How To Guide: TDM900 EFI onto a TRX (warning, many pics)

Quan-Time · Post by **Quan-Time** » Thu Oct 25, 2007 11:42 am

This "How To" guide is more or less because endless searching doesnt yield a complex "in depth" look at how this conversion could be done.

This isnt a 100% completed guide at this stage, but it will hopefully be a very good source of information on the subject.

TDM900 THROTTLE BODIES

As you can see, its possible to bring these out to 42mm ID fairly safely. Remember that stock TRX carbies are 38mm, and FCR carbies are 41mm (unless you buy the 39mm version). By increasing the bore size, you would need to change the butterfly to suit. You might also want to make a bevel adapter before the air filter.. Think "tunnel ram" or funnel to increase air velocity. This can be measured on a flowbench, but keeping the same angle which its currently tapered at (im guessing 5degs included angle) should be plenty.

THOAT TO HEAD ADAPTORS

The top picture shows how the injector is placed and how some blending of the angles into nice radius would most likely help with CFM flow.

The middle picture shows how the height of the 2 adaptors is vastly different. The throttles are the same length as the carbie models, but are located closer to the head, this would mean the stock airbox would require lots of modification, or simply adding on some pod filters would work fine, which is a much better option.

You can also see that the larger of the 2, the TRX stock item DOESNT have the "dip" in the bottom for the injector. The bolt holes are also different so the TDM adaptor DOES NOT bolt directly to the TRX head. The TDM adaptors provide no room for new holes to be drilled in.. Either a spacer plate or custom adaptor will have to be fabricated to fit, else an entire new head will need to be sourced.

The bottom picture shows how the dip would sit when bolted to the head.. Many hours of die grinding would need to be done to make this flow properly. The picture does not show it, but the walls on the head are approx 1mm bigger all around than what the adaptor plate is. This will help with air flow a touch, and means no further work other than the gouge for the injector need to be made.
You can also see that the bolt holes dont match up

THROTTLE TO HEAD FITMENT

The spacing for the throttles matches up well which means they can be used safely if new adaptors are fabricated, or a way to fit the TDM adaptors.

Even after having all this, you need to control the injectors. Theres 3 methods that i know of for this.

1: use the stock ECU module for a TDM900. It should be fairly straight forward, but you will need need to make a wiring harness or try to adapt the 2 together

2: Buy a PC3 for a TDM900 (power commander 3) and do the same as above. Will be flexible and dyno-tunable

3: Make a "Megasquirt" ECU unit. Megasquirt also provides the added bonus of being able to control spark, thus being even more tunable than the PC3 option.

MEGASQUIRT ECU (MS2V3)

The front of the MS has a data port (for linking to a laptop) and 3 LEDs, for quick reference (injector fire and or spark fire)
The back has a larger port to which you make your wiring harness, the small black plug is a barametric pressure guage, which means its capable of elevation correction (riding in hills or sea level) in realtime with no user / retune needed, and is also capable of 40psi turbo boost (auto rich / lean cycles).

Controlling spark would require you to machine the CAS (crank angle sensors) and making a 36-1 encoder wheel. This would be possible and i already have the CAD drawing ready and am halfway thru modding a pickup to accept this.

Thats about all for now.. I hope ive shed some light on the subject.

wicky · Post by **wicky** » Thu Oct 25, 2007 8:31 pm

looks and sounds good Quan even though it's all above me and looks loike it'll be very complex to complete

will you be using the TDM 900 cold start mechanism ?? as they can seize up and cause problems, so it might be worth stripping and cleaning it b4 ya get around to fitting the throttle bodies. Wick

ekoja · Post by **ekoja** » Thu Oct 25, 2007 11:15 pm

So, will it get me to Aldi any faster?

Quan-Time · Post by **Quan-Time** » Fri Oct 26, 2007 4:07 am

All the electronics will be handled by the MS2v3. It has built in auto rich cycles for start etc.. One of the probes you hook up to temp, set a condition for it "if below this temp, run THIS map condition ELSE run OTHER may condition"

It has "live" timing to some extent if you have a detonation detection sensor. Basically turn bike on for the 1st time.. runs like a pig. uses 12:1 (very rich fueling) for a few seconds until the O2 sensor in the exhaust starts to automatically lean it back to a safe level.. usually 13.7 / 14:1 depending.. but manual control is also an option. Timing is started at fairly retarded rate. As the bike starts to idle, it advances the spark in 0.1deg incriments each 100rpm of the engine. Once it starts to detonate, it backs it off 1degree.. After doing this 3 times, it "learns" what idle should be at. start to rev the bike, hold it at 2000rpm, it will smooth out, hold it for about 1-2secs, and slowly climb to max rpm. Thats it,, from there you will be able to ride and its fairly well tuned.. Obviously a dyno run would help and manually fine tune it, but for being able to ride down to a dyno tune place, it works perfectly.

Many ppl put a MS on cars, its possible to handle 16 injectors and multi spark setups. Lots of ppl dont even do dyno runs to tune, they just let it auto adjust.. I remember reading about one guy who got a tune up after 3 or so months on his car, it was running so well they didnt actually change anything except some fuel lines and pump or something as it showed it was very much on the limit (he turbo'd his)

Quan-Time · Post by **Quan-Time** » Fri Nov 23, 2007 2:20 pm

just a quick update.. even tho its a bit slow, im still working on it.
Ive fitted the adaptors and slightly "relieved" the roof of the port for the injector to spray thru..

So basically the head is modified to accecpt fuel injection now.

I still have much work to do on it,, But least i can get it on a flowbench and start getting some numbers from it (cfm, velocity for starters) and work from there.
The throttle bodies are being machined out to 41.2mm which is a big as i dare go as the walls get a bit thin at one point. Im also custom making some funnels / trumpets / rams / whatever you want to call them.. More on that later.

davesta · Post by **davesta** » Sat Nov 24, 2007 2:17 pm

Always very impressed with your work, your skills are certainly something to be proud of. Keep up with the updates, i'm sure i'm not the only one interested in this project

billy budge · Post by **billy budge** » Sat Nov 24, 2007 6:14 pm

Damn Quan, you are my friggin hero!!!!

Is there nothing you wont tackle, your definately something to aspire to.
Do you fancy popping over to blighty and finishing my bike off, all beer paid for of course!!!

philk · Post by **philk** » Sun Nov 25, 2007 12:13 am

We're not worthy........We're not worthy........

Nice mod this, really wish i was capable of this level of modifying.

dirty trix · Post by **dirty trix** » Sun Nov 25, 2007 12:58 am

Looking great so far. Was hoping you'd put it on stock and dyno it first, but makes sense to do it all in one go.

Quan-Time · Post by **Quan-Time** » Sun Nov 25, 2007 2:54 am

i probably could put it on stock, and see what figures it would pull. BUT it wont be much higher than a stock carb setup.

If you look at one of the top pics, it says the throat size is 38mm, which is the same as the mikuni TRX carbies.
Also note that the TMD900 is almost identical power figure to the TRX which is 50cc LESS, its the limitation of the throat, BUT you get a touch more torque because of extra CC size.
So by machining out my throats to 41.2mm i should be on par with FCR41s, or if anything, a touch better as the velocity of air would be a little quicker, which helps premote rev pickup and fuel mixture.

wait and see what happens in the next comming weeks.

dirty trix · Post by **dirty trix** » Tue Nov 27, 2007 12:07 pm

I'd have thought the biggest benefit of the FCR's and also the fuel injection though would be the speed of response of the slides (or even quicker with the butterflies), rather than waiting for the CV's to suck their way on up.

If stock they give me faster throttle response, first and foremost that would be a great improvement.

Extra airflow would be just an added plus.

Quan-Time · Post by **Quan-Time** » Tue Nov 27, 2007 3:20 pm

quicker response is a different kettle of fish.

Your exhaust has as much to play with response as your air box does funnily enough.

if by response you mean "go from 4000rpm to 7000rpm while in neutral gear FASTER than it does currently" theres so many things that can effect that its crazy. you REALLY need to plan what you want.

Seen on car after market parts they claim "3500 - 5000rpm power range" for a cam and 4000 - 7000rpm" for a inlet manifold ?
you CANNOT mix and match.. you need to have them all done the same, and running in the same range. This is why some ppl can do a mod and get nothing, because something else is limiting it. Then you do something else, and you think OMG THAT WAS A HUGE DIFFERENCE..
tell the next person to do the 2nd thing you did, and they dont get the same power / impression.

Someone stated (cant remember who, but they were right

) mikuni carbs prefer "still" air rather than "rushing" air.. ie, they prefer a lower velocity. FCR on the other hand prefer more velocity to work efficiently, its why pod filters work on FCRs but not on stock mikuni's.

Theres a huge ammount of theory on this, and im still trying to learn it.

HansJ · Post by **HansJ** » Tue Nov 27, 2007 6:17 pm

Interesting work! A big thumbs up from me.

And a question re the flow, You don't see any problems with the porting, I mean that You probably won't get an equal flow over the valves (the mid one probably will get more flow?) Any thoughts on why not adding a thicker adapter and thus getting better possibilities to smothen out the inlet flow?
Which in term would give a longer distance between throttles and valves, I know...

Quan-Time · Post by **Quan-Time** » Wed Nov 28, 2007 4:39 pm

i own a flow bench

fear not

You are correct in saying that each side is different than the middle. You normally flow all 3 together, but you can seperate them.. But final flow testing must be done on all 3 as the cam would do it.

When i get a chance ill start pulling more numbers off the head and adjust accordingly.

Having a 3valve inlet head, you actually reduce velocity, because the air is spreading out in the head to the valves. If you can re-speed the air up, or keep it as fast as its going already, you can help the head work better.

please note, by velocity porting you DECREASE your potential CFM.. Because you are taking up room, BUT you are increasing the rate at which air moves.

Think of it like this.. Run 100m with your mouth open as wide as you can, now run with a straw in your mouth (please stop after about 2m, else you will prolly pass out)
Now run it normally.. Your mouth tends to adjust automatically to what your lungs tell your brain what you want. So you close your mouth, but you are still getting required amount of air in, but because your mouth is smaller, the air must move FASTER down your throat to your lungs.

Take a garden hose, make sure there is nothing on it, just the open hose.. Turn it on a bit so its flowing out.. This is your raw CFM number. You can measure the velocity via the diameter of the hose. now LIGHTLY start pinching the hose. you will see it shoot further, you are STILL getting the same CFM but the velocity is increased.. if you squeeze it too much you loose cfm and velocity will eventually drop, unless you have more pressure, ie: turbo or something (which is why stock cars respond to turbos really well).

Theres a whole bunch of stuff you can do here.. Porting your head for a certain lift state, certain RPM, HP, torque, response, economy and efficiency (i think thats all of them)..

when most ppl want "best HP possible to brag to mates" you actually LOOSE HP in usable range, but at top end right on the rev limiter is where you get max power.. such as drag cars. Because the air CFM / velocity is just not right for the mid range where most of your actual work is done.

Remember, in one full valve opening / closing cycle, it is closed longer than its open, and its only open MAX for a split second,, its 0 - 35% open more than 35% - 100%.
By calculating your total cylinder size, rpm you want (say 3 - 5000) you can work on this as a base. ie: 850cc = 425cc per side, 4 strokes, so 425cc of air per 2 revolutions. Say 4000rpm (for a nice start point) would be 425cc and 2000rpm, you can start establishing real world CFM, and start velocity porting from there.

I know a NASCAR engine builder (actually head porter

) who prototyped a max velocity possible on exhaust side, he 100% IGNORED CFM.. he did everything he could to get max air speed..
Put it on a flow bench, he DROPPED like 15cfm over a STOCK head..
That engine won its next race.. go figure that one out

HansJ · Post by **HansJ** » Sat Dec 01, 2007 9:47 pm

Quan-Time wrote:
...
Remember, in one full valve opening / closing cycle, it is closed longer than its open, and its only open MAX for a split second,, its 0 - 35% open more than 35% - 100%...

True, and that's why flowbenches don't tell the whole truth. Resonance and all other evil stuff... Who knows, it might even be better with more flow in the middle inlet valve?!

The guy who did my head has decreased the size of the inlets quite dramatically, after they split int three sections. He has also done some work for Husaberg, and has surpsrised the constructors with his ideas/results

Nevertheless, interesting stuff, keep up the good work!