ESE's works engine tuner

[Flettner]

If a carburettor has one advantage over fuel injection, this is it.
When the quantity of fuel is controlled via the open/shut time of an injector, it is practically impossible to create a homogeneous mixture at the injector. And I'm not even talking about vaporized fuel here; I'm just talking about every part of air getting the same number of fuel droplets.
Fortunately for us much of the necessary homogenisation takes place in a two-stroke's crankcase, with the con rod stirring things up and transferring its heat to the yet-unvaporized fuel droplets. But with direct fuel injection you loose this advantage, and the con rod bearing looses most of its cooling.

Keep it simple, Neil. Spraying a constant flow of fuel in the inlet will do fine. I've been doing some work on 6,5 cc engines (google F3D and MB40) and since those little bastards rev to 36.000 there was no chance in hell that I would find injectors small enough and fast enough, so I designed a constant-flow injection; really simple, with an electric motor driving a gear pump and a spring-loaded conical-seat injector that would open according to the pump yield.
The pump is no problem: they are readily available for people building there own miniature jet engines. Controlling the pump rpm was less easy because the motor that comes with the pump, simply won't run slowly enough. Those litte jet engines are real guzzlers, even compared to a two-stroke, and the motor's stall rpm would still drown our piston engine while we were trying to start it.
Stepper motors were out because their bearings could not handle the gyroscopic forces in an F3D-airplane cornering at over 40g, and because stepper motors can miss steps without giving any feedback other than the engine running too lean. So I had to pulse-width-modulate the original pump motor in order to keep it running at a low, but to all intents and purposes regular rpm. I tried to measure its rpm by having a Hall sensor look at the teeth of the pump gears, but the teeth were so fine that the sensor couldn't distinguish them. The solution was to drill a small number (say 4) holes in the gear flank and have a Hall sensor look at those.
This is where the project came to a halt because of other, more pressing activities. But it may give you something to go by.

Yes I'm familiar with pulse width modulation, most car / motorbike electric throttles work like this. It's how my 360 RV throttle needs to work. I think the LINK ecu will drive this as an injector.

[Martin1981]

Wobbly, i have a question regarding the reed setup you posted once.

you said, that one should use the softest carbon reeds that one can find. i have reeds of 0,25 mm thickness. is that ok or too thin?
and can i use the 0,25 mm thick carbon material for the revplates and the backupplates also? thanks.

[mr bucketracer]

Wobbly, i have a question regarding the reed setup you posted once.

you said, that one should use the softest carbon reeds that one can find. i have reeds of 0,25 mm thickness. is that ok or too thin?
and can i use the 0,25 mm thick carbon material for the revplates and the backupplates also? thanks.

the hulk is a 2 stroke

[TZ350]

How the Rocket Three is built. I like their approach to the small details, like handling and lunch breaks.

[TZ350]

.

The last roll of the Dice.

Sat down and took some time over checking out the EFI injection end angle.



First step was to set the Ignitec up so it had zero advance and then magnetically connect a spare trigger for the Ignitec to a solenoid coil energised by the pulse from the EFI as it fires an injector.

The EFI is triggered of the original rotor ignition timing trigger and the EFI in turn triggers the Ignitec and timing light so we should be seeing the true timing of the EFI injection cycle. By swapping the solenoid wires I can see the injection start and end point.

With the wires the right way around I can easily see when the EFI is ending an injection cycle.



Next step was to position the piston at different places, mark the flywheel with a marker pen and then use a timing light to see whats happening and then adjust the Injection end map until the timing mark lines up.



Kept notes on Transfers Closing, Inlet Opening, Inlet Closing and BDC. So now I am confident that I can accurately position the injection end time.

Ran the bike up on the dyno, started Ok and ran fairly well pulling 11k+, runs well up to 9k'ish but still has the deto problem over 10k. So will put a little bit of time into seeing if we can tune that out. If that does not work then I guess injecting into the transfers will never work and I will move the main injector to the inlet tract and see how that goes.

Its the last roll of the Dice and if I don't get any joy there then this EFI thing is destined for a Turbo 4T CVT project.

[wobbly]

Reed petals natural vibration frequency ( rpm ) can be determined easy enough in EngMod, but when adding backup stiffeners
especially with rev plates it all gets way too complex way too quickly.
Using reeds that are too thin shows up in the sim as the tip lift going spastic on successive iterations, one run the reed will open all the way and hit
the stops.
The next run it will open early but nowhere near all the way.
Due to this the inlet wave goes mental as well, and power drops.
On the dyno the thin reed will make good mid power then loose it all up top.
But using a rev plate and a backup stiffener you can use a reed that is in effect too soft but control the errant lift profile.
I have always used 0.4 or 0.5mm rev plates, and have found that for some reason,thick fibreglass backups seem to make more power as well.
The 0.5mm spacing lets the a thin reed open quickly, but the backup then takes over and damps the uncontrolled oscillations and excess tip lift.
The only way to test any of this is to dyno combinations of thickness, and varying backup length.
I have used very this 0.25 carbon main petals, but they require long and thick backups, and ultimately slightly thicker ones have worked better.
But that was with the specific engine I was testing on,they all respond differently.
Great thing about the revplate and backup idea is that you can also change the stiffness on top and bottom for intakes that are bent down due to the carb bowl angle.
But also you can vary the backup length across the reed block, to bias the flow back toward the inner petals in an engine that has the manifold bent to the side - like most MX setups.

[seattle smitty]

while im thinking of it, whats everyones preference for texture on the walls ? ive heard alot of various opinion on this subject, mostly from noobs that dont know much, so i figured i would ask the experts here. ive always just used the carbides and pretty much left that texture on the walls but maybe theres something better ?

I'd like to hear some educated answers on this, too. My "swag" (scientific wild-ass guess) on this is that the long-side turn and the sides of the transfers (and also the whole inside surface of the crankcase) ought to have a textured surface similar to what gets left by glass-bead blasting. However, since glass-bead is very dangerous to the engine since it sticks in every crevice and embeds itself in cast aluminum, another media, say salt, would be safer. The inside turn or short-turn of the transfers should have the textured surface, but golfballed with small dimples. How you accomplish this in cylinders without sleeves (and many that have sleeves) is, um, an interesting challenge. And of course this is only a guess. AND, you first have the shapes and sizing of the transfer passages, and the aiming and timing of the port-entries into the cylinder just right to begin with, making any worries about the surface treatment secondary or thirdendary. But again, this is all rank (very rank) speculation, presented in hopes that the smart guys will provide the straight dope here.

[Martin1981]

thank you for the detailed answer, wob. as usual very interesting.

so i will make carbon reeds with 0,3mm thickness. the spacer with 0,5mm and backups with 0,4mm (you said that 0,5mm backups are too thick.) but probably one can not totally nail it down to one perfect setup for all engines.

by the way, sweapatrick tried the setup you posted and said that in comparison to that the vforce 2 can stay home. do you think that this custom made reed setup can even be better than the vforce 2?


you mentioned the bend intake on mx engines. indeed, i am working on a 1986 CR 125. either i make a dead straight intake or i will have the bend between reedcage and intakeport to fight the flow issues. but i think even with the bend not between the reeds and the carb, there maybe problems. so the best way should be a straight intake without any bends.

[Drew]

[SIZE=3]
Its the last roll of the Dice and if I don't get any joy there then this EFI thing is destined for a Turbo 4T CVT project.

That would be a terrible shame, but I think no one would begrudge the decision.

From what little I've retained reading these pages, det can be managed by backing timing off and feeding in a shit tonne of fuel.

Is there a field in the map that's not doing much at the moment, you could set to open an injector over certain rpm and just leave the bastard squirting straight through the transfer port at the piston top to cool it when it's down?

Or have I not grasped the basics again?

[wobbly]

The CR125 will make around 2Hp with a dead straight intake, over the best bent one.
An old RS125 short manifold one if you are reving to 13.000 or the longer late model RS125 one if you need more mid power.
The VF2 works very well in a CR125/RS125 and I have not been able to get a stock cage of any sort to make as much power.

[peewee]

you mentioned the bend intake on mx engines. indeed, i am working on a 1986 CR 125. either i make a dead straight intake or i will have the bend between reedcage and intakeport to fight the flow issues. but i think even with the bend not between the reeds and the carb, there maybe problems. so the best way should be a straight intake without any bends.

or you could do what ktm did, although it may be alot of work

both cylinders have the same total angle of 20*. the honda on the left has the full 20* in the manifold which makes a kink so to speak. ktm on the right has 10* angle in the cylinder inlet and the other 10* in the manifold. who ever said them guys at ktm were a bunch of rubes

[husaberg]

thank you for the detailed answer, wob. as usual very interesting.

so i will make carbon reeds with 0,3mm thickness. the spacer with 0,5mm and backups with 0,4mm (you said that 0,5mm backups are too thick.) but probably one can not totally nail it down to one perfect setup for all engines.

by the way, sweapatrick tried the setup you posted and said that in comparison to that the vforce 2 can stay home. do you think that this custom made reed setup can even be better than the vforce 2?


you mentioned the bend intake on mx engines. indeed, i am working on a 1986 CR 125. either i make a dead straight intake or i will have the bend between reedcage and intakeport to fight the flow issues. but i think even with the bend not between the reeds and the carb, there maybe problems. so the best way should be a straight intake without any bends.

or you could do what ktm did, although it may be alot of work

both cylinders have the same total angle of 20*. the honda on the left has the full 20* in the manifold which makes a kink so to speak. ktm on the right has 10* angle in the cylinder inlet and the other 10* in the manifold. who ever said them guys at ktm were a bunch of rubes

This is the intake Wob refered to.
http://www.swedetechracing.com/parts...ke_pre-95.html


They also do these

http://www.swedetechracing.com/parts...reed-cage.html

[wobbly]

No husaberg the old RS125 one is very short and dead straight.
The KTM does reduce the asymmetric flow within in the reedblock, but then that reduced asymmetry is made alot worse by having
the flow exiting the reed tips made to run off to one side of the cylinder/case,no matter what is happening within the reed.
This will mean that no matter how clever the reed tensions are made to equalise the flow exiting the tips,the flow into the transfer duct entries
above the crank will always be asymmetric.
Not so clever at all IMHO.
As I said in a CR125 or a RS125 having the dead straight manifold is worth a couple of HP, as the flow is dead straight horizontally as well as vertically.
In the KTM that is actually impossible.

[husaberg]

No husaberg the old RS125 one is very short and dead straight.
The KTM does reduce the asymmetric flow within in the reedblock, but then that reduced asymmetry is made alot worse by having
the flow exiting the reed tips made to run off to one side of the cylinder/case,no matter what is happening within the reed.
This will mean that no matter how clever the reed tensions are made to equalise the flow exiting the tips,the flow into the transfer duct entries
above the crank will always be asymmetric.
Not so clever at all IMHO.
As I said in a CR125 or a RS125 having the dead straight manifold is worth a couple of HP, as the flow is dead straight horizontally as well as vertically.
In the KTM that is actually impossible.

Really this is the part number they refer to in the pic?
P/N 16220-NF4-900)

So you mean the one on the Left with a post 95 (NX4) part code?
16210-NX4-000
looking at the bottom
27mm plate to carb both sides
http://www.risingsuncycles.com/image..._isolators.htm

[wobbly]

Yep, thats the trick shit part for the last of the MX based RS125s - NLA,except a few lying around that were used in open class CR125 engines before they stopped them in USA.
Thats worth a couple of Hp and then stick on a short PWM or better SPJ and thats another 1 1/2 at the top ( over PWK ) in the MX or early RS125.