China CDI Reviews on ATC350X

[ps2fixer]

Well I've been poking around China CDI's a bit and figured I'd find my findings, atleast for using them on an ATC350X.



First thing first, they are all wired based on the 350x pin out, so it's a nice stright plug in and test vs stock. My stock CDI works great, so I can do ABA testing. These reviews are purely based on listening to my engine and basically dry reving it. The stock CDI revs out the full rpm range well, but I do know the top end is a bit on the lean side (need to rejet for the DG exhaust I put on it).


256120256121
First one up is the most expensive China CDI I bought a "Racing CDI", still stupid cheap compared to even used CDI's. This fancy blue housing CDI looks to be built fairly well. The red LED light blinks per spark, after a little bit of RPM it appears to be solid on, probably just blinking faster than my eyes can see. It started the engine and idled well. Low to mid range it ran just like stock. Had high hopes, but high rpm it doesn't rev out fully. I suspect the timing is too far off from what the 350x engine demands. If you never get above around 3/4 throttle, this CDI could be a cheap alternative if you're in a pinch. China part number on this one is H048-014.

256122
Second CDI is this ornage housing one. I've read people were having luck with them working out well. The China part number on them is H048-016. This one ran the same as the above CDI for idle and mid range. The high rpm was even worse though, not a huge difference but probably 500-1000 less rpm (I need to get a tach to be more scientific lol).


Both of the above CDI's made the high rpm sound like it was leaner than the stock CDI, maybe it's just from a stronger spark and retuning would fix the problem? I'll have the CDI's around when I rejet the engine some day.


256123
The last CDI is the typical black one that looks similar to the stock one. Looks exactly the same as the orange one, but the plastic housing is black and the epoxy fill was done horribly (not flat at all). This one wouldn't fire up the engine. I gave it a little gas and it puffed a nice ball of smoke out of the carb. I guess this is a 2 stroke CDI or something. The China part number for this one is H048-024.

256124
Bonus CDI, in the past I bought this one with high hopes, but like the ones above, it falled short. It was hard starting, but I could get it going with a little throttle. It advanced the timing WAY too fast, at around 2500-3000 rpm it would hit a wall and not rev any higher. I don't have the part number for this one sadly. It's slightly smaller than the above one though.


256185
Bonus #2!

Stock 1999-2004 TRX400EX CDI part number 30410-HN1-003. One of the replies to this thread suggested this was a suitable replacement. The 400ex is just a tiny bit smaller than the 350x CDI, I can get measurements if someone needs/wants them. I warmed up the engine with the stock CDI again, double checked it revs out well and all is good and smooth. Popped on the 400ex CDI (service manual calls it a ICM [Ignition Control Module] btw) and it fired up exactly like OEM. Low to mid range seems to be a perfect match. However once again, top end it's not quite right. I had a couple back fires though the exhaust, nice flame effect but not running perfect. I'm wondering if it's hitting a rev limiter or something. So once again stock OEM seems to be the thing to stick to.



Mikes ATV Fix CDI

Took me forever to get my carb cleaned, reassembled and reinstalled, but I'm back at it again. My machine is an 85 350x with what we suspect is the "low voltage" stator design, so we pretty much expected there to be running issues since other's have had issues too. From my understanding the 86 stator's exciter coil puts out more power and runs this CDI just fine but I don't have the newer stator on hand to test the theory out. Anyway, as normal, fired up the 350x and warmed it up with the stock one to make sure all is running fine, carb is a lot better now that it's been though an ultra sonic cleaner. It's a tiny bit lean so likely need to raise the main jet a little, but it's pretty close to being right. I popped on Mike's CDI and tested the "easy start" feature which seems to work well, I can hear the timing change after the first fire, just took enough rpm to get a spark. Stock takes a tiny bit more rpm, but not much more when warm at least. Low to mid rpm sounded the same as stock, however high rpm it seems to hit a wall like spark just dies right out, almost like a rev limiter but it doesn't fire at all till I let off the throttle. If I bring the RPM up very slowly I can get it almost as high as stock before it starts to break up.



I built myself a special adapter so I can tap into the CDI wires while it's running to take readings. This test isn't standardized at all, but might show a bit of what's going on. Probably best to take this section with a grain of salt since it's only one metric to base things on. The idea is I can use my home built peak voltage tester to take a reading of the input voltage from the exciter coil and compare to the output voltage to see how efficent the CDI is at passing though voltage. Amps/current is the other half of this and I'm not sure how to read that matric with this kind of setup. I performed the test at idle, but the idle was a bit high, roughly 1600-1800rpm.

Anyway, base line reading with the stock CDI was as follows:
Exciter (CDI input power) - 207v
Ignition Coil wire (CDI output power) - 92v
Note: Output voltage increased with higher rpm till higher rpm then it dropped a little. Not going to red line my engine long enough to get actual numbers though.

Mike's CDI tested as follows:
Exciter (CDI input power) - 76v
Ignition Coil wire (CDI output power) - 29v
Note: Output voltage dropped with more rpm

Orange CDI from above:
Exciter (CDI input power) - 79v
Ignition Coil wire (CDI output power) - 29v
Note: Output voltage dropped with more rpm

I think the next step is to buy an ignition test light and check what the spark curve looks like for each CDI that makes the engine run at least ok.

[floydechoes2000]

Thanks so much for testing these. I have a 400EX CDI on my 350X and it plugs right in and works great. They're way more common than the 350X ones. I'm curious how it would have fared in your test.

[ps2fixer]

My test is nothing special, but I've grown up working on engines and such so have a bit of an ear for if it's running right or not.

I've pulled timing spec numbers from a bunch of service manuals, but I think the 400ex (and a lot of newer machines) don't list hard numbers for timing specs sadly. If you have a timing light, you can see if the timing is within the range it should be. There's two screw caps on the side of the engine, the top one is to see the timing mark. There's generally an "F" mark for idle fire timing, and two lines that show full advanced timing. I don't have a timing light right now, but I'd like to run though the CDI's again to see just how far off they are.

From my little bit of poking around, it seems the idle/low end timing is fairly consistant between 4 stroke engines, but the high rpm it changes a fair bit. Newer engines seem to advance more aggressively, and sometimes have a bit higher stock idle timing advance too (might be mechanical based though).

Anyway, if someone wants to send me a 400ex CDI, I'd try it out and see if I notice any difference. I suspect it might advance a tiny bit too much, but probably not a deal breaker. I'd think more advancement would be ideal for a built engine though.

I've heard some people having problems kick starting their 350x, mine starts generally first or second kit, but my compression release cable is adjusted spot on, I don't even feel any compression while kicking, just feels like I'm kicking it over with no spark plug in. My other 350x feels like you're hitting a wall on the compression stroke lol, it's decompression cam/shaft is broken so it doesn't work at all.

[Ranvier]

Great work. It would be awesome if someone had a ricks cdi in addition to the 400ex and oscars to test.

As far as hard starting, I am curious if a grounded plug would have a visually stronger or better spark?

If you find one that does seem to rev ok at higher rpm, a test under load would be useful.

Thanks again for taking the time to do this and write it up.

[BOB MARLIN]

People that have a hard time starting their 350x just aren't going through the starting sequence right. It has nothing to do with spark. I struggled with mine for years until someone (vegas 250rr) showed me how to do it right. You have to get the piston just past tdc on the compression stroke before you kick it or your just wasting your time. I have one 350x with a high compression piston that has a broken decomp system and it still starts easy.

[ps2fixer]

Great work. It would be awesome if someone had a ricks cdi in addition to the 400ex and oscars to test.

As far as hard starting, I am curious if a grounded plug would have a visually stronger or better spark?

If you find one that does seem to rev ok at higher rpm, a test under load would be useful.

Thanks again for taking the time to do this and write it up.

Oscar said he was going to send me one for my 350x. Once I get it I'll test it out.

I ebay'ed for a 400ex CDI, and dang there's a flood of China CDI's everywhere there. Sorted by used only and cheapest first, and like the 3rd listing was a CDI + regulator/rectifier. Never hurts to have more parts on hand so went for them at $10.45 shipped on both. The listing says the parts came from a 2004 TRX400EX (CDI part number is 99-04). $10 for a back up known working CDI is cheap assuming the 350x runs good with it. Also Honda still makes them, but they do cost more than Oscar's CDI's which would be mapped properly for the engine and possibly has more modern tech in them.

https://www.partzilla.com/product/honda/30410-HN1-003?ref=2d617943d65df5c006b19168c7d632f9b6ad7bef

Yea a load test would be doable, my tank is screwed up atm, have to fix it, and one of my foot peg bolts are broken off. It's cold here so I'm a little bit lazy on the working outside in the cold lol. I'm using a dirt bike tanks currently, doesn't exactly fit the machine right lol.

Out in the open air spark doesn't mean a whole lot. It's a good way to see a very week ignition system, but strong vs stronger it makes no sense. My dad has a spark tester for Chain saws, basically it runs inline with the spark plug, you set the gap to a min spec and start the engine. Then you widen the gap till the engine dies. This test basically tests how much voltage (spark distance) the machine has. The chainsaws my dad works on can get somewhere around 1/2 an inch or so gap + the spark plug on OEM coils. I doubt I could get a good photo of the process, but I suspect I could borrow the tool some day. Would tell me if my coil is healthy and the CDI/Pulse generator are working well. I believe my while stator assembly is more or less new, the guy I bought the engine from bought one of those NOS engines likely from ebay for his machine and I think he swapped the stator assemblies because of the cut wires. That was one of my first tasks after owning the harness making tools. I'm also thinking about getting some sort of tach, atleast for testing things, setting idle etc. Would be interesting to see what my engine's max rpm is. If I recall correctly Oscar said the 350x stock CDI has a 9k rpm rev limiter. My dad also has a "wireless" digital tach that works really well on chainsaws. I've seen it peak out around 14,800rpm when the engine went super lean, seems like the carb isn't getting enough fuel flow and running out of gas. Piston is lightly scored as well so needs cleaned up too.



People that have a hard time starting their 350x just aren't going through the starting sequence right. It has nothing to do with spark. I struggled with mine for years until someone (vegas 250rr) showed me how to do it right. You have to get the piston just past tdc on the compression stroke before you kick it or your just wasting your time. I have one 350x with a high compression piston that has a broken decomp system and it still starts easy.

If the decompression release cable is bad/worn/not adjusted right, or the cam shaft that goes in the head is worn/broken, then yes you must get just past the compression stroke and kick it. My beater/ugly 350x has it working, I can't even tell where the compression stroke is, I just kick it and it fires up. My other machine's decompression shaft is messed up so it doesn't work at all, so I have to get past the compression stroke and kick. It's a lot more effort to start, but it generally starts in 1-2 kicks as well. I'm more of a computer tech guy, so maybe my leg is just weak, but I much more prefer the compression release working right machine and the nicer one I have I plan to replace the bad parts on it so it's like it's brother. If my internet wasn't crap, I'd get my 18 year old sister to start each machine, I'm sure she could start the one with decompression release working, the other even with the right technique she probably couldn't start it (she's skinny as a bean pole).

I'm not sure if any work has been done to either 350x engine, but the one with decompression not working, I can nearly stand in the kick starter on compression stroke and not turn the engine over. I'm around 150lbs or so. My ankle hurts if the engine gives me fits. My last ride with it was out in state land, and the carb decided to flood on me out of the blue. Had fun turning the gas on, waiting for gas to pour out, then turn it off and ride and try to fiddle with the gas valve to keep it going to get back to the truck.

The process of getting just past compression stroke is pretty typical for with higher compression. Most of the time I ignore it on my 250es/sx since they are easy enough to start. Never checked if they have a decompression system, but the compression stroke doesn't feel like a wall you have to overcome.

[Ranvier]

Haha at my sister can start your trike.

So the healthy blue spark dogma is crap? It’s all or nothing? I learn alot from your electrical posts so thanks for taking the time.

And agree with bob, the technique on the X makes all the difference.

[ps2fixer]

Yes and no for the blue spark thing. Orange/Red is 100% weak spark. Blue/White can be weak spark too, but it's likely to run on it in most situations. Make the engine too rich or lean and it might not run. High rpm could have issues etc. Think about it, open air spark is atmosphere pressure with minimum air movement, inside the cylinder it's like 200mph winds, and it's 150psi (over 10x more dense air). If you have 2 stroke, you also have oil in the air too. Of course there's fuel in the air too.

It's always best to have the right tool for the job. This the exact tool my dad bought and it seems to work extremely well vs the ground the plug and check for spark method. It also gives a nice performance measurement. If something claims a stronger spark, this tool can measure if it's true and by how much. Pretty sure you can do the math and calculate what the voltage is based on the total gap. Anyway, it's a Oppama PET-4000.

And here's a related video with a chainsaw that's no spark with great compression, good spark, and good fuel supply, this is the special case the tool is actually made for. I believe you can buy them either Echo brand on it or the "generic" Oppama. A quick google search looks like it's around $60-100.

https://www.youtube.com/watch?v=dipmevU5cjU


Also, the idea about my sister starting the 350x came from this video. My dad has one of the same series chain saws (4200, 43cc I think, professional series). We like Dolmar and Mikita chainsaws. I own a bunch of 6400 series (professional 64cc), and a 5100 series (51cc professional), this is the one with the carb issue I mentioned before. My dad also has a 7900 series and 9000 series (79cc and 90cc). Dolmar/Mikita model numbers just make sense, 2 number is home owner grade, 3 number is farm and ranch grade, and 4 number is professional. First 2 numbers are the engine size, or really close to it.

This Dolmar 420 has the "easy start" system. It starts better with an adult pulling of course, but, she can get it to start lol. These saws have a bit of a process to start them, choke and pull till fire (max 3 pulls), take off choke and it will run, easy start or not. If you're not careful with the choke, it will flood, and it won't run on full choke most of the time. If you're good, you can pull it on full choke and have it fire and slap it off choke and have it going. Of course this is cold engines only, warmed up ones is just pull start it, if that doesn't work choke and unchoke it (sets a high idle).

https://www.youtube.com/watch?v=s2Jv9I1jGkk

[BOB MARLIN]

I agree that having the decomp working properly makes life a bit easier on kicking, but almost all the ones I see aren't working right anyway or are flat out broken. I guess the point I was trying to make is people are asking if these CDI boxes are going to solve their starting problem or not. They will if your current CDI isn't working. But they won't if you just don't have the sequence down.

A few years ago I bought an aftermarket CDI for my 250R that was supposed to produce a " hotter spark" from "Ricky Stator". It didn't make any difference in the way it started or ran. The spark plug looked the same after running also. I just carry it now for a back up.

[ps2fixer]

Yea, Oscar/Mike claims the same thing. I think the wording might be a byproduct of language translation, atleast in the case for Mike. He's working with a Japanese company that manufactures the actual product in Taiwan. Of my understanding, they use a larger power storage capacitor (something like 4x larger), so when it sends the signal to the coil, it's a stronger "hotter" source signal for the coil to use. The coil does the leg work of upping the voltage high enough to actually spark the spark plug. Just my theory, but I think a modern "hotter" CDI + a proper hotter coil would net a stronger spark (higher voltage). I suspect mis-matching the two parts would net you with near stock level performance in the spark department since a higher output coil would likely need more input power, and the high output CDI would just make the stock coil put out more amps, not more voltage, unless the signal voltage is higher.

Of course I'm not an electrical engineer, but I do enjoy learning about electronics/physics.

Now my interest is perked on if there's any disadvantages to a higher output coil/CDI, or what the real need is vs just saying it has better spark.


Also, I agree, a CDI change isn't going to fix starting problems, Honda engines generally start easily, if not there's an issue. If the issue is the CDI, then replacing it clearly would fix it, otherwise, someone needs to get in there and diag the root problem.

[ps2fixer]

Looks like the folks with scooters are doing about the same thing I'm messing around with. The blue and orange CDI looks just like the ones above, probably the same manufacture, but there are different models that all look identical for the orange/black ones. I kind of wonder how he got the hard timing numbers with the timing light. My guess is looking at service manual specs (base timing + max advancement) and pixel measuring the video. That's the same test I wanted to perform myself with a timing light. I think these CDI's are actually designed for the moped/scooter engines, so it makes sense that their spark curves are very similar. To compare, the stock 350x timing notes in the Honda service manual says, base timing is 10 degress BTDC, and full advanced is 30. Sadly that doesn't give the extra timing info of the actual timing curve, but the graph in the video seems to graph it pretty well.

Based on his graph, base timing is more advanced than stock 350x by a lot, just like his stock unit vs the China ones. The interesting thing is, they loose advancement as rpm increases, around 6-7k rpm it shows around 24-25 degrees advancement, vs the stock one which should be around 30. That explains why the high end seems off to my ears. It's kind of funny, all of the China parts that guy bought, and end of story was stock is still best. The pickup/exciter coil mod was something I had in mind, but I was basing the idea on what people do with chainsaws, they use a degree wheel, and advance the actual flywheel (no key installed) to achieve the advancement. In their case the pickup/coil/"cdi" is all one part so not too ideal to grind away at it lol. The pickup coils on the other hand are something like $20 new. I wonder if there's much of a performance gain advancing the timing with the stock coil on a more or less stock engine. Wish I had a dyno locally to get hard numbers.

Also note the use-age of the large gap spark tester to give proof the China hot coil is junk =).

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


Little more info that seems to make sense. The CDI doesn't actually advance the timing, it works in reverse, the pickup coil is at or near full advancement already, and the CDI delays the spark signal to get it's idle/low rpm timing and as it increases in rpm, it reduces the delay. In other words, if you want more advancement at max rpm, you have to physically change the pickup location, or make it get the signal sooner. That's not my direct goal, but figured it would be worth mentioning. There seems to be a lack of actual tech details about CDI's, ignition advancement and such with the searches I've used.

So in theory, the China CDI's are just more or less a 5 degree delay and don't do a whole lot while the stock CDI is delaying the spark around 20 degrees. Makes sense why they are cheaply built, they do less than OEM. In that video, you can see how nicely the OEM unit works. Doubt there will be a cheap alternative (aka Chinese) that will duplicate that kind of results.

[fabiodriven]

Of course I'm not an electrical engineer,

.

I call BS on that.

[350for350]

I call BS on that.

He kind of sounds like it to me too. Of course since I'm an electrical dummy, my opinion may not count for much here.

[ps2fixer]

Haha, I wish I knew electronics more. It's kind of like computers (yea, I'm actually a professional with computers lol). The end users use a mouse and click around in apps/programs. I know how to program, which on it's own is a HUGE field of knowledge to be had, I'm pretty good, but minimum actual proper schooling on it. Next step deeper is knowing assembly language, I've poked a tiny bit around that, that's literately to the point of 1's and 0's, meaning switches on or off. Beyond ASM, is the guy that designs/builds the boards. I'm far from that point, but I know a couple of the basics like capacitors, resisters, ac vs dc, diodes and such, but I wouldn't know enough without googling to build like a flashing LED light, I just know it uses a 555 IC timer, same thing used in like alarm clocks (I think).

Anyway, I like electronics, just like I like computers. I hit a lot of topics and such, so I have a pretty broad understanding of things, but I can't pin point other things exactly. I kind of view electronics as part of knowing about computers, so it's kind of part of the same field =).

Not to throw real life problems into this, but I'd love to work with an Adriano, which is a programmable micro controller. In theory a CDI box could be built based on one and be fully programmable and such. I can program, but I don't know the hardware side of things well enough yet. There's a lot of things that a Adriano could do though, like an electric controller for my wood stove to burn it at peak efficiency (most heat off the wood) when I need heat, and dampen it down when I don't need the heat. Even could go off the deep end and burn something small like wood pellets, and have an electric feed it controls. Anyway to get to the actual real life issue, I'm low on funds, so little fun projects like that I can't fund right now. Either need bitcoin to boom again (probably won't happen for 3-6+ months), or get more effort put into my business to make more sales, which in turn makes more profit. Been focusing on the business clearly which ironically is making more side projects (neutral/reverse sockets, speedo sockets, the project with Mike/Oscar, etc).

I kind of wonder how things would have changed if I took electronics classes in 11th-12th grade or even in college. Heck it's really is an option yet once I have the spare funds. I wouldn't be going for any degrees, just the math and such classes to make more sense of things. Never took advanced math like calculus or trig, I think the most advanced math class I had was just algebra lol.

I just did a little thinking, and if programming an Adriano is anything like I'm used to, building a CDI (programically/logically) should be pretty easy for me. The hardest thing I was thinking of was how to control spark advancement (The delays in ms), but you'd have rpm (ms time for 1 pulse), so not too hard to calculate 1 revolution into 360 degrees and math out the delay time to have the advancement point you desire. Like idle (1400 rpm) = 1400 spins per minute/60 = 23.333 rps (per sec), or 42.858ms per revolution. One degree would take 0.11904ms or 119.04us (I think that's the correct letters for 1000th of 1ms). probably a lot more technical than most people want to read, but it was an interesting thought, never hurts to have the idea wrote down encase I ever do it lol.



Of course I'm not the first one to think of this, here's a video of someone that's done it already. Their code is opensource/public, but it's overly simple. It just fires based on the pulse signal (they used their own pickup instead of the stock pulse generator). That means no spark advancement at all or any fancy mappings like I'd want to do. It seems like their project is more of a proof of concept. Have to love the computer power supply CDI box though lol.

https://www.youtube.com/watch?v=xYR3e3lAR70

There's more videos too if you search for them on youtube.

It seems like a neat project to take on, in theory there could be two mappings, one for cold start (ideal timing for running in cold weather at first startup), and fully warmed up temp mapping. I've seen a pretty simple thermoresister that bolts on with the spark plug and gives a head temp, not sure how reliable they are for the harsh atv world though.

[ATC King]

if you want more advancement at max rpm, you have to physically change the pickup location

Bingo. The max timing is physically fixed, no CDI is going to change that, even the expensive adjustable ones. The length of the pulse coil trigger also affects timing. In the case of the 350X, it would mean modifying the trigger on the flywheel.


Back to your first post; I've been running that blue, finned CDI on my 200ES for at least five years. I forgot it even had the blinking LED because it's in the stock location under the tank, in the stock rubber strap. I just took a picture of it this morning.
256140
256141

I changed the round to a square plug at the time too, and it's completely reversible if needed, back to the round plug by just releasing the terminals and installing them into the OE round plug.

I also tried a few different Chinese CDIs at the time, some just ran (not right) and others worked OK, and yet others ran good. It's a crapshoot, but they're cheap enough to try several. One thing to note, the CDIs meant for smaller displacement engines typically have a higher rev limit or none at all (which is a bad idea anyway). Consequently, I've got a couple of new CDIs that I don't want to throw away, but just hang around in my tool box.


Also, if you want to test actual spark energy, there are some relatively inexpensive tools for that.

I've used these https://www.ebay.com/itm/Snap-On-MT2700-DIS-KV-Ignition-Probe-Tester-MT-2700-Diagnostic-Tool-Tools-Case/143046414886?hash=item214e3b1226:g:jOsAAOSwQKlbAkR m

They're still pretty basic, but it provides more valuable data compared to the adjustable gap type.

I know oscilloscope prices used to be out of reach for the home mechanic, but they're much more affordable now and if you're really wanting to dive into ignitions, they're invaluable. Adding an assortment of transducers and cables as you go makes an oscilloscope the go-to tool for just about everything. Want to actually see the pulse coil waveform? It's not happening without an oscilloscope. Being able to interpret secondary ignition waveforms is monumental to understanding what is actually going on during the combustion process.

[fabiodriven]

I might be misunderstanding this so please bear with me, but if the total timing is fixed and the CDI just delays for lower RPM, then anyone who claims more total advance than stock would be lying right?

[ps2fixer]

Based on the ignition results in that video I posted, the blue and orange CDI would work extremely well for the 200ES, or really any of the 200 series that uses the mechanical advancement. Might have to adjust the pickup coil slightly (maybe mod it not sure), but the ignition timing basically doesn't change all that much.

That Snap-On KV tester seems like the same test as the gap based one, just measures in KV instead of gap. The gap is a rough KV tester, just have to do the math. Gap in mm * 3,000, so 6mm gap should be around 18kv. Math is based on a google result, but the number seems to make sense. Of course air conditions effects the spark etc, so it's only a rough number, guessing the Snap-On tester is more accurate and consistent. Both tools are similar priced though. Guessing the physics behind the KV tester is similar to how the wire wrap tach gauges work.

Ironically I have an old oscilloscope, but it's pretty old. Never had proper schooling for electroincs, so don't know 100% how to use everything on it. It was an ebay special for like $50 shipped like 10 years ago and I've used a few times to try to look at a PS2 laser signal (to get it adjusted within spec). It's an analog meter, and I think it's refresh rate or whatever it's called is too slow. I just pulled it out and it's a Hitachi V-212 20Mhz Oscilloscope. It has a small CRT tube inside it. Not super sure where my probes are, but would be neat to poke around things a bit.

I believe all of these Chinese CDI's are more or less analog. I suspect with a digital (micro controller based) CDI it would be possible to advance the timing beyond the physical set point by basing the timing on the last revolution instead of the current one and only above x rpm where that much advancement would ever be needed. It would have a slight advancement delay, but nothing like turbo lag.

I've seen a few people on youtube that build their own digital (aka programmable) CDI, but nearly all of them does nothing with spark advancement. I did find one that was running a boat engine that showed their spark advancement with a timing light. Clearly that one ran the best, the other's were slow to rev and didn't seem to sound quite right. Kind of like my 200es with the spark advancer stuck at min advancement.

It's awesome you confirmed the CDI operation (delay for idle, less delay for high rpm). Always love having multi-able sources for things I'm not sure on. Also nice to see that blue china CDI has stood the test of time quote well. Besides being a little dirty, looks like new yet. Mine's a slight reversion, it's LED is red instead of green.

Also, yea I agree, a rev limiter should be in place as a safety for the engine. That's kind of an interesting point that goes back to my dad's chainsaw building. He desires coils with high or no rev limiter, like the 5100 series he tested and it went to 14.8k rpm, it's stock rev limiter should have been 13.5k. Might have some special coil, or the limiting part failed. He tunes the carb to be the rev limiter with his tach. Just fine for someone experienced with engines and can tune them etc, but end user it's not so safe. Like myself, I don't know the 2 strokes that well, so I scored the cylinder a little using the saw last year even though I knew it felt a little off (goes lean after a few seconds full throttle).

Pretty neat this "review" thread turned into discussion on electronics, and I love it lol.



I might be misunderstanding this so please bear with me, but if the total timing is fixed and the CDI just delays for lower RPM, then anyone who claims more total advance than stock would be lying right?

I kind of addressed this before seeing your post in my reply above, but I think if the CDI is micro-controller based, it could in theory advance more than stock, but I'd guess it would be a bit inaccurate if rpm changes while in that range. It might have to do some averaging and guess work to see if the rpm is steady, increasing, or decreasing and by how much.

[Dirtcrasher]

Thanks! Excellent reviews you are providing us with here bud!!...

[Rob Canadian]

Very cool and thanks for the info. Years ago I raced gokarts (4 stroke) and ran a 'Digatron' set up as far as a Tachometer. It had tach and head temp. Basic unit. Might be something to look into getting. Test rev limit ETC. Upper units had EGT and water temp. I honestly never have checked timing using a timing light on a small engine.

Am I correct that all ignition advance/mapping is done in the CDI on the 350X? (Sorry if I missed that)

[ps2fixer]

Yes, all of the CDI based Honda 3 wheelers does the spark advancement inside the CDI box, the only exception are the machines with the mechanical spark advancer on the cam end, generally under the CDI cover. Basically that covers the 200 series engines (185/185s/200/200e/es/m/s/x). That mechanical advancer was used on dirt bikes for ages too, the springs I sell are sourced from like a 2002 bike. There still might be some extra advancement logic in the 200 series CDI's, but I believe they are just a dummy box, it sees a trigger signal, and it sparks which is why the China CDI's work pretty well on them.

I might be wrong, but I think points based engines are all static timing, atleast for the Honda 3 wheelers. If I was able to build a CDI for the machines that normally use a CDI, I suspect similar logic could be used on a points engine, just the circuit to detect the signal would have to be beefed up or take out the condenser and jump the wire so an open/close signal could be read. Not sure if it's faster to read voltage or resistance though, I suspect voltage is faster which is why most sensors are coil based, or on cars, it's a 5v signal + ground + signal (resistance based, but computer uses the return voltage for it's calculations I'd guess).

There's two ways to adjust the whole timing curve on a CDI engine w\o modifying the CDI, one is to move the pickup/pulse generator side to side to change when it picks up the signal, or the other way is to modify the metal (generally on the flywheel) it uses to pick up. Generally oblonging the mounting holes on the pickup coil/pulse generator is the easier way. With a degree wheel, you can measure how much of a change you're doing. Remember, this is moving the whole spark curve, so if you desire +2 degrees advancement at high rpms, this method would also increase the low rpm by +2 degrees too, aka harder starting generally, but not nearly as bad as like +15 degrees advancement like the guy had in the video lol

I've attached a screen cap of the graph for easier referencing. I wish every CDI had a graph like this, so someone with a performance/race build could pick and choose which new spark curve fits their needs/wants better.

Also fun thing to point out, the scooter the guy is using is a Chinese scooter, so the actual OEM CDI could be a good source for other Honda machines, however who knows if you can even buy them anywhere. Here's the first google result looking for a Tao Tao CDI. It says the brand is Tao Tao, so maybe if a Honda CDI matches this curve fairly closely it would be a good *cheap* CDI to use. The Honda service manuals generally states the spark advancement starts at about 3k rpm, the graph shows the stock one started around 2.5k. Also most 4 stroke based Honda engines are max advancement is 28 to 32. Based on the graph tat Tao Tao engine is somewhere around 27 basing off the orange (guessing dummy box). Also most machines have an idle/low rpm advancement of 10 degrees, a couple have 13 like the 250es/sx. Most of the Honda timing specs are + or - 2 degrees, so that puts the stock Tao Tao CDI in a pretty good sweet spot for working fairly well for most of the non-200 series Honda 3 wheelers. The big difference would be the actual middle points that Honda doesn't give which would effect the 3k rpm to max advancement point performance (aka racing, taking off, wheelies, fast cruising etc).

https://www.amazon.com/Scooter-Moped-139qmb-TaoTao-Roketa/dp/B00GMQ1MTY

This kind of makes me want to get a quality test light (aka accurate), and modify my 350x flywheel and sync marks/scratches to a degree wheel to measure the ignition advancement of different CDI's. If I built an actual CDI, it would give a good way to measure the exact advancement the end calculations are receiving, the actual math, signal time etc would all delay the timing ever so slightly, not sure if it's enough that it would need the delay accounted for in the code or not.

Of course I'm far from being able to build a CDI, I need to learn the basics of micro controllers and how to interact with hardware before even trying it. You know, like a simple flashing LED lol.

[ATC King]

Thanks ps2fixer for working on this. People could just keep taking chances on random CDIs, but you're going to nail down just which ones.

This thread should relate well to the 350X, since the systems operate similarly.
https://advrider.com/f/threads/xrl-ignitech-programmable-cdi-thread.1140612/


It's a very long thread. The thing to take away is so many people jumped on board spending $$, to only realize they have control over advance only up to the physical max designed by Honda. People were programing in what they though was a higher maximum advance, only just because it shows it during the programing doesn't mean it's really happening. Without doubt, most of the people who bought those $$ ignitions had stock or very mildly built engines, of course, the price of a OEM CDI for those bikes isn't far off of the aftermarket price and the OEM ones are known to have failures. Meh, in that case, buy the fancy one when price is nearly equal.


Something else worth mentioning, is without more input parameters (manifold pressure, throttle position, intake/engine temp), the timing never changes from what is programed. It won't change with engine load or even engine temp. It's not a dynamic timing curve, and responds to engine speed alone. Two selectable maps is as good as it gets without the capability of more live input data.


As ps2fixer mentioned somewhere, the CDI simply wears out. That's why having a quality, stock style replacement is so important when Honda doesn't offer them anymore. It doesn't have to be super high tech or expensive, just work as good as the original. I like fancy stuff just as much as the next rider, but at some point it just starts becoming silly wasted money, like a $500 ignition box on a stock Ford 300 I6 in a '79 F150.

[ps2fixer]

I'll have to check over that thread, assuming it's interesting enough I'll probably end up reading the whole thing lol.

Also, yea I've bought the China CDI boxes from my supplier, so I got them at "whole sale" prices, and was already in orders for my own stock so shipping was free ($200+ orders), so I'm spending as little as possible on them still =). In the scooter CDI vid, the guy mentioned he paid something like $40 for two, I have 5 different models, probably cost me around $20-25. The cheapest of the cheap ones are $3-5, blue one was $10 if I remember right. Of course more hands in the middle, more price markup. For me it's Factory -> whole seller -> me.

It's probably worth mentioning, Mike/Oscar is paying to have CDI boxes made as well, for $70 if they perform as good or better than stock, it will be well worth the money. I know some people are either too cheap for those prices, or flat out can't afford the much for a beater machine. Knowing what works, even if not perfect, for say $25 shipped would still cover another market point. CDI's are fairly universal in function, it's that timing advancement that makes them all slightly different. He's still having tweaks made to the CDI's, the next batch he's getting is going to be more sensitive to detect the pulse generator pulse. What I'm figuring is the CDI was designed after a modern CDI, and the old machines made less voltage for the pulse generators (less ohms per spec). I think it was something like two people had a bit of an issue with them, one of them swapped the CDI to a friend's machine and it worked great on the other machine, so it's not directly the CDI's fault, but likely an under spec part.

I wonder if an arduino could be epoxied in a plastic box just like a CDI box and not have issues, or better yet, a aluminum case to mount the heat-sink to for the power supply (12-60v+ AC -> 3.3v or 5v DC). Just a quick google search shows the arduino can have a bluetooth adapter added to it to program it wirelessly, so a programmable "CDI" could be possible. I think that would require my code to be opensource/public though. Good and bad, bad being someone in china could copy my work and do it cheaper with lower quality parts and cheaper labor. Even if I don't make a product out of this, it would be a great learning experience for myself.

The hurdle that's been keeping me kind of out of these is knowing which one to buy lol. There's so many options. I'm thinking the Arduino Pro Mini which runs on 5v and has a clock speed of 16mhz and has 8 analog inputs. Would be neat to have a LCD output and a few buttons to navigate a menu and "reprogram" though that interface, but everything is ideas at this point, keeping it simple first is ideal. It on it's own is around $10 + shipping. The pin header to add in to the epoxy is like $12 for the Japanese connector most likely used in OEM CDI boxes. Chinese copies of them are like $1-2. Might go for the China knock offs for now since I haven't heard of people having too many problems with the connectors. If this project nets to anything interesting, I'll probably start a new thread on it, but maybe just a thread title change would be enough. Of course my test dummy engine will be the 350x. I also have an 85 250es, 85 250sx, 200es (will run with a little work), trx350d (350x style top end, 250es style bottom end), 81-84 250r, and for a points engine I have a TX500 engine is a blaster frame that I need to build the harness for. Would be cool to make it run with a CDI style box.

Hopefully the readers of this thread like long posts, I generally make mini-books on subjects of interest lol.

Also for the input sensors, I think bare minimum one could get away with basing some of the mapping on head temp (sensor is available to bolt on at the spark plug). It wouldn't be quite as responsive ad proper EFI systems with TPS and such, but I think the code could estimate TP based on RPM change, of course it won't know the load so I guess it couldn't take full advantage of the estimates. Would be simple enough to enable/disable features like that. Kind of the same idea of making more advancement than physically set on the pickup, in my mind the coding makes sense, it would just have a 1 or 2 rpm delay (~6-40ms depending on RPM, it shouldn't kick in till probably 5k+ rpm anyway, so that's around 12ms). Clearly estimating things won't be nearly as accurate as a system with all the sensors in place, it would be more or less the equivalent of Limp mode for cars except it wouldn't pull the timing, it would try to add a little timing (or pull in some situations). I'm thinking anything like this would be default disabled since it likely would need tweaking per machine, assuming this ever becomes an actual product. 32KB isn't a whole lot of data for the program, so can't go too crazy on the design, and the ram is only 2KB. If I run out of ram, there's a 2.5KB version, but nothing larger. Wouldn't want to move up to a raspberry pi since it's an OS based mini computer basically, lag etc could be a problem with it running an engine.

Anyway, I'm open to advice on the circuit designs, I've ordered a few arduino's just encase I burn some up with my little understanding of electronics (I have just the bare basics down) lol. In my mind, there needs to be a trigger circuit for a peak in voltages to trigger it. Service manuals spec the pulse generators as low as 20 ohms and as high as 400 ohms, not sure what that is in voltage though. Maybe just a way to read voltage and check it often enough to see the peak would work, I know my pc can run though millions of lines of code per second, 16mhz would probably be 1000's per sec or so though (for the whole loop).

I suspect the standard CDI circuit would work for the capacitor to charge and discharge, basically a diode, capacitor, and a gate with a trigger pin (don't know much about these or what they are called exactly, I suspect a transistor is about the same thing?). Not sure if the arduino has enough signal strength to trigger the gate though.

Clearly over sizing the capacitor is ideal to take in as much extra power as possible, not sure how to make the power supply for the arduino work with both it's power needs and the capacitor needs to fire the coil. Maybe having an extra wire on the box to tap into the lighting circuit for power would be ideal so the maximum energy can be put into making the strongest spark possible, heck with that wire it could power the capacitor from that too and for sure have more energy stored than any stock CDI, wonder if it would make much of a difference though, I'd guess the coil would output more or less the same spark voltage but maybe more current, or a longer duration.

My first project will be something simple like a flashing led (good way to signal data to the user like error codes). I guess the second project would be to add a button and detect the button presses and light up the LED, then add a small delay on it if the "rpm" is fast enough (for human button press speeds), and move up from there. I legit have a shed full of electric boards, so I have probably anything and everything I'd need to play around with this.

Oh, I just noticed, high clock speeds are available too, it was in another section. Highest one appears to be 180mhz with 256KB of ram and 25 analog inputs. I'd guess that's enough to run most/all sensors on a car lol. The one I ordered is probably plenty good enough I'd think. Maybe I can hook up the pulse generator to a chain saw and see what the peak RPM I can get out of it is. One of my dad's saws is a 14.5k rpm saw, and I have a bunch that are 13.5k rpm. If it can handle max rpm on them, it should be fine with pretty much any atv engine I'd think. Not sure what the rev limit on the 250r would be, guessing around 12k rpm?

Another fun side project of this would be to get accurate timing numbers for the pulse generator location per engine model so the timing curve would be accurate for them, or base it on what the CDI actually does and make it delay/retarded degrees vs advanced degrees (if my idea/code works right) as the CDI sees it.

Anyway, I think the book is done for now lol.

TLDR;
Anyway to sum this post up, I'd like to see if I can build a CDI off an Ardunio to get it to atleast stock level performance, then add in extra features after that to toy around with and learn from.

[Dirtcrasher]

I do know that last year, I tried 3 different CDI's, and one of them seemed to work the best. And it was also the most expensive model at 100$.

This is not the time of year for me to test things, Oscar did send me one to try out, but, it will be awhile here in the Massachusetts.

You'll probably see me building a pole barn this spring before I can test these things out fully...

Shorting Bitcoin has been lucrative :lol:

[ps2fixer]

Haha, yea bitcoin was a fun pass time at my midnight job, made a bit playing the market. Job was boring as hell, ended up getting to the point I had to do something to stay awake, the job was for 24/7 IT coverage at a factory, and our team got things so relaible it was dead overnight.

Did you get a 200x or 350x CDI from Oscar? I'd be interested in hearing feed back from others as well.

[fabiodriven]

Lmfao my $750 in bitcoin turned into like $200.

[ps2fixer]

Lol, that's the nature of bitcoin, when it drops, you target increasing btc holdings (sell and buy when it drops more) timing is everything though, I haven't had time to focus on it so I'm just holding though the dip. Ironically I was thinking about selling when the price was around $18k per coin, but ended up being greedy and holding, missed out on selling my collection for a sizable chunk of money and good profit. History repeats itself, it's very likely it will recover to atleast 10k, and probably beyond, might take 6-9 months though :-/.

This kind of reminds me when I had the bright idea of buying silver when it was increasing in value, bought right around $50, aka the highest it's ever been. It's now around $14/oz, not going to see profits on that investment for a long time I suspect. Not a huge deal, I made the investments when I had a good paying job and was "money to burn", I basically treat it as a CD or bond, can't really touch them till they mature aka make a profit.

Basically the rule of thumb is, if everyone is going crazy buying something, it's time to sell, and when it looks like all hope is lost, that's why you buy.

[ps2fixer]

Did some more reading about the arduino project, and it sounds like this is a very doable project with a 16mhz arduino. To do the extra calculations and fancy stuff, I'll probably need the higher speed units since the math involved and such might take the arduino too long to process. Basically, attempt to mimic the OEM CDI's function more or less exactly is about the perfect goal for the project. I believe the code can port over to the other boards with no issues, maybe slight config changes. I'm clearly not the first one to want to do this, I've seen a couple threads now with people with old motorcycles with out of production and very hard to find CDI boxes that have failed and they need a replacement.

I'm kind of excited to get to poke around with micro controllers and such. Might have to disassemble a cheap used CDI to check what components were used OEM so I don't have to fully reinvent the wheel.

Sounds like I get to dive into C programming, never done normal C programming, I had classes on C++ and C# though. From my understanding, C++ is just an extended version of C. There's also the arduino IDE that gives a higher level type of programming language that compiles likely to C code then assembly/binary. Not sure if I'd want to use C (better control, lower to dev), or the free IDE.

Once basic CDI function is working right, in theory I could build a dash (like LCD display for example), and display RPM, Temp (if sensor hooked up), vehicle speed (based on rpm, so would have to be based on highest gear ratio, stock chain sprockets and stock tire size). Could probably be fancy have add in timers like run time (current time since engine started), total run time (since the unit's been installed), hour meter, etc. The timers would be simple code triggers at the startup process and shut down (say under 100 rpm it expects the engine to die, and loose power, write updated times etc). MPH and RPM would be a little extra math, but I think it's simple enough it wouldn't be too much. Not sure how much processing is needed for an LCD display though. Thinking of what's possible is kind of neat with my limited knowledge.

[ps2fixer]

Well, I was wondering what exactly made up the China CDI's and if there was anything special about them. Looks like this one with the teal connector (known good working one for a 200es) is legit a dummy CDI. I don't know electronics well enough to say what exactly is going on, but there's no micro controller clearly, so it's analog based. Capacitor is nice and large, but I don't have an OEM CDI cut apart to compare either. It was a rear bear to get into the CDI, had to cut the 4 corners of the plastic out and use vice grips to break it free from the epoxy. The black top layer is slightly rubberish while the clear is very glass like (brittle), and the white type on the bottom is similar to the clear but a lot stronger, guessing it's just sand mixed in to make it stronger like normal cement has sand and gravel added in.

The capacitor is a 105J which based on google is a 1uF 400V capacitor.

The big black box which I was guessing was some sort of transister might be the "gate" that dumps the capacitor to the coil. Here's a spec sheet on it, it's commonly used in motor controllers so it seem to make sense.

www.cldkj.com/pdf/THYRISTORS-LOWIGT/TSE2P4M-TO126.pdf

The small black boxes with 3 numbers are resisters, nothing to special there. The gray box with nothing on it is a small capacitor. Not sure what the M7 black boxes are, but based on some multi-meter probing, they seem to be some sort of diode. Based on some more googling, seems my guess is accurate.

Based on my understanding of the parts and the basics of how a "simple" CDI box works, this one is truly a dummy box that fires as soon as the pulse generator sends the signal. There's no electrolitic capacitors to fail, so even though it's a very simple design, I suspect it would likely last as long as OEM assuming the solder work and parts used are within spec for the application. I suspect if I pulled apart a 200es CDI, I'd see a similar circuit with 1980's era components (though hole instead of surface mounted). The connector was the first thing to break free from the epoxy, so if that happens and it vibrates enough, it could break the solder connection and cause issues over time.

This was fun to check into, but a real pain to open lol. I did loose one resister from beating this thing up trying to get that epoxy off. Anyway, enjoy the pics =).

For those with the China engines that use this CDI or similar, I suspect you could gain respectable performance gains by running a Honda OEM like CDI on it. Might have to adjust the pickup coil/pulse generator mounting location to match the Honda OEM placement. Reason I say this is because the CDI doesn't do advancement (static timing), and most/all China clone engines I've seen are modernized with a basic pickup coil, no mechanical advancement, like the 200 series Honda engines. I could be wrong with what I'm thinking, but might be a fun project for someone to try out. I have a couple tiny China engines I got cheap (like $50 for 3 that need work), never poked around them though.

I might have to follow the pin out and traces to see if I can understand that CDI's circuit better. Seems like in theory I could use the same components, or very similar for the digital CDI idea. I also thought of a way to power the Arduino with out loosing effectiveness of the spark power. Unless I'm misunderstanding it, CDIs only use a half wave rectifier (just a single diode) to charge the capacitor, so the opposite AC wave isn't used at all. Should be able to have the reversed half wave rectifier be used to power up the Arduino with some voltage stabilizing parts (capacitor and such) and be good to go.





Also I don't think I posted about this publicly, but since the CDI idea seems to be possible, it should be possible to build a CDI tester and have it trigger a spoofed pulse generator signal and read the output to the coil to test if it's working and what the timing is. With some data logging, could test every 50 or 100 rpm to plot out the advancement curve quite accurately assuming the CDI is in good working order. I'm not sure how the analog ones delay the timing, but the capacitor in them could get weak over time, still could read a weak signal though. Assuming this works out like I hope, it should be possible for me to have people send me CDI's and I can show them the spark advancement curve. I think it would be nice to know just how different the spark curves really are model to model. Also could compare China and other 3rd party CDI's to see how accurate they are compared to OEM.

Another side project/idea I have with the Arduino is an automated wire cutter/stripper. Could make my life a bit easier assuming I can figure out the psychical controls for the wire (which sensors needed to detect wire placement, stepper motor for measured movements, etc).

I also heat with a wood stove, sometimes the house is already warm/hot, so might use an Arduino to control the fan on it. Could get all fancy and control the air dampener and such too. Right now it's fan kicks on when the fire box gets hot enough (I have an adjustable temp switch on it, OEM one failed). I suspect a basic thermostat would work good too, just my main one is hooked to my backup furnace. Maybe control both wood stove and backup heat from the same controller would be a nice project though =). So many possibilities.

Oh yea, there was a little bit of talk about Li-Ion batteries, in theory the Arduino could be used as a "smart" charge controller too with the right supporting circuits.

[Shep1970]

Ps2fixer, Pretty sure you’ve seen these on your searches, figured I’d post so others could see too. China 4-stroke cdi - two different ones- clear epoxy used to encapsulate it/ kinda interesting to see what’s involved in there- just eBay pics-
256182 256183
The little blue box comes through the epoxy for manual timing adjustment “they claim”
That’s all I got- just passing it on....
Shep

[ps2fixer]

Awesome stuff. I never seen it, never searched to see if others have taken the CDI's apart lol.

The CDI in that photo has some sort of micro controller it looks like, so very possible it truly does have an adjustable timing curve, question is, how does it adjust it exactly. I'd assume the micro controller reads the pot (blue adjustable resister) and adjusts it's programming based on that. I could see it doing the adjustment across the whole rpm range, like delay everything 1 degree, or maybe it does some sort of percentage thing where the lower rpms are adjusted more, and the higher rpm stays fairly close to the same due to our understanding of how CDIs work now (pulse generator is psychically mounted as max spark advancement and the CDI does the least delay for high rpm).

Interesting thing on those two CDIs, it clearly looks like the exact same board is used. Nothing super major changes that I can see except the pot moves a little (second set of pins?), the transistor looking part has a heat sink on one of them, and "ZD9" componet is populated on one but not the other. Assuming that's a zener diode. Can't really say what the difference would be exactly, but I'd guess the micro controller programming is the main difference. Seems almost silly to have a heat sink embedded in epoxy, there's no air movement, so if it's really needed, it just delays the part from overheating. My guess it's more of a looks/marketing design.

That would be a fun CDI to play around with though, can you post or pm me a link where its for sale? If I get my CDI tester made I'll have to buy one and see what it does exactly.

Also looks like they stole my idea lol, one diode from the power supply pin appears to charge the capacitor, while a diode in reverse powers the rest of the circuit (micro controller and such would be my guess). I'm not aware of any AC to DC power supplies that only uses two diodes, it's 1 or 4 generally. Nice to see things I'm thinking of seems to be good solution to problems, maybe I'm learning electronics more than I thought lol.

[Shep1970]

Sure- https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com%2Fulk%2Fitm%2F33 2893683261

eBay- gy6 cdi There’s about 1000 items and many of the clear ones stating 50cc-200cc bikes

Shep

[ps2fixer]

Ahh that explains the transformer and extra didoes I was seeing. It's a DC CDI, won't work on our 3 wheelers, they are all AC based (atleast for the Hondas). If someone added in a voltage regulator/rectifier and ran the dc power to the CDI as well it would work though.

Really, engine size doesn't mean a whole lot. Based on the service manuals, the timing specs for an ATC110 matches an ATC200X, and also an ATC350X. The only specs the service manual gives is idle advancement, and full advancement though. Of course the 200x has mechanical advancement, but the overall spec matches.

Talking about the 200 series engines, it would be neat to actually read the timing curve of say my 200ES engine (I have one that seems to run really well, smoker though too), and install the spark advancer springs I sell in it and compare new vs old but working well OEM. Would require a little bit different logic since I couldn't directly control the RPM. Since the pulse generator timing changes, I'd have to figure out how to detect rpm and calculate degrees from that. I'm guessing tapping into the AC generator and reading the wave forms would be the best route, just no clear indication of top dead center except the spark timing. I bet if I made some sort of meter it would be super useful for those with this type of engine to tune the advancement. Different springs would give different curve points (it's an actual curve too). I guess a basic timing light would give a fair idea on how it's performing though, just no hard numbers.

Seems like just about everything is possible now with programming and a micro controller lol. I think I might have a new passion.

Also the GY6 engine is a scooter based engine, old ones were 2 stroke, and I think newer ones are 4 stroke. Since Chinese parts never seem to have any real specs, you have to be sure it's for a 4 stroke atleast if you plan to run it on a 4 stroke engine.

Here's one I found that says it's AC powered and also is the adjustable timing curve style. It says it's digitial, but I don't see a micro controller on the board. Either it's hiding on the other side, under the huge brown/redish cap, or they are flat out lying and it's analog based. I see a few sizable caps on the board, so I'm thinking it stores the pulse signal, and delays it somehow still but really don't know more than that.

Also it's worth nothing, I don't see anything in the CDI boxes that would generate enough heat to ever need the housing to have a heat sink, again probably just a marketing tactic for the general public that doesn't know the difference. I suspect there's no Honda OEM CDI's with heat sinks, I could see once fuel injection came into play that need for them due to the higher clock speeds and such the computer needs to run at (aka ECU).

https://i.gyazo.com/06c6d8b8146c28072e58b6ed140d86de.png

[ps2fixer]

Updated first post, got the TRX400EX CDI today. I'm kind of shocked it didn't work out like I was hoping. I'm really interested to see what's really going on with the high rpm timing. Maybe the new/china CDI's are faster than the stock one and it's getting too much advancement at the high rpms.

If anyone else has a CDI they want me to test out, I can see what the effects are, just pay the return shipping if you want it back ($4). I can adapt my harness to any CDI if needed.

I've bought a lot of harnesses lately, so I probably should put some focus on getting reproductions made. I'm spending more than I'm making, so I can't keep buying stuff. That reminds me, I have to order more heat shrink for a project I'm working on -_- lol. R&D is a fun pass time though.

[fabiodriven]

Have you done anything with the SX and ES? I think I saw an SX CDI in my garage the other day. I'd have to confirm that's what it is before I sent it but if it's any help you could definitely use it. I don't remember if you've already done the SX or not.

[ps2fixer]

The 250SX and ES are wired different, not a big deal for me, but they are not plug and play for the 350x. They also have a neutral safety switch that needs to be grounded to get spark. I'd defo have to make an adapter to test this since the Green and Green/White wires on the 350x would need to be connected together, and ran to both pins on the CDI to bypass the safety circuit.

I have an 85 250ES CDI, don't have a easy to connect connector right now, but I think I could modify another connector just for testing though. Would you need the SX CDI back? I'd like to get the stock advancement curve off it if/when I figure out the CDI tester stuff. I probably have an 86 250ES CDI around somewhere, but I only had one of those machines, not sure where the electronics from it are at. Pretty sure the 250ES and SX used the same part number for the CDI.

[ATC King]

Have fun building your own CDI.


All kinds of stuff is possible. Doing all this CDI thinking, reminded me about using a GM HEI module on a 200ES. I had to drag through some pictures to find it. Of course, this only works on trikes with batteries because it's an inductive ignition instead of capacitive.

That's part of the heat sink from an old Boss amp that died. GM HEI modules require a heat sink.
256212
256213

[ps2fixer]

Haha, ironically I used the exact same module to fire up a low miles Toyota 22R engine that was missing the "ignitor", maybe it was a ford module, can't recall 100% for sure, same shape and such. On the Toyota engine, I could tell its timing was slower than stock, maybe it has no advancement at all? Not sure exactly how those work internally. Of course I did things wrong, I didn't put a heat sink on the module lol, it was only a test run for about 15 seconds though and the guys that were looking at it instant bought it for $400 and didn't even try to haggle the price. If I remember right it was an 83 engine and I sold it in the late 90s. Engine only had something like 30-40k miles.

Anyway, even if I fail at the CDI project, I'll atleast learn something from it, really already learned a bunch and haven't even gotten the boards yet lol.



Also kind of an update on the thread, I might be getting a Yamaha CDI to play around with too. Unsure what the source machine is though. Poking around on ebay, it seems the Yamaha CDI's run $50+ used, so an adapter + known working CDI's from the Honda/China clone world might be a thing people would like.


Also, my dad has one chainsaw that doesn't run right, timing isn't right at all, hard starting etc. Based on what I've talked to him on these CDI's, he's thinking maybe the advancement circuit failed in the coil and it's not delaying the signal, aka full advancement while trying to pull start it. He mentioned that he should be able to test that theory with his test light and starting the engine and pulling the side cover off. If I built a fancy CDI tester, I might be able to figure out some way to measure the chainsaw advancement properties too. Same issue as the 200 series though, no sensor to tie into to know where top dead center is located. Maybe a simple magnet switch opposite from the coil could be used to calculate it (since the flywheel is very close to the outer case). Just ideas/thinking out loud.

Also, if there's any other micro controller ideas out there, I'm all ears. Might not be an expert on these, but having more project ideas never hurts, even if it's just "simple" ideas like a harness/atv tester that plugs into the CDI connector and checks the circuits (resistance when engine off, maybe peak voltage when running, etc). I'm half thinking of trying to figure out some sort of semi-universal dash cover that could have this micro controller in it and display rpm etc to the rider. Just would need a long enough lead wire to plug it into the CDI plug (if my CDI project works, I should be able to replace the CDI + display the outputs with a powerful enough micro controller). Big ideas, but not much funds right now to put into them, so everything will be more proof of concept than actual prototyping/production building. Maybe something like this already exists? I know there's one for scooters, but it doesn't replace the CDI and requires user to mount it and wire it in using their own methods.

Before I know it I'll be wanting to put a 7 in display on the machine for digital gauges, fuel injector and related engine sensors for EFI lol.

[TexFest]

here's a performance racing 350x cdi i found online, maybe you could try it, www.mikesatvfix.com

[ps2fixer]

here's a performance racing 350x cdi i found online, maybe you could try it, www.mikesatvfix.com

Yea, that's Oscar/Mike. I'm working with him for making adapters for other models. At some point he's going to ship me one of his CDI's for me to test out the product, and if I figure out this CDI tester/timing curve reader thing I'll compare his vs my stock one.

Also there's this CDI linked below, it's $130 though
https://ricksmotorsportelectrics.com/OEM-Style-Honda-CDI-Box-15_622

[ps2fixer]

Another fun thought, has anyone tried the 87-92 TRX250X CDI on a 350x? More or less the same engine so I suspect the timing curve should be very close. The 250x cdi is still in production but $110... Also the pinout is wrong of the 350x, I think it was just the black/red and black/white wires that have to be swapped.

Also, since I'm logging the machine/CDI specs and such from service manuals, I figured I'd also log the little tag number for each CDI. Seems like that's a solid way to ID them and there isn't a good place to look them up yet. If anyone wants to post them up that would help me out and validate what I have logged so far, can use any machine including TRX models. Some CDI's are printed with the numbers instead of a sticker/tag.

Example is the 87-92 TRX250X CDI is labeled as HCO CI536, and the ATC350X is HA5 CI-92. Some machines have more than one CDI part number, so I suspect those special cases will have their own numbers too.

I have a couple CDI's I'm unsure what machines they are from.

969 CI-42 (1984-1985) - nvm on this one, it's a 200es one
HA0 CI-93 (1984-1985)

[ps2fixer]

Well a little stumped on the 87-92 TRX250X CDI. The reverse switch is a strange design, and if I'm understanding things right, it needs a +12v on the ground pin from the lighting coil (after a bulb/resister). Does anyone know if the gray wire is critical for the CDI to work? I figured not hooked to anything would bypass it, but I see now it's a little more complex than just a normal ground switch. I'd think the two circuits are completely separate though and the ground switch is just to light the reverse light up, and the CDI side is just to test if it's grounded or not. Might have to wire up a 3w bulb inline and hook it up and see if I get any spark out of it. I'll poke around a bit more to research this out. Could just be a dead CDI I have too, or the 350x's exciter coil voltage is too low for some reason (it's about 80v peak at kick start speed, most other machines are 100v min).

Look at the 350x vs 250x stator, looks like the 250x one matches the 350x one with less coils. I wonder if that's a possible interchange for these 350x engines. Probably an idea to test out another day. If this is a solid interchange, then in theory the Goki starter kit could be used on the trx250x engine, even if the 350x flywheel was required. Maybe they made a kit for the trx though, never looked into it.

Reverse switch is just left of the CDI box in this diagram.

I dug up the wire harness I have from the TRX250X machine and the wire diagram is spot on, white/yellow regulator plug to gray/black gives me 10k ohms reading, so that's the internal resister I'd suspect. Kind of hate to do it, but probably should pull the harness apart to locate it and see how big the resister is (1/4w, 1/2w, 1w etc).

https://i.gyazo.com/699d47ed463a8aeaccf7a79c1614f7e2.png

[ps2fixer]

Really unique setup for the resistor. Found it located near the tail light plugs. It's one of the weird greenish colored resistors which I'm not 100% sure what's different about it vs a standard resistor. Color bands match my readings, 10k ohms with 5% tolerance. Kind of funny that it's crimped to bullet connectors, seems like it would be easier for them to directly crimp the wire to the resistor. It has a hard plastic tube to protect the resistor and then meltable tubing covering the whole thing and melted to the wire at both ends. Anyway, pretty sure it's just a standard 1/4w resistor.


Anyway, photos to see some weird Honda engineering for the TRX250X.

[Ranvier]

Still no objective test of this “racing” cdi?

[ps2fixer]

Still no objective test of this “racing” cdi?

Sorry, I still plan to test it. Weather got super cold in my area (lows of -12F) and I don't have a good shelter to work in. After that my carb needed work since it wasn't running right even with the OEM CDI. I got the carb off, cleaned out and tossed in an ultra sonic cleaner to get it cleaner. After that I needed carb parts because the bottom plug was corroded so badly it had a hole in it. Carb is all set to install now, but I had a drive failure in my computer, so that delayed me on things for about a week or so. I'm self employed so my ebay listings are critical for my income, and I'm running out of stock atm on a lot of stuff so that's been my priority atm to get fixed.

After I catch up on things, I have to re-pin my 350x CDI connector to be stock again since it's still swapped around for the TRX250X CDI, install the carb and tune it on the stock CDI and probably run though the CDI's again.

I also thought of a new metric to add into the mix, but I'm not sure if the numbers will even mean anything. If it gives interesting results, I'll post my findings.

I do have one of Mike's CDIs on hand, it's just still sitting in the box till I get a chance to test it. From what I've heard (though him) most people love it and say it's better than stock, but there's no hard numbers that I know of to cover the claims. I know the buttometer isn't always calibrated right and can give biased results.

[ps2fixer]

Well I'm back at it after quite a long break from this thread. Carb has been cleaned though an ultra sonic cleaner, had to replace the bottom main jet access bolt, and I rejetted a step richer for both pilot and main since my engine has a DG exhaust. It's still a little lean, but probably close enough to be safe to ride it this way. It's just a tiny bit cold blooded now for reving, before it was worse for reving cold and had to run it on choke longer, so have to adjust the needle height and such to see if I need to go up another jet or not.

Carb tuning aside, I finally got to test out Mike's CDI, results are in the first post.

I also made an adapter so I could take voltage readings while the engine is running. Right away I can see a common thread on CDI's that were not designed for this machine, they are drawing more power from the exciter coil (more amps), so the peak voltage is reading lower. For Mike's CDI, I'm not sure if this is an effect of having too large of a capacitor, or if it's the fact it's a digital CDI and the micro controller needs too much power at higher rpm. My guess is the capacitor though since the Orange China CDI had similar readings and effects but I'm pretty sure it's an analog style CDI.

Anyway, sorry it took me so long to get my machine back together, life kind of happened and got side tracked for a while.

Long story short, at least for the 85 ATC350x, stick with the OEM CDI for best performance unless there's another CDI that comes out on the market that's truely designed for this machine, or Mike works out the kinks with his CDI (I've heard it works fine on the 86 style stator).

Speaking of the 86 stator, I'd be interested in the stator + fly wheel if anyone has one laying around. Bare min I'd like to borrow it to get some voltage readings and such to see exactly what the differences are, and stator wire size measurements and resistance to estimate exciter coil length (aka wraps) vs the 85 stator. I think the big problem is that all these CDI's are designed for more modern machines, which standardized at a higher exciter coil output spec than what the 85 350x was designed with, but sounds like the 86 350x received the updated design. Weird thing is, the 85 and 86 CDI are the same part number.