Coils in the caps

[Ruby Racing]

Has anybody got rid of the coils on the FZS600 and replaced them with plug caps with the coils in them?


If so can you tell me which model you got the coil/caps from please? Fancy doing this mod on my project bike.

[schlumpf]

I did that. No FZS600 but on a FZS1000, by using COPs from R1. Maybe that gives you some informations anyhow.

Electrically it's tricky. The COPs have about have the primary resistance of the original coil. And as you have to put each 2 in parallel you end up with a very low effective resistance. That might blow your ignition box. I solved it by making a interstage box with some power transistors. So check the rated values of your ignition box/coils with the values of your selected COPs.

(COP=coil-on-plug)

[Fazerider]

Superfour Tom did it a while back.
He too was concerned about the resistance and "solved" it by putting the coils in series.
If I was going to try it I'd make some inductance measurements first, I suspect the resistance may not be nearly as important.

[schlumpf]

I did a lot of measurements on my single COPs. Best way is to run them in a single configuration with separate driver. Even then it's not optimal as they are designed for a single coil system that only fires every second rotation. 

Putting the coils in series makes it much worse than original. You'd end up with less than 1/2 of the original spark energy. That still works most of the time. But at higher revs you risk dropouts.

[Fazerider]

I suppose that without knowing what the current capability is of the output transistors in the existing ignition unit, buffering the output as you've done is the safest way to do it. Tom seemed unaware that he was running with feebler sparks than before, but it worked so he was happy.

[Ruby Racing]

Thanks for your thoughts guys. I have a "how to convert to stick coils" from a magazine so I'll be following that when I do it.


Sounds like nobody has done this yet for the FZS600 then?


I have a mate who is a top sparks, so I'll be enlisting his help when I do this. Will let you know how I get on, although it's some way down the modifications route yet.

[midden]

What are the expected benefits to be had from the conversion?

[Ruby Racing]

The reason I want to do it is to free up space under the tank for my project Fazer.


According to the article:


"Coils and their HT leads tire with age and can be prone to water ingress.

Each cylinder has its own coil, so the tens of thousands of volts produced is delivered to a single plug instead of two and travels a shorter distance. It's a more efficient, stronger system and fitted properly should be more reliable than original parts."


You also need to get hold of a sub loom to attach to the stick coils as most Jap bikes and Triumphs use Denso sticks and you need the correct block connectors. No mention is made in the article about the issues you guys have raised, but I do seem to remember reading somewhere else about checking their output is compatible with the existing set up.

[midden]

   cheers

[keratos]

I dont see what the benefits are , certainly from an electrical point of view, indeed ... this may reduce performance and disrupt the ignitor unit which relies on a predefined resistance in the coils.
There are two coils , one per 2 cylinders. Each one presents a load within the circuit by means of a resistance that is factored into the math that the ignitor unit firmware uses to program the High voltage circuit
By putting coils into each plug you are disrupting the electrical circuit
I dont know what you are fitting but what you need to do is measure , using a good multimeter (I use Fluke), the Ohms resistance of each coil. Then, measure the resistance of each COP. you then need to ensure that the resistance for each plug is the same be it served by COP or by the stock coil pack.
As far as I am concerned, there is no benefit whatsoever of fitting COP - fact. Indeed what you may do is foobar the entire ignition system and at best, make it perform underpar.
If you insist on your "Project Fazer" then good luck and dont be surprised if things don't work out

[schlumpf]

You are fully right. Changing the iginition without some knowledge might end badly.

Luckily the FZS ignition is not that advanced as in modern car engines. It is only a simple electronic switch with a timing map on top. The move towards single coils came with the demand for sufficient spark power with increasing revs and lean fuel mixtures.

Here are some pictures about what I'm talking.
First is taken with the original FZS1000 coils.
Second one is taken with R1 coils and additional driver stage. You see the advantage at highest revs. You might make use of the better spark in case you have modified the carbs an/or exhaust. Gives you some more room with the fuel air ratio.
For stock configuration I don't see any significant benefit. But would be worth a thought if you are anyway on the way to change coils because of a defect. Failures in coils are tricky to detect. A new one gives you confidence.

[keratos]

All those charts tell us is that in isolation the spark is more pronounced
It does not give the full spectrum of effects when combined with extant electronics and dies not give any indication as to the overall circuit performance. As I said, there is no perceivable difference to the'normal' rider, moreover one could damage or impair the performance...OVERALL

[Fazerider]

Interesting, at least for the geeks amongst us.
The mod clearly shows the advantage of the smaller inductance at high revs.
I'm puzzled by the variation in dwell angle though: ignoring the labelling errors, it looks as though the ECU halves the dwell at 1000rpm compared to 700rpm... increasing the dwell angle at higher speeds makes sense as the current needs time to build, but why increase it at lower speeds?

[keratos]

do you mean charge time or the actual angular curve? the dwell angle looks - almost - the same at 700/1000, as it should be; but the charge time is reduced, as you'd expect at higher RPM so I dont see the issue ???

Anyway I dont know what box this guy is using or where he got the labels from?

as an aside, who cares - its a wasted mod for the UK road rider. track days perhaps, but on a detuned FZS600 road riding at 20,30,50,80 mph whatever - who cares

[Fazerider]

Agreed it's not an essential mod, but the dwell angle oddity may be common to our machines (I was going by the waveforms rather than his labels, that "12ms" looks more like 24).

[Ruby Racing]

Appreciate the technical warnings chaps. Don't worry, I'll be getting a grown up to help me with this.


As I mentioned I'm only thinking of doing this to free up space under the fuel tank, as I'll need to relocate stuff from under the seat once I start changing the shape of the bike. I'm not after any particular performance gain as I don't envisage modifying the motor other than a jet kit for the carbs to work with pods.

[keratos]

Agreed it's not an essential mod, but the dwell angle oddity may be common to our machines (I was going by the waveforms rather than his labels, that "12ms" looks more like 24).

Ah.OK. that case ... agreed
The spark at 10k is held for quite some time on the mod. Not bad. But my riding on the commute is more like 3 to 4k

[schlumpf]

Sorry, the related explaining description is missing. I just picked the pictures out of the long report.
The ms in the charts indicate the time the coil needs for a full charge. Afterwards the current goes into saturation. The static current is given by the ohmic resistance. The smaller the inductance the quicker the charging. The possible spark power is proportional to the inductance and the current.
Yes, geek stuff.

My aim was to offer my support and share my findings if anyone wants to go this route, for whatever reasons. Not to convince that this is an essential mod. But it should be done the right way.

[keratos]

edit...
Aside,  does the the ms durations not represents (a) - flatline duration  -is the idle time, and (b)the curve - is the dwell time. the idle time may potentially be of a longer duration than that required for electricity to develop to max capacity (notwithstanding self-inductance) through induction, to a given capacity ,  potentially. Getting technical  ....

[schlumpf]

Sorry for the word "static", forgive a non-native speaker. "Settled" is a better expression.

Referring to the very first graph, original FZS1000 configuration, taken at 700 RPM:
The given voltage is taken accross a small sense resistor, looped into the ground line of the ignition coil (lacking a current clamp). The red labled peak current is calculated from this sense voltage.
The max current settles at 2.6A. It drops to almost zero when the ignition occurs.
Dwell time is about 22ms.
After about 12ms the current has reached its peak. Means, it takes 12ms for the coil to charge completely.
Going to higher revs the ECU decreases the dwell time more and more. This is due to the limited available time. At 10000 rpm the crankshaft needs 6ms for one cycle.

Next picture, first graph, with F6T558 coil, additional driver stage, at 1000 RPM:
This coil has lower inductance and lower resistance (~8mH, 1.3 ohm) than the original (~14mH, 2.2 ohm).
It is fully charged after about 3ms.
Even at 10000 rpm it can almost be fully charged.

With this configuration you get a copy of what's used in the FZ1 (or many other single coil ignitions). Only difference is that single coil systems only fire at each second cycle, allowing much longer dwell time and with this more backup for highest revs.


I don't see the problems in copying an existing system. And my pratical results are all fine, running on 2 bikes and since several thousands miles.
Do you have more informations about the FZ1 ECU's insides?



 

[keratos]

And..?

[Fazerider]

And..?

… and it's a simple, robust solution for anyone wanting to convert to stick coils.


Shlumpf: you might have a market for your interface box, not everyone has sufficient knowledge of electronics to make one.

[keratos]

...no perceivable  benefit to riders like me and riders I know. Were all scratching our head. Hats of to the guy for his engineering prowess and Well done with the' prototype'. And good luck with marketing !