I had an issue with the turbo starting to leak oil. Upon investigation I found that the turbo was heat spiking, causing it to wear prematurely. It appears that this is pretty common when you track after-market turbos that don't have water cooling.
So I decided to switch to water-cooling. You can order the turbo casing with or without the water coolant connectors. It's not really that difficult. You split off the tubes from the main line; they enter/exit 90 degress from the oil connections. You can see the red & blue connectors on the sides of the turbo in the picture below:
The trickiest part was figuring out how to attach the braided lines into the existing coolant hoses...I finally found a reduction fitting that works well. And part of the tubing rises & falls right at the level of the overflow tank. So I adapted a gas sampling connector to act as a bleeder port. It works perfectly.
The other issue I've had is that the ignition coil is too close to the turbo (see picture above). The connector & parts of the casing are beginning to melt. And my attempts to shield it have just made it worse by blocking the air flow around the turbo.
I've wanted to switch to one-coil-per-plug to provide more individual control of each cylinder anyways. This also solves the heat problem by moving the coils far away from the heat source. I looked into changing to coil-on-plug, but I could not find ones small enough to fit in the R53's spark plug tubes. However Motec mates a coil-near-plug variant:
I purchased these & struggled with how to mount them in the Mini's very cramped engine bay. I finally came up with a simple but effective mounting scheme tying them together with small strips of aluminum:
Because I don't have the top-mounted intercooler, this *just* fits between the head and injectors and attaches with a small bracket to the header bolts:
Here's a picture of it mid-way through wiring them up:
And here's what it looks like with all the wiring:
And finally everything together:
This has not only moved all the wiring away from the turbo, but it has opened up a better air path back to the turbo:
And I'm thinking of mounting a curved metal sheet where the coil used to be to help direct the air flow between the valve cover & the turbo.
Ok, the car is running like a dream now....but I did have to fix two (of my) stupid mistakes first:
My first Homer-esque mistake I discovered immediately when I went to run the ignition tests with the ECU. I hooked up a spark tester to ignition output 1/cylinder1, told the ECU to fire the spark (2400 RPM @ 6ms dwell) and it worked great the first time!!! Same thing with output2/cylinder2, output3/cylinder3, output4/cylinder4!!..... dope! They're supposed to be wired in firing order!! Sigh. Check the book...no, I can't tell the ECU to change the firing order. Fortunately I just had to rearrange one control wire for 2,3 & 4...sorted!
Ok now everything works great! Great spark, everything in the right order, injectors running great, etc. etc. Finished reassembling everything, pushed the car out into the driveway, fire extinguisher handy, turn everything on, push the button - car turns over fine....but will not start. Nada. zip.
Hmmm....well, I know the ignition & injectors are working OK. Let's pull out a spark plug & check the spark - perfect! (actually the best I've ever seen it). Battery low? Ok charge it...no change. I was running a bit rich last time I started it, maybe I'm flooding it. Trying cutting fuel by 10%, 20%, 30%...nope. Ok, add fuel by 10%, 20%...nada. Tweak timing...zilch. Sudden fear: maybe I left a rag in the air intake!!! Frantic disassembly...no, nothing there...ack!! Bad fuel? Worked last time I was running it & haven't added fuel since then...rogue settings value change in the ECU? (it's happened before)...methodical search...nope. Major frustration.
As is often the case when I let my brain work on a problem in the background (i.e. when the silly conscious part is asleep) it quickly comes up with the solution. And as I was driving to work the nextday it explained it to me:
As I've mentioned before in a 4-cylinder, wasted spark engine there are two equally correct timing points that are separated by 360 degrees. Setting the GRIP to either of these two values will work equally well. However now that I have one-coil-per-plug, this is no longer the case.
I was 99% positive I had the correct value for the GRIP, but it's easy enough to check. The minute I got home from work, I pulled out the computer, changed one entry in the settings, flipped the switch...and it purred to life! Success!
My impression thus far is that the spark with these coils is amazing. It may be my imagination, but the spark is far more intense than I remember it being before. Just as an example, when I was running the initial tests I moved one spark plug wire from coil-to-coil. At one point I forgot to switch to the next coil. When I started the test, the spark jumped from center post of the previous coil to the closest metal contact point..over 1.5" away!