Ignition Timing

[cicone]

Jaguar---Dude, you are out of control!!! Wish I could pull off that sort of thing. Really looking forward to you posting the timing curve---as it looks like there is a never ending debate regarding adv/ret and it's effects. FWIW, a two stroke is NOT the same as a four stroke. The static timing has little to do with the dynamic timing of the CDI. With the "modern" two stroke there are some counterintuitive constructs that effect ignition timing. Some of those include---the effect of port timing vs. valves---the ram effect of the pipe---and and the lowering of compression at high RPM(unlike a four stroke), which creates a combustible mixture with a different density and different burning rate. THERE---hope that clears things up! :eek:

 

[canyncarvr]

Jag: Is the CDI not an epoxied/potted unit? Are these pieces surface mounted?

So, if it was retarding the spark 2.7º @ 3KRPM (and presumeably a progression from that point) it is now (with the 5.7uf) and stator plate change doing...what?

You're running more initial advance (stator plate change) and starting the retard higher/later and that rate of retard is slower?

Your 2.5mm stator plate move (.100") got you what..about 4º?

What is the overall performance impact..bottom to top?

BTW...I get a 'page not found' on your geocities links..

 

[cicone]

Jag: Is the CDI not an epoxied/potted unit? Are these pieces surface mounted?
BTW...I get a 'page not found' on your geocities links..

try truncating the url at .html

 

[canyncarvr]

Thanks!

I didn't realize the links were caps sensitive too.

..and the point of looking...to see the 4.7 as an electrolytic.

Isn't it amazing what some folks know? ..and can accomplish.

Hats off to you, Jaguar!

I can't resist...........
To the folks that think the 'green room' is a waste of time..that it is not much more than misinformation..I'll bet you will find nothing close to the information available on this board anywhere else.

What is far more common other places is the senseless BS you find similar to this . It started out great but turned into pages of tripe. If 'they' enjoy it that's fine.

I've asked simple questions more'n once on other boards to get zip.point from anyone.

Point is, this (Jag's) sort of post is what makes this board of more value by far than most.

Well...it founders from time to time...but you can find some good stuff here.

 

[jaguar]

timing curve

Seeing that it was designed as an enduro bike they probably made the curve so that the spark timing hinders top end performance. Now that Ive changed the curve, I have both low end and top end power! Low end power is the same, good, but the top end really woke up.
Canyon wrote:
---Is the CDI not an epoxied/potted unit? Are these pieces surface mounted?

sealed with silicone with stones mixed in to make it easier to dissaemble. regular components not surface mounted (super small).

----So, if it was retarding the spark 2.7º [at] 3KRPM (and presumeably a progression from that point) it is now (with the 5.7uf) and stator plate change doing...what?

Changing the timing cap moved the whole curve 2.7degrees later. so the plate had to be advanced the same amount to keep the low rev timing the same.

---You're running more initial advance (stator plate change) and starting the retard higher/later and that rate of retard is slower?

---Your 2.5mm stator plate move (.100") got you what..about 4º?

no, I guesstimate 2.7degrees

---What is the overall performance impact..bottom to top?

 

[G. Gearloose]

Changing the timing cap moved the whole curve 2.7degrees later. So the plate had to be advanced the same amount to keep the low rev timing the same.

Great work..
Do you mean low, mid, and above 6K all moved 2.7 degrees, or just everything below 6K. I suspect the latter..., no?

 

[jaguar]

I believe the whole range is moved since all timing is affected by that capacitor.
But my 2.7 degree estimation is just that. I need to verify everything with a timing light.
Actually, just below 3000rpm (that I need to verify also) the timing curve starts to have its second curve for high rpm, so my bench test at 3000rpm isn't completely valid.
Once I have the timing light I plan on removing the extra 1uf capacitor and setting the stator plate at its standard position. Then use the timing light to see what timing it has at some rpm, say somewhere steady between 1000 and 2000. Then add the 1uf capacitor and recheck the timing with the light. That will tell me exactly what the shift between the two is.
Then I will have to note the timing at various rpm's in order to make a graph for the two CDI's (std and modified).
That graph should validate or discredit my seat-of-the-pants analysis.
Since that timing capacitors solder points are close to the top of the CDI box, it should be feasible to cut off a section of the box, clean off any silicone there, and then solder onto two described points a 1uf/10v capacitor. Then seal the cap and everything that was uncovered with silicone sealant. I plan on making a drawing showing what part of the box to cut off and where to solder and how to place the capacitor with the correct polarity (not reverse which would blow the cap). That would make the job a lot easier than a full CDI repair and feasible for the average Joe. Then we can have more input on the subject than just from me. What is unknown is whether the compression/porting/gas desires a spark timing curve to match, or whether all KDX's equally benefit.
:-)

 

[G. Gearloose]

please let us know

Also beware it is possible to get almost +/- 1 degree of timing error every time the flywheel is installed, depending upon how you do it, the keyway slot can permit some variation in alignment before cinching it down.

So for the most scientific evaluation of moving the stator plate, I suspect it is critical to use the same technique (twist the flywheel in one direction against the keyway while tightening) same direction every time.

 

[jaguar]

Today I played with different stator plate settings and concluded that the 2.5mm CW position was the best. I should be getting the timing light around the 24th. stay tuned!

 

[jaguar]

An interesting swist to the story:
I had a rewound stator coil on my bike up until yesterday when I installed a stator coil from Elextrex in England. The OEM coil should have 304 – 456 ohms according to the manual. My locally rewound coil only has around 250 ohms. The Electrex coil has 400 ohms. I think the rewound coil has wire that’s slightly thicker than OEM which would account for its lower resistance although the coil body is the same size as original.
When I installed the Electrex coil my bike lost some of the top end power, a fair amount. It just wasn’t impressive anymore. So I reinstalled the rewound coil and shazam, the magic was back! After thinking about it I believe the Elextrex coil (which I assume is equal to OEM) lacked current delivery ability at high revs due to its very fine wire. And the rewound coil is able to better maintain the >300 volts going to the CDI at high revs due to its thicker wire. It could be that a delayed spark at high rpm is of little use without a strong spark to quickly ignite the gas/air mixture. So in effect, this “fix” could be dependent on two things; a more aggressive timing curve, and a stronger spark at high rpm’s.
I looked at the voltage waveform with my oscilloscope on the timing capacitor with the two stator coils. The Electrex coil produced a peak –4 volts at idle. The rewound coil produced –3.2 volts peak. That would make sense if the wire used was a slightly thicker gauge which would limit the number of “turns” of the wire around the body which would limit the voltage since voltage out is directly dependent on number of turns. But the thicker gauge would be able to deliver more current to the main capacitor which means it would have less of a hard time fully charging it in the small fraction of time it has to do so at high revs.
But of course I need to evaluate the two timing curves (with the two stator coils) to see if there is enough of a difference to account for the power difference.

 

[cicone]

But of course I need to evaluate the two timing curves (with the two stator coils) to see if there is enough of a difference to account for the power difference.

Jaguar----I'm VERY interested in your findings regarding timing comparison of the coils. When I replaced my stock stator coil with the Electrex C31, my bike started kicking back at me while starting it---thinking the coil was causing a spark advance. I talked to Electrex about it and couldn't get an answer from them. Guess I'll have to break out the timing light. Please do report if you note a difference---I think it could end up being a couple of degrees! :worship:

 

[jaguar]

hmm, I was thinking they were selling a coil thats the same as the original. Could be that it does put out more voltage which would cause the timing to be more advanced which would give it more of a tendency to kickback when kickstarting.
that means now I have to obtain an OEM stator coil to be able to fully research this subject.
I will look on EBbay

 

[tedkxkdx]

This is very interesting stuff and since Jag is getting techy on us I would just like to add that
An Electrex, ENGLAND purchased, coil and CDI combo had arching while I put an icat on it. The stock system did not arch with icat on and so I don't use the combo unit any more.
Also I found the kickback problem on several eletrex stator coils. It is very annoying.
Would the thicker wire on the coil help prevent shorts from occuring?

 

[jaguar]

Probably what was happening is that the higher voltage from the Electrex stator coil was making the voltage spike out of the CDI and to the ignition coil primary higher, which would make the output voltage from the ignition coil higher which could cause arcing. Did you detect where the arcing was occuring?
side question: did the Electrex CDI come with selectable timing curves?
If the stator coil was wound with thicker wire then it would generate less voltage which would prevent the arcing problem you mentioned. (but its able to maintain the voltage better at high rpm's)

 

[tedkxkdx]

The arcing was all along the ignition wire. I did note that the wire was not rubber but some cheesy plastic, so that could be a consideration too.
attached is the HT-CDI which fits almost all jap 2T bikes. I don't know why they don't sell it in the USA. The label on the outside of the plastic wrap says Italy on it so maybe electrex just buys the unit from there or the the outer sheath is from Italy. The electrex guy would not say. The unit is less than $80 so it is much cheaper than a oem CDI and works well in a pinch if you break your cdi or ign coil.
http://www.electrexworld.co.uk/

 

[canyncarvr]

re: thicker wire would prevent arcing.

Maybe. Maybe not. It depends on considerably more than that single issue.

It is true that more wraps means more voltage (capability) and larger wire means more current (capability). The ignition parts and pieces as a system need to be considered if there is a breakdown/arcing problem.

BTW, my answer to the 'thicker wire...prevent shorts' question is, 'No, it will not.'

But that would come from understanding a short to be a low-resistance connection across a voltage source. That is the definition of a short circuit.

Arcing is a different matter altogether. When technologies changed coil activation from negative (grounding a saturated coil) to positive (driving spikes/waveforms/pulses into a coil) arcing came to be considerably more prevalent.

It's still true (last I checked ;) ) that voltage will take the path of least resistance. If for whatever reason a charge is allowed to build past the point of which the system can contain it, it will find someplace to 'get out.' Normally that happens every time a plug fires because the plug is the path of least resistance. Obviously the charge will go someplace else if the plug is not that path.

 

[jaguar]

Well, I received the timing light and put it to use on my KDX. What I saw is that the timing progressively advances from idle all the way to maximum advance around 3600rpm and then starts retarding (going more toward TDC) more and more with higher revs. The following data is from my KDX with a rewound stator coil which affects the timing, but probably the same amount of degrees all the way thru the rpm range. I will be receiving a used coil to test in a couple weeks. But anyway this test does show the relative effects of timing cap change on the timing curve.
4.7uf is the standard value of the timing capacitor in the CDI. I added .5uf and 1uf to it for the following 3 values of timing cap. The readings (in degrees BTDC) under SR are from a standard CDI for my international version of street-ready KDX which uses a 4.7uf cap but a lower value of thermistor (which is in series with the cap).
rpm 4.7uf 5.2uf 5.7uf SR
-----------------------------------------
idle 19 19 19 17
3600 23 21.5 21 21
6000 19 19 18 19
Ideally I should of tested the timing at 9000rpm also but I didn't have the timing light that also reads the rpm. Instead I used a circuit and then a frequency meter which I had to the side and didn't want to hold the revs so high long enough to get a reading which was more complicated from having to look at two places. Also it was hard to rev it to 9000 and keep it there.
You can see from the results that the biggest change was midrange although from experience either of the modified CDI's gives me better midrange and top end. I was practicing a couple weekends ago with the 5.7uf CDI and the bike was running hot and so I raised the needle 1 clip position to richen the mixture a bit. After that I experimented with using 5.2uf and on street runs I felt it gave me equal results and probably doesn't let it run as hot on the MX track.

 

[cicone]

Jaguar---I'm surprised at the numbers. This is from the tech tips section under "KIPS power valve": The CDI advances the ignition from about 6 ° BTDC at idle, to 21° BTDC at 6000rpm, allowing the flame front sufficient time to travel across the chamber, producing the greatest possible pressure rise. The CDI and KIPS Valves work together to produce the best possible low-end performance.
Hmmmmmmmmmmmmmmmmmmmmmm. :think:

 

[jaguar]

What was said is true, except they left out that there is a curve in between those two points.
Basically the timing advances linearly from 0 up to around 2000rpm, where the timing cap kicks in and starts to lessen the linear advancement. The greater the value of the timing cap, the sooner the departure from the linear advancement. I calculate a difference of 300rpm for every .5uf. I have drawn a graph but I need to go to town and scan it and then post it here and on my web site. Hopefully by this evening it will be viewable.

 

[jaguar]

timing curves

Here's two graphs of timing curves, one std and one with 1uf added to the timing capacitor in the CDI. Theoretically the two lines would come together at a very high rpm. The curve for 5.2uf would fall between these two curves.

 

[jaguar]

My international model (SR) was designed to have 19BTDC as the timing at 6000 rpm. I think all old KDX's are that way (probably pre-'94). So if you have a modern KDX with 21BTDC timing then you shouldn't advance the stator plate at all with this modification. That would keep the timing at 21 at 6000 rpm with this mod. I think anything more advanced would make it run too hot.

 

[tedkxkdx]

Jag and cicone, I had my electrex stator coil just bun out on me after reading this thread. funny how things like this happen. I went to a local rewinder and the electrex wire was .112 and I had him wind it up with .125 and am getting 315 ohms. I have yet to ride it but it was a bear to start since the timing is off as cicone spoke of and it is kicking back something fierce.
So how do we get to a non-kick back situation and I am willing to lose top end performance for easy starting?
I really wonder what the oem wire thickness is?
I have another coil to be rewound and I think I will try thinner than .112 maybe? Thoughts.

 

[jaguar]

If anything, you should try even thicker wire on the second coil. You just need to turn the stator plate CCW to retard it to offset the advance that you have which is evident by the kickback. It won't kill your top end power. If it still has a bit of a tendency to kickback and it is running hot then you'll need to turn the plate a little more CCW.

 

[jaguar]

latest graph

Below is a graph showing my current setup using the 5.2uf cap with the stator plate advanced 1mm (CW), and of course the standard timing curve. This shows that the modified timing is more advanced at low rpm's and less advanced in the lower midrange and more advanced in the upper midrange to maximum rpm's.

 

[cicone]

Have been talking to Ritzo over at Electrex lately and expect to have some information regarding the dreaded stator coil kick back issue experienced by so many kdx owners after installing the C31 Electrex replacement ignition stator coil. I'm hoping this will help many of us out. I will post his response soon, as he related that he is awaiting feedback from certain sources. No, I haven't actually checked my timing, but this weekend looks good for that. :cool: