Fluctuating (& high) voltages on different circuits - causes?

Been looking at few videos and reading some topics here, seems a good place for advice, and hopefully someone can help in diagnosis. Eternally grateful for help that gets it going again.

2006, Peugeot 207, 1.6 petrol

What the symptoms and issues are:

High voltage 9v (should be 5v) on single wire from the ECU going to the pedal. Testing all with ignition on, this voltage fluctuates, for example, it will change every second from 9.49 to 9.55 to 9.58, then cycle again randomly. Youtube video of it:

Voltage on the TPS motor ground, is the same value as the positive when backprobing. When the connector is disconnected, the positive reads correctly 5v.

On a seperate circuit, internal central locking switch, I also read the strange flickering voltage when probing a live feed. Can also find it on two wires from the engine fusebox that lead to the internal on, the BSI

When operating various functions, say the hazard light switch, reverse light, the central locking switch or the brake light switch, voltage is seen to drop on the affected circuit, such as the pedal. It also affects the throttle ground.
When the fuel pump runs, voltage decreases on the pedal - if I disconnect the fuel pump fuse or wire, it increases, also in above video link.

All sensors otherwise receive correct voltage and report no faults. The ecu is tested and good (have three of them, all same). Only two faults actually exist, high voltage on signal track 2 and throttle stop programming fault. Virtually fault free otherwise.
New ground wiring from the ECU and fusebox and chassis to battery, no resistance to battery post. What I have found if I run a wire from the ECU case directly to a chassis ground, voltage on the TPS motor ground drops slightly.

This all prevents the engine running correctly, no throttle control with a misfire. My diagnostics can however see the position of the throttle valve angle.
Testing with a test light and pulling fuses didn't reveal much in way of finding a short on any circuit.

I've had this post saved to submit, but didnt get round to it, in the meantime though, I found voltage on ground wiring when the fuel pump runs, up to .20v. So its like the voltage reduction from the pedal for example, gets loaded onto the ground.
Is this a positive to ground short or positive to positive short possibly?

When you disconnect the plug at the tps motor an the 5v dc is read, dose it fix all other anomalies in system such as voltage fluctuations an correct voltage.

That seems you still have a ground issue.

I suggest you to check chassis voltage (negative voltmeter test lead to (-)battery post and positive voltmeter test lead to chassis). You cannot read more than 0.1v with engine running. If my suggestion is what you are doing, you can track voltage dropping to the point where ground is 0V.

Same result if you Jump (-) battery post to different places of chassis or ground wires and check voltage dropping.

9v on a 5v regulated output means pcm voltage regulator is not finding a good ground. Same method mentioned before is useful to find where pcm ground wires are loosing contact.

Agreed, high voltage on a ref. circuit is usually the result of a faulty ground.
I would check the grounds at the the PCM connector by using them to light a headlamp bulb.
If all checks out then the next most likely cause for high voltage on the reference circuit is a failing voltage regulator. Considering the problem exists with 3 computers, you can probably count that out for the time being.
I suppose that a partial short to voltage could cause elevated voltage on the ref circuit, but that is the least likely scenario in my opinion. There would have to be harness damage (thinking about the loom laying on exhaust or being pinched somewhere) or possibly water intrusion in a connector bridging circuits.

Jesse c wrote: When you disconnect the plug at the tps motor an the 5v dc is read, dose it fix all other anomalies in system such as voltage fluctuations an correct voltage.

No unfortunately, no changes, just like disconnecting/connecting various other bits like sensors or the alternator exciter.

guafa wrote: That seems you still have a ground issue.

I suggest you to check chassis voltage (negative voltmeter test lead to (-)battery post and positive voltmeter test lead to chassis). You cannot read more than 0.1v with engine running. If my suggestion is what you are doing, you can track voltage dropping to the point where ground is 0V.

Same result if you Jump (-) battery post to different places of chassis or ground wires and check voltage dropping.

9v on a 5v regulated output means pcm voltage regulator is not finding a good ground. Same method mentioned before is useful to find where pcm ground wires are loosing contact.

Finally got to checking all after some rain and freezing cold, all check out to 0.03v at most. There is 4 ground wires only from ECU, crimped into a larger wire all to a single point. Have actually fitted a negative post that links to the earth strap, was meant for a alarm system and extra lights, moved the ecu ground to this and voltage reduced to 0.01v, it seems for certain the ECU is grounded without issue now.

Noah wrote: Agreed, high voltage on a ref. circuit is usually the result of a faulty ground.
I would check the grounds at the the PCM connector by using them to light a headlamp bulb.
If all checks out then the next most likely cause for high voltage on the reference circuit is a failing voltage regulator. Considering the problem exists with 3 computers, you can probably count that out for the time being.
I suppose that a partial short to voltage could cause elevated voltage on the ref circuit, but that is the least likely scenario in my opinion. There would have to be harness damage (thinking about the loom laying on exhaust or being pinched somewhere) or possibly water intrusion in a connector bridging circuits.

Seems the ECU is absolutley grounded good now.
I'm not sure how many voltage regulators there are fitted on the ecu but was thinking that all or half the sensors would be affected, but this is what is confusing, since only the throttle control, pedal and throttle motor have the faults. Unfortunately no diagrams seem to exist of the ECU, which would I could see where it links to.


Thanks for the help, at least I know the grounds are good now.
Majority of fuses removed, except ones for ignition. When I removed the ECU fuse in the engine, put on ignition, got 0.45v, on the faulty output to the pedal.


Was meant to reply sooner, but looked at more things. Seems that there must be a wiring shorting, maybe. I turn on the hazard lights, and see each click of the relay affects voltage on two ABS wires that connect to the ECU (dropping and rising, just like the pedal), these two wires in turn are spliced that go to the BSI and power steering. Bit burnt out by it lately, may start by disconnecting everything and just re-connect bit by bit, see if the circuit can be isolated.

Hi, good news about ecu voltage drop.

The fact you are reading ground voltage raising in different areas with different loads, makes me think is a general ground issue.

Did you check main ground cable? The one which goes from (-) battery post to chassis (not the one which goes to engine block).

guafa wrote: Hi, good news about ecu voltage drop.

The fact you are reading ground voltage raising in different areas with different loads, makes me think is a general ground issue.

Did you check main ground cable? The one which goes from (-) battery post to chassis (not the one which goes to engine block).

Finally got time to go over it, its a strong connection right from the chassis terminal up to the battery.

After a bit of investigation, when any ground connection is made to to the BSI, voltage increases and is unstable. The locking switch sits stable at 6v, increasing to 9v soon as a ground is connected, (the switched live side begins to go up and down from 0.5v to 4v with the switch on).
The rear wiper motor connection to the BSI also acts as a ground, so removed pin 3 (diagram below) and it would be fine, though of course nothing works since the BSI is not grounded after all of this. I keep believing that there is a faulty connection to the BSI, as so far all the errant voltages are readable on connections from here, and pretty much everything links back to the BSI by some means. I may be wrong, but am completely lost with what is happening.

Unfortunately not yet fixed it, but have found near battery v
voltage on the 12v accessory socket ground, with a 3v on the courtesy light ground.

But, I traced the grounds physically, after metres of fabric tape peeling, only to disconnect the rear seatbelt detection connections, voltage gone, or at least 0.03v.

Thanks to all the help though, just seems this issue was crazy elusive, let's hope.