cdi baseline info - wires, waveforms, timings, substitution
Close
    
    
Page 1 of 4 123 ... LastLast
Results 1 to 10 of 31
Like Tree2Likes

Thread: cdi baseline info - wires, waveforms, timings, substitution

  1. #1
    Senior Member slowmod's Avatar
    Join Date
    Apr 2014
    Location
    nova scotia
    Posts
    112

    cdi baseline info - wires, waveforms, timings, substitution

    for a wide view, here is a good summary of the various types of "powersport" ignition systems.
    Powersports Ignition Evolution

    "making spark is easy- making it at the right time is difficult"

    overheard on the internet.

    -

    hi

    This thread is to provide and build baseline information about the tw ignition system,
    so that testing, replacement, and substitution becomes easier. not to mention possible.

    -

    my bike is a 1990, so some of this initial information is specific to the 1988-97 tw's.

    lets keep general cdi troubleshooting - "help my bike won't start !" , "hey friend have you checked that sidestand switch?" off this thread please.

    corrections, updates, overlooked resources, independent verification (with a real oscilloscope!?), data for other years, special skills, and confirmed sightings of successful tw cdi swaps with part #'s are all very welcome!

    thanks,
    slowmod

    -

    once upon a time,

    my bike ran erratically, yada yada, ended up being a bad cdi, yada yada. oem = $430 cdn plus tax. no.

    after months of combing the internet, it seems that cdi substitution attempts seem to fall into 3 catagories: 1) the many who tried valiently but eventually had to give up, 2) those rumoured few who lucked out, hacked something in and it worked, but it's not easily replicate-able on the tw for some reason, and 3), those who bought expensive programmable ignition units to squeeze out max performance, or went to truly great lengths to build their own, or swapped entire systems from similar bikes.

    good tw reference data at that last link, by the way; cdi #s, years, previous attempts, links, etc.

    however, none of these outcomes is of much use to those of us who just want something resembling a reliable stock working bike for cheaper, before the summer ends, ideally.

    so i set out to find what the tw offers up to a perspective substitute cdi. i am really hoping not to end up in catagory 1.

    -

    this first round deals with trigger and power; finding a cdi that also has a similar-to-stock timing curve- starting near 10deg btdc advancing up to 30, (a straight line, a rise then rolloff, or?) is the next layer.

    -

    scalpel please

    if one pulls out a dead tw cdi, what's left? (ignoring the interlocks and all that crap for the moment).

    upstream; the spark coil, plug wire, cap, and spark plug.
    downstream; the cdi power/source/charge coil/stator/magneto, and the pulser/pickup/trigger coil/stator.

    finding out exactly what those coils are putting out, and when, is kinda key to matching in a substitute cdi.

    as well, the possibility of powering a new cdi with 12v exists, although the older tw charging systems don't have much/any headroom for power consumption. swapping out the headlight with a new led type might free up enough power to run a modern 12v cdi.

    focussing on the pulser: different cdi's look for different types of triggers, (separate from when the cdi's should have them, and how long they delay once they have the trigger signal..).

    some cdi's look for a rising-plus-voltage-when-it-hits-1v (ie the briggs and stratton magnatron which also provides some timing advance, as the 1v occurs earlier with more rpm/energy); some cdi's look for the plus, or minus peak; square wave, sine wave...

    some cdi's look for a rising-minus-at-zero, and some look for a falling-plus-at-zero.

    (the gy6 cdi's, for example, according to the good folks at buggymart), who offer plug-and-play retrofit cdi service, though they haven't done tw's yet. i am in contact with them and will report back if we figure it out for the tw.

    -


    cdi/pwr/pulse schematic
    tw_cdi diagram_v1.0.JPG

    stators
    tw_coils1_web.JPG

    magnets
    tw_magnets_web.jpg


    testing, testing

    all tests done at cranking speed.

    first, i determined and marked actual physical tdc.

    led-as-timing-light is powered directly by the coils (with resistors as needed to drop voltage).

    shows the signal peaks.

    pulser coil

    grn+
    tw_grplus_wneg+web.JPG

    w+
    tw_wplus_grneg_web.JPG

    there is also wasted spark trigger occurring directly opposite, but the led is not pointed there so not visable.

    (continued in next posts, only 5x images allowed per post...)
    Last edited by slowmod; 07-22-2014 at 05:37 PM.

  2. #2
    Senior Member slowmod's Avatar
    Join Date
    Apr 2014
    Location
    nova scotia
    Posts
    112
    continued

    these are the source coil outputs, timewise. (peaks)

    brn+
    tw_brnplus_grndneg_web.JPG

    brn - 3172
    tw_brnneg_grndplus_web.JPG

    (red is very similar, as is brn to red. see waveforms)


    -

    waveforms:

    these pulser/stator signals are recorded as audio then "looked at" in audio mix program (details at end).

    voltage levels are not to scale, ie the pulser puts out ~5v peak, and the red wire is ~100v+ peak.
    the signal strength levels have each been adjusted while recording so they "fit".
    time (ie peak to peak, zero to zero, etc should not change).

    -

    for a tdc reference signal, i hooked up a mechanical wiper switch contacting at tdc,
    electrically isolated from the bike,
    switching a 1.5v bat to the audio.

    ref wiper bench
    tw_ref_wiper1_web.jpg
    the the two wires at the left are bridged by the contact point at the right when it comes around on the guitar, er, camsprocket.

    ref wiper installed
    tw_ref_wiper2_web.jpg
    the reference switch contacting (in this case, switching 12v to directly power the led) (w/450ohm inline).
    at cranking speed, ~350rpm, so that's ~10x individual contacts/light pulses we're seeing! pretty consistent.
    the image would be a little sharper but the camera lens is focussed past the switch.

    ref check @start
    tw_tdc_reftest1_web.JPG
    rechecked at end of session, had not strayed.
    (switching 12v to power the led).


    (continued next post)
    Last edited by slowmod; 07-22-2014 at 06:02 PM.

  3. #3
    Senior Member slowmod's Avatar
    Join Date
    Apr 2014
    Location
    nova scotia
    Posts
    112
    continued

    these may not be the actual polarity of any or all of the waves- if we figure out it's a different configuration i can generate new views.

    tw_statortest2.1_wideall_web.jpg

    tw_statortest2.1_wavetiming1_web.jpg

    -

    for a cdi looking for a falling-plus-at-zero, or a rising-minus-at-zero trigger, the tw pulser will serve up either ~63 deg btdc or ~2 deg btdc.

    -

    analysis/ thinking out loud

    we don't know what the stock tw cdi was looking for polarity-wise for signal or power.

    focussing on the source coil-

    keeping in mind that "ground" is not really ground, it's just a big wire with lots of different ac and dc electrical potentials with different polarizations passing thru it concurrently and, mostly, independently..

    12vdc; unregulated/unrectified 15-18vac, 30,000vdc (spark "return" to coil), at least 30vac (brn "return"), and at least 150vac (red "return").
    _that's why the little but important 5vac pulsar signal is not referenced to ground, it has a wire on each end going right into the cdi.

    in terms of talking about polarity, ground/frame and the ac output of the source coil, the frame of course would have alternating polarity.
    however, the peaks would occur at different times as the load "sees it" depending on how it is hooked to the coil output.

    the cdi would probably be looking for certain polarity or particular chunk of the wave at certain times; getting a blast/peak of source juice _just before or during spark output to optimize the cap discharge is the one i've seen mentioned a fair bit, makes sense.

    because electricity flows from neg to pos, spark plugs apparently fire better if the tip is neg; the energy bottlenecks there in a focussed jumping-off point.
    not sure how much that is practised, but i am thinking that the tw cdi may treat the frame as plus, and harvest (only?) the neg half of the stator wave from the brown and/or red wire to charge the cap. or not.

    the cdi may draw (ac) power from the both ground to brown, and ground to red. maybe as low and high speed, maybe also using brown for some sort of timing advance reference, maybe..

    or, perhaps the cdi's main power draw is from between brown and red, as a form of isolation, and the windings between brown and ground just drains off interference or such? are they "floating" the whole cdi power supply except for that winding/shunt to ground?

    interestingly, the factory manual only says to test brn to red, it is not concerned with ground...

    the newer tw's have an extra wire from the source coils, i wonder if it is an independent "ground"/common back to the cdi?

    some discussion of tw waveforms here.
    Another CDI question... - Yamaha TW200 Trailway Discussion Forum

    -

    because i had recorded the source coil signals via a transformer, i couldn't be sure or the "polarity" of those signals.
    that's where the led timing marks off of the brown wire come in - the polarity of the diode is a known; one can see, timewise, where those peaks happen, and cross reference that back to the waveforms.

    (the led (light emitting diode) blocks current flow in one direction, and makes light when it lets it flow thru in the other direction. that's it's polarity. essentially one is seeing half the ac signal: above or below zero). if one hooked up another led (oowh, a different colour..) oposite to led #1, one would see both neg an positive spikes at once on the camsprocket.


    -

    how these tests were done.

    i cleaned the surface around the cam hole and used water soluble kids glue-stick to attach a small piece of bond paper. then i found tdc with a wire in the plug hole and turning the crank with a socket wrench. made a mark. checked again. and again. and again. and again. good. and dead on at the crank T line. ok. lined up the paper degree-wheel cut-out with the tdc mark and glued it on.
    here's a printable cam degree-wheel pdf. kinda hacky, i mistakenly deleted my first one.
    tw_cam_degreewheelv1.0_web.jpg

    the cam wheel timing marks were lit by a ~4.5v led hooked directly to the green and white pulser outputs; or brown to ground(frame) for the stator test (with 500ohms added inline to protect the led).

    time exposures of 3-10 seconds while cranking. tired starter, no plug.

    the waveform signals were recorded onto a computer as several .wave files, via a small audio mixing board for level control and isolation. i used extra inline resistance (up to 200kohm as needed), and also used a small transformer from a 6v "brick" power supply (which gave 7.7vac from 115vac house power for example) to further isolate and drop the outputs from the brown and red stator leads.

    audio equipment is looking for levels that are nominally ~.5vac mic level and 1.5vac line/headphone level.

    the mechanical tdc switch was hooked up to a 1.5v battery and into one of the mixboard channels. it shows as a 1-2deg spike (or dip), depending on polarity. a lot of filing and fiddling to get it thin and narrow and accurate. pretty happy with the result, but "if i had to do it again" i would probably go for something bumping a microswitch, or ign points or such.

    i only have one mono channel into my computer, so i mixed signal (ie pulser) and the reference pulse into mono, and recorded that.
    the next signal i again mixed in with the reference pulse to mono and recorded that.

    then i imported the audio files into an audio edit program, lined up the reference marks, and looked at the waveforms.
    then i did screengrabs and added notation in "photoshop".

    i arrived at the waveform btdc values by measuring 360 deg worth of pixels, then the other points, then doing the math.
    they matched pretty good with the cam timing marks.

    this whole timing thing has maybe a plus/minus 2deg overall accuracy, which, if true, is not too shabby for a bubble gum-pantyhose-and solder type job.

    -

    that's what i got so far. thoughts?

    -
    Last edited by slowmod; 07-22-2014 at 06:07 PM.

  4. Remove Advertisements
    TW200Forum.com
    Advertisements
     

  5. #4
    Senior Member Landser's Avatar
    Join Date
    Aug 2013
    Location
    Martinstucky,Indiana
    Posts
    254
    Thank you,slowmod. I'm trying to understand all of this,with the intent to make some sort of successful substitution using another system(or parts of) to provide my TW with just a functioning ignition system. As you mentioned ,there have been(allegedly) successful substitutions ,but either the author failed to properly explain the procedure in enough detail,or the components used weren't sufficiently identified as to allow anyone else to know exactly what to source......yes,it(may) have worked for them,but replicating the substitution is still difficult because of that lacking details.....

  6. #5
    Senior Member supachip1's Avatar
    Join Date
    Apr 2013
    Posts
    299
    this has to be the most 'technical' post I've seen….i think it will take a few reads to take it all in..certainly more than a browse i just did…good info

  7. #6
    Junior Member mntw200's Avatar
    Join Date
    Jul 2014
    Posts
    17
    Slowmod - let's try to find info on the other years cdis we go through this process.

    Mine is an '87. I bought the bike recently not knowing TWs and thought it was a carb problem. Did the "The '87 Dance" as lizrdbrth calls it. Timing light showed a bad cdi, tried to solder any cold joints with no success. Need to find an inexpensive solution as you mentioned above.

    Thanks for your work so far.
    Last edited by mntw200; 07-24-2014 at 12:40 PM. Reason: spelling
    littletommy likes this.

  8. #7
    Senior Member slowmod's Avatar
    Join Date
    Apr 2014
    Location
    nova scotia
    Posts
    112
    Quote Originally Posted by mntw200 View Post
    Slowmod - let's try to find info on the other years cdis we go through this process.

    Mine is an '87. .
    that's perfect! you can check yours and then we'll know for the '87's!

    the led timing tests are really easy, you don't need to photograph (though that's not hard either).
    for either eyeballing or photos, you'll want a somewhat dark space, it's not bright like a full up timing light.

    any electronics type places can sell you led's and resistors, or you can find a cheapy led flashlight that takes 3x1.5v batteries (=4.5v), bypass any logic switch (one push on/bright, two push dim, etc), note the polarity the bulbs are getting from the batteries, and there's your tester for the pulser.

    adding 400-500 ohms resistance should let you test the cdi source coil output timing (lowvolt/brown wire, if there is one) as well, and also the resistors let you hook the led directly to 12v should you like.
    i am not sure of the wiring config on the '87, so meter it first. worst that will happen is the led stops working. don't lick the stator output wires while cranking.

    i seem to recall the 87 pulser wires are also green, and white. i think all the tw's are.

    if the pulsar timing is the same across the years, then we know that a 12v cdi, anyways, which works on one tw would work on all tw's.

    -
    Last edited by slowmod; 07-24-2014 at 07:44 PM.

  9. #8
    Senior Member slowmod's Avatar
    Join Date
    Apr 2014
    Location
    nova scotia
    Posts
    112
    here's a new "view" of the waveforms.

    tw_cdi_stators_v2.21.jpg

    in thinking about this some more, i realized that the tw source coil is grounded at one end, so thus it can only "present" to a cdi in that polarity, or phase i guess in ac terms. the new view reflects this.

    the fact that the tw pulser is not referenced to ground is good, it runs a cleaner signal, and polarity and thus trigger point can be swapped. many of the newer/cheaper bikes and cdi's seem to tag one end of the pulser to ground to save on a wire, and the cdi's have only one terminal for the pulser. with the tw pulser, one can choose which end of the pulser to tie to ground.

    timing of the source coils may not be important. the cdi may just store up what it needs powerwise whenever, it's getting 2x pulses per spark.

    question: how do "they" get twice as many pulses out of the source as the pulser? both coils look to be sitting similarily next to 4 spinning magnets.

    -
    Last edited by slowmod; 07-26-2014 at 12:19 AM.

  10. #9
    Senior Member Landser's Avatar
    Join Date
    Aug 2013
    Location
    Martinstucky,Indiana
    Posts
    254
    What are your thoughts/suggestions regarding (component) substitution,slowmod?

  11. #10
    Junior Member PushingBoulders's Avatar
    Join Date
    Jul 2014
    Location
    Virginia Beach, VA
    Posts
    8
    I was recently given a pile of TW because the previous owner just gave up on it. I reassembled the bike, rebuilt the carb, cleaned the tank, changed oil, and finally got the bike to fire. Barely. After following a bunch of threads on the subject, testing the pulse and source coils, testing all safety switches, I was fairly certain it was a CDI issue. In fact, after knocking the CDI around a bit, the bike actually ran well enough to zip it around the block a few times. But after the bike cooled down, it went back to its old ways. In desperation, I poured over the schematics for the 87 and subsequent models to see why the 87 CDI system is fabled to be so different from its later iterations. The biggest difference I noticed was that the later pulse coil white wire was bonded to chassis ground somewhere in the harness, whereas the 87 remained isolated. This means that universal AC cdi boxes that require a single-wire pulse signal seemed to be out of the question. With nothing left to lose, and a cheap $10 universal CDI in hand, I bonded the white wire to ground and was able to measure voltage between the green pulse coil wire and chassis ground while cranking. I wired in the universal CDI and the bike fired up and I was able to rev it through the entire RPM range with no stumbles or hesitation. I have been starting the bike and taking quick rides around the block for the past two days. It fires quickly and idles perfect every time (87-90 degrees F ambient temp). With only one previous ride on an "OK at the time" CDI to compare it to, the bike seems to run about the same. Perhaps I got lucky. Perhaps this may be a viable stop gap for those who want to keep riding while they save up for a pricey OEM unit. For sure it was a good option for someone who got a non running TW for free and had nothing to lose by trying it. It does come with concerns though:
    1) Non model specific timing curve. These cheap units are clones sold to fit anything from 50-250cc Chinese knock-off ATVs scooters and motorcycles. This may mean a loss of performance or perhaps even too much timing resulting in spark knock or predetonation (I have not noticed any spark knock or detonation and I don't have much of a comparison point performance-wise).
    2) Possible damage to pulse coil. I am no electrical engineer, so I have no idea what the long term ramifications of bonding the neutral(?) side of the pulse coil to chassis ground are. From what I can tell from schematics, the 88+ systems came like this from the factory with no ill effects. Only time will tell.
    3) Component longevity. Who knows how long this CDI will last, but at $10 shipped to your door, keeping a spare around isn't gonna break the bank.
    Last edited by PushingBoulders; 07-27-2014 at 06:08 PM.

Page 1 of 4 123 ... LastLast

Sponosred Links

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

Similar Threads

  1. Carb substitution works great! Cheap, too.
    By cnc200 in forum Technical Help
    Replies: 126
    Last Post: 05-02-2017, 07:44 PM
  2. more info please?
    By tazewellteedubb in forum Performance and Customization
    Replies: 19
    Last Post: 10-24-2012, 08:29 PM
  3. a little info please
    By Hutty in forum General Discussion
    Replies: 11
    Last Post: 04-04-2011, 06:59 PM
  4. Rats nest of wires behind light? Pics anyone?
    By brownie2 in forum General Discussion
    Replies: 3
    Last Post: 11-07-2010, 11:56 AM
  5. Stator wires
    By dawgs in forum Technical Help
    Replies: 9
    Last Post: 08-13-2010, 02:12 PM

Search tags for this page

250cc with pulse coil with red and green wire
,
cdi output wire ground signal
,
chinesse atv 150 cked stator good pickup good cdi good coil good still no spark why
,
dc cdi advance
,
gy6 blue cdi issue
,
gy6 new racing cdi will not fire engine
,
gy6 programmable cdi
,
gy6 spark knock
,
how to test gy6 cdi box
,
no battery cdi waveforms
,

swappable cdi

,
timing map blue racing cdi
Click on a term to search for related topics.