We are testing an capacitive fuel level sensor in field. Its mounted on top of a heavy construction vehicle (colmar) with a 24v battery system. Our sensor runs for few weeks and damages. Every time a particular track is damaged together with all the components (mcu+sensor) following that track. So far we have only found a potential ground loop, but the track (burning repeatedly) is tested for up to 12amp continuous current. And if it had been a high voltage surge it should burn other tracks and component as well.

enter image description here So far our findings are

  1. nothing except marked in red is damaged
  2. track in red box is evaporated. Only that portion of ground track, as if some heavy current passed from battery-negative to chasis ground.
  3. Vehicle under question is a heavy duty digging machine, with all sorts of electrical noise.

Is it possible that a ground voltage variation between a) battery-negative and b) chassis-ground could develop, capable of burning 10-12amp rated PCB track? Even in that case, why rest of the circuitry following that track will get damaged?

Let me recheck battery configuration with field engineer, this way their is a constant 12v difference in my negative and chasis ground. I hope input is taken from another battery. battery checked I was wrong, correct topology updated now.

  • \$\begingroup\$ Where in your diagram is the 24V? \$\endgroup\$ – Dave Tweed Jan 10 '19 at 17:57
  • \$\begingroup\$ Vehicle has two batteries 12v each, a 24v system. For this sensor , power input is taken from a single battery. \$\endgroup\$ – neeraj Jan 10 '19 at 18:16
  • \$\begingroup\$ @neeraj - (a) Please improve your diagram to accurately show the design regarding both batteries and their relationship to chassis ground. (b) You say that your equipment works for a "few weeks" and then becomes damaged. Is it possible that there is a long-term "over-current" situation (e.g. just above what the PCB track can tolerate), through the track which becomes damaged? If so, you can measure it at any time and progress your troubleshooting. If not, and the over-current only occurs at the moment when the PCB is damaged, then something changes at that time - investigate what changes. \$\endgroup\$ – SamGibson Jan 10 '19 at 18:27
  • \$\begingroup\$ Your question says 24 V but your diagram shows 12 V. You can make this difficult for everyone or you can make this easy. Please make it easy. Your circuit needs to show any other (external) connections between battery negative and chassis. \$\endgroup\$ – Transistor Jan 10 '19 at 18:30
  • \$\begingroup\$ I have added full battery setup, vehicle under question is a construction vehicle(colmar colmar-rail.com/t10000fs) . 1) its own electrical system has 2 batteries in series, 12v each. My sensor is only connected directly across one battery. 2) My input negative is coming directly from battery negative of that one battery. 3) As this sensor is screwed on fuel tank roof, another ground is introduced as shown gnd2 on right end. \$\endgroup\$ – neeraj Jan 10 '19 at 18:46

If you really must use only 12 volts as the power supply for your sensor, you should use the battery that is connected to vehicle ground, not the "top" battery. If you connect your sensor as your drawing shows, any connection between your sensor ground and vehicle ground will produce a short circuit across the grounded battery.

Better practice would be to use the vehicle 24 volt supply with a 24V->12V DC-DC converter to provide power to your sensor. If you only connect your sensor to one of the 12 volt batteries, that battery will be discharged more than the other battery.

  • \$\begingroup\$ Let me check which battery is used. Never really bothered to check with field engineer. And its very much possible as so far they were using a sensor without chasis ground(plastic housing), where battery selection is immaterial. \$\endgroup\$ – neeraj Jan 10 '19 at 19:05
  • \$\begingroup\$ Checked I was wrong, right battery was used, so was not shorting any battery \$\endgroup\$ – neeraj Jan 11 '19 at 14:39
  • \$\begingroup\$ then your two chassis grounds aren't well connected together, and when something, like the starter motor, uses a lot of current, some of it is flowing through your track and killing it. Why are you connecting to chassis ground again anyway? \$\endgroup\$ – james Jan 11 '19 at 16:56
  • \$\begingroup\$ sensor housing is aluminium based and one of the inner pipe which is part of capacitor is electrically connected with it. And as per design in capacitive sensing that pipe is kept at ground. Hence the connectivity. as per your comment either inner or outer chasis has say some resistance. In first case I don't see any issue as current won't prefer route via sensor circuit, in second case its possible somehow sensor is providing a less impedance path for huge current to pass. But even in that case only the track should suffer I don't understand why rest of the component will get damaged. \$\endgroup\$ – neeraj Jan 11 '19 at 18:30

Now you've drawn the complete circuit it's more obvious what the problem is:

you've shorted the battery on the bottom right through that track: current flows out of the +ve terminal, through the track into the chassis, out of the chassis and into the -ve terminal.

Once the track is blown, the aduino is getting 3.3v + 12V from that battery, well over it's maximum voltage of 6V.

  • \$\begingroup\$ Let me confirm the battery configuration once again. \$\endgroup\$ – neeraj Jan 10 '19 at 19:09
  • \$\begingroup\$ checked and correct configuration updated in diagram. \$\endgroup\$ – neeraj Jan 11 '19 at 14:39

The most likely cause is something shorts the +ve 12V battery terminal to the chassis ground. When that happens the >100A that the battery can provide will burn out the weakest part of the track that connects battery -ve to chassis -ve.

After that happens the VDD of your arduino will be at 3.3V, and it's GND pin will be at 12V. That means instead of it getting +3.3V it'll be getting -8.7V which will kill it.

  • \$\begingroup\$ if internally something shorts +ve 12v to chasis ground, then other tracks, on +ve voltage sides will get damaged as well, but its just a portion of ground rails that is evaporated. as if their was a substantial voltage difference across battery-negative and chasis-ground. BTW i didn't mention but their is a 3A rated schottky diode in series at +12v input. And it was intact. \$\endgroup\$ – neeraj Jan 10 '19 at 18:18
  • \$\begingroup\$ I meant directly from the battery terminal to the chassis, for example, by it moving and touching the chassis, or the insulation on the +ve cable rubbing against a part of the chassis. \$\endgroup\$ – james Jan 10 '19 at 18:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.