TRX EFI conversion *updated video*

[Quan-Time]

Innovate LC1 WBO2 sensor.. Bit of a pig to get going, but so far im happy with it.. Others hate it.. Really up to you.

90hp on this tune, Most ive heard of was 102.3hp with a 878cc kit and other tweaks. I strongly believe i can get close to that, on a stock 850cc. Id need a cam tho obviously.
Please remember that ive got 907cc pistons, which are higher compression, and i fully intend to do some fairly extensive head mods. I believe 100hp should be EASY to hit, and should be looking higher. Wait and see what happens in the long run.

[redkiwi]

I find the tech side of things interesting but a bit too tedious to follow it into too much detail.

I'm dead keen to hear your thoughts on the end result (how it rides etc) and hopefully a back to back comparison with a stock (or perhaps stock +FCR41's) TRX850 when it's all up and going.

[Quan-Time]

Went for a quick ride today. Bike runs and works fine. The spark curve is quite passive, it felt REALLY flat and unresponsive. Yes the bike accelerated, BUT not like it should, felt more like a 250cc or something, didnt have the urgency id like. Point is, it runs and works.
Using some unifilter pods on it, and will take it to a dyno this week or next week to tune it properly, ill post a graph when ive got more details.

[lenny]

Thanks for sharing this info Quan Time. Can you explain your tuning procedure. Will you be using the data logging capabilities of the megasquirt? I would love to set my bike up to log throttle position, rpm, a/f ratio ,timing and speed as it is ridden as a tuning aid. A dyno is great for getting into the ballpark but nothing beats logging with real world airflow especially with ram air, even the auto tune function sometimes gets it wrong. I did the '95 supersport season with Troy Bayliss and that bike never went near a dyno, relying on Gavin Cosway's very experienced ears and eyes for tuning and playing with airbox/ram air volumes and it was probably the fastest Kawasaki of that spec in the world that year. We knew of competitors who had spent mega dollars on the dyno and their bikes were only as fast as a stocker with a good exhaust. I realise your not "tuning" on the dyno but I am interested in how easy the megasquirt is to work with.

[Quan-Time]

Depends how far you want to go really.
MS can do most anything you want, you have multiple tables on each setting.. Spark, VE (fuel load) and AFR. You also have multiple choices on your VE tables, Speed density and Alpha_N. If you dont know what they are, id suggest looking them up.
http://www.77e21.info/mstuning.htm

Have a good long read of that, should explain it for some people, but it MAY confuse many. Summary is this:
Speed density takes MAP (manifold air pressure), MAT (manifold air temp) and calculates fuel load based on that.. When tuned, it can be REALLY effective.
Alpha_N uses your TPS (throttle position) and RPM (duh) to calculate fuel load. Can be tuned REALLY well, but requires dyno time to get a good setup.

Look into "palm pilot" data logging. MS3 has a built in SD card for logging, which makes things a bucket load easier. With a old skool palm you can use a serial cable to log data and output it in real time. Light weight, clean, and does the job.
http://www.blessedmemories.net/Palm/PalmLog.htm

I still need to fit a knock sensor. Its not 100% needed, but it does help things. MS has a nice option of "advance spark until you get a knock, then back it off 2deg, advance again till knock, back off 2deg.. compare those numbers, get average, set to that number minus user defined amount". This basically gets the bike REALLY close to timing, and match with a WBO2 sensor, you can get some really good results.
With a MAP sensor, even a flow meter, you could monitor ram air quite easily. Depending how you want to set it up.
Most of my tuning will be via me, or via the MS unit itself. Long as you give it restraints, its fairly straight forward. Even tuning my base fuel load was reasonably easy. Get your target AFR, and get your injector stuff set (req_fuel) and start fiddling with some VE tables.. Get it close, and your already on your way. You can see things alter in real time, but you need to have the bike under load to get it sorted properly. At this stage the MS has trouble doing it without a computer. I wasnt prepared to put the 17" laptop in a bag and ride around with it to log stuff. I had the main computer disconnected atm for other things which im doing. Ill get all that sorted at a later stage.

Dyno is for a few reasons. Mainly cos i want to see what benefit, if any, this whole project gives. Secondly, how close the tuning is to what it should be. Im no tuning expect, ill freely admit that to anyone. Reading dozens of online guides and experimenting is far from the skill level of an experienced tuner. Ive set everything really passive and on the safe side. Spark is WAY retarded to what it should be.. Yes the bike runs fine and doesnt stall, but i didnt want to advance it without having concrete proof. In time ill sort that too.

AS for others who wasted money on dyno time. Yes i know how that can go, BUT im going to say that the dyno operator, or the person logging the data was a complete moron. How on earth can you not get a decent tune ? You can simulate ram-air to quite a reasonable level surprisingly, and if you cant get a decent tune, you really need to look for work as a ... anything other than that. Saying that, 95 to now im guessing there has been a fair bit of improvement in dyno tech and measurement stuff. Could be a thing to concider.

I spent a season with a 125GP bike (Aus series) and the bike we had, did indeed have a bunch of dyno time. It had done it before we got it. It was in a simulation chamber. They could change airflow, air pressure (baro), humidity and temp. Made for some GREAT info. We had a weather station and 3mins before the call, id read it, compare what jets i should be using, and what setup, do it up, and just make it every time. We would then record what happened after the race and compare to what we expected, and what we got. We got it to the point we could be VERY consistant on things, and went on to win many times. You could always hear when a 2smoker is on song or not.

Im still scared of people who only tune by ear and knowledge.. Yes its a skill to have, but its been dis proven time and time again now. Im not saying it cant be done, but a GOOD dyno session will always be better. Its like one time i read a post on a youtube video.. something along the line of "your bike looks great, timing sounds a bit off, just tweak it a few degs and you will be loving it". Err.. how the HELL can you hear its off by a deg or 2 from a video on the net.. Audio recordings are not a true signal at the best of times.. I think it was just some fruit-loop trying to sound like he knew what he was on about.
As i said, the guys who couldnt tune a bike on bulk dyno time obviously had some HUGE problem they didnt want to fix or completely ignored, or the dyno operator was a complete nugget. They could have only been looking at top end bang, and only changing needles to suit, rather than looking at mid, which will help drag their ass out of a corner.. Ive seen that a few times.. Many people only look at the max output, or max rpm.. How often do you sit there in a race.. only about 25% worth. you spend like 60% of your time in mid to upper mid. Such is life i guess.

Getting back to MS.. by default it logs EVERYTHING. Anything that is hooked in, its logged. You dont have to display it, but its there if you ever want, even the battery volts are logged.

Hope you made some sense of my messy post.

For the record, the dyno im going to, he does most of the dyno work for ASBK teams fairly smart dude.. Jim from Superbike Dyno in Adelaide. Another one is Nevil, down south in Victor Harbor. Does all the drag bike stuff.. He is one smart cookie. Met him years ago.. Supposidly one of the best around, but cbf going that far yet.. Wait and see.