ESE's works engine tuner

[TZ350]

what carb i should put on my 50, is that one 28mm?

It is a 24mm OKO, nearly everything we have is 24mm equivalent one way or another.

[TZ350]

Page 1490 ........

I have used the Lambda in the front muffler mount for ages on the dyno and at the track in karts ( its allowed during practice ).
The end of the rear cone is a very hot spot and the sensor seems to like this ( its heated as well of course ).
The readings are very stable and I have only had to replace a sensor once in hundreds of dyno runs.
But generally i am only doing full throttle pulls on the dyno so cant use the sensor for part throttle adjustment.
For that you would need load control, or have full data logging at the track, which we have.

I am puzzled that this actually works. As a good 30% of the delivered air to the engine goes straight out the exhaust port never to return to the cylinder, this must result in 30% of delivered oxygen going down the exhaust pipe. This is equivalent to a four-stroke running 30% lean of stoichiometric afr. Now a wideband sensor will indicate the correct lean afr for the four-stroke so how does the sensor give useful information for the case of the two-stroke?

I must admit I was very skeptical about using a lambda on a 2T, especially as several dyno operators told me they wouldnt use
the sniffer on a Dynojet ( even with its inline filters etc ) as it continually shit the probes.
But in 07 when we first went to USA racing the SKUSA CR125 box stock class,I saw a probe mounted thru an alloy bung welded
right at the front of a muffler.
The Lambda read the A/F going past inside the perf tube.
I asked to look at the data on the PI system, and there it was, a perfect readout varying with throttle and track position, between 12:1 and 15:1.
The fact it was 12.8:1 at full noise down the shute,told me it was saying the right things, as the egt was 1280*F, just as we were running - right on the edge.
The Yanks were obsessed with trying to get the fueling number correct the whole lap, and this made the corner exit ratio " correct " but slower.
And that jetting was too lean down low to get the kart launched at the starts as well.
So I ignore the numbers in this area and do all the bottom end response against the rear brake, on the stand.
If it will snap off the bottom when given instant full throttle from idle, then the pilot/needle/tube/slide are correct.
If it snaps from 1/2 throttle,without hesitation, or blowing a heap of rich smoke,then the mid jetting is correct.
We shot for a specific egt number in accordance with the weather, and this always corresponds with a A/F around 12.6 to 12.8.
So as we can use the egt ( but not Lambda ) in races I can then track the weather and jet change on egt alone.

This fits exactly with what ive found... I use a wideband o2 on the dyno and it is extremely reliable... made up a simple vacuum setup using a funnel held up against the main extraction fan for the room and it pulls exhaust gasses up through a 1/4 inch diameter copper tube thats pushed up the silencer. Reading are basically taken from the beginning of the silencer and drawn up past the sensor via rubber hose. The numbers are highly repeatable and after doing enough bikes you get to know exactly what AFR each bike likes at each rpm/throttle position. My dyno is an eddy current so i hold it at 4 or 5 rpm points and then test each point at 1/4, 1/2, 3/4 and full throttle... end up with a matrix of AFR's and its then very easy to make adjustments to pilot, needle or main to alter the AFRs at each point. Before using this setup i went through about 3 sensors a month ($$$ ), now i use 1 or 2 a year and the dyno is used 3-4 days per week. I also find 12.6-12.8 to be a good range to aim for up top, then slightly leaner around 13.5 at mid throttle and mid rpm, and leaner again around 14 at low throttle and low rpm.... this is for offroad and MX bikes though. Im about to setup a KTM330 with an Ignijet so we'll see how this method goes with EFi....

My guess is the CHT follows the actual temperature in the cylinder by a few seconds. Obviously the heat generated has to heat the head which heats the probe and this takes time.

I presume that once I tap off coming to a corner that the cylinder heat will go down straight away.

From the graph it looks like the delay is around 7-8 seconds. Does this sound correct?

The highest temp (190 Deg C) is midway round the hairpin at Ruapuna. (See the dot)

Attachment 320232

Here is another plot.

See how the temperature drops with the 'high' speed down the straight. The max is still behind the corner.

So I suppose a question is, does the bike seize from overheating or over cooling? Would be quite a different story with a water cooled motor.

Attachment 320233


This is very interesting. I would have thought the cart guys would have looked into this. I hope someone can shed some experienced light on it.

Maybe this is the first time we have any real publicly available aircooled cylinder head data. I would love to know the interpretation and what the data means.

If this is truly the way it is in realty with a good motor then there is nothing much to fear from running a well tuned big hp aircooled motor hard down the straights its just what happens in the corners that matter.

PS, please tell us more about the data logger you are using, I would like to use one too. I have a 4 channel unit but it does not record temperature relative to the position on the track.

If you are getting lower temps at high rpm,then high temp at lower rpm ( throttle opening ) then I would say the main jet is rich and the overun jetting ( that is the pilot/needle/tube combo is lean ).

You need alot of fuel ( rich ) in the bottom end jetting to stop a lean condition on the overun, when road racing - even more so with an aircooled engine.

It needs to be actually rich ( smoking ) off the bottom when free running - and having a case full of rich mixture as you open the throttle on corner exit is always faster.

The high corner temps would also be an artifact of the over advanced midrange we talked about previously.

I discovered this years ago when initially using a straight line motoplat ignition on an air cooled YZ125 kart - on Methanol back then.

When we changed to a later model retarding unit, it was way hotter mid corner, and lean on corner exit.

It needed way more fuel to keep the CHT/EGT under control at lower rpm with a ton of mid advance.

I missed this, Wob must have posted as I was writing mine.

Two strokes like an ignition curve approximating to an inverse of the power curve.
With around 28* in the mid range, that ramps up from an easy kick starting 10* at idle to about 1500,2000 rpm.
Then as the bmep rises and the pipe gets in sync with the port we pull out timing in relation to the dynamic compression created by the wave action, to prevent detonation in the end gasses.
Most engines like around 15* of timing at peak torque, we then drop this away after peak to increase the heat released into the pipe.
We have a finite amount of heat energy to use to heat trapped combustion gas,heat the water, heat the piston/head, or release it into the pipe.
The ignition delay sets the amount dispersed into each element and how much this creates power at the crank is set by the interaction of each component part.

Attachment 249567

In working out a curve to programe into the Ignitec we did various dyno runs with a straight line setting but different base advances. The runs were cut short as we were only looking at the lower part of the curve.

Attachment 249566

Same runs compaired to my previous KX80 fixed ignition 16* BTDC (Red Line).

WTF ….. The Beast screams around the track on full song for a whole lot of laps then quits when taken out for an easy trot around in the two warm up laps for the next race, what is that all about?????

Attachment 261714

Damage is in the exhaust port area, and thats a steel ring thats cleanly snapped in two.

I will dig out the dyno graph and ignition curve.

Rob,re the Ignitech curve "issue". I have found that the engine will not make much if any more power, as shown by the dyno load it will pull, but by adding a heap more timing, it will for sure
operate much , much better on part throttle response, on track. Here is a 125 kart engine curve that made 50RWHp,and won 3 back to back 125 National titles easily. This thing would deto like mad on part throttle at 6000 if held there, but on track it would pull out of 1st gear hairpins hard enough to spin the tyres on full throttle from the same rpm, then rev out to 14200. A TPS and a 3D map would achieve a better result, but I didnt have a RS125 Honda carb at the time.

.........

Attachment 261718

To be on the safe side we had richened the mixture up (Blue line)

Attachment 261717

And in case the piston might be to close to the head we tried an extra 0.5mm base gasket (Red Line) but we didn't like that and took it out again.

But it is interesting to see what lifting the barrel only 0.5mm does.

Attachment 261716

There is no evidence the piston was hitting the head, just a lot of derbis crushed into it. I regret we didn't lift the head for a look after the first race, then we would know for sure, but it was going so well I didn't want to disturbe it.

Attachment 261719 Attachment 261720

The original dyno graph and ignition curve.

With a straight line ignition curve the engine made max power around 16 degrees, the advanced portion improved the lower power curve. The ignition curve is 16 degrees at 12,750rpm and retads 10 degrees by 13,500 rpm for max over rev. This curve was carefully developed on the dyno using full throtle runs and took a lot of work and time.

10 degrees retard, on full throttle is that excessive? is it OK to hold it on full throtle in the max retard area at 13-14,000 rpm for longish periods? like holding out to the end of a short straight.

Any way the problems happened at lower rpm and lesser throttle openings when the motor wasn't being driven at race pace.


I am wondering if this could be the problem? and I might need to use a TPS and go the 3D map way, or at least switch between 2 maps depending on rpm and throttle position.

The one and only reason you get a destroyed piston like that is DETONATION.
And looking at the Ignition curve you have posted it is immediately obvious why the thing keeps on blowing up.
NO 2T race engine making any sort of power can take anything like the advance numbers you have except in very fast, short burst, transient testing.
If at any time you had held the engine on part throttle at 10,000 rpm in 6th gear on the dyno, then done a roll on test it would have seized instantly.

I have never seen in 30 years of dyno testing any engine making reasonable Hp that would take 22 degrees of advance at peak torque, 12200 , then 20 * at peak Hp, 12400 in this case.
You have the wrong end of the stick completely with the comment about holding the engine in the overev region.
Its safe as houses up there as its BMEP is low ( little dynamic com ) plus there is not much advance, alot of the heat energy is going into the pipe - not the piston and head .

The real issue is exactly as per the comment of mine you posted re that 125 kart engine - the edvance curve it liked when used at full throttle in the mid range,would deto its tits off when held
for any time at those low rpms, be that on part throttle for a fast corner, or slow acceleration due to corner loading.
In a kart, part throttle is very rarely seen in a corner, as they have tons of grip.In a bike you run part throttle in corners most of the time, especially in the area just under the peak torque point
and having 27* of timing at 11,000 is going to kill the thing every time unless jetted well rich.

Using fuel to cool an over advanced engine looses power big time and it is impossible to set up an ideal ignition curve on a rolling road with no instrumentation at all.
With no deto sensing or being able to track EGT plus CHT you have no idea about what is happening with a particular ignition setup at all.

You used the port timing map off the sim I did as well as the head geometry, that put the overall engine efficiency quite a bit higher, but then ignored the combustion file that had an ignition curve tailored to suit the bmep
plus the overriding fact that it is an air-cooled engine.
If you take the sim as I last ran it then do a Turbulent Model with your ignition curve,the TuMax will be thru the roof,and the program will probably scream DETO WARNING at you in the mis range.

The way to go is pointed to by the fact that the RACE box has two curves that can be used in all sorts of ways - one of which is as per Honda RS125 setup, using a shift rum sensor that has allows alot of advance
in gears 1 thru 3, then uses a retarded map in gears 4 thru 6th when the acceleration rate is alot lower.
But really you first need to get a handle on a single map, that can be used in any gear at any TPS without it destroying itself.

These issues have nothing to do with an inherent problem with the 2T engine as some are joking about - I went to the kart nationals over Easter, with over 300 entries, only one siezure of a 2T all weekend that I heard about,
and that was caused by some idiot running so low on fuel it sucked air.

[speedpro]

I'm thinking of proposing a rule. We have to consider handlebar width, I see no reason not to have to consider how far out the carb sticks.

[TZ350]

I'm thinking of proposing a rule. We have to consider handlebar width, I see no reason not to have to consider how far out the carb sticks.

I am already on to it, I have proposed a bike can be no wider than its length or the number of dollars it took to build it, whichever is greater .....



Anyway my carb sticks out the front ....

[ken seeber]

https://www.youtube.com/watch?v=Bp4tGTNNi1I

Any Rygers in this lot?

Fletto, get a move on !!!! Before it's too late.

[JanBros]

I am already on to it, I have proposed a bike can be no wider than its length or the number of dollars it took to build it, whichever is greater .....



Anyway my carb sticks out the front ....

what are the plates with springs for (on the side of the cylinder above carb entrance ?
and there are 2 versions ?

[ief]

I am already on to it, I have proposed a bike can be no wider than its length or the number of dollars it took to build it, whichever is greater .....



Anyway my carb sticks out the front ....

If it works to it's full potential or not (yet) I so do like the way you came to this construction

[Frits Overmars]

https://www.youtube.com/watch?v=Bp4tGTNNi1I

Did I mention KISS before?

I'm thinking of proposing a rule. We have to consider handlebar width, I see no reason not to have to consider how far out the carb sticks.

Talking about KISS: if you think you need a maximum width rule, stipulate a maximum width; no need to mention the handlebar or any other part of the bike.
As for that width being asymmetric (two feet of carb sticking out on one side): don't bother; things will sort themselves out or prove to be unpractical.
The best rule is an absent rule; the second-best one is a short and simple rule. So let's try to Keep It that way: Short and Simple.

[TZ350]

what are the plates with springs for (on the side of the cylinder above carb entrance ? and there are 2 versions ?

It is the blow off valve. There was combustible mixture in the plenum chamber. The first version blew the plate so hard against the studs that the plate bent. The second version used the Scotch Bright as a cushion, works very well.

The plenum bike has actually been test ridden in the drive and for several practice outings at the track. Testing was stopped when we realised there was considerable fuel dropout in the plenum and we became concerned about how much oil (how little) was actually getting to the engine.

EFI is an attempt to overcome the lack of mixture control through fuel drop out and the doubtful oiling problem.

Couple of video clips, first ever startup and driveway test. The clips are quite old, when we first tested the plenum idea it was to overcome the perceived restriction of a 24mm carb but it became quite clear that at 20hp the 24mm carb was not the problem.

Now we are regularly knocking on the door of 30hp the idea was worth re visiting, but it is doubtful that even at 30+hp the plenum is really necessary. Fun idea though.

Starts Easy.



Crisp as under power, throttle response was Ok on the track too. Motor 125cc plenum 1,200cc but should possibly be 3,000 according to EngMod2T.



We found some interesting results on the dyno. It may be possible to make the plenum with two chambers the smaller one to resonate for a low end off power torque boost then open the plenum up to its full size for full on power.

[ief]

Allways thinking it must be me or something I misunderstood but... latest test update and turbulent runs great, Neels the man.

[adegnes]

Allways thinking it must be me or something I misunderstood but... latest test update and turbulent runs great, Neels the man.

Nice!
Go Neels!

[marsheng]

My guess is the CHT follows the actual temperature in the cylinder by a few seconds. Obviously the heat generated has to heat the head which heats the probe and this takes time.

I presume that once I tap off coming to a corner that the cylinder heat will go down straight away.

From the graph it looks like the delay is around 7-8 seconds. Does this sound correct?

The highest temp (190 Deg C) is midway round the hairpin at Ruapuna. (See the dot)

[TZ350]

The highest temp (190 Deg C) is midway round the hairpin at Ruapuna. (See the dot)

That is very interesting, not what I would have guessed.

Hairpin least amount of cooling air flow and greatest amount of engine heat soak, or maybe part throttle detonation heating the head. It will be very interesting to know what is happening there.

Is it similar for the other corner apex's coolest on the straight hotter at the apex?

[AndreasL]

Allways thinking it must be me or something I misunderstood but... latest test update and turbulent runs great, Neels the man.

Is this with the version released some days ago?
Anything special/different how you ran your sims compared to before?

I'm still not getting what is supposed to be the most accurate way to do the sims or how to interpret the results when comparing older and newer runs.

One turbulent run to generate a turb file.
Very much deto.
Run the newly generated file in prescribed.
No deto.
Different results then from before the latest update.

I think it's the same as others have reported.
But is the way described above still "the one to be trusted for now"?

[adegnes]

Is this with the version released some days ago?
Anything special/different how you ran your sims compared to before?

I'm still not getting what is supposed to be the most accurate way to do the sims or how to interpret the results when comparing older and newer runs.

One turbulent run to generate a turb file.
Very much deto.
Run the newly generated file in prescribed.
No deto.
Different results then from before the latest update.

I think it's the same as others have reported.
But is the way described above still "the one to be trusted for now"?

I asked Neels the same question, and what you describe(run turbulent, create combustion file, use file in prescribed mode from then on) is the best way to go about things for now(and the best way anyways, if you want to save time).
He is of course on the case!