Oil Pressure/Water Temperature Gauge Voltage Chart

[ayates]

I have a Caterpillar/Perkins diesel engine with standard VDO oil pressure and water temperature gauges with restive sensors. I am wanting to do a little data logging. I can easily log the voltage seen at each of the gauges. However, for converting to the appropriate unit, I was hoping for a little more accuracy than just looking at the gauge and taking a guess of the reading to establish conversion factors.

Does anybody have any accurate figures for this? Or even detail of the internal schematic of the meters, and I could reverse engineer from there.


Thanks,

Allan.

[Brewgyver]

IIRC, Imperial sending units are reverse acting resistor sensors, with resistance dropping as the variable rises, and I believe both have an operating range of 240 to 33 ohms. Typical Imperial pressure sensor is 240 ohms at 0, 33 ohms at 80 psi. The change of resistance is pretty linear within that range, so the conversion in easily calculated. I would think if you're using a Raspberry Pi or similar unit, should be easy enough to connect the senders to an ADC chip.

[GordMay]

[ayates]

These are both great information. But using the latter diagram and the 0-80 PSI oil pressure as an example, I am looking to know the voltage at S on the gauge, when that terminal is connected to the sender, and the pressure is at 40 PSI. Any PSI/voltage pair would work.





Thanks,


Allan.

[Wotname]

IIRC, some VDO 'marine' oil pressure gauges require specific VDO senders i.e. they are neither a USA or European standard - although ATM, I can't lay my hands on the data - so I might be wrong...

[Brewgyver]

I'm just waking up, first cup of coffee half done, so I could be wrong, but pretty sure the voltage at the S connection will be your battery voltage.

It seems counter-intuitive, because we call the resistive sensor the "sender," but it's not really sending, it's just a variable resistor to ground. The S connection on the gauge is always the same voltage as the I connection, but the current flowing through the gauge circuit varies with the resistance of the "sender," and the bimetal attached to the needle bends with the current, moving the needle.

To put it simply, both the oil pressure and coolant temp gauges actually measure current flowing through their individual circuits. If you want to log voltage, you'd need current sensors. Given the low currents involved, you'd probably need to use flow through chips, not CTs. Maybe one of these: https://www.sparkfun.com/products/14544

[ayates]

The voltage at the S pin definitely changes as the sensor changes resistance. This is a bit of data logging for the water temperature gauge and the oil pressure gauge.


My understanding is the gauge has a constant voltage source, and one resistor in a voltage divider circuit. The sensor being the other half of the voltage divider. Thus S will have voltage that varies in proportion to the sensor resistance.





Allan.

[Maver]

Originally Posted by ayates

I have a Caterpillar/Perkins diesel engine with standard VDO oil pressure and water temperature gauges with restive sensors. I am wanting to do a little data logging. I can easily log the voltage seen at each of the gauges. However, for converting to the appropriate unit, I was hoping for a little more accuracy than just looking at the gauge and taking a guess of the reading to establish conversion factors.

Does anybody have any accurate figures for this? Or even detail of the internal schematic of the meters, and I could reverse engineer from there.


Thanks,

Allan.

One possiibility is to convert to digital signals with f.i. Actisense, and than logging the data. For VDO there excists a resistant table for pressure (10-180 Ohm) and for temperature (app. 345-20 Ohm)

regards,

Martin Versteeg

[ayates]

Just found these two threads... Haven't digested them yet... The latter has a nice graph...

https://www.cruisersforum.com/forums...sting-645.html

Oil temp & pressure gauge / sender operation - Pelican Parts Forums


Allan.

[hzcruiser]

I'd say you're better off "calibrating" your ADC using the gauges, if you trust them, rather than relying on the sensor to be exactly "in spec".

For temps you could also use an IR meter to double check, for the pressure it's not so easy to verify them, but then again, all you really want (I guess) is logging and alerting you if it's outside the "normal" boundaries, which you would have to establish empirically.

[ayates]

Found this useful video:

https://www.youtube.com/watch?v=QHHRiXa0IVI

I could replace the sensor with a potentiometer, then adjust the resistance so the needle in the gauge lines up exactly on the divisions, and measure the corresponding voltage. It does assume the gauge is accurate, but at least it will be no less accurate than the gauge.


Allan.

[ayates]

The Actisense unit (senders/gauges to NMEA) has all the voltage conversions built in for various gauges. I though their customisation utility might show the voltages, but it turns out the Volt column is actually displaying the sensor resistance.


Allan.

[IOM]

Originally Posted by ayates

I have a Caterpillar/Perkins diesel engine with standard VDO oil pressure and water temperature gauges with restive sensors. I am wanting to do a little data logging. I can easily log the voltage seen at each of the gauges. However, for converting to the appropriate unit, I was hoping for a little more accuracy than just looking at the gauge and taking a guess of the reading to establish conversion factors.

Does anybody have any accurate figures for this? Or even detail of the internal schematic of the meters, and I could reverse engineer from there.


Thanks,

Allan.

If you have a system to read voltage and convert it to a unit value, then its really not difficult to build your own tables. Your bigest challenge is will your data logger accept the voltage levels that the sensors are using. Some loggers only accept 0to5v and you may need 0to14v input.

Your interest to use tables over measured values is probably related to accuracy, however if the reference voltage is not supplied, measured and compensated for by the data logger, then the accuracy issue is a mute point. If you keep your current gauges in operation then the reference voltage is subject to substantial variance which you may not be able to compensate for.

The pressure sensors I have used from VDO were all linear in response, as are most pressure sensors, so you only need to measure the zero point and the voltage at any known pressure . So read the zero point in with the engine off, and then read the voltage at idle and pressure at idle and you will have a fully functional pressure conversion table.

The temperature curve may not be so easy, and may require you to plot a few more points to get usable data. The points you need to be concerned about are around boiling point and a few either side of that. With the temperature sensor, you also need to be sure your conversion table can understand a inverse or negative table...ie voltage goes down and temperature goes up. Some data logging systems cant comprehend this.

Hope you come right.

[ayates]

Thx for the details in your reply. As an electrical engineer, I get what you are saying.

Data logging the voltage is under control. I have 0-10V logging with a Datexel DAT-3017V for analogue to digital, and that is logged with a Sensaphone Sentinel.


Definitely the temperature table is more challenging than pressure. Would be much simpler if that was linear too


Thanks,

Allan.

[IOM]

Both curves you have generated look typical to what I have used in the past.


If its any consolation, when using some sensors (VDO were one) we trimmed the pre-configured tables in order to get more precise readings to a calibrated source. We later did away with the VDO sensors because they were not so marvelous.



Because you are using a foreign reference voltage, and you are adding additional loading to the system (however small it may be, it is still something) you will be better off creating your own table than using some pre-configured table that you assume to be correct (and therefore don't check), but probably is not .


Be interested to see the final application in the boat