| HOME | Software | Lotus Cars | DWARF | Kindle | Solar |
|
|
I began my Megasquirt adventure (after not spending enough time reading, but oh well) with buying stuff from diyautotune.com. Below, MS stands for MegaSquirt
I should prefix this with the caution that I am leaving out a lot of adventures to save me time writing this up. Do you really care when I spent a week agonizing over something that was not working just to discover some wiring mistake I made? I think not. Maybe I'll add more later.
Well, you could buy the Stim and the ECU (EMS) as kits, but I was not going to spend that much time assembling and soldering. You really need the Stimulator to ensure your Megasquirt is working. It lets you power up the MS on your desk and pretend you are pushing the throttle and lets you watch the blinky lights demonstrate it's firing the spark plugs! No engine, engine wiring, or even car needed! Counting the wiring harnesses (I bought two different ones) and a relay-board I wound up not using I spent just over $700. The Megatune software is free.
As shipped, the MegaSquirt had old firmware. Upgrading to 2.88 firmware was an adventure made much more challenging by the fact I run GNU/Linux, not Windows. The first challenge was that my laptop had no serial port, but Megasquirt needs one. Well a USB-to-serial dongle fixed that. Then after 1 minute or so the Megasquirt stopped responding at all to the serial line. Well, I finally tracked down (on megamanual.com) how to check the serial line chip in Megasquirt A few simple measurements on the board said 'this chip is dead'. So I unsoldered the 16pin MAX232 chip and ordered a new one from diyautotune.com for about $7. The web page listed no sockets and I wanted one of those so I would not have to desolder that chip again. On the order I asked how to get one. diyautotune.com sent a socket with my MAX232 chip and charged me nothing for the socket. Pretty nice. Once I soldered in the socket, plugged in the chip, and stuck things back in the MS case my computer could talk to the chip and (again with struggle) got the firmware updated.
Some of what follows won't seem sensible unless I talk about my engine setup. My TWM2000 intakes don't have a vacuum takeoff anywhere so I chose Alpha-N engine control. Other people seem to use Jenvey throttle bodies with good success and make up an intake plenum for the MAP sensor on MS, but my Caterham source was using TWM2000.
Thus I (and MS) ignore the MAP sensor on the MS. That also means I'm running open-loop, meaning I am not using the lambda sensor to control my fuel. So everything is really up to me getting the VE table correct. And to guide the tuning of MegalogViewer the Air-Fuel ration table has to be sensible. The key to my tuning going from impossible to control at start to seeing some control was realizing I needed to make the low end of the tables fine-grained. It takes very little throttle to run around town and so I needed fine control there. Up at higher rpm and throttle things are not so fine-grained. At least with my engine.
So now I tried to start the engine using a basic .msq file from a Zetec-using friend. My misunderstanding of a part of the air path of the TWM2000 lead to a week of very slow progress. The engine would fire but not idle. Frustrating.
I had many adventures (mistakes in wiring EDIS and The TPS (throttle position sensor) and other things and watching megatune dials on the laptop while trying to start really was crucial. When I saw there was no spark signal from EDIS I knew something was wrong in the EDIS wiring. By the way, in spite of my EDIS mistakes and shorts I never seemed to harm the EDIS itself, it seems pretty foolproof.
Then I realized some problems were due to not using enough shielded wire. So I bought several 10foot lengths of shielded wire from diyautotune.com (if you need it for MS they have it!). Using that for the spark sensor and TPS lines helped. It's good stuff, though making wiring mistakes and redoing things is very time consuming...
Once it would idle I discovered I could not actually release the clutch. I removed the transmission and realized I needed to modify a few parts there (parts I designed and had machined).
Once you can idle and work the clutch one can drive around locally. I would drive around the block (flattest route possible, though that's difficult where I live) and on a hesitation I would back off. Doing data logging using Megatune. Then I would use Megalogviewer to figure out where it was lean and richen it up. Iterate over and over and over. To the point one can even drive on the freeway with appropriate air fuel ratios (lambda).
This only gets you so far, it gets you no where near full throttle. With safely over-rich top end (from the friend's msq file) I did a little full throttle work. And a lap or two at a track day. When I did briefly do full throttle the clutch seemed to slip. So that meant removing the transmission again and getting a high-pressure pressure plate (CenterForce). Which could not work till I again changed the throwout bearing (doing some machining using a milling machine). So, put the transmission back.
Now I had no clutch slip. So at a track day I took the car to MCE Racing's dyno and Kevin Murray got my Wide Open Throttle (WOT) calibrated. Kevin on the dyno and throttle and me watching my numbers on the laptop and making changes he suggested. He basically suggested one change on WOT (across a specific range) and air-fuel was perfect for 3000-7500 rpm.
This work is licensed under a
Creative Commons Attribution 4.0 International License.