hydractive ecu fix query

[PIT]

This is the file you will need if you want to implement the hydractive ECU features.
https://my.pcloud.com/publink/show?code ... ykMYo2gdFy" onclick="window.open(this.href);return false;

I used it to implement the steering wheel sensor algorithm in an Arduino, used for adaptive lights with the front fog lamps.

[CitroJim]

Dei, the electrovalve is driven with a 12v pk/pk 50:50 square-wave at around 1KHz after a 500mS initial application of full 12v to initially operate it to the soft mode. The 50:50 squarewave is a holding condition and it's what causes the hum you hear..

Coil resistance is quite low and current consumption at full 12v is about 3.5A as far as I recall... Back-EMF protection is essential.

Again, the VN05 is not a straight transistor. It's an amplifier/driver IC. You could certainly fabricate a substitute from discreet components if push came to shove...

Whatever you do as a substitute electrovalve driver it does need protection from shorts...

Only a pukka Lexia or ELIT (clone or otherwise) will read the Hydractive ECU properly although some more expensive third-party tools like the Sykes will to an extent.

[dei9]

PIT, thanks a lot for that, I had one years ago, and it got lost, and then I forgot it until now! It has additional info that i'd need.

Jim, I'm just going to use a VN05N (since I found a spare here), so I can experiment with the Arduino with it. So I'll input the wave into pin 3 and output it to pin 5, along with a diode to stop the back emf, pin 3 takes a 12v input, so i'll nick that off the existing plug pins, as well as the inputs for the arduino, and then i'll program it! but first task is it to see how if the electrovalve fire up, then take it from there. Thanks for the info too. Isn't the wave a 1kHz 25% duty cycle not 50%? It's the initial application I wanted and the currrent draw of the solenoid at first stage, so thanks for that.

Now, if I can get my Linux to talk to the Arduino! (it works on windows)

[CitroJim]

Yes, the waveform is about that Dei, here's a picture of it but 50:50 will be OK for the purposes of keeping it operated if you're generating it yourself with the Arduino.

[dei9]

sure, i get u now, we just need to keep it open right (as it defaults to close (hard mode))... so does a longer duty cycle keep it open 'stronger'?

[CitroJim]

sure, i get u now, we just need to keep it open right (as it defaults to close (hard mode))... so does a longer duty cycle keep it open 'stronger'?

No, the duty cycle is set such that the solenoid is energised just enough to hold the shuttle in soft mode.. The coil would be overloaded with full 12v applied all the time but remarkably they do survive it quite well..

A while back there was a bodge carried out on many XMs where the electrovalves were wired direct to an ignition-switched 12v feed to hold the car in permanent soft mode, By and large all electrovalves so modded I have seen survived unscathed so the duty cycle by extension is not too critical!

The valve is operated by a combination of hydraulic pressure and electromagnetism. The initial switch of modes is achieved by the 500mS initial pulse.. Provided it operates OK on that it will not take too much to subsequently hold it operated.

[Chris570]

Forgive me if you've thought of this but with your idea for a replacement ECU will it be able to sense changes in body roll etc? I ask because a Hydractive car stuck in soft mode can be quite dangerous......

[DickieG]

Forgive me if you've thought of this but with your idea for a replacement ECU will it be able to sense changes in body roll etc? I ask because a Hydractive car stuck in soft mode can be quite dangerous......

Not if you fit flat Hydractive sphere's

[Chris570]

Forgive me if you've thought of this but with your idea for a replacement ECU will it be able to sense changes in body roll etc? I ask because a Hydractive car stuck in soft mode can be quite dangerous......

Not if you fit flat Hydractive sphere's

Now you see I didn't think of that....

[dei9]

Forgive me if you've thought of this but with your idea for a replacement ECU will it be able to sense changes in body roll etc? I ask because a Hydractive car stuck in soft mode can be quite dangerous......

Good question, yes I have, this whole plan evolved from me wanting to test the electrovalves using an Ardunio, I have the Arduino Due (it's has a 32bit ARM Cortex M3 with an array of different types of input and output), some pins can do PWM natively, and the frequency can be set, two pins are 1kHz by default (25% duty would be 256/4), so it's as simple as a:

analogWrite(pwmPin,255); // full 12v
delay(500); //for 0.5 seconds
analogWrite(pwmPin,64); // 25% duty

and the output of these get fed into the VN05N.

I will feed the inputs into it too once I've confirmed that the electrovavles work nicely and it goes nice and soft as it should. There are 8 inputs, and they will all be monitored, all I need to do is scale the sensor outputs of 5V to the Arduino's 3.3V, do the maths and call the function that switches the electrovavles in and out. Upon looking at the Hydractive 2 Technical manual it doesn't seem to difficult. Quite a bit more maths than one anticipates, but very doable.

At the moment I think it works, I connected my prototype onto the car and the electrovalve fired up briefly, why briefly? Well it's because the wires probably can't give the solenoid the proper current! Annoying. So what I will do is get an Arduino Protoshield and wire it all up on that with proper wires that can output the (possible) 3A surge when the solenoids 'kick' in.

I might start it's own thread when I'm more onto it and it looks tidier.

One thing it won't have is the diagnostic interface onto the LEXIA, but new features I will give it is monitor the sensors graphically using an LCD display! and put a multicoloured LED where the switch is and the switch will be multifunction depending on the number of presses, etc. So the plan is for the LED to light green when the system is on, and flash orange when soft mode is active. So if you press and hold it, you can then choose the mode, always on, always off, normal or sport, etc! (maybe not always on). Similar to how the Traffic Master button works, multfunction.
One thing I'd like it to do is to raise and lower electronically, but I can't see how I would over ride the manual control yet. It may have to be removed and some kind of other solenoid put onto the height control mechanism. And there would have to be safety failsafes built in. But that's WAY later after a lot of thinking. I'm actually thinking of doing this as my third year project at Uni.

One concern with the height control is, some cars will lower if they go above a certain speed (I think the C5 does this Hydractive 3?), but it seems the height is critical for maximum comfort so I'm not sure how that is implemented on the C5 and whether it would work on the Xantia?

I've always wanted to do this kind of thing with an automatic gearbox, so if anyone has an auto xantia car to swap with, holler me

[PIT]

All sensors are quite good described how they work and what signals they produce in the document that I posted above. It is possible to implement them as I did with the steering wheel sensor, which has an algorithm for determining the straight position.

The adjustment of the height can be done with something like stud bolt which will be moved by rotating a nut with fixed position. You will need to identify the original 4 positions and add some switches to stop there, or markers to identify where you are and to stop moving the stud. Then you can implement any additional stop markers.
If you do this, you will have to remove the original lever, so no over ride is needed. This will be the only mechanism to change the height. Just with buttons

[dei9]

Thinking to keep the original lever inside and use it as the 'button'. and have the position of the lever hit some swithches. But the problem is, I would need some way to know how high the car is for the logic to work. And also to program how much delay there is before the height moves up after pushing or pulling on the height corrector, this delay could be measured and memorised for knowing what condition the hydraulics are as they go through wear and tear.
I like your idea of the nut and bolt, it seems a good way to get some fine tune movements.
The algorithm for the steering wheel position is quite complex, but not too bad, it would take a few brain hours to translate it into ardunio compatible code tbh.
All in all, there is a lot to think about, tomorrow I have someone coming over with a Lexia to check my hydractive ECU, then I can confirm if my electrovalves work or not, (they are a bit lifeless with my arduino setup, so I've got my mate coming over with an oscilloscope too so we can check the outputs on from the ECU and compare it with what I have on my Arduino, once that is done, programming the sensors shouldn't be too bad. The wiring is the easy part.
Thanks PIT

[PIT]

I think that you don't need to know the height of the car.
You just need to make the mechanism enough precise.
Then you need to calibrate the four factory levels, then you just need to test and calibrate the fifth for example, which will be 10 mm lower than the normal height.
All these 5 levels are just markers where you will stop moving your mechanism - the bolt for example.

In fact, the factory mechanical system works the same way. The four positions are just some notches where the lever falls down and prevents the mechanism from moving front and back, which movement commands the height correctors.
Here is a genuine document that shows how to lower the car a little bit on the third position, so the suspension is not so hard.
https://my.pcloud.com/publink/show?code ... PeoXeWnoxX" onclick="window.open(this.href);return false;
They just show how to modify the notch and the spring, so that the lever will stay in the new part of the notch.

So all you need to do is to calibrate your "notches"

[dei9]

LOL, thats interesting, I need to get my antiroll bar clamps set up exactly first, plus one of my rod clamps aren't holding on too well, and to make sure she is 100% working hydraulically.
But you're on the ball there.

[dnsey]

C5 height sensors would provide an easy feedback system if required.