CT70 Lighting Coil Exposed

Discussion in 'Tech Area' started by allenp42, Sep 4, 2018.

  1. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    As a pre-winter project, I wanted to revisit the lighting coil dilemma to see if there is an easy way to convert a CT70 to a 12v full wave bridge set-up, and back it up with a few tests to make sure that it's safe and will work long term.

    My first step was to unwind a 2 wire lighting coil (Green & Yellow) and try to put a "number" on each winding. Here's what I know as of today, which includes a few surprises that I was not expecting.

    The coil consists of 2 coils, connected at a common point (ground), asymmetrical by design, with 2 outputs. The yellow supplies the juice for the AC components (headlight, speedo lamps), and the green wire for the DC circuit (battery, TL, stop light, horn).

    The "yellow winding" (AC circuit) is the outer layers on the coil. I counted ~149 turns, 33.5' of wire, DC resistance somewhere between 0.5 & 0.6 ohms.

    The green winding (DC circuit) is the inner section. ~252 turns, 43' of wire, 0.6 to 0.7 ohms. I put a range on both readings because my meter would sometimes dither between the 2 numbers. Makes sense because my fluke 77 will only read to 1 decimal point.

    The wire measures 0.68mm diameter on my not so trusty HF caliper. The wire size is the same on both windings. In looking at a few magnet wire charts on-line, the wire is 22 gauge or very close. I checked the math on resistance per foot of 22 ga. and it's in range with the resistance measurements.

    Lighting Coil Sketch.jpg
     
  2.  
  3. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Here are the pics of the disassembly:

    From start of the unwinding:
    Stator 1.jpg

    Closer view
    Stator 2.jpg

    With the outer layer of tape removed:

    Stator 3 tape removed.jpg

    Unwound to insulation barrier between the 2 coils. This is splice between the 2 coils, which is also grounded with the tab under one of the stator mounting screws.

    Stator unwound to 2nd coil 2.jpg
    Stator unwound to 2nd coil.jpg

    Stator with 1 turn remaining on the green (ac coil):
    Stator with 1 turn left.jpg

    All done. With the wire wound up on a section of PVC in case I need to take another look at something.

    Stator unwound.jpg
     
    wanrep and racerx like this.
  4. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Some reference materials, and what I found surprising.

    * More windings in the DC circuit than the AC side

    * The total number of turns is almost a dead match from a page in the service manual

    * I measured the wire using a HF caliper is VERY close to one of the charts that shows the dia for 22 ga. with insulation. It may be 22 ga., but was expecting it to be between a couple of AWGs which is typical for wire that I assumed had a mm2 (squared) size.

    I am by no means an expert on magnet wire used in motors, transformers, coils etc., and it has been 30 years since this was covered in school, so I have some learning to do. However, I have a friend who is a geek on this this of stuff so I'm going to pay him a visit in the next few weeks.

    Good reference on a Mitsu stator
    https://lilhonda.com/index.php?threads/stators.21608/#post-170970

    And what started me down this path.

    https://lilhonda.com/index.php?threads/cl-70-voltage-regulator.21827/

    Screen Shot 2018-09-04 at 2.38.37 PM.png
     

    Attached Files:

    #3 allenp42, Sep 4, 2018
    Last edited: Sep 4, 2018
  5. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    Thinner guage wire would increase voltage, but at reduced amperage. More feet of wire, paired with higher impedance...a narrower highway for those electrons.

    Thicker wire would deliver less voltage, at higher amperage...the opposite of thinner wire.

    Based on what I've noted with 12v alternators, thicker wire seems to be the engineers' choice. I think in terms of "cubic windings", the total mass of copper wound on the armature. Any increase in total wattage output will come from increasing the total quantity of copper used, in the optimal guage. That may take some trial-and-error testing. There is room to accomodate a larger coil OD...and it's worth exploiting
    It will be interesting to see if you "hit a home run" with the initial trial rewind/reconfigiration. That could open the door for an inexpensive electrical upgrade/update...and all the newly-available options it would make viable.
     
  6. kirrbby

    kirrbby Well-Known Member

    Joined:
    May 28, 2009
    Messages:
    8,734
    Likes Received:
    1,076
    :) :) :) YAY!

    I'm excited about this thread. It's awesome that a electrical dude is doin it too, so you can use some science to back up and kinda confirm what you find. I'm gonna be all big ears and eyes... look, listen, and think...twice, then try to ask only GOOD questions :)
     
  7. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Yep, you're spot on as always.

    Would be my choice as well. I have a suspecion the current design provides much higher voltage than is really needed....and is compromise solution for providing "some" voltage at low RPMs, trying to keep both circuits satisfied, and at same time, keep the cost low.

    What's bugging me is both circuits have 6 volt loads, yet the turns ratio is different, so why not just use both wires, remove the ground and be done with it. It may turn out that way, but want to do some experimenting first.

    We'll see. Once I get that second opinion, I should be ready for cold and rainy days to see what I can figure out.
     
  8. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    I believe the answer is...cost, simple as that. Compare the CL70 electrical system, more complexity & cost. At the OEM level, we're talking million$.

    Fast forward a half-century. Now, we have viable LED technology, SLA batteries and everything is 12v flavor. With everything powered from the battery, it won't matter if the system runs at a loss, at idle. As long as the battery can charge, above ~3000rpm, everything will work...well
     
  9. Kansas Corey

    Kansas Corey Member

    Joined:
    Jul 12, 2018
    Messages:
    37
    Likes Received:
    2
    allenp42 - awesome project!! Good luck
     
  10. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Thanks KC. I had hoped to have some rough testing done by now....and did. But, I found out that one of my DVM's is WAY to generous when measuring current flow. So all I have right now is bunch of voltage measurements, which I trust, but no current measurements. I'll get back on it during another rainy day.
     
  11. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    A crude-but-effective way to test current is adding known load(s) to the system, then measuring voltage. If you've rewound this coil and it's now a single, length of wire...as much as the armature will hold...I'd expect to see a little more than double the stock output. That should be somewhere in the 50-60W range. That also happens to be close to the current draw of most automotive headlight bulbs. If this setup will power, say, a 55-65W bulb and still deliver 12v+, who cares how much more capacity it might have? Under that scenario, I'd call it a success.

    OTOH, if it comes up short of that mark, then it does matter what the actual numbers look like. 35W+/- is still enough for some really good lighting, way better than stock, when it's DC.
     
  12. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    That is exactly what I did to verify my 2nd DVM is way off on current. I just happened to have a 7.4 ohm 50 watt resister in my parts bin. Right now, I am using a test plug I built from from scraps that inserts in-between the stator and main harness on HK0. I can check E and I easily in the shop. When I saw 10 amps DC side @ about 6000 rpm, I knew something was wrong. It's reading about 5x too much as compared to my Fluke.

    The next time I get back on this project, will probably just use the resistor, and maybe another BAR as the load and calculate the current...and power.

    Just trying to get an idea of what I'm dealing with to know if it's worth the time and effort to rig up a cheap and dirty fixture out of some trashed cases and an old crank.
     
  13. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    24W vs 120W...ummmm...yeah.... something has a foot in more than one reality.o_O
     
  14. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    I've had an interesting morning and one of those days where I leaned something new. I've made a ton of checks, but the easy part first.

    Will a K0 stator plate work on an H flywheel with a H crankshaft? Short answer - Not on my H crank and flywheel. To explain further, the points sit further out toward the end of the flywheel to the tune of about 5mm. I mounted both in a set of cases with an H flywheel and the points were riding up on the big end of the points advancing cam. They never closed. I made a few measurements, but the easiest way to explain is by showing a pic of of the advancing cam and the 2 plates.

    I put a piece of tape showing where the points had rubbed the cam. The tape is cut to the same width as the phenolic block on the points. There is ~2mm between the edge of the wear marks and the end of the grind on the lobe.

    I remember reading somewhere that someone had success. Can't dispute it, but it would not work on any of my parts.
     

    Attached Files:

    #13 allenp42, Oct 15, 2018
    Last edited: Oct 15, 2018
    racerx likes this.
  15. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Next up are the coils used on a K0 and HK0. I'm 99% sure they are the same...but the wiring that leaves the lighting coils appears to be different (HK0 vs K0) leaving the coil up to the 4 pin plug.

    Based on the parts manuals I have and Partzilla, the K0 and HK0 share the same Primary coil. Honda p/n 30540-041-005. This coil also fits many other models. The coils look the same, measure the same resistance, mount the same, have the same dimensions. Pretty strong evidence they are the same:)

    Lighting coil, just the coil and not the short wiring harness that connects to it - Look the same, same dimensions, same resistance but different part numbers. I suspect the different p/n is because the wiring is not the same. The parts manual I have does not show the bullet connector for the primary coil being used on K0 as it is on the HK0. Partzilla does not it either. This may explain the different p/n?

    Can't do much more poking around beyond K0, HK0, and HK1 because none of the manuals I have nor does Partzilla show a breakdown of the stator coils from K1 - 79.

    I always leave room for error, hence 99% is about as high as I go.
     
  16. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Voltage and current measurements. I am going to sleep on this one before I post some prelim values. I have said "what the hell" more times than I can count today.
     
  17. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    I have tried to embed as many comments, thoughts, etc. into the results I obtained while making a few baseline checks on stator from my HK0. To make these checks, I made up a patch cord so I could disconnect the Y & G wires at will, make checks under load and no load and about every combination in-between.

    Not sure I trust my cheap tach, nor is all that easy to maintain a steady RPM.... For that reason, the RPM readings are ballpark. I've been "pondering" the results and still plan to reach out to a close friend who has more experienced than me on coils and transformers.

    Bottom line - There may be a way to convert to 12v with the existing coil using LED lighting.....but you'll have to be careful with current draw.

    For sure, will need to a test fixture before I do much more testing! It's a pain doing it live on a bike - I need more hands, or more accurate meters, or both. I learned a lot, and also learned that I have 2 meters that don't measure AC current very well:)
     

    Attached Files:

  18. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    If I read correctly, these look like reasonable values, for the stock configuration. Rewound, as a single/continuous winding, with a floating ground and as much wire as real estate will allow, I think you'll see more than both of the AC outputs (HL & charging) combined, coming from an electronic reg/reg unit. IOW, the two ends of the coil windings would each feed one input to the bridge diode rectifier, with the reg/rec unit, itself, providing the ground. This type of full-wave reg/rec unit requires a floating ground.

    Just to be clear, the stock coils (there are, in fact, two coils wound on the same armature) are both run to ground, each with a single output. Thus, they can only be used as two coils and the only practical way to get DC is with a (highly inefficient) half-wave diode rectifier. By "highly inefficient" I refer to the fact that <50% of the AC output would be lost as waste heat.

    Props to you for the great start on this project.(y)
     
  19. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    I think you are reading them correctly Bob.

    My intent in the "Test #1" was just to see how high voltage goes with no load. Serves no purpose other than to satisfy my own curiosity.

    Test # 2 was to see how much power could I extract from the lighting/charging winding with no HL. Honestly, was a little surprised to see values that high.

    Test #3 was to see how much power was available when both windings are under load. I wanted to see them individually and how they look combined. In this example, I think I overload the DC side (green) a bit, I think. I would guess the max power available for the DC side to be in the range of 8-10watts, and total ~28 watts.

    Test # 4 was just to see the how much power could be extracted from both coils in series. Was a little surprised, but as reality sinks in, I think I'm beginning to understand why. I know the coil has 400T of #22ga. wire, so the sweet spot is probably going to be x numbers turns with 22 ga. or 20 ga. Got to have the voltage go high enough in 3000-4000 RPM range to deliver some current, or else one would have to run around all day at WOT to keep the battery charged:)

    For what's it worth, I'm slowly becoming a believer that some designer in Japan did their homework when he/she designed this system....including the use of 1/2 wave selenium rectifier.

    FYI - I already have a coil with a single winding. Now as soon as I stop "testing coil on a Stihl 025", will get back on on this project.
     
    kirrbby likes this.
  20. racerx

    racerx Administrator
    Staff Member Administrator

    Joined:
    Jul 10, 2005
    Messages:
    13,093
    Likes Received:
    1,168
    I agree. That said, the reasoning involved more than lighting performance alone. The CL70 is the proverbial fly in the ointment, full-wave DC powered everything...using selenium rectifier technology, albeit in a different configuration. Selling tens of millions of units (the same 1/2 wave alternator + magneto ignition was used on 49, 72 & 89cc variants, worldwide), every dollar saved at OEM level translated to seven, at the dealership(s), we're talking about megabucks. Most of these electrical systems went into bikes that really didn't need lighting so, from that perspective, the corporate decision made sense.

    Without going too far off the deep end, yet again, for the moment suffice it to say that this boils down to wattage. If there's enough to power X number of watts total current draw, with the stock setup, then there should be the same number of watts available as full-wave power. It's almost certainly going to take some experimentation to find the right wire gauge, to realize the desired voltage. FWIW, I think you're right about the existence of a "sweet spot"; I'll define that as the total mass of copper wound around the armature lamination stack. Based on your results, thus far, that'll probably mean more turns of slightly thinner wire, to "tune the system" for higher voltage...analogous to tuning an engine for hp vs torque. Oh yeah, this would be easier if 6V electrics were still viable...with aftermarket parts support. Unfortunately, 1968 was a long time ago...
     
  21. allenp42

    allenp42 Active Member

    Joined:
    Apr 1, 2014
    Messages:
    686
    Likes Received:
    178
    Still plugging away on the little project. Picked up a precision ohm meter for low resistance values. This is shedding a little light on some differences. Have rechecked everything I have and documented for future needs. Working on the test fixture at the moment. Thus far, I know of 2 motors that won't drive the test fixture :(....but I just stumbled across one that will. Need another pulley, which will be here in a week or 2.

    Have a few coils wrapped with X number of turns just so I can (hope) get an idea of "which way to go". Regardless, I plan on getting baseline info from several models (H, K0, K2, and a 77). Trying to test 2 of each to get more of a "warm fuzzy" feeling.

    Trying to do this on the cheap is slowly coming together. Will probably be January before I have anything to share.
     
    kirrbby likes this.

Share This Page