What problem is the most difficult to diagnose? The one that does not exist when you are looking for it. I think I am particularly unlucky with those because I possess some sort of magic touch. I’ve seen a whole lot of cars that were doing all kinds of troubles for their owners but once they get to me – they are working perfectly fine. I think I should charge at least some small fee for doing that 🙂

2012 Ford Escape 3.0 V6 got brought to the shop on a tow truck and left on a parking lot.

Customer described concern as “engine shuddered when driving, then battery light came on along with other warning lights and then it stalled altogether”. My first thoughts were “OK, that’s typical, the battery died because alternator was not charging then engine shut down once voltage dropped below operating range, should be simple enough”. I connected battery booster, to the battery, started engine and drove the car into my bay. Then while engine was still running with booster connected I checked charging voltage.

Surprise! Charging voltage was at 14.6V – perfectly fine. Disconnected battery booster, turned engine off, restarted again – starts just fine. Only at this stage I realized that battery in fact was fully charged with voltage at 12.7V, it couldn’t get any better. Not so typical anymore :-(. If there were any problems with charging then battery would be low on charge or completely depleted. Tiny chance still existed that someone, somewhere got this battery re-charged and then sent this car to the shop, but this was extremely unlikely. In twenty years I never came across this kind of kindness, always had to start with dead battery having to charge it first before proceeding with diagnostics.

Let’s check what fault codes we have here. First thing I noticed was a code P1000 stored in PCM. This code sets in PCM memory if someone had codes cleared and engine control system have not got a chance yet to go through all checks required to verify correct operation of different subsystems (these checks require specific conditions to be achieved that might not happen for some while during normal driving). Oftentimes code P1000 in memory is a strong indication that someone was already trying to do some repair – and eliminated all the useful evidence for the next diagnosing technician!

Also codes U0100 (Lost communication with the PCM) were stored in few other modules: 4×4 control module, ABS, instrument cluster (IC), power steering control module (PSCM) and airbag module (RCM). Finally, code C1963 (Stability control inhibit warning) was stored in ABS as well. It might look like a lost of troubles happening, but in reality all codes could have been set for example if someone had engine started with nearly discharged battery at some stage quite a long time ago. It would be so handy if every code would have some time and date stamp associated with every code, but there is no such a thing in cars, not for now at least. So all these codes could have been set minutes ago or few months ago – there is no way of knowing it :-(. As a matter of fact I am seeing plenty of U-codes in different modules on different vehicles, some models seem to be impossible to get them without a single U-code in any module, so this is not what rings an immediate alarm bell for me.

Customer got contacted with questions and stated that no one, including tow truck guy, connected any kind of diagnostic equipment or cleared codes recently – and meaning of “recently” in this case was in the last two years since customer bought this car.

For sure, I took this car for a longer test drive, including some rough roads in my route and monitoring battery voltage as one of the PCM live data parameters. I was hoping to see some hiccups in this data, that would give some clue about possible problem with charging or power supply to the PCM – nothing was wrong.

Finally, I put the car on the lift and inspected it for any evidence of damage down below and under the hood – nothing again. Actually this particular car was in a pretty good shape – fairly clean and seemed to be looked after reasonably well.

What I could do? Not much really. All I had was reported concern that only happened once (!) and have not left much evidence apart from codes that might not be associated with concern at all. If its not broken you can’t fix it?

Sometimes you can :-).

One of the possibilities was that power supply was interrupted to PCM or/and some other modules at some stage. Then all those codes could have been set when this power supply incident got to happen – but this was only one of the multiple possible reasons to have all those codes in control units memory. Charging system or battery issue appeared to be unlikely so I looked into PCM power supply circuits. First of all I printed out diagrams and then step by step removed and thoroughly checked PCM power supply circuit components: relays, fuses, wiring harness, connectors and terminals and so on.




Here is what I found. Ground point for PCM ground connection G109 was quite corroded


20151026_092656 20151026_093006

But even more suspicios was the fuse F5 that supplies permanent battery voltage to the PCM (KAPWR voltage at pin C175B-62)


It had some corrosion on one of its legs



It did not look as very bad on the first sight (it looks worse on the picture because it is greatly enlarged). This fuse, though, is quite critical as it also supplies voltage to PCM power relay coil

PCM power relay

Therefore if power flow through the fuse gets interrupted power to another PCM power input (VPWR at pins C175B-67 and -68 through the fuse F29) also cuts off


Although ignition ON power to PCM (also VPWR but supplied from entirely different circuit throught the fuse F23 to pin C175B-21) would remain


Lets take this as a hypothesis for what had happened. How to check this if this theory is right?

I cleared all the codes except P1000 (this one can not be cleared with scan tool and requires to go through the specific drive cycle in order to be cleared) and connected variable resistance in series with the fuse (new fuse was installed and terminal where its plugged in got cleaned). Started engine, engaged drive and let wheels spin in the air on the lift. Then I added a bit of resistance into the circuit simulating by this the fault that supposedly had happened. Engine stalled and all the warning lights came on on the instrument cluster. I rechecked for codes: all the same U0100 codes in the same modules were detected again!

Looks like the theory was correct as the car never came back again.