I also encountered the UCH problems. The car was a 2003 Renault Clio with a UCH-N3.
Intermittently the following faults were appearing: knocking relays, wipers not working, lights not working, windows not working, central locking not working, instrument cluster backlight not working, etc.. It was obvious, that it had something to do with the whole of the UCH unit not receiving power at various points in time.
The features my UCH fault presented:
1. When you first started the car, the car was intermittently in disco mode (all electricals going on and off approximately two times a second). For some seconds everything is working fine and then for some seconds the disco mode and then again back to working. This would keep on happening for a good few minutes. After some time all the faults always stopped appearing. Evidently it thus had something to do with a component on the UCH warming up and starting to work properly.
2. Tugging/pulling/twisting on the 40-pin black connector had an apparent effect on the fault. When pulling to a certain direction, the fault goes on, twisting to the other direction the fault goes away.
3. Hitting/tapping the UCH housing/case/plastic container with mild force would also randomly turn on/off the fault.
At first I was sure that the 40-pin connector was the culprit. Must be a bad contact. I removed the whole connector and substituted it with some new wire, which I soldered straight into the PCB (printed circuit board). To my surprise, this had absolutely no effect to the fault. So clearly it wasn't the 40-pin connector at all, that was causing the problems. At this point, grabbing the freshly installed set of wires and pulling/jerking/twisting them had the exact same effect as previously: fault goes on/off depending on how you twist them. What the heck? So it must be something related to the fiber glass of the PCB itself bending way or another.
At first I wanted to believe that the people on the internet know what they are talking. As in some thread there was this guy "bil_liser" saying, that ALL the UCH problems are caused by the decoupling capacitors. So I went and changed all the UCH capacitors around the 40-pin connector (although not all of them are capacitors, as you will see in the attached pictures in the end).
So, I had changed all the decoupling caps to fresh ones. Still, no effect. The fault remains exactly the same.
I did some more research on this topic online and read that BBA-Reman does a "full-rebuild" when the symptoms match what I was encountering. I don't know what a full-rebuild is but I started with trying to find out if there were replacements available for the main microprocessor. I didn't find even a datasheet for these poor buggers. So I concluded that BBA-Reman probably is not changing those either. Also, from my experience, ICs do not fail that often at all. Okay, so replacing the ICs was discarded as a solution. So what might a full-rebuild for 100 pounds be in reality? Maybe a reflow of the board in a reflow oven/gaseous phase oven. That would although require the removal of the through-hole mounted components, i.e., the 7 plastic relays on the PCB as they would melt at such extreme temperatures.
As I didn't want to detach the relays, next on the list was a re-solder of the board using a soldering iron, some flux and some new solder. The area of interest would logically be in the vicinity of the problem causing 40-pin connector. As the symptoms described in the beginning of this post (points 1...3) matched perfectly to a dry solder joint, off I went.
Before heating up the soldering iron, I used a DMM (digital multi meter) to measure some resistors. Didn't find any with a short-circuit. Then some capacitors. They also had enough resistance, no short-circuits here either. Then onto the diodes. Using a DMM in the diode mode I measured the diodes. One of the diodes seemed to give very intermittent readings. At first I wasn't getting any reading, then when I just held the measurement leads in place for a second or two the DMM would display the breakover voltage for a very short amount of time and then it again said that I am measuring an open circuit, although, I was measuring a diode. Aha! There must be something wrong with the solder joint(s) of this diode. Maybe the power to the microprocessor goes through this diode? That would explain why I was losing all the UCH related functionality whenever the fault was on.
I re-soldered all the components inside the yellow box, shown on the attached picture. Special care was given to the diode which was giving out intermittent readings. The intermittent diode is marked with a red outline in the picture. And what do u know, the UCH has been working like it came out the factory the very day.
I must give a word of warning to anyone trying to remove the black 40-pin connector from the PCB. Without using a hot-air soldering station or a suction-pump soldering station, the job was not the easiest one. First, I tried to remove the connector using solder wick. I put four hours into this and no luck. It just wasn't coming out. The next trick was to flood the pins with solder. I put a very liberate amount of solder on the pins such that I had two long rows of solder on the board covering all the 40 pins. Then I used 4 soldering irons simultaneously and gave it some good heat. Having some locking pliers/vise grips already in place around the 40-pin connector, I yanked it out when the solder turned liquid. Yay, I got the connector out without breaking the PCB. But boohoo, the connector was now a melted piece of plastic. So, I had to order a new connector from Mouser.
I am attaching some pictures, maybe these will help someone in the future when encountering similar UCH faults.
First, a picture of the temporary wiring loom I had used to figure out if the 40-pin connector was causing the UCH problems:
Next, the overall look of the board after I had re-attached the 40-pin connector. It can be noticed that the connector is sitting a little higher then originally as I inserted an "adapter block" between the connector and the PCB, might I ever need to detach it again in order to reach the SMD capacitors and such beneath it.
Next, is the picture taken by andy_casey and edited by me to include my measurement data. I had to measure all the capacitors to find out their values. Also, note the red frame pinpointing the location of the intermittent diode I was talking about, above in the text.
And the looks of the PCB, after all the capacitors and MOVs were replaced with new ones:
Also, I had to redo the wiring loom at the car's end because, as you might recall, I had removed the original connector as the very first step of this troubleshooting process:
No BBA-Reman for me. And more importantly, no new UCH from Renault dealer, price estimation (as installed onto the car) was from 750 to 850 euros. And from an independent car electrics specialist I got an offer of 620 euros. Yikes!
Overall, it took me nearly two months to get this baby back on track with all the time included waiting for parts from overseas.
But oh boy, what a personal victory it was when it finally started working again.
