A few questions regarding I/O and sensors

[stevieturbo]

You CANNOT wire 2 wire sensors to two different devices and expect them to work as normal.

Sorry,I believe I didn't explain it very well.

If the temperature sensor is wired ONLY to the gauge, hot fluid - 55ºC - is read as 3,75v (according to a working voltmeter hooked to the sensor wires).

If the sensor is removed from the gauge, then wired ONLY to the ECU, heat se same fluid to the same 55ºC as before and put the sensor in it, the reading changes to 2v (according to eot1V), even the temperature being the same as it was when tested with the gauge.

The 2 wire sensor was not wired to two or more devices at the same time (as you instructed before - to choose only one of them gauge OR ECU ), and I get the different readings above.

Thanks,

2 wire sensors are purely resistance based. This is the only thing that matters when taking measurements.

How any voltage is then created will depend on the device it is attached to. This may vary with each device depending what voltage it applies across the sensor to take its readings.

You cannot compare the gauge to the ecu, or indeed a voltmeter. They all measure the resistance of the sensor, but may do it in slightly different ways

If you know for sure the temperature is 55degC, then you can view the voltage on the Syvecs/EOT1V and this will create one point on your linearisation scaling.

Do the same for as many different temperatures as possible and build your scaling for that sensor.

[pavlo]

Start here:

viewtopic.php?f=6&t=401

[TimH]

The two wire sensors are (usually) thermistors - devices whose resistance varies with temperature. In order to convert this to a voltage, a pull-up resistor is needed which, in the case of the Syvecs, is 3k. If you are reading 2V at 55C on the Syvecs, then the sensor is 2k at 55C. We can deduce from this that the gauge must have a 1.5k pullup (assuming it's to 5V), not that we really care about that!

You *could* run the gauge and sensor together if you turn the pull-up off on the Syvecs input, or use an input with no pullup - you may get a slight inaccuracy on the gauge due to the input resistance of the Syvecs. I've not tried it, but I would hope the error would be small enough to ignore.

Worth a try, anyway as you have nothing to lose and if it works it will help with the linearisation as you can read temperatures off the gauge and enter them in live via SCal as the oil warms up.

[macromraco]

These explanations really helped a lot, I appreciate the help and patience guys!

TimH I already gave up using the gauge for the oil temp as the sensor is working now wired to the ECU. I was going to make the linearisation using an external high temp thermometer I have here, but just out of curiosity I'll try to do what you said and check if I can successfully run both, temp gauge and Syvecs - I'll let you know how it goes .

So oil temperature and fuel pressure are both closed cases, thanks again for the help. The only one left is the one I mentioned a little before about the 3-wire 5v oil pressure sensor that I tapped into the signal wire, kept it connected to both gauge and ECU and got a really weird behaviour:

For the oil pressure, I only tapped into the oil pressure sensor signal wire and kept the gauge. First time I turned the ignition on the gauge reading got skewed - reading 2bar of pressure with the engine turned off. Then I assigned it to the AN07/AU3 input and the sensor voltage reading immediately dropped to -0.9v as well as the gauge pressure reading. I read in the link below that some offset or noise is expected but that it will only affect the gauge:

http://www.syvecs.co.uk/forum/viewtopic.php?f=6&t=340

Now I'm not sure if I did something wrong or if I can go ahead and start the engine that the sensor is going to get back to its default 0-5v reading. Any ideas?

With this one that should be straight forward from the previous explanations I have no clue what could be wrong. EOP1 was with default configuration, 5V input etc when I tried to configure it. Should the readings - at least the ECU voltage readings - get back to normal 0-5V when I start the engine and the sensor start giving signal readings back through the signal wire?

[pavlo]

You tapped into the signal wire...

You will need to connect the sensor ground wire too, that is likely your problem. If you connect to a resistive sensor you will need to do the same although there is a small offset created by the input impedance of the Syvecs, it's about 2ºC or thereabouts (varies with measure temp)

The GP wiring manual covers the use of the various inputs with and with pullup resistors to a basic level.

[C. Ludwig]

That's because the gauge and the ECU have different pull-up resistors. The output from the sensor is pulled up to 5v through a resistor. Different resistor equals different voltage. If you can turn off the pull-up in the ECU, you should be able to wire the sensor to both the gauge and ECU.

[pavlo]

The pressure sensor doesn't use a pullup.

[macromraco]

Ok, I got it regarding the resistive sensor vs with/without pullup inputs and the installation options. Thanks.

You will need to connect the sensor ground wire too, that is likely your problem.

I see your point here, but grounding both gauge and Syvecs sensor grounds together wouldn't create a ground loop, and may affect all Syvecs sensors readings?

Instead of tapping the ECU input into the gauge/sensor wire, would it be better to power the sensor using Syvecs +5v and GND, and tap the gauge into the sensor wire? Does it make more sense or may I end up with the same grounding issue?

Btw, thanks - I took a look at the wiring guide, very useful info but doesn't cover tapping into existing sensors, only dedicated sensors configuration, right?

I really appreciate the help.