Monday 20 February 2012

BMW R1200GS fuel pump controller

It's a known issue on certain bikes and mine has just died. I missed a rendezvous to ride out to Ace Cafe Adventure Day because of it! It was totally out of the blue; worked fine when I put it away and was dead when I went to use it 48hrs later. At least it didn't die at 70mph on a motorway!

I won't repeat the info that's on the forums and a particularly good site, motorcycleinfo.co.uk, but considering this issue is well known (as well as a work-around) I thought there must be a fix out there. Apparently not, at least, I couldn't find it.

So what are your options?
  1. Buy a replacement fuel pump controller (approx. £85 at time of writing).
  2. Bin the controller, after saving the connectors to make a bypass cable.
  3. Buy a pre-made bypass cable from forum members/eBay.
I didn't fancy #1, unless I really had to. #2 and #3 don't sit right with me, they'd be great as a get-you-home solution but BMW went to the time and expense of putting the controller in there; it has to be for a reason.
So I decided to open the thing up and take a look-see.
Scrape out the rubber potting compound and carefully prize out the board.
With hindsight, the better way to do this is to de-solder the 3-pin connector and the electrolytic cap (which you don't know is there until you open it - now you know).
I think the cap was fine before this. The connector snapped back together.
The cap is a Rubycon 470uF 35V AL ZLH series capacitor rated for 105degC. It is AEC-Q101 qualified for automotive use and has a life of 10000 hours. As with all electrolytic caps, their life if significantly reduced if operated at high temperature. Could this be a cause of failures reported by many? Interestingly, there is a recess down in the heatsink area, made for a thermistor for monitoring temperature, that never made it into the final design?
Water seems to seep in through the base of the connector.
Water marks and corrosion inside suggest water seeps in through the base of the connector. The seals on the mating half seal against the inside of the connector body, but water is free to sit against the base and can seep in under the rubber potting compound. The water itself won't really be a problem but the corrosion and contaminants would be.
Pretty cruddy!

So I examined the circuit board and took the numbers of the power components (first suspects). What looks to be the power switching FET: 2N06607, the number was difficult to read but comes up as 17A 60V N-Channel logic FET, seems right to me. Next to it is a similar but smaller device: 12CWQ03FN, a 12A 30V dual common-cathode Schottky diode.

I ordered replacements for the cap (busted), FET (suspect) and diode (measured ok but I'm placing an order anyway). MOQ was 5 for the FET and diode and 10 for the cap. They should arrive tomorrow and we'll see what happens...

Meanwhile, whilst drawing out some of the circuit, I noticed that the FET had a lifted pad on the Gate lead! Trouble is I don't know if this is the root cause of the failure (mechanical stress from curing of potting compound or from vibration?) or if I did it when probing around. 
Pad has lifted, but when?
Pad and via is now open circuit.
It's getting late but I might just try repairing that...

2 comments:

  1. Very interesting analysis, can you update with the result? I am also considering repairing mine. Thanks for the report.

    ReplyDelete
    Replies
    1. Sadly, I ended up buying a replacement as I needed a quick repair to get the bike ready for a trip. I then had no motivation to fix the old one.

      Delete