CDI ignition system upgrades, thoughts?

[Trike_crazy]

So I've been thinking about this for a while and noticed that after all of the normal power upgrades like a bigger bore, stroker setup, carb, exhaust, cam etc... that one of the only parts I never see upgraded is the ignition system. Maybe a coil but never the cdi

My thoughts are more along the lines of ignition base timing and total advance. I know that there is some adjustment but I don't know if there is enough to make a difference. I know that regardless of the amount of advance you adjust into the pickup that the total amount of timing stays the same. Would there be any benefit to more total timing on a small engine like say a 350x? Or as small as 110? Lets say stock is 10° base timing with 25° total. Would 10° base and 30° total (example) make a large enough gain to make a difference? Or a custom timing curve? I know in the pursuit of power that people want every last ounce they can get. Has anyone experimented with stuff like this?

Basically I want to build a custom engine combination that I'm pretty sure the stock curve wouldn't be adequate for. But I'm not sure if using a cdi from another atv/atc would help either. I'm thinking of going from 125cc to about 190cc. So would a 185 cdi work? 200x? Xr200? I want all of my ducks in a row before I just start throwing parts at it. I'd hate to throw $350 at a programmable cdi but if there might be a noticeable difference or benefit I'd consider it. It also has a launch setting and can be switched between 2 different ignition curves if desired.

The Chinese racing CDI's are junk from what I've experienced. I mean they work but they really don't work as advertised. The reviews I've seen were they actually tested the base and overall timing say the racing cdi's are locked at 25° or so timing regardless of rpm and whereas a stock scooter cdi has 10° base with 25° total. And with that they seen no evidence of extra power or performance.

Sorry for the long winded post. It's just interesting stuff to me lol. https://uploads.tapatalk-cdn.com/20181031/47121bb3b2493016cad097b9bceeef80.jpg

Sent from my SM-G965U using Tapatalk

[Gabriel]

Ooh that MSD would be awesome.
That would HAVE to be mount conspicuously just because

[Trike_crazy]

Here is some good quick information on ignition timing on a larger engine.

Everyone gather around. It's time for today's mind blowing technical tip.

When we talk about ignition timing, such as 34 degrees before top dead center or BTDC, it means that your igniting the combustion charge 34 crank degrees before the piston reaches TDC.
Center being when the piston is at the dead top. Not before or it will push the piston back down the hole killing HP and killing the bottom end.
Well why are some engines 32 BTDC, and some 38 BTDC?
Because of the efficiency and burn rate of the combination chamber. Some combination chambers burn the charge sooner so it needs less led time to ignite the charge.
So your starting the charge 34 degrees before the piston is at the top so the time it reaches the top you are getting the maximum charge.
Well let's look at the piston speed. If the piston is moving at twice the speed at 6,000 rpm vs 3,000, why does it need the same timing? It dosen't!
The faster the piston speed the sooner you have to ignite the charge because the piston is going to reach top dead center sooner due to the speed travel of the piston. This is the purpose of the mechanical curve in a givin distributor. So ask yourself, is locking your distributor the correct way. I will say that we have found between 18-28 HP and ft pounds of torque through the entire curve on the dyno everytime.

Facebook link for credit

https://m.facebook.com/story.php?story_fbid=1854523094666597&id=621247104660875

Sent from my SM-G965U using Tapatalk

[Trike_crazy]

I'm taking it that no one else has ever thought about timing curves on a small engine when doing custom small engine build? Lol.

I know that there are a few places that make quality aftermarket cdi's for 350x's and 250r's but they are just generic curves that are a "one size fits all fix".

Sent from my SM-G965U using Tapatalk

[christph]

There is a guy in the Tecate group (Rubbersalt) that reverse engineered the Tecate CDI. He made the internal components more stout but I'm not sure he's got it 100% yet. I think CDIs are beyond the know-how of most people on this site--you have to have some background in electrical engineering. I've rewound ignition coils with success and got more spark energy out of them, but I wouldn't know how to design a CDI. If you go to HPI they sell ignitions for a lot of older machines and some of them have programmable curves. Here's the link. They are out of Italy but the Tecates guys have been using them with success.

https://www.hpi.be/

[Trike_crazy]

There is a guy in the Tecate group (Rubbersalt) that reverse engineered the Tecate CDI. He made the internal components more stout but I'm not sure he's got it 100% yet. I think CDIs are beyond the know-how of most people on this site--you have to have some background in electrical engineering. I've rewound ignition coils with success and got more spark energy out of them, but I wouldn't know how to design a CDI. If you go to HPI they sell ignitions for a lot of older machines and some of them have programmable curves. Here's the link. They are out of Italy but the Tecates guys have been using them with success.

https://www.hpi.be/I had seen where he was building those. I'll research their site a bit and see what I can find out. I'm not at the point of even having a running engine but I'd like to figured it out before that point comes so I'll be ready to make a decision. I appreciate the link. Thank you

Sent from my SM-G965U using Tapatalk

[Arky-X]

Interesting.......

I wouldn't know where to begin with one of these but I wouldn't know why there wouldn't be some improvement.

I have a tune in my Chevy 5.3 and I can tell a difference. It was not put on a dyno but I logged 3 different runs and emailed my tuner and he dialed it in. I don't have anything more than intake, exhaust, and transmission servo (and run on 91+ octane) and it makes a difference. One day I decided to put the stock tune in it and OMG......what a turd! Back to the custom tune.

Biggest advantage I would see is being able to move off a generic curve. Every engine is different and as far as I know, stock CDIs back in the day were not self-learning.

[Trike_crazy]

The more I think about it the more I want to buy it just to mess it one. The launch control feature would be nice for sitting on the line waiting for the flag to drop. You could hold the button while holding the throttle at the full throttle (or enough to be over your cutoff rpm) instead of feathering the throttle. I guess consistency would be easier in that regard. It wouldn't be as important for the type of riding I'd do but with the ability to change between 2 programmed ignition curves you could have one set for max power and another that pulls power out of you were having traction issues. I like to see one used back to back against a stock CDI, a hot shot or similar CDi and one of these MSD boxes with a tuned curve.

[ps2fixer]

The CDI's on these 3 wheelers are defo no self learning, there's no O2 Sensors or anything to give it any feedback, it just knows RPM and Timing, it doesn't even know if the engine is warm or cold.

I made a post with a bunch of stock CDI specs from service manuals. Newer machines have more advancement, aka poorer combustion chamber design? The post above describing timing is worded well to understand why more advancement is needed, but doesn't make sense to me why a newer machine needs more ignition timing both at the base and max.

Just to give some generic numbers, nearly all of the 4 stroke 3 wheeler engines have a base timing of 10 degrees BTDC. The full advance spec ranges 28-32 degrees. If you jump to a 93-2000 TRX300EX, it's base timing is 17 degrees and full advanced is 33 degrees. Newer machines quit posting full advancement specs from the few I looked up.

I don't have any way to read the curve or measure the advancements, but I'd be interested in seeing just how different the CDIs are for the Honda 3 wheelers. Seems most are more or less the same for the basic specs.

Also, the size of the engine doesn't make a massive difference, even 51cc chainsaws benefit from more timing advancement when being ported. The way chainsaw builders do it is remove the key out of the flywheel and manually adjust the timing base on the flywheel vs crank. I suspect 3 wheeler engines could be done the same way.

A programmable CDI would be neat to play with, but not really practicable for general riding, racing of course it makes sense. If I had one, I'd personally set the second settings to a "cold" setting and the main setting tuned for when it's warm/hot. Might be somewhat simple to add a temp switch in to switch between them depending how the programmable CDI functions.

The CDI circuit isn't overly complicated from the little I remember looking up about them, There's a couple parts you swap out to change the curve and advancement, so if you have a programming setting in mind, you could in theory build a test CDI by hand, then once you dial in the settings, have it made professionally and sold as a true performance CDI for x mods on x engine. I suspect the market is a bit too small to support that kind of dev work though.





Here's the CDI thread I mentioned with the service manual timing specs.

http://www.3wheelerworld.com/showthread.php/187179-ATC-Harness-Connector-Pinouts

[Arky-X]

Forgot about all the sensors that would be needed to maximize the results.....but it would be cool as H E double hockey sticks to play with.

[ps2fixer]

Forgot about all the sensors that would be needed to maximize the results.....but it would be cool as H E double hockey sticks to play with.

It's possible and could be adapted from a newer quad likely. Basically what you're talking about is EFI, or fuel injection. Like for car talk, the Toyota pickup from 1985-1988 had a carb option, however it still had an O2 sensor, temp sensor etc, but wasn't self learning because it had no way to adjust the carb electronically. The trucks that came with EFI of course were self learning and the ECU was much more complex since it did so much more.

EFI is kind of a double edged sword, it's awesome in the face with a programmable ECU you have a lot of control with timing, air fuel ratio etc, but there's so much more that can go wrong with sensors etc that could render the machine dead on the side of the trail.

If I was wanting to go EFI, I'd target a 650/580 rincon or similar machine to snatch it's EFI system from. Mod the intake to fit say a 350x, figure out how to mount the engine temp sensor so it's somewhat accurate (water cooled vs air), and trick it into not caring about the transmission shift solenoids not being connected (like taking an automatic car ECU and putting it into a manual, I've done that once, but the truck ran/functioned fine so I didn't care about the check engine light, no inspection in this state).

Anyway, I'd say the EFI system would benefit over the Carb builds since it can adjust timing etc based on temp, so it can have prime tuning at full operating temps with no concern with how hard it is to start etc. In theory it should be more responsive too, big time since there's no accelerator pump on our carbs.

Basic requirements off the top of my head for an EFI system is listed below, might miss something but gives an idea what sensors would be needed.

ECU/Brain of the operation of course
Air intake metering (so it knows how much gas to give the air charge)
Fuel injector(s)
O2 sensor (to make sure the tuning is burning correctly and at the right air fuel ratio)
engine temp sensor (cold vs hot needs are different, aka so "choke" can work in the EFI system)
crank/cam sensor (so the ECU knows when to spark etc, basically the same setup as what a CDI does, pulse generators might even work already for this need)
Coil might have to be changed, depending how the ECU fires it, it's more of a computer, so it doesn't have raw power running though them normally, like our CDI's do.
Throttle positioning sensor - this tells how much gas you're giving it, so if you go from a low rpm and floor it, it can give extra fuel charge to make up for the sudden vacuum loss. It also can cut injection when going from high rpm to idle to save fuel, most modern cars do this when engine braking in gear.

I think that's about everything, of course you have to power the ECU off 12v, so you would probably have to add a DC power kit and battery if the machine doesn't have a battery factory. This also means likely a dead battery = no start. If I recall correctly the Rincon does have a pull starter though, nearly impossible to start off it though I bet. I know the 450 foreman (carb for the one I have experience with) is a heck of a pull to get going.

Probably should mention, the EFI systems don't really care how big or small an engine is, but going smaller than design makes it harder to meter the fuel as accurately for the over sized injector(s), same goes in reverse, going bigger than designed hits the limit of the injector not keeping up at high rpm/max fuel charge situations. The biggest thing to have accurate is how much air is flowing into the intake and it will adjust accordingly. Matching engine cc size would help keep it more accurate though. Changing the injector(s) might throw off the ECU map, but matching the engine size might make it more correct than having the oversized injector(s).

Oh another spot to look for EFI systems could be modern motorcycles. First one that came to mind that probably wouldn't be a first pick is the 600rr bike, it has dual sets of injectors and multi cylinder so kind of a different beast from what our 3 wheeler engines are deigned after. Would make a wicked fast 3 wheeler with that engine in it for in a straight line race with the 110hp or so and if I remember right 15k rpm red line.

[Trike_crazy]

It's possible and could be adapted from a newer quad likely. Basically what you're talking about is EFI, or fuel injection. Like for car talk, the Toyota pickup from 1985-1988 had a carb option, however it still had an O2 sensor, temp sensor etc, but wasn't self learning because it had no way to adjust the carb electronically. The trucks that came with EFI of course were self learning and the ECU was much more complex since it did so much more.

EFI is kind of a double edged sword, it's awesome in the face with a programmable ECU you have a lot of control with timing, air fuel ratio etc, but there's so much more that can go wrong with sensors etc that could render the machine dead on the side of the trail.

If I was wanting to go EFI, I'd target a 650/580 rincon or similar machine to snatch it's EFI system from. Mod the intake to fit say a 350x, figure out how to mount the engine temp sensor so it's somewhat accurate (water cooled vs air), and trick it into not caring about the transmission shift solenoids not being connected (like taking an automatic car ECU and putting it into a manual, I've done that once, but the truck ran/functioned fine so I didn't care about the check engine light, no inspection in this state).

Anyway, I'd say the EFI system would benefit over the Carb builds since it can adjust timing etc based on temp, so it can have prime tuning at full operating temps with no concern with how hard it is to start etc. In theory it should be more responsive too, big time since there's no accelerator pump on our carbs.

Basic requirements off the top of my head for an EFI system is listed below, might miss something but gives an idea what sensors would be needed.

ECU/Brain of the operation of course
Air intake metering (so it knows how much gas to give the air charge)
Fuel injector(s)
O2 sensor (to make sure the tuning is burning correctly and at the right air fuel ratio)
engine temp sensor (cold vs hot needs are different, aka so "choke" can work in the EFI system)
crank/cam sensor (so the ECU knows when to spark etc, basically the same setup as what a CDI does, pulse generators might even work already for this need)
Coil might have to be changed, depending how the ECU fires it, it's more of a computer, so it doesn't have raw power running though them normally, like our CDI's do.
Throttle positioning sensor - this tells how much gas you're giving it, so if you go from a low rpm and floor it, it can give extra fuel charge to make up for the sudden vacuum loss. It also can cut injection when going from high rpm to idle to save fuel, most modern cars do this when engine braking in gear.

I think that's about everything, of course you have to power the ECU off 12v, so you would probably have to add a DC power kit and battery if the machine doesn't have a battery factory. This also means likely a dead battery = no start. If I recall correctly the Rincon does have a pull starter though, nearly impossible to start off it though I bet. I know the 450 foreman (carb for the one I have experience with) is a heck of a pull to get going.

Probably should mention, the EFI systems don't really care how big or small an engine is, but going smaller than design makes it harder to meter the fuel as accurately for the over sized injector(s), same goes in reverse, going bigger than designed hits the limit of the injector not keeping up at high rpm/max fuel charge situations. The biggest thing to have accurate is how much air is flowing into the intake and it will adjust accordingly. Matching engine cc size would help keep it more accurate though. Changing the injector(s) might throw off the ECU map, but matching the engine size might make it more correct than having the oversized injector(s).

Oh another spot to look for EFI systems could be modern motorcycles. First one that came to mind that probably wouldn't be a first pick is the 600rr bike, it has dual sets of injectors and multi cylinder so kind of a different beast from what our 3 wheeler engines are deigned after. Would make a wicked fast 3 wheeler with that engine in it for in a straight line race with the 110hp or so and if I remember right 15k rpm red line.Ecotrons has small engine efi systems for gy6 scooters and universal systems up to 800cc. I've thought about buying one of these too but I can never justify the cost with as much as I actually get to ride (not enough).

https://www.ecotrons.com/small_engine_fuel_injection_kit/400cc_to_800cc_engine_fuel_injection_kit//

Sent from my SM-G965U using Tapatalk

[ps2fixer]

Neat product, I think I might have seen the page before. Maybe I'm just an idiot, but what's the point of listing a product like that, having a cart button, but no way to add the item to cart, or even post a price? Would be nice to atleast see what the price range is. Either case, I'm sure it's more than I'd want to spend with my current fund situation and such, but maybe my business will boom some day and I can sample a products like that on my 350x =). Being able to play with the ECU programming would be a neat learning experience.

[Trike_crazy]

I believe the listing that's for the 125-250cc engines is about $599. That's about the range of cc that I'd be looking at but I don't have the money for that kind of project at the moment lol

Sent from my SM-G965U using Tapatalk

[ps2fixer]

Not too bad, I was expecting $800-1000, still more than I'd be wanting to spend though lol.

[Trike_crazy]

So fast forward to today and I've been watching lots of motortrend on demand. There was an episode recently where they did a comparison of timing vs. air/fuel ratio and they found that timing adjustments make a bigger difference than fuel mixture. So with that I think I'm going to bite the bullet and buy one of these once I'm closer to coming home from this deployment. It might be more than a pain in the butt than its worth but I think it will help in the end to get the most from what I want to build.

Sent from my SM-G965U using Tapatalk

[ps2fixer]

You thinking about going with the programmable CDI, or the EFI system?

On the subject of ignition timing, the whole curve you can adjust by physically moving the pulse generator/pickup coil. In the chain saw building world, people remove the flywheel key and time the flywheel instead of moving the coil. It's cheap and you get 99% of the timing benefits with out the cost, but it's very hands on and technical. Too much advancement isn't good either. Less advancement is more safe and easier to keep within the limits of making it run well. Go back to the carb based cars, you had direct control over the timing with adjusting the distributor which was effectively moving the pickup coil location in relation to the cam. Generally the most consistent way to adjust them back then was with a timing light. Same story for atvs, you can use a timing light to check the timing and if you do any mods, you can actually see the timing change in real time.

Just another topic, but another huge factor is cam profile, it can make an engine run like a tractor (tons of torque but doesn't rev out well), or it can be closer to a high strung race engine with no bottom end grunt, but it revs out now. Air fuel ratio needs to be within the acceptable range, too rich/lean will cause issues in the long term. Timing and cam profile effects the characteristics of the engine. Depending how far you go with timing, I'd dare to say cam profile has more effect than timing, but with a timing curve like a CDI driven engine has, it really gains a lot of benefit from the timing change based on rpm.

Of course cam profile effects how the engine breaths, so intake/exhaust mods might be required to take full advantage of a given cam.

Anyway, what exact machine are you wanting to upgrade? You mentioned a 350x before, but I don't see one in your list of machines, or maybe that was an example and the atc110 is what you was going after?


Basically with the pickup coil adjustment, what you're effectively doing is changing the initial timing and full advance numbers. So say an engine idles at 10 degrees advancement, and 3k+ rpm it's 30 degrees, if you advance the pickup coil 1 degree more advanced, you changed to 11 degree initial and 31 degree full advanced. The curve doesn't change at all, just physically moves to more advancement. The CDI box/pulse generator doesn't know or care about where the pickup coil is, it's only job is advancement vs rpm.

Either case, if you do go down this rabbit hole, keep us updated on what you discover. Always interesting to see things through someone else's eyes. Would be great if you got some sort of numbers to baseline your efforts against such as a GPS speed test in highest gear starting at x mph and topping out the speed it will go. Just make sure tire size, gear ratios, etc isn't changed in the mods and it should be a fairly good AB style testing. Take off to max speed isn't reliable since shift points and such becomes a large factor, you want to keep to unbiased as possible tests.

Here's a pretty good video how a guy did AB style testing to get the best power, performance, and mpg out of a Chinese scooter. Same basic concept of what you're doing but with a scooter.

https://www.youtube.com/watch?v=p63JXW29tVo

[1babanshee]

Far as 2 strokes go stick with stock CDI. Most builders will say the same thing.

[ATC King]

A programmable CDI is very likely going to need 12DC to operate.

The smaller Honda engines that have a CDI also have a mechanical advance. You'll have to fix that, as in, physically lock it in at full advance. Otherwise, it'll screw up any CDI that advances itself. The CDI can't actually advance anything past what is set mechanically, it just retards the timing until a set RPM.

The pickup on the 200cc and smaller Honda engines is on the camshaft. That means half crankshaft speed. A 350X ignition fires off the crankshaft, which also means they fire twice, on the exhaust as well as the compression stroke. Is the aftermarket ignition adaptable to either system?

If the camchain, guides, and adjuster aren't like new, then they have slack and timing will vary. Inaccurate timing will not allow for increased power, no matter how fancy the CDI. Exactly why people throwing 'self learning' throttle body fuel injection on old V8 engines will see less performance increase than if they went DIS over distributor. The DIS also being smoother running. That's automotive engines, for comparison.

Stable ignition timing is paramount to producing power, not just total advance. Total advance is not a power adder, it's a variable. Advance under light load is different from under a heavy load. Any of the engines mentioned don't have nearly enough sensors to walk the fine line of maximum advance under all situations. That computing power and sensor input would definitely mean a 12v DC system capable of handling it.

Having switchable ignition maps is nice, but I don't think it'll make a difference on sub-10hp machine. That is beneficial for high powered machines that have traction problems.

It'd be interesting to see the adaptation of an aftermarket CDI, but I don't think the benefits would outweigh the cost. A flat slide pumper carb would likely produce more power for the same cost and increase fuel efficiency.

[ps2fixer]

I don't claim to know everything about CDI's, but the above is true, only thing I'm not sure of is the crank pickup coil vs cam based. I suspect they electronically work exactly the same, but the timing curve accounts for the wasted spark system vs normal/cam driven.

The OEM CDI's from my understanding are more or less analog, like I don't think there's a micro controller in them. Never taken an OEM one apart to validate though, but the Chinese "dumb" CDI I took part was literately a circuit to charge a capacitor, and dump it when the pickup coil sent signal. It had zero ignition timing circuit.

The automotive note is interesting, I figured EFI converted engines would run smoother due to the automatically adjusting settings if the base line was setup well. Maybe that's something you don't get in the DIY world vs OEM though.

Interesting you mention load vs spark advancement, that is completely true and I suspect a TPS from a newer quad could solve that problem, of course then there's more electronics etc to deal with and when an ECU is involved, it has to get power somehow, so the 12v DC is likely true for most cases. I do know someone that sells digital CDI's that are powered directly off the AC exciter coil power. The chinese CDI only used a half wave rectifier to charge the capacitor (a single diode), so maybe the digital CDI uses a full wave rectifier so it has enough power to power the micro controller.

This is an area I'm really interested in but haven't had a chance to fully dive into it yet.

I agree that any gains made by better ignition timing over an otherwise healthy engine/ignition system will be minor. Not sure how much things change when you build up an engine, say 12:1 compression, high octane fuel, porting, bigger carb, free flowing exhaust, etc. Cam profile probably effects what you can do with ignition timing as well I'd think. Everything effects everything else, huge balancing act.

The "easiest" power gains with out opening up an engine in the automotive world is throwing a turbo/supercharger on it to cram more air in the engine, which demands more fuel with a byproduct of making more power as if it's a larger engine. Lots of heat and such to deal with for issues though if not done right and the engine might not be strong enough for much boost. Has anyone slapped a small turbo on a 3 wheeler yet? I know the carb system is tricky but carb'ed turbo engines exist, just not sure how that works exactly.

[ATC King]

The stock CDI is already a huge improvement over the previous points systems that ATCs had. On a highly modified engine that already has $$ in it, it makes sense to eek out performance by upgrading the ignition. I'm only trying to contribute to an understanding that spending $$ on an aftermarket CDI is most likely going to be an engineering exercise only. Overall satisfaction on a project starts by having reasonable expectations.

It's also hard to argue the durability of the stock CDI system.




I don't claim to know everything about CDI's, but the above is true, only thing I'm not sure of is the crank pickup coil vs cam based. I suspect they electronically work exactly the same, but the timing curve accounts for the wasted spark system vs normal/cam driven.

Yes, the pickup coils serve the same function. The question I propose is whether an aftermarket CDI can be manipulated to adjust advance based on half engine speed parameters. Maybe the ones on the camshaft, which have a mechanical advance, were just an early design so the CDI could be made more simply, or they didn't have a reliable enough design through R&D yet. Probably a transitional technology too, staying with the mechanical advance until the basic CDI was proven enough to incorporate more electronic control. Maybe running it off of the camshaft allows for more coil saturation since there isn't a wasted spark.


The larger single cylinder Hondas that have the pickup on the crank, do have advance built into the CDI and sense true engine speed. The lack of any other inputs means it's only changed by RPM, which doesn't leave much room for increased performance. To add tunability, a TPS and/or a MAP sensor are needed. Those CDIs also come in 12v AC and 12v DC.

Here is a link to an aftermarket ignition manufacturer in the Czech Republic. They have other designs, but the link is to one for possibly this application.
https://www.ignitech.cz/en/vyrobky/dccdip/dccdip.htm

Here is an ADVrider link to 77 pages of that ignition being hashed out on XR650Ls.
https://advrider.com/f/threads/xrl-ignitech-programmable-cdi-thread.1140612/


The rest of this post is related to general automotive igntions.



The automotive note is interesting, I figured EFI converted engines would run smoother due to the automatically adjusting settings if the base line was setup well. Maybe that's something you don't get in the DIY world vs OEM though.

A properly tuned carburetor will run really well, they just don't compensate for weather and altitude changes. Distributor based timing is simply not close to the accuracy of DIS, and the most important function of a spark timed engine...is spark accuracy. ICE engines will run relatively well at a multitude of air-fuel ratios and those can bounce around a little, undetected, during idle and cruise. They can also burn a multitude of different liquid and gas fuels. They will only run with an electric spark that is precisely timed.

Take one those carbureted, OHV V8s, with a distributor ignition, that people have in their car or pickup. That engine will have a timing chain, which will have slack, even when new, that drives the camshaft, which drives the distributor via a gearset, which has slack. The camshaft flops around under valvetrain pressure, exaggerating the timing chain slack. Everything is bouncing around by the time it reaches the distributor cap. Inside of the cap, there is a micro environment, where moisture and high electrical energy arching across gaps can create plasma. It's really a mess, and that's on a properly functioning system.

That is done away with in a DIS, in part by it's direct electronic sensing of crankshaft angle (no chain and gears) and a direct coil to plug path, not a gap before the sparkplug gap. Even a waste spark DIS is multitude times more accurate and precise than a distributor system.

For old automotive engines, DIS is the best drivability upgrade. The engine will idle smoother and run much better. Junkyards are full of OE parts too, for the budget minded DIY. A modern ignition is grossly overlooked by the typical hot-rodder.

[ps2fixer]

I think your observations are correct. The ATC models that use the mechanical advancer tend to work fine with the Chinese "dumb" CDI boxes. There's a whole balancing game with the capacitor size too, too big and it doesn't charge up fast enough, too small and it gets fully charged but doesn't contain enough energy to make a good spark. OEM likely has 100% the best balance for capacitor size since they engineered the whole machine, they know exactly what the ideal numbers are.

I might be crazy, but I wanted to build my own CDI using an Arduino as the main processor. I wasn't going to worry about the power source since it's just a hobby idea I'd like to play with. My first target would be the mechanical advancer version where the CDI just stores the charge and dumps it based on the pulse generator as a proof of concept. After that base design, adding the ignition timing curve shouldn't be too hard, just a delay based on what the rpm is and an "easy start" system could be made by making the timing extremely delayed and below idle rpm. From what I've been told, the early Honda OEM CDI's only have 4 "points" to the curve. Most service manuals give you atleast 2 of the points.

Like for an ATC350x, initial timing is 10 BTDC at 1400 rpm, full advanced at 30 BTDC at 3500 rpm. I suspect under 1400 rpm it would be 10 degrees BTDC and clearly higher than 3500 rpm it can't advance any more so there's only 1 or 2 points between those rpm ranges that the CDI adjusts on. It's a bit of an odd idea, but I think I could build a device to record the ignition timing of a CDI box using reverse logic on the Arduino CDI setup, or since it relates to data logging, maybe a Raspberry Pi would be more ideal.

Another interesting thing I've indirectly seen is some aftermarket cams mechanically advance the timing slightly (cam vs pickup coil location). One of the people I know that sells aftermarket CDI's has had issues with his CDI + aftermarket cams because the cam adjusts the pickup timing vs cam plus the CDI is trying to advance a little more than stock too so both together is too much.

[ATC King]

Your 'China CDI Reviews on ATC350X' thread is a good resource for anyone thinking of cheaping out on a CDI. The basic conclusion is that sometimes, the OE part is the only one that'll work correctly and anything less doesn't cut the mustard.

http://www.3wheelerworld.com/showthread.php/187411-China-CDI-Reviews-on-ATC350X

The golden info in that thread is the live voltage readings, which reveal the significant drop in input voltage for all the aftermarket CDIs. They're simply not matched to the system, and not any kind of bargain except for maybe an emergency spare that is carried on the trike.


I think the OPs initial question is virtually impossible to answer because a custom engine will need a custom tune to run optimally. With available and affordable (relative) technology of today, my attempt at an answer would be to find an aftermarket ignition that uses throttle position, manifold pressure, and engine temperature inputs. Otherwise, stick with a stock ignition and build the engine to work with that.

It's going to be tough to find anything remotely close that doesn't need 12v DC. That if hurdle #1 I think.

[ps2fixer]

Yea that was a fun post to make lol. I want to expand on my testing abilities with the ignition curve reading to get a number the says how similar it is to OEM timing. Also a spark gap tester to measure the max gap the ignition works on since the CDI is very much apart of how good spark will be.

The 12v DC requirement can be solved somewhat easily on most machines, but I'm a guy that build's harnesses, so what's easy for me isn't exactly common place for everyone else. Simple answer is add wiring from the lighting system, like how someone would add extra wiring for an LED light bar, the catcher being it's not that easy on like an ATC350X since the terminals isn't a common type consumers have access to, but there is extra wiring where the tail light plugs in to tap into the AC power. Since it's AC on that machine, the DC CDI bare min would require a rectifier which most likely has a voltage regulator too, so the OEM regulator should be removed. Depending how picky the DC CDI is, it might require a small battery or a power smoothing capacitor (internal or external) to keep the DC power clean. Long story short, it would be possible to make a kit to make the CDI plug and play if there's enough desire for this kind of mod, wouldn't be super cheap, but I feel it wouldn't be overly expensive either.

I should try to find some super cheap Honda CDI's on ebay that's from the late 80's just to take apart and learn more about them. I really don't understand how the spark advancement system can work, like how do you delay a signal with out a micro controller. Guessing it's something to do with capacitors and triggering when they are under x volts or something like that.

I think you're right on needing extra sensors to really push the ignition timing to see the best gains. Seems like it would be the best of both worlds if it was pulled off, EFI style throttle response, but still carb'ed so not super complicated and less sensors to go bad to cause problems. Using OEM or major brand parts from newer machines would be ideal in my mind for the reliability, the physical sensors aren't anything special, they just operate within a range of parameters and there isn't too many parameters to set.

[Trike_crazy]

Ok, here's what I'm trying to accomplish without getting into a lot of specifics. 1986 ATC125m bottom end, 65.5-67mm bore with stock 54.5mm stroke (183-192cc, big bore-short stroke with room to go bigger it if can stroke it), crf230 cylinder head with the roller bearing cam they use. I'll have to build a custom intake more than likely to fit a 200x carb and I'll probably mod a 230 exhaust to fit once I get it in the frame

As you can see my build doesn't really fall in line with anything stock. I'll have 12v to run the CDI so that's not a problem and with that model of engine the pulse generator is under the lower case cover and not up by the cam like previous models so it would be harder to adjust if that model is even adjustable.

You might ask why I want to build something like this? Mainly just for fun and the wow factor but I do also race this same 125 at Trikefest and I've placed 3rd the 2 years I've run it. I have everything ready to be modified I'm just not in q position to do alot where I'm at currently.

I agree that a cam and carburetor that are properly set up will make the best power. But I think that there's a little bit to be gained in the ignition area. I'll probably start with a cdi for a 200x and see if I can make it work. If not I'll work on the aftermarket cdi.

Sent from my SM-G965U using Tapatalk

[ps2fixer]

Well here's some quick feed back from me. First thing first, don't even attempt the 83-85 200x CDI, it uses a mechanical advancer, so the ignition curve will be completely wrong. You'd be better off with like the 350x stock CDI as a base line.

Here's the specs for the 86 atc125m

https://atvmanual.com/honda/atc125m/1986-specs

Under ignition, you have the initial timing and full advanced, 10 and 28 respectively. The 350x had similar ignition numbers but was a little more aggressive (the physical pickup coil is more advanced). So assuming I'm understanding the effect of using a different CDI on another engine, it means that on the ATC125M it would have an initial timing of 8 and full advanced of 28 (you can't advanced more than the physical location of the pickup). The middle two points of the ignition curve will be good for the 350x engine, I'm not sure if they will be good for the crf230 top end.

https://atvmanual.com/honda/atc350x/1985-specs

I don't know what year you plan to use and know next to nothing about the CRF230, but based on this diagram it's a 2 valve engine while the 350x was 4 valve. My guess is the ignition timing might be too aggressive for the head design.

https://www.partzilla.com/catalog/honda/motorcycle/2016/crf230f-ac-crf230f/camshaft-valve

Since I have the specs in a database, here's a list of machines that have ignition timing based on the 10,28 numbers (for what I currently have in my database). Alright just did the search, only one machine came up....


86-87 ATC200X

Maybe you're onto something with the 200X CDI if you was talking in the context of 86-87. Like the 86-87 ATC125M, it's one of the models I don't know a ton about, so back to partzilla to check the diagrams.

https://www.partzilla.com/catalog/honda/atv/1986/atc200x-a/left-crankcase-cover

Alright, pickup coil located on fly wheel check, 2 valve engine check. Just guessing, but it appears the 86-87 200x CDI is likely your best option if the OEM one doesn't work to your liking. If you want more agressive ignition timing (more advancement), you'll have to grind out the mounting holes in the pickup coil so you can physically move the coil. Check the direction the flywheel goes, and move it in the opposite way to advance it (it triggers sooner). Only grid on one side so you can move back to stock. It isn't an easy thing to adjust since you have to take the side cover off every time, but if you move it a set amount each time you adjust it and mark what the best setting is, it shouldn't be that bad. The pickup coil should be accessible directly right after taking the cover off, like the fly wheel doesn't need to be pulled on all the engines with that design that I'm aware of. For testing, you could line up the fly wheel to the timing marks and put a mark on the fly wheel for the stock location and adjust from there.

The secondary thing in the specs sheets is the alternator output, 120w for the 86 atc125m, so more than enough to run a DC aftermarket CDI if you wanted to run one. Stock the machine had a battery, not sure if you're keeping it though? It might be required to make the aftermarket CDI work but I've recently came across a voltage regualtor/rectifier that has a capacitor that removes the requirement of a battery for the DC CDI's. I bought a test unit to put their specs to the test on a 350x (200w single phase AC). Either case the stock CDI is powered directly off the exciter coil (AC based) and the same story with the 86-87 200x one. They differ in the safety department, the 125m has a gray wire for a reverse safety (can't start in reverse even if manually pull starting), while the 200x cdi uses the 6th wire as a second ground (one grounds at the frame ground ring terminal, and the other goes direct to the alternator). I'm not sure if internally the CDI shorts the two ground pins together or not, if they do, no big deal, if they don't you will have to rework the wiring. I don't have the pin outs for the two CDI's in front of me right now, but I can check if ignoring the safety system if the pin out matches up or not. They both should be the 4+2 style connectors.


Anyway, you have an interesting project, as long as you can pull off the mechanical side of things to get that top end swapped over, it shouldn't be too hard to atleast getting it running off the stock ignition. Like it should atleast idle fine even with the completely wrong CDI assuming it has some sort of ignition curve in the CDI (base timing like 5-15 I'm pretty sure it will idle fine on). The actual curve is an area of interest for me because I don't know enough about different mods and how it effects the ideal ignition timing. Generally speaking, going with slower (less advancement) timing is safer for getting things running, advancing the timing just makes it run better, like quicker throttle response, quicker rev ups, all the desired stuff in an engine. I'm sure you're aware already, but too might ignition timing can cause spark knock, so be mindful of that too. Higher octane fuel should help prevent that issue, might be a good idea for the first fire up and go from there. Like a lot of modern cars can benefit from running higher octane fuel, but they are advertised to run on regulator gas. Like my T100 pickup runs fine on 87, but the owners manual actually suggests to run 91 in it... the truck is from 1998 btw. From what I've read the ignition timing is pulled for the lower octane fuel, so there's a power increase and possibley a mpg increase with the higher cost fuel, but probably not enough to justify switching purely for making $ per mile traveled cheaper.


Sadly I don't have much for the CRF230 model for service manuals and such, but I suspect if the stock engine has the pickup coil on the fly wheel like the 125m, you might be best off running the CRF230 CDI since the ignition timing should match the head design the best, but I'm not sure how the stroke length difference would effect the ideal ignition timing.

I'm not sure if I'll ever complete my project, but I want to build a device to record CDI ignition timings to better understand this stuff and maybe some day build my own CDI. It's more of a hobby idea than anything, but I think it would be really neat to have access to that kind of data. In theory I could do the same for just about any ignition system if I get it working. Basically a timing light that gives out rpm and advancement numbers (well relative to when the pickup coil/pulse generator signal is sent, so really how much it retards at low to mid rpm).

Also, the exciter coil in the stator is effectively paired up to the CDI, it seems they generally have similar specs, but are not always the same. Could be another thing to compare to check for CDI interchange (ohms spec says how long the wire is in the coil assuming the wire size is the same on both).

I'm kind of glad we got talking quite a lot on CDI's, since it seems to be in the area that you need info on. If you need anything searched up in my database of specs, let me know. Not everything is searchable but a fair amount of it is. Most/all details are pulled from a Honda manual of some sort with the hopes of the info being as accurate as possible.

[Trike_crazy]

I appreciate your help and advice ps2fixer. Ignition seems to be one of the last topics when putting together an engine combination. I know that its probably not worth a lot of hp on an engine this small but I'd like to see if it would be. I have an extra 86 200x cdi and a trx200sx cdi to try also. One of the features I'd like to use on this MSD is the launch rpm as basically a 2 step to hold rpm at the line. If I get used to using the software I might try to work in using a TPS to control the curve.

I'm also gonna have to search some other forums and see what applications they mainly use these for. I've watched one video on a guy showing how he set it for a Kohler engine. Maybe I can get a better idea of any other gains that can be had. I would think that after 30+ years that you might still be able to improve on the ignition even on a more stock setup. Idk. We'll see once I'm closer to leaving.



Sent from my SM-G965U using Tapatalk

[ps2fixer]

It might sound crazy but the people that designed these engines and machines really knew what they were doing. The biggest advancements EFI has over carb is mainly the automatic fuel trim (self learning based on O2 sensor and other readings), and mechanical upgrades such as VVTi, or intakes that have long tubes for low rpm and short tubes for high rpm. The concepts and basic core ideas haven't changed much. I guess another way to look at it, how much progress did we have from the early 1900 engines to say the 1960's engines. The biggest thing we've done is just replace mechanical based parts with electronic versions and made things self adjusting so a "tune up" today doesn't involve any tuning.

I'd be leery for an ignition system that claims to be a one fits all setup. Being DC powered is an advantage since it has the potential to be well balanced in it's design, but the coil is still an external factor that they can only generalize in their design.

It would be neat if you could measure any changes in power. Ideally a dyno would be used, but clearly not the cheapest option. Some sort of speed test might work well if gearing isn't changed. Problem is, taking off from zero has a lot of factors, and shifting has a lot of factors. I'd think using a gps app on a cellphone, maybe one that records a trip or whatever could be used to get a slow target speed for highest gear, then floor it and see how fast you get to a near max speed number. Graphed over time I'd think would show the effects of the ignition map too since if the acceleration dips down at a given rpm, the ignition timing might not be ideal. Of course the engines have peak rpms that they work best at, they aren't electric motors with a flat line torque curve.

Either case, should be a very interesting project.


For the typical small engines and lawn mowers and small garden tractors, I doubt their ignition system is too advanced. I'd dare to say they run off of static timing for anything old at all and modern stuff might have effectively a built in CDI like the Dolmar ignition coils have, atleast for the easy start models.

I pulled up my document of CDI pin outs, the 86-87 125m I don't have the pin out info on. I'm pretty sure I have a harness though, so I'll have to dig it out and log the pin out. The 200x and trx200sx are not direct plug in swaps. The kill wire and power wire is wapped (the 2 pin connector is reversed), also the 200x has two ground wires, while the trx200sx has a ground wire for one of the pins, and the other is for the reverse safety detection (grounded will mean no spark, so needs that pin disconnected). According to my notes, I don't have the ignition numbers for the TRX200SX, so can't really say what the ignition timing would look like for sure, but if I recall correctly it has a similar engine to the 86-87 200x so should be similar ignition.


EDIT:

Dug up the harness, the pin out is a perfect match to the 86-88 TRX200SX CDI. For the 200x CDI, you'd have to swap the 2 pin connector wires around on your harness, and you might have to follow where the gray wire goes to the reverse switch and ground it (or just manually pull start it in reverse).

It's so nice having a near complete collection of 3 wheeler harnesses on hand for this kind of stuff lol.

[pennington2002]

This is awesome! I was hoping to find some info on the atc125m wiring situation. I'm taking a trx125 engine and electrical and putting it in an atc110 frame. I'm looking to eliminate as much of the busy wiring harness as possible and do a really clean install. Any chance you could hook me up with the pin out?

Sent from my Pixel 2 using Tapatalk

[ps2fixer]

As my site says, it's all in the details, I'll need to know years.

85 vs 85 requires rewiring (reverse safety on the TRX125 CDI, and also the power/kill wires are reversed).

For 86-87 ATC125M, according to the 85-86 TRX125 manual I have, the pin out matches right up. Only thing that might not match up is the ignition advancement curve since the part numbers are different.

Speaking of part numbers, the 86 vs 87-88 TRX125 CDI's also are different part numbers. I don't have the 87-88 manual to validate the wiring or what exactly changed.

As for the busy harness, there's really not much going on, it's the electric start and neutral/reverse lights that are the main things extra vs the most basic of harnesses. Like the ATC110, there's no battery, electric start, neutral/reverse lights, starter solenoid, safety switches (to run the lights) and voltage regulator and depending on the year, it could be a points engine with no pulse generator and CDI.

Anyway, why not use the TRX125 engine with the TRX125 CDI and TRX125 harness on the ATC110 frame? Lengths probably won't match up well, but you can mount stuff fairly well with quality zip ties and such. Once you switch to another harness, you have to make sure everything that plugs into the harness matches, not just the CDI. Sadly atv's aren't wired in a standard way except maybe the battery and starter lol.

Good luck with the project.