After encountering the following symptoms:
- Hesitation
- Stuttering
- Stalling
At seemingly random intervals I decided to dive into what was causing this. Now previously many in the community have pointed to the Fuel Pressure Sensor (which I replaced) but then later had the same issue return within 1 month. The other, as we’ve come to know well, is the PEM unit which is responsible for sending fuel signals to the fuel pump from the CEM. Now in almost every previous case i’ve seen people simply replace them. Here is the cause of failure and fix i’m proposing.
Now this may be specific to the 2004 model year S60R and V70R models with the Bosch 30636431 — other modules may have different control boards. But if you encounter this issue you can check yours as a last resort. As you may also know the 04 year has a specific PEM unit, whereas at some point in the future the PEM’s were seemingly interchangeable on a number of model years. Add to this that the 04 model years had the PEM mounted EXTERNAL to the vehicle, which allows the following condition to occur.
You may have already noticed (in some previous posts) that the box, when mounted externally, will sometimes crack around the heatsink and allow water ingress. In northern climates where salt is used on the roads, this will allow a salt crust to build up inside as water enters the casing and then is dried (due to heat in the case?). You will see this crust if you open the box (and hopefully don’t break every damned clip on the thing ).
Now lets get inside and see what i’m going on about.
The chip circled in red is BUK9606 which is responsible for sending/controlling the fuel pump (from what I can gather). The most important take away from the data sheet is that it is THERMALLY SENSITIVE. These two black chips have vias throuigh the board to the heat sink on the back (which takes up the whole face of the case). Now when I say it is thermally sensitive, I mean that the current it is able to output is determined by temperature. It has a cutoff of about 160c where it will then drop off like a rock. Now let’s look into why this might happen.
After removing 2 torx screws you will see the thermal pad by GENTLY prying the board up (I broke some pins, but whatever). Be careful as you are bending the 5 feet that go to the harness (in center of image). But on the left you will see the pink (contaminated to all hell) thermal strip responsible for allowing the chips above it to interface with the heatsink. As you can see, water and salt ingress into the case has made this filthy and corroded the aluminum slightly underneath it. the holes you see above it on the board are the vias that allow some thermal interfacing from the chips, to through the board, then to the thermal strip, then to the heatsink. This interface must be clean.
Solution in summary then is to remove this strip, clean it, remove the corrosion on the face of the aluminum heat sink underneath it (scraped it off with a flat head screw driver, ran some alcohol over it to wash off and repeated until clean). There are some holes in the strip, be careful to align them to the screw holes upon re-assembly. Ensure the strip does not wrinkle during assembly, it must lay flat against the aluminum and between the board to create good contact and allow the cooling to happen. Also be sure to tighten those screws down good and snug so the thermal pad is tight with the board and heat sink. I made the mistake of doing this up gently and ran into the above issue immediately again. I mean if you hear a crunch you probably went too far lol.
So what do we learn? When the chip on top does not make adequate contact with the thermal strip (interrupted by salt, grime, corrosion etc) it will not function well. This leads to the chip overheating and at 160c will drop off current to the fuel pump significantly to the point where you will see stuttering and bucking and then stalling.
On VIDA I did note prior to the repair that the fuel pressure was going up, and then dropping, then going up in a sort of sawtooth pattern. This suggests to me, a brief cooling cycle while the pump is unable to function with lower current. This allows the chip to recover slightly, but then is ultimately going to enter a thermal run away pattern until it cannot provide current to the fuel pump and then shuts down or the pump shuts off and the car stalls. Sort of a built-in recovery mode. This also explains why restarting the car seems to “reset” the car. It does not, it only allows the chip to recover for a short while.
Ultimately, however, this does not explain why the vehicle can and does enter normal operation at some point. It’s a very strange problem, but this is my running theory as best I can track it. Please excuse my verbal diarrhea as this is my first write up.
I hope this helps some of us R guys keep these wicked cars on the road longer. God what a pain in the ass this was to figure out.
Link to Data Sheet BUK9606-75B – N-channel TrenchMOS logic level FET