This process took the rest of the year and it has to be said I’ve had a number of problems trying to tune my map. I must state at the outset though, that none of these have been the ECU or the software, it has mainly been a lack of time on my part.
The basic premise is that you connect a laptop to your ECU. The ECU can then be instructed to dump all the inputs, outputs and resulting settings to a file as you drive around. You can then sit and read through this file in the comfort of your home. Contained in the datalog is the O2 value from the wideband, along with the map value used, the rpm, etc. You can therefore see precisely where the engine is running rich or lean and what value in the map was used. It is then a case of working through the datalog adjusting the map values according to the O2 values you saw verses what you want. Simple, huh!
The great thing about using an open source ECU is that there are people out there who know far more about these things than I do and who obviously have the time and skill to write little applications which does all of the above for you. The standard version of this is called MSTweak3000 and it does precisely as you would do, you feed in your target O2 values, your map and your datalog and it manipulates the values of your map in an attempt to achieve your target O2 values.
In reality this wasn’t homing in very quickly, although I later found a good reason for this which I shall come onto later.
The biggest problem I had was a matter of time. I simply ran out of it during the year. My work had taken a major U-turn and had really picked up, which I’m pleased about. However, the knock on effect was that I was a lot more tired when I did get time to myself. The other problem was that we are in the process of decorating the whole of our recently purchased house. This year we finished the small bedroom and then concentrated on the back garden. We shifted two skip loads of soil in an attempt to level it, I built a deck and laid two tonnes of gravel and slate. It felt like every spare evening and weekend was spent out there breaking up our horrible earth and shifting it around to form piles ready for the skip. Anyway, that meant that I didn’t get to drive the car much, and you can’t map it if you are not driving it.

My early attempts to map the car did not go very well. I had initially set my acceleration enrichment to next to nothing so that it did not influence the map, but somehow this setting got lost, either I didn’t send it to the ECU, or I reverted to an earlier map which didn’t have the changed settings. I didn’t figure this out and thought that the map was to blame for the car going rich when I attempted to accelerate. I kept on pulling fuel out of the map and couldn’t figure out why the engine was still going rich, but then lean…. It’s obvious now, of course.
I found the problem whilst out for a drive. I went out for the drive after I had loaded a revised map from a previous drives datalog. I hadn’t gone far when I knew this was even worse than before and was cursing inside my crash helmet the day I decided to embark on this whole upgrade. As you do. I found a place to pull over and made some adjustments which also made no difference, it was still going rich. At my next pull in point, I reduced the fuel yet further and found I couldn’t even pull away as it was going rich and then incredibly lean. I stopped again and decided to give up with my latest map and reloaded the previous version. This was better, but was obviously not a step in the right direction. As I made my way home I thought about things, trying to work things out and pulled over again to make some tweaks to the map. I don’t know what triggered the ‘check the enrichment’ thought, but for some reason I had a look. I have four settings for throttle application, for gentle through to slamming your foot down. This is set my means of a voltage vs time value and then an extra amount of fuel is specified which is over and above the value in the map. All of the throttle speed settings were set to the lowest possible, which mean that they were always being activated and all at once. The values against these were default values which effectively meant that when I applied the throttle even slightly I was trying to add 60ms of duration to my injector opening, or basically more than 6 times the fuel it wanted. This immediately made sense as to why it went rich when I moved my foot, but why it would then show up as lean. I was leaning off the map trying to compensate for the enrichment. At the side of the road I set all the values to the lowest they could be and pulled away – the difference was incredible, the car felt fantastic, or at least in comparison to before. Still, I had only wasted 3 or 4 days of mapping, unfortunately that equated to around 2 months!
With a new found enthusiasm borne from finding the above problem and a feeling that maybe I could get this all working after all I booked a few days annual leave to really get stuck in – my holiday year runs from July to June so I usually have days left over. This happened to be when the temperatures were particularly hot and after a couple of days in the garden I got to go out for a drive and datalog a run. Towards the end of the run, sitting at some traffic lights, I noticed that the wideband readout was showing the engine running leaner than it was at the start of the run. It wasn’t a vast amount, but more than enough that I would have tried to map it out, around 0.5 lambda if I remember correctly. As I made my way the last couple of miles home, I was thinking things through and couldn’t come up with an answer, other than fuel pressure. If the pressure had dropped it would cause the engine to run lean. The other factor was temperature – the ECU determines the fuelling based on the manifold pressure, rpm and temperature of the inlet air, and therefore its density. I know, however that my air temperature was reading correctly as it was reading out on the laptop screen next to me. Once home I removed the bonnet and saw that the pressure on my gauge had fallen from the 43.5psi to around 35psi. Not good. I left the car to cool down and after an hour or so I flicked the ignition on and achieved closer to 40psi, another hour later and it was back to the 43.5psi setting.
This is where I made a mistake. I decided that it must be my fuel hose expanding and causing the drop. I had bought cheap'ish standard rubber hose, albeit high pressure, for the run from the tank to the TB’s. My reasoning was that over such a length I could be seeing a drop due to the heat soak. I decided to swap it over to high quality Aeroquip braided hose and fittings. The pressure rating of the hose was far higher and the fittings were required. I sent my fuel rail back to Jenvey who swapped it for a -6 fitting version for the cost of the postage (which was great service). I went to my local motorsport shop, Autocross, to get the hose and fittings. This is where things went wrong. They didn’t have any hose as they had had problems fitting some fittings and the rep was due in to see if there was a problem. I bought the fittings and called a few days later to find out the outcome of the problem. The rep couldn’t find anything wrong and it seemed to just be that they were not following the instructions (although they had never needed to before, they claimed). Anyway, they had my hose so the following weekend I made a trip to collect it only to find out that they had bought 3m (the amount I had told them I needed) and then cut 1m off to do their trials. Another week wasted. During the next week Mandy called in and collected a full 3m and I fitted the first fitting. When I got the next fitting out I found it was different and wouldn’t fit. Mandy took this back and swapped it to the type I had before (i.e. a proper Aeroquip straight compression fitting). This replacement didn’t look the same colour and whilst I was sure it wasn’t the same manufacturer I couldn’t afford more lost time so tried to fit it. I got the fitting into the end of the hose and the thread started fine. With around 2 threads to go the thread sheared off and I was left with a fitting stuck in the hose and no fitting. I was just thankful that I had left a good amount of spare length, as this fitting now had to be cut off, loosing a couple of inches. Autocross will not be getting another recommendation from me, let alone any business as from this point on they were rude and arrogant to Mandy who was the only one of us able to visit as she almost passes them on her route home. The best thing was that when they were being particularly rude about her not knowing what a proper Areoquip was she was able to lay in to them about working for a nearby race engine company – and therefore knew what she was talking about. Whilst it is useful having a company like them nearby, I will not be treated like that and I have not used them since, and nor will I.
I went to ThinkAuto, and wondered why I didn’t do so in the first place – I got precisely what I wanted, a 'proper' Areoquip part, delivered to my door in less than two days! I got the fitting in place without any problems and the hose routed through the car. I had in the meantime had plenty of time to think things through. Unfortunately I came to the conclusion that the hose probably wasn’t the problem. If the hose expanded, the pump was more than up to delivering the extra capacity at the required pressure. I had begun to suspect that it was more likely to be the pressure regulator. When I then started researching this aspect I came across a number of people who had tried using the same FPR and had pressure loss problems. I borrowed an alternative from a friend, Steve, and fitted that and set it to the required pressure. This is where it was useful to have selected a nice round 3bar pressure as it meant that I could then buy a fixed pressure unit which would be more reliable. A combination of the new fuel lines and Steve’s alternative regulator solved the problem and the car ran consistently from the start of a run to the finish.

That is pretty much where I am currently at with the mapping process. I made a few more runs and put the datalog through a neat application that allows you to set the target AFR values, and adjust the map – instead of aiming at a single ARF figure and trying to do the whole thing in stages. Time, however ran out and I currently need to work on the fuelling from around 5k onwards. Over the 2005/06 winter I have found that one of the later versions of MegaTune incorporates an autotune function which uses a target map for the AFR and adjusts the map realtime while you drive around. At the end of a drive you then hit save and burn the updated table into the memory – I can’t wait to try this out as I definitely found the process of data logging and then interpreting the results and making changes a slow one. As I write this, it is February, around 5 degrees outside, icy with salty roads, but I’m still looking forward to hitting the road again soon (waiting for an uprated starter motor).

24/06/2006:
For the last year and a half I have had a misfire on my car. As I moved the mapping range higher and higher I started to suffer with a misfire that was always at the same rpm, ~4250rpm, and then again at ~500rpm gaps. This was making it hard to map, although I had only just got to that point by the end of the year. I was also suffering from a weak pulsing on my shift lights - taken from the ECU. I spent quite a while chasing this pulsing, fitting voltage regulators in the circuit, trying to stop it - I gave up.

Over the winter I pulled my loom apart in the engine bay and decided to re-route the screen for my crank sensor to the ECU, move the ignition module earth back to the same place, etc, etc...

At the start of this year I was all freshly motivated to get the car sorted out - Doh, the misfire was still there, just like before. I carried on playing around with the mapping (this is when auto-tune was implemented, and I set it up to self map against an AFR table). The mapping was starting to come together but the misfire wouldn't go away. I even tried routing the ECU earth back through the ignition module and straight to the battery, rather than under the dash. No change.

A chance grumble from the wife last weekend after I was testing the car in the garage and the car interfered with the TV meant that both Bill Shurvinton and Kevin Wood suggested that it sounded more like an HT problem. That evening more testing in the garage and with the lights out I could see a light blue glow tracking along the longer HT leads... a quick check with Bill, Kevin and Steve Foster, and apparently, no, they are not supposed to do that.
Because I am using the Raceline coil mounting plate, I can't use standard ignition leads - they are too short (a lot to be said for keeping things standard!!). As I had a Friday and a Monday booked off I really tried to arrange for some new custom leads made, but failed. So on te Friday I climbed into the garage loft, rummaged around in a couple of boxes and found the standard bracket that, for some reason, I kept because 'one day it will come in useful'. Swapped over the mountings, fitted the coil, nipped up to Motorists Discount for some 'never heard of brand' £18 leads for a Mondeo. It was with some trepidation that I started the car - I didn't know what else I would do if this didn't fix it.

Way hay!!! Car was instantly happier ticking over and more willing to rev when blipping the throttle via the linkage. The nosecone and bonnet went back on and 20 mins blat around the lanes. It's Gone and boy does it feel fantastic to drive again. I had almost forgotten. It also showed that whilst the problem seemed to be at certain rpm, it was actually effecting the whole rev range, as it is now better everywhere.

This will teach me to not always suspect my own work - I couldn't see past the fact that I had built the ECU and loom and therefore it must have been my fault somewhere. £18 spent last year and I'd have been so much more motivated towards the end of the year when I was getting frustrated and regretting the upgrades that felt like downgrades.

A massive thank you to Bill, Kevin and Steve for their help - always proves what a great bunch se7eners are, always willing to help others with a single similar interest, same 'brand' or not. Cheers guys, it's going to be a good weekend - who knows, I might even finish mapping it - I hope it doesn't start raining now.

Update:
It wasn't fixed, only improved on and after many frustrating efforts to find the problem I threw in the towel and changed the ECU to an Emerald

 

 
Copyright © Webmaster