Voltage regulator. Not a common failure mode but some do fail that way.

 

Why saying the bad VR with 14v when engine running and 12.8v engine off?

I blame the DME system that measured it wrongly.

 

You can just replace the voltage regulator on the alternator in most cases. Not a big job.

 

And you too ***8212; why a bad vr when it is perfect at 14v with engine running?

 

I'd check your multi-meter for accuracy, it might be the more accurate of the two testing systems.

 

Tested but it gives out the normal voltage. Only the cluster gives out an over voltage issue. And as well as when scanned.

Sent from my SM-N950F using Tapatalk

 

It could be just the cluster if it has its own voltage regulator chip on the cluster board. Maybe ScottJoh has seen this before?

OP, which control module has the over-voltage error code? The IKE is the cluster module.

 

I'm wondering what prompted you to check the system voltage, in the first place.

 

Had my car scanned for a speed sensor error. Overvoltage came up also. But the scanner is reading normal voltage levels. Multitester shows normal reading also. So we tried testing from the cluster and that came out at 17V

Sent from my SM-A705MN using Tapatalk

 

Do you know which module is showing overvoltage?
Obviously your cluster is over-reading.

 

Do you know which module is showing overvoltage?

Obviously your cluster is over-reading.

Will check again. But tried replacing the cluster i borrowed from a friend. Same reading.

Sent from my SM-A705MN using Tapatalk

 

Are you sure you're on the right cluster test? Nothing personal, but I know I've tested temp many times thinking it's voltage and vice versa...because I can't remember if I want test 7 or 9 and those are the only two I've ever used. Just checking.

 

Im sure im on test 9

Sent from my SM-A705MN using Tapatalk

 

Let's take a moment and be sure. Although protected, some control modules don't like voltage near/at peak all the time.

Might I suggest you use a higher level scan tool and while running, take a look at the voltage inputs of the DME and EGS. INPA and PA soft should be able to call these up on the individual tests.

That's one way to be sure.

 

INPA shows normal voltage on its "dashboard" (if thats what you call it lol) but gives out an overvoltage error.

Sent from my SM-A705MN using Tapatalk

 

Does a scan tool actually measure voltage or just report the voltage a microprocessor somewhere in the car is reporting?. On my E46 the voltage reported by the scanner never matches the voltage I measure with a DVM. Electricity is pure ___ magic. If everything is working don't go to far down the rabbit hole. If I wanted to check something it would be ground blocks/connections.

 

The e46 has no pid for voltage. Not sure how it works with INPA, but the elm 327 obd2 bluetooth/wifi adapaters have a microprocessor or circuit in the adapter itself that calculates the system voltage that OBDFusion displays. It likely calculates voltage across pin 4 (chassis ground) and pin 16 (battery power positive) in the obd2 port. The elm327 I have reads 1 volt too low relative to OBD port pins 4 & 16 (battery voltage), so I always need to add +1 volts to whatever reading OBDFusion displays.

Fyi, here is jfoj's explanation from a few years ago:
https://forum.e46fanatics.com/showpost.php?p=17662609&postcount=7

Here is the problem using OBDFusion, Torque or any other App for monitoring the Battery/Charging Voltage. The E46 does not have a PID specificially for the DME Voltage, so the Voltage is derived from the OBDII interface itself. The issue is many of the cheaper OBDII interfaces just use a basic Voltage divider circuit and they are rarely "calibrated" and could also vary over temperature.

Even 0.5 Volt shift is enough to cause some concern and I have seen some interfaces off by over 1.0 Volt. What you need to do is measure the Voltage at Pin #4 (Ground) and pin #16 (Battery +) with a Voltmeter while the engine is running. Then compare this value to what the OBDII App indicates the charging Voltage is. If there is a difference, you then need to make a "mental" correction factor and possibly label the difference on the interface.

So lets say the interface reads 0.5 Volts low, you then need to mentally add 0.5 Volts to the displayed charging Voltage reading. Usually these interfaces tend to be off by a fixed amount and are not usually impacted over temperature, but they could be.

I have even seen some of the more expensive interfaces have this Voltage offset or inaccuracy problem. Newer, full CANbus vehicles tend to have DME/ECU Voltage PID that typically tends to be more accurate than the interface Voltage which will still be a displayed option regardless of the DME/ECU Voltage PID. But even the DME/ECU Voltage PID needs to be verified if it is out of range. You could run into a DME/ECU that provides poor data for the Voltage PID or there could be a Voltage drop problem to the DME/ECU.

As they say, "Trust, but verify". I almost always use 2 measurement verification on any strange or important signals or Voltages. I recently ran into an older Ford that had spikes in the 5 Volt reference that a very expensive Snap-On scan tool was seeing. After a bit of research and then finally locating the 5 Volt output of the ECU and monitoring it will a Fluke Scopemeter and determined the 5 Volt reference was stable from the ECU, but for some reason was reporting or showing with spikes on the Snap-On scan tool.

I rarely take the first readings at face value, I prefer to always get a sanity check on anything I am diagnosing

 

I'm with Terra's grounding issue.

And it's going to be brutal to troubleshoot, as Sapote's right, there's hardly anything that's trying to pull ground below the chassis ground. Or V+ above battery.

So, to be clear, WHICH electronics are reporting overvoltage? Cluster? ECU?

Once you narrow that down, then start metering both the supplies and grounds for that module.
For example, if the dash is getting -3v on one of its grounds, this will make it see 'overvoltage'-
it's also possible that something floating ABOVE ground on a pin that's supposed to be a ground reference.
will make the converters think that they're seeing more voltage than they really are.

When the ins and outs meter ok, then it's time to start swapping assemblies, because it's something inside one of them.


hope this is at least a place to start...

t

 

Have tried swapping clusters. Same problem. Just replaced voltage regulator still same, have not checked if i need to try and replace the whole alternator assembly. Dme or any modules havent tried any

Sent from my SM-A705MN using Tapatalk

 

Wow, 0.2ms injector opening latency from 14v to 12v Vbatt. This is critical to time the injection at high rpm.

Re ignition coil dwell time, I sure remember reading each coil has a current sense resistor feedback to the DME coil driver to avoid saturate the magnetic field. Maybe the table is default during starting the engine, but it's just my guess. You're right. Thanks.

 

Injection TIMING's not all that critical, as in, when the injector pops open. What's more important is the duration, especially at low duration times.

If the ECU thinks it'll take .4ms to get the injector open but it really takes .7ms, that's going to deliver less fuel than the ECU thinks.
It'll trim up to compensate, of course, but it'll be off.

...or is that what you meant?

t
short duration, low stamina

 

No, the injection duration (ON time) is calculated and independent with Vbatt, but due to the latency (delay) from the time the injector received the voltage to when the injector completely opened, the DME needs to turn on the signal earlier with lower Vbatt and delays the turn on signal with higher Vbatt. It's true that injection timing might not be that critical but with today very stringent emission BMW wanted to optimize the mixture at best.

 

Here is what the scan tool says

Sent from my SM-A705MN using Tapatalk

 

Anyone again? [emoji16]

Sent from my SM-A705MN using Tapatalk

 

I’m having the same issue. Did you ever find the problem?