor wheel, second one was Alec_t's Idea to building a circuit for signal simulation and the third was using function generator. using function generator finally ends to building custom circuit too.
with your comments I go to test electronic circuit and switch to electric motor coupling if the circuit does not work.
for faster response I will use an embedded ready to use board with DAC and PWM outputs.
Thanks.