We have ignition

[blitz11]

After 6 months of work on nights and weekends, I finally have a CDI with which i am happy.

I had it done last week, but I blew up an ignition coil. I was testing at 11,000 rpm, where i am able to deliver 350 volts to the coil (the stock delivers ~200 at that speed). After about 10 minutes, the coil blew up. Too much voltage/energy.

So, a new change, and I now limiting voltage to 300 volts. I've been running at 11,000 RPM for an hour and the system is rock solid.

The ignition is in a breadboard stage now. After I help an 8'th grader with his go-kart project (I'm his mentor), I am going to solder up a proto board, and get that running.

I am thinking of writing a semi soup-to-nuts theory of operation on ignition systems, from the rotor, stator, pulsar pick up, and CDI fundamentals. I am also going to have some circuit boards made; 3 for my bikes, and two (or more) for members here who are willing to test. GMP and Kenndunn have expressed interest in the project; I'd be willing to send them each a test CDI if they're game. I'd ask them to test it on their bikes so I can gather some durability information. (It runs cool on the bench even at 11,000 RPM, but it's not yet been in a bike.)

The ignition is fully programmable. I need to build a library to simplify the process of building your own advance curve. That will greatly simplify that process.

Once I have the protoboards done, I need to specify connectors for the USB (for programming), for the map switch, and for the connection to the bike. This will take me a week to specify the connectors, and find a suitable switch for the map, and an enclosure.

When I started this, i thought i'd just replicate designs which are on the web. To be honest, NONE of them work in this application. However, as I worked through this, i hit on a design i'm really proud of. Looking at what Vortex gets for their programmable ignitions, I am actually thinking of starting a business to build and sell these things. Might as well get something out of my work. I know that this performs better than the CDI that came on my bike; programmable, much higher spark energy/voltage, and multi-map.

(I already have a job, so I am looking for something to do for fun into the winters. My day job is described here: www.bus2.me.umn.edu .)

Anyway, if people want a full theory of ignitions, I'll start writing one as I document my work. I won't go into my final circuit design as I have not seen anything like this, and I think it offers me a competitive edge on what else is out there. I'll look for some feedback here.

I also promised LEDs for diagnostics on this. I have changed my mind. What I'll offer is a small diagnostic box with LEDs which you can connect to the stator wires, and it will tell you whether your alternator is putting out sufficient juice, and whether your pulsar is working well. This is a much cleaner approach.

This would work on all bikes. Might be handy, knowing what I know now.

Basically, if you have good signals, your CDI is toast. I blew up a few of mine working this out.

I have a few photos and a video of the ignition winding down from 11,000 RPM (it shows the advance go from 0 degrees at 11,000, up to 30 degrees at 3000 RPM, and back to zero at ~2000 RPM. (The 30 degree number is exaggerated for demonstrative purposes).

I'll post those once i figure it out. (If anyone has picture/.avi file ideas, I am all ears.) I am not sure of the best way to do that. If people want at theory of options, please let me know. (IT might be a couple of weeks until i finish that.)

Thanks for reading.

blitz

 

[widebear]

quote (The ignition is fully programmable. I need to build a library to simplify the process of building your own advance curve. That will greatly simplify that process).

To plan on using a dyno to quantify baseline settings and establish parrameters? Sound like in the wrong hands it could possibly, you know.

 

[blitz11]

I have the baseline from running the stock ignition on my bench. (I'll document that in the write up.) We have a friend with a dealership with a dyno. We can document that as well.

Ignition curves for two-strokes are pretty well documented. People don't have to program their own; the tools could allow only a small change (or levels of changes) based on how far people want to take it.

The other idea is to make the standard ignition fixed (middle switch setting), then allow people to vary the "mild" and "wild" settings.

Good points. Thanks for those.

blitz

 

[Neil E.]

Very interesting subject. I'd like to hear more about it.

Basically, if you have good signals, your CDI is toast. I blew up a few of mine working this out.

Does this refer to the "process of elimination testing" that most of us would have to do to find problems?

 

[GMP]

Blitz,

I am envious, congratulations on your work! I have lots of questions, but I'll let you provide only the information you want first.

 

[blitz11]

I uploaded some pictures and video to photobucket. I've not done this before, so let's see how it works.

1. Here are the circuit breadboards on my bench. One is the low voltage supply, one is the high voltage supply, and the center board is the signal processing (timing) circuit.

http://i1229.photobucket.com/albums/ee479/blitz111/CircuitBoards-Copy.jpg

2. Here is my test rig. I have a Taig mini-mill, and a Taig mini-lathe. The lathe motor spins the right way, but only will spin the ignition to 4000 rpm. Using the mill motor, i can spin the system to 11,000 rpm. With my design, i don't care whether the rotor spins forward or backward...South pole rotating clockwise creates the same pulse shape as the north pole rotating counter clockwise.

The stock ignition won't work that way.

http://i1229.photobucket.com/albums/ee479/blitz111/TestRig-Copy.jpg

3. Here's the coil i toasted when running at 350 volts at 11,000 RPM. Surprisingly, it would still spark in this condition. Not very brightly, but it would still spark.

http://i1229.photobucket.com/albums/ee479/blitz111/ToastedIgnition-Copy.jpg

4. Here's just a screen shot of the timing and ignition signals. In the write up, i'll document this in more detail. The red peaks are 150V (i use a voltage divider), so the voltage into the coil is now 300 volts.

http://i1229.photobucket.com/albums/ee479/blitz111/Plot-Copy.jpg

In the video, when (if) you watch my timing mark, it looks like it jumps. It looks wrong, but when you do this based on timing (and not "bins of rpm"), it appears to jump. i could average speeds and time off of that, but I think that produces less precise timing. When you watch this on the scope, it is dead nuts. Anyway, this is all timing based, no averaging, at this point. It hasn't been in a bike yet, but i may change this once i ride.

The jumping comes from my test rig. IT uses a v-belt, and the v-belt sets up a standing mode. The motor has much more inertial than the ignition rotor, so it exhibits some torsionals. Anyway, this might change once it is in the bike, but my gut tells me it will be more precise this way.

I've got 16 MHz of 8 bit CPU (for $5), so I might as well use it.

5. Now, video.

A. sparks at 6000 RPM:
http://s1229.photobucket.com/albums/ee479/blitz111/?action=view&current=Sparks-Copy.mp4

B. 11000 RPM wind-down:
http://s1229.photobucket.com/albums/ee479/blitz111/?action=view&current=blitz11000RPM-Copy.mp4

(I've gone to 30 degrees of advance here...stock uses 18 or so max.)

C. My ignition at 4000 rpm, then winding down. Note relative quiet to stock, and how much longer it takes to wind down. This is the manifestation that the stock unit appears to have a resonance or dead short. I only have this as a reference, so maybe my cdi box is bad.

http://s1229.photobucket.com/albums...on=view&current=BlitzIgnition4000RPM-Copy.mp4

D. Stock winding down from 4000:
http://s1229.photobucket.com/albums/ee479/blitz111/?action=view&current=kukosan4000RPM-Copy.mp4


Anyway, now i have to build a circuit board, solder it up, find connectors, then put it in the bike.

 

[Keg]

Well done! That seems like a fair bit of trouble you have gone too. Keep us all up to date with it.

 

[waycrazy]

I am jealous of people that can do things like this. Your way smarter then me . I would end up making a huge tazer by accident and zap my self to death! I would think there would be money to be made in this though. There are not a ton of GasGas Mo better parts out there. And the GasGas bikes are getting more popular by the day. People love gizmo's they can hook to there laptops. So I would think it would be a hit!

 

[GMP]

It would potentially be not just a GasGas accessory, but for any bike that runs a Kokusan.

 

[widebear]

Thats very cool. Seeing the nature of advance under the strobe. Makes me want to get some of that special tape that car guys use on the harmonic ballancers.

 

[blitz11]

Thanks for the feedback.

My motivation was the huge expense of the CDI boxes. we have had friends with a rash of CDI failures. The price seems to have dropped recently. KTM quotes about $280 for a Kukosan box, whereas GG seems to get about $325 (see gofaster's website).

I recall a $427 charge last time I looked.

Vortex charges something like $500 for the CDI, then charges another $500 for their programming interface. Ouch! See

http://www.vortexcdi.com/model_product.php?product=1&model=94&mp_id=289

I figure that there are a lot of guys out there who face ignition difficulties, and may want another (less expensive) alternative. At the very least, I understand this better. Hopefully, my forthcoming write up takes the mystery out of this. Once you understand the basics, you'll be in a better place to diagnose your own problems.

I've spend about $500 developing this, so it's about a wash in hardware costs.

I hate when I don't understand things, and I hate it more when things break. Hopefully, we'll kill 2 birds with one stone here.

Thanks for following and comments.

blitz

 

[Fred1956]

I am jealous of people that can do things like this. Your way smarter then me . I would end up making a huge tazer by accident and zap my self to death! I would think there would be money to be made in this though. There are not a ton of GasGas Mo better parts out there. And the GasGas bikes are getting more popular by the day. People love gizmo's they can hook to there laptops. So I would think it would be a hit!

+1 I think that stuff is amazing & super freakin' cool! I wish I had the time to work on things like that. Awesome stuff at the least.....Thanks for the updates and please keep us posted. WOW !

 

[Swervin]

ignition progress

How are you coming with this project?
Curious if you have any working prototypes yet?

 

[blitz11]

I did this in early 2011. Then my real job became very busy, including a trip to Alaska to install head up display systems in Snowplows on the Thompson pass. Busiest 5 weeks of my life.

(You can see our stuff on Hulu from the speed channel: http://www.hulu.com/watch/139658 It was also on the Discovery channel, but they took it off of the web.)

Then my real job blew up, starting in August 2011. (I have to write a book about this...)

I spent a year looking for a new gig which would be close to what I was doing (and loved to do); I started that gig in September of this year. My wife's sister had a stroke, so we had to sell her house in Oregon, move her to Minnesota, and do the whole assisted living deal.

So, I tabled it. It rattles in the back of my head (I have a new idea for a different topology which would improve efficiency even more), but I haven't worked on it since April 2011.

It was fun as hell, but when i did the numbers, it would have been tough to make any money on this deal. The job search in 2011-2012 sort of sucked the life out of me, and now I am trying to carve my niche' in my new job. It's been ok (it only took three weeks to file my first patent application there), and hopefully in a year, I'll be able to show you something way cooler than the ignition (or the Head Up Display stuff). I just have to convince a customer to go ahead with it.

Thanks for asking. I had big plans, but life got in the way.

 

[Swervin]

Ignition

Works that way most of the time!
I hope everything comes up Aces for you.

 

[GMP]

Craig, good to hear from you, very sorry for your troubles. Hope things are looking up, and keep in touch.