Some numbers needed

[the-elf]

from experience its the standard ignition that holds back a highly tuned KR1s engine. Getting the ignition optinized to the tune, expansions etc is hard work but worth it.

[TwoStroke Institute]

I'm intersted of Zeeltronic ignition (CDI-RZ1,VCDI-04,PPV-KR+handheld programmer).

I doubt the standard ignition will be any cause for concern, as it will be gathering dust somewhere

[mj43]

....I'm interested of Zeeltronic ignition (CDI-RZ1,VCDI-04,PPV-KR+handheld programmer). What is the real advantage with that and do you have any objective tests made (dyno curve) comparing the stock ignition? I remenber that someone had said that there is no huge increase in power or driveability for stock bike.....

I am not aware that anybody has done any tests and posted results. Until something appears from a reliable source I would treat everything as rumour and speculation.

I have seen some good results posted for the VJ22 engine. I can't find the link but the best I have seen was a series of tests from Holland on an Aprilia RS250 (think it was the result of about 180 dyno runs!!!)

Becareful if you start messing with ignition curves especially if you start pushing advance as you may stray into detonation territory.

On ignition curves I couldn't tell you what a stock KR curve is like. I know various RGV boxes have been measured but I am not aware that a stock KR box has been mapped. As a consequence where do you start with your programmable?

[kr1stian]

My intress for Zeeltronic rose after hype I have read here in this forum. Sure I am very very keen of all boost witch has good prize/throttle response factor . I am still thinking for buying one but is it worth if in MOTA I can only stretch modified powercurve few hundreds over stock ignition? Really, I dont know if this is anything to do with real Zeeltronic, but by with this map I got my still unborn cylinder and pipes work best so far.
1200/13.5 8200/25.0 9800/24.7 10200/22.2 10500/16.7 10800/15.5 12000/16.9 . Would it be worth with Z at all?
Thing I fount interesting since I got my first self made decent pipe tested in dyno (not official calibrated one but well built by my race orientated friend) I made about two years ago, I bought MOTA cause I wanted finally to compare that real curve against the simulated one. Those were matching suprisingly well and I don't think I had only roughly right input in the program.
By the way, it's sad that 250cc won't be seen anymore in GP2...


kr1stian

[TwoStroke Institute]

Setting up and ignition curve is really simple stuff, and there is nothing to fear. Just begin with something that works(the Zeeltronic should have a pre-programmed curve anyway) a TZ 250. A TZ will have peak power at 12,500 a KR-1 at 10,500. So just shift your curve 2000 rpm to the left. Ignition curves don't vary that much in on pipe/in phase rpms.





LINK

[scooble]

I have run lots of ignition sweeps before (all be it mostly on fuel inejcted development 4 stroke engines). It is fairly simple as long as you have a good dyno that can accurately control engine speed (not inertia dyno).
I would go up in steps of 500rpm at WOT (wide open throttle) and advance the ignition afew degrees at a time and watch the torque output. The point at which the maximum torque is first achieved for that RPM is called MBT (Minimum ignition advance for Best Torque). The compression for the engine may be limiting factor and you may run into detonation before reachine MBT, in which case you set the ignition at the best torque for least detonation called DBL (Detonation Border Line). I used to use special acclerometers bolted to the engine rigged up to headphones. DBL is subjective though. alternatively, I would use an in cylinder pressure transducer rigged up to an AVL pressure analysis application which would output the pressure data into a realtime graph giving a good visual indication of detonation - less subjective.
In order to keep detonation at bay, you had to retard the ignition, unfortunatley it would make the exhaust port and downpipe very hot as the mixture was continuing to burn as it exited the engine.
An alternative method was to add more fuel, sometimes as rich as 10:1 in order to keep the combustion chamber cooler.
When the ignition curve was optimised, I would then move onto AFR swings in order to optimise the fueling. I found most 4 stroke engines I worked on gave best torque at around 12.5:1 give or take, although I have never set up a 2 stroke on an engine dyno.
It was also noted that very high power ignition systems gave good combustion at very lean mixtures.
I found that factory ignition curves are quite often very conservative and quite a bit more can be eeked out of an engine with ignition optimisation, therfore, I would expect an optimised zeeltronic to be an improvement over a standard KR ignition module - although there is no evidence yet to prove this

[TwoStroke Institute]

Evidence is things like Fery Brouwer running Zeeltronics on his works Yamaha's, I wonder why?
What do you do if you only have a inertia dyno? Tell me more about the pressure transducers, I would love to look at these traces to see how closly they matched my simulated pressure/time histories.

[scooble]

in order to run any kind of sweep, the engine needs to held at a steady speed and control any fluctuations in power. My development work was always on engine dyno's not chassis ones, so the control system had direct control of the dyno speed, throttle position, air intake temp/pressure, ignition, fueling and exhuast back pressures all controlled by PID ( Proportion Integral Derivitive) controllers with manual over ride. Inertia dynos weren't able to hold the engine speed like water brake or eddy current dyno's.
The cylinder and exhaust pressure transducers we used were made by a company called Kistler and the incylinder data logger analysis was by a company called AVL - both of which are industry standard and hideously expensive. The software used for engine prediction was called WAVE - fairly close to real results but not exact. No matter what the 'boffins' came up with as engine management maps, all of them required 'real' optimisation. For a very unstable engine once the predictions were so poor that the engine actually blew itself up on the dyno with thermal run-away (Det).
Once the steady state conditions were optimised it was a case of knitting it all together under dynamic conditions, sudden changes in speed, load, boost and throttle position - this was kind of tricky, sometimes required some fiddling about of PID values between different maps.

[mj43]

Evidence is things like Fery Brouwer running Zeeltronics on his works Yamaha's, I wonder why?
What do you do if you only have a inertia dyno? Tell me more about the pressure transducers, I would love to look at these traces to see how closly they matched my simulated pressure/time histories.

Mmm... Zeeltronic a couple of hundred Euro's a blown cylinder probably 3 to 4 times that to repair if it takes the crank.
I am going to learn to walk before I run - only KR with a non standard ignition fitted that I know about has gone bang everytime it has run. Might not be ignition related but KR's don't go bang... or at least shouldn't.

Scooble very interesting - looked at special plugs that allowed cylinder pressure to be measured but decided that mortgaging the house to but one wasn't really sensible.

[TwoStroke Institute]

But what went 'bang'? As I said a hell of a lot of engines across a broad spectrum are running with Zeeltronics and Ignitechs now, none of them are going 'bang' at all.

Ah Kistler well I can forget about buying one of those.Knitting it all together is done on the sim as well the only thing I find needs to be optimised after the ports/ pipe/ignition etc is done is the fueling, which can only really be done in the field.

[scooble]

I was running a turbo charged GDI triple at about 18:1 AFR (non stratifieid), with a CR of about 13:1. The calibration was so bad, that the combustion process was so eratic that the control system attempted to take control of fluctuating dyno speed whilst also compensating with ignition and boost. The system got so out of phase so fast that I didn't have time to manually intervene, what resulted was the engine speed shot up, boost increased and ignition advanced all at the same time - result = very bad detonation. Upon realisation, I cut the ignition but it was too late as it was already doing a diesal and just ran away - BOOM!
Losts of smoke and the little thing decided to 'put a leg out of bed'. Not very popular with the customer.

We had a team of statisticians working on creating the maps with the aide of Matlab and various CFD software tools, however, from past experience, they didn't always work. its a bit like predicting the weather and you end up by sticking a wet finger in the air early on in the development process. However as time went on and more data was gahtered and 'crunched' more accurate maps could then be generated.
Once thats all done, its then opotimised on the road/test track in real world conditions.

[mj43]

I am not questioning the ignition systems I am warning that care should be taken before bolting one on. For most people fitting a different ignition system without knowing how the curve matches a stock curve is playing with fire.

To much advance may result in detonation, retard it and you could over heat. How do you set it up on a dyno - max power on a dyno may be right on the edge of failure. Without the tools that Scooble has talked about you may not know. With variable conditions on road/track, a change in the weather, different fuel and bang goes your engine.
I am just counselling caution - not, not using the ignitions (I am about to try one).

Before fitting my ignition I will hopefully run some tests on the stock ignition this weekend and if successful will post the curve. Ignition will be logged using a Pico oscilloscope with ch A logging the spark and ch B the trigger. From this I hope to be able to calculate advance and rpm - just testing the control software at the moment. As sample time of the scope is fast, it can grab a reading every 333nS, I am pretty certain I should be able to get fairly accurate results.

12000 rpm = 200 cycles/second.
1 degree resolution = 72,000 samples per second (forgotten his name - was it Nyquist - says you need to sample at twice the frequency) so that is 144,000 samples per second which is 7microseconds per sample. As system can sample at 20 times that rate resolution shouldn't be a problem. Hopefully more accurate than a strobe and a degree wheel.

I can then fit the programmable set it up with a stock curve and easily check that it matches the stock curve both on the oscilloscope and the dyno (I have heard stories of instability). From there I can start experimenting though I will have EGT fitted as it is one tool I can afford.

Edit - just checked up on the Nyquist sample rate. Got that bit right just massively oversampling but that isn't a problem

[KR-1R]

was that a M.E. or M.E. PHD?

[ScottaKR]

was that a M.E. or M.E. PHD?

Yeah, my brain hurts too...

[Sheik Yerbouti]

We used those kistler pressure transducers in injection moulds where I did my apprecticeship. I think they were about 3 grand each for the sensor itself ( I could be wrong there but it was a flippin lot ) and the connecting wire.
Anyway, whats wrong with temp sensors under the spark plug and hanging in the exhaust to help you guess when its gonna go bang?