Problem with reverse voltage on old truck dashboard

Projects General Guidance
1

Hi,
I've a numeric dashboard project for my old truck that has a faulty dashboard at the moment (fuel level and water temp has wrong value, the sensors are good).
In this project, I've 4 sensors that are basically a resistor that changes with the value of the content (NTC400 water temp sensor, fuel capacitive sensor and 2 pressure sensors for air tank). But today I've a problem with some reverse voltage in measuring sensor ohms value. This project use a voltage divider to get the sensor resistance value.

For more context, I've tested air tank pressure sensor to get the value in ohms @ X bars. The sensor is orginally connected like that on the old dashboard (see attachement for drawing):
+24V (702A "11 pin)=> Gauge => dashboard sensor output (702A "14) => sensor => ground (I assume, it's an ohms mesurement with a voltage divider as the old dashboard gauge is only made from resistors and other small stuf, not ADC or anything else from what I can see inside).

All tests was done with a calibrated multimeter and not with the Arduino (the value in Arduino was not good until I set an offset, but like explained below, the reverse voltage value is not always the same).

I test without the dashboard, I test only the sensor that haven't any power supply like shown on manufacturer drawing (only two wire on this sensor).

With the master relay OFF I've correct values from 10 to 140 ohms (0 to 7.5 bars) and when I engage the master relay, the value became totally broken (50 ohms to 80 ohms @ 7.5bar)... When I check the voltage between ground and the dashboard input of the sensor (702A "14) I get 65 mV (Ground => Voltmeter => sensor input => ground) like shown in attachement.

This values doesn't change when engine is running but change as I engage more or less equipments from 50 to 95 mV on this input (with NTC400 water temp sensor the value isn't the same, about 20 mV when air pressure one is at 65 mV).

So, how can I "suppress" this reverse voltage (I don't know if it's the good name for this) to get the real good sensor value (ohms) like when master relay is not energized?
Or another way to get my sensor value maybe ?
I can set an offset in the arduino (it work as the equipement activated not change).
However, the old Analogic dashboard work perfectly for the air pressure values (the water and fuel gauges are broken and gave wrong value..., that's why I've this numeric dashboard project) but I can't use it anymore and a new one is few thousand euros...

Also it's difficult to explained very well the problem, let me know if it's not clear or need more informations.

Capture d’écran 2022-09-06 130332
Capture d’écran 2022-09-06 130332977×475 25.7 KB

2

You can’t really use the sensors if they are also connected to the existing dashboard .
What’s wrong with existing dashboard ? There may be a voltage regulator that feeds the gauge circuit that is faulty .
Anyhow …..
You really need to power a voltage divider with a stable voltage ( or current source ) and use the internal 1.1v reference to allow for reasonable voltage divider resistances . ( the sensors used are low impedance )

3

Like I've explained, on drawing and text, actual dashboard is disconnected when i test the sensor. I just power up other equipements :wink:

I don't know what's wrong with the actual dashboard. The gauges work sometimes and for some unknow reason the value become very unaccurate (not relative to humidity, temperature, or vibrations, of course i cant check inside the gauge itself unfortunatly). And back to normal a minute or more after. The old dashboard take rain for some years and it has a lot of corrosions inside. i've cleaned up evrything many times but it doesn't solve the problem... And of course I can't check inside the small "motor" of the arrows of the gauges.

My main problem is this millivolts when i try to check the pressure (dashboard disconnected) when the master relay is ON.
For the arduino test i've made later, I used the +5V output of the arduino. It's not stable enought you think ? The ref resistor is 1k.

I can use the 1.1 ref when I power my voltage divider with 5V ?

Thank you for your thoughts on the reference voltage !!

4

Your mV is likely due to poor earthing at some point and could solved by using local separate earth connections to the sensors . You may be able to just ignore it though .
You need to be getting a higher voltage signal from your transducers - the low impedance of them is an issue , a smaller divider resistor helps. Do the maths and supplying from 5v aim to get 0-1v output into the Arduino . If the value gets below say 270ohm the 5v line will not be able to supply the current .
You could just fit an oil pressure switch and say use a ds18b20 as a separate water temp sensor.

The old fuel sensor is likely to be a variable resistance with a float , probably broken - these can be fixed .

Personally I would restore the existing dash , less hassle , and nicer result . Test the gauges with a low voltage supply ( careful) and see if they work ok . Gauges may all be supplied by some form of regulator , which u can replace Then you just have wiring and sensors to check .

1 Like
5

Yes it's my tought too but i cant find it... On this olb truck (in fact a Bus...) all ground connction came to bar bus in the electric panel. I've cleaned and check it but it doesn't help.

I'll try what you explained for the voltage divider if it work.

The old fuel sensor is a capacitive one and not and a electromecanical with a float, the sensor is new and the value of the sensor is good at the dashboard plug. it's really the gauge where the problem is.

I've tried to fix it many times but I never success... The gauge was supplied "directly" by the +24V thrue a 220ohms resistor. No regulator or something else inside. The gauge itself (for the fuel one) is only made of some resistors (I've changed all of them) and a motor for the arrow, nothing more (i think the motor is damaged or something else). An new gauge is about 200/250€ each... and the complete dasboard about 2000€...

6

You aren't measuring the resistance on your DMM while current from the dashboard is flowing through it, that is with the relay ON?

Tom... :smiley: :+1: :coffee: :australia:

7

Yes of course, that's why, like I sayed in explanations and on the drawing and also in a response to Hammy, the dashboard is disconnected when I test, I only have buttons to control equipements like master relay.

The sensor is completly disconnected fron any power source, just connected to ground.

I've changed the diagram in main post :wink:

8

The only way the voltage drop can be created is By current flowing through the ground wiring between sensor ground and where you measure the voltage ground and or /battery voltage change .

As said you can run separate grounds for sensor circuits and make sure your sensor output is in the volts range ( battery volts will change by more than 65 mV anyway ) so small volt drops are irrelevant . You can’t rely on chassis ground or long cables with small voltages on a truck

9

Yes, thank you for your help ! I've tried with a smaller ref resistor and i get a really accurate value without setting an offset.

For the separate sensor ground cable, I didn't say earlier but all ground come to one point, a bus bar, this busbar is directly connected to the battery. I'll made more tests but with your help I think I've a good solution even if i've some milliVolts :wink:

10

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.