1995 R100RT Classic - Sudden engine cutoff

Discussion in 'Airheads' started by Muravey, Aug 14, 2019.

  1. Muravey

    Muravey Folk you.

    Joined:
    Jun 22, 2011
    Oddometer:
    29
    Location:
    Bucharest, Romania
    Alright, I'm back. I have a hunch the Hall sensor is fine. More on that at the end. So today I did the following:
    1. checked the rectifier bridge and it's ok. I got between 509 and 533 Ohms on each of the six big diodes and none of them let current through from the other end:
    2. checked (with my multimeter set on VAC) the alternator output between every two of its three AC spade connectors. It checks out, as the reading quickly jumps over 13VAC after tickover rpm:
    3. VDC measurements (it should be said that my tiny multimeter isn't a very quick reader, as it has some latency)
      • rectifier bridge output: 12.72V off, 12.70V main switch on, engine dies at 12.64V
      • battery: 12.72V off, 12.70V main switch on, engine dies at 12.56V
      • ignition coil positive terminal (i.e. switched voltage): 12.63V with the switch on, so I've got a 0.07V drop compared to the above readings. Is this a significantly high drop to worry about?
    4. I just found a wire with torn isolation. I don't know what it does, but it goes together with the blue wire into some insulation which also houses the regulator wires above the engine. The black and blue wires go through a white connector. If I unplug the connector, I'm apparently disconnecting the starter.
      IMG_20190912_180118.jpg
    5. Found out that the blue wire goes into the rectifier bridge. What does it do?
      IMG_20190912_180658.jpg
    6. The black wire goes to the starter solenoid. I'm on the wrong lead here, but I'll fix the wire nonetheless.
      IMG_20190912_181609.jpg
    7. Repaired the wire in a slightly overkill manner:
      IMG_20190912_183704.jpg
    8. So I decided to test the Hall sensor's sensitivity to the alternator's EM field: I unplugged the regulator, ran a DF line directly from the battery through a 21W bulb and fired up the bike. The voltage rose slowly (should I be concerned? I thought it was supposed to jump rather quickly), but it did so all the way up to about 13,8V (after which I stopped the bike to prevent overheating). The bike ran fine. No sign of spark cutoff. I don't claim to understand how EM works here. But I feel that the sensor should've cut off from the EM field at that charging value, had it been excessively sensitive. Opinions?
    Obviously, if I plug everything back together, the old symptom comes back: ignition dies when the charging light goes out.

    So what the heck is going on?...
    #61
  2. shaner1100gs

    shaner1100gs Been here awhile

    Joined:
    Nov 23, 2005
    Oddometer:
    494
    Location:
    Laurel, MD
    I think the blue wire goes to the charging light.
    #62
  3. Muravey

    Muravey Folk you.

    Joined:
    Jun 22, 2011
    Oddometer:
    29
    Location:
    Bucharest, Romania
    You're right, sir. I just checked the diagram, even though some minor stuff isn't exactly what I've got.

    R100RS-RT Wiring Diagram - public.jpg

    Life is easier when you look at diagrams before taking the world apart. Somehow, I've been troubleshooting this problem ass-up. Anyway...
    #63
  4. ccmickelson

    ccmickelson MonoMania Supporter

    Joined:
    Dec 11, 2013
    Oddometer:
    2,778
    Location:
    Fanta Se
    In case you want it, here’s the scheme for the monolever RT from the Haynes Manual.
    B2F3DF95-8264-4856-97F4-9192A3928C41.jpeg
    #64
    Muravey likes this.
  5. Warin

    Warin Retired

    Joined:
    Aug 30, 2012
    Oddometer:
    11,704
    To test the diode board you need to test ALL the diodes .. not only the 6 large diodes but the smaller diodes that supply D+.

    The blue wire is D+.

    D+ gets supply from;

    1) The alternator/battery warning lamp - this should be at start up only.
    2) The diode board - when the engine is running

    D+ supplies;

    The voltage regulator.

    So the 'blue wire' as D+ has 3 connections.
    #65
  6. boxerboy81

    boxerboy81 Stay Horizontal

    Joined:
    Jul 22, 2009
    Oddometer:
    4,069
    Location:
    Melbourne, Oz
    That's a fantastic diagram. There are some differences in the headlight wiring..Euro v USA, otherwise I think it's pretty much bang on.
    #66
  7. Warin

    Warin Retired

    Joined:
    Aug 30, 2012
    Oddometer:
    11,704
    The battery, alternator, diode board and regulator are exactly the same. Variations on the key switch number of positions 3 or 5 depending on market.
    #67
  8. 100RT

    100RT Long timer

    Joined:
    Nov 4, 2009
    Oddometer:
    1,846
    Location:
    NEWTON FALLS, OHIO
    Are you a member of the Airheads forum?
    #68
  9. Muravey

    Muravey Folk you.

    Joined:
    Jun 22, 2011
    Oddometer:
    29
    Location:
    Bucharest, Romania
    If the question was for me, I - unfortunately - am not. Will look into it, thanks!

    I also replaced the regulator relay with a Mobiletron VR-B191. No change.

    So, after exhausting all the tests I could understand, afford and carry out myself, I finally decided to let karma work instead of my wallet - which is powerless at the moment anyway. To this end, I asked an acquaintance with an R65 to lend me their bean can. I had forgotten this to be an option. As irony would have it, their bean can looks horrible compared to my shiny unit. Here it is on the donor bike and on the sidewalk, next to mine:

    IMG_20190914_192309.jpg IMG_20190914_192316.jpg IMG_20190914_193744.jpg

    Well, whaddya know? @Ola A was quite right and it turns out that my sensor was indeed inadequate. Check out my bike working fine with the donor bean can installed:



    Karma figured I've had enough stress for the past few weeks, so the kind gentleman allowed me to keep the bean can until I fix mine. I already ordered a Siemens HKZ121E sensor from AliExpress, cheap Chinese knock-off or not. This and the HME301 seem to be what people use to replace the original Honeywell 2AV54. At $10 to be delivered in about a month at my door, it's worth a shot whatever happens.

    So I don't claim to understand what the problem is. As I figure it, magnetic interference should have disallowed me from generating 13,7 VDC with a separate DF line running directly off the battery with a 21 W bulb on it. But it is what it is.

    My bean can now lies disappointingly - in all of its aluminium shine - on my living room table. Will take it apart again and see what I can find. Maybe I pinched some wire during reassembly, though I find that hard to believe. I will keep you posted.

    IMG_20190917_013616.jpg

    Thank you all for your help!

    P.S.: The replacement bean can seems to have one heck of a sticky advance. Tried to spray some lube on its counterweight lobes through the maintenance port as a very-temporary fix, but it only had limited effect. Might be the springs. We shall see.
    #69
    durtwurm and UnclePete like this.
  10. Warin

    Warin Retired

    Joined:
    Aug 30, 2012
    Oddometer:
    11,704

    Once you fix yours, you can fix theirs. You can even clean it up for them :-), thought that might show up the rest of their bike :augie

    $10 is cheap for a HES. Please let us know how it goes..
    #70
    Muravey and durtwurm like this.
  11. Muravey

    Muravey Folk you.

    Joined:
    Jun 22, 2011
    Oddometer:
    29
    Location:
    Bucharest, Romania
    Weeell, it's been a very long time since I last posted. Sorry for that! On the flip side, I did get around to taking my friend's bean can apart, cleaning it out and lubing it nicely back together - it too had had its sensor replaced at some point, by the way.

    ulmeanu_bean_can_collage1.jpg

    Although I haven't got a photo of it cleaned, it did work flawlessly afterwards. So, for those of you with "sticky high idle syndrome", get your bean cans cleaned and greased!

    Since we're on the topic of good news, the aftermarket Hall Effect sensor that I'd bought off AliExpress arrived in a few weeks:

    AExpress_HKZ121E_collage1.jpg

    Keep in mind that I ordered a HKZ121E unit (cheapest on that posting), but I got a unit labeled HME101. Apparently, they're all the same. Anyway, I liked the design of the new sensor already for two reasons:
    1. The unit is fully enclosed in plastic and has got a rather simple shape, meaning that it's easy to clean off any magnetic shavings that might've wandered about.
    2. The mounting posts are made of rather brittle soft aluminium, but they are hollow. Makes for a very easy threaded mounting solution.
    So, exactly one month after you posted, @Warin, I got around to taking my bean can apart again and fitting the new sensor in there. I kind of broke the "new old" one off, although - in hindsight - I could've been more careful with it. Let's just say I wanted to let off some steam for all the trouble. :D

    I first tried to rivet the new sensor in place, but found the aluminium prongs to be too brittle for that. Instead, I broke the protruding bits off and found a pair of computer case screws that tapped themselves in. Check it out:

    AExpress_HKZ121E_collage2.jpg

    Keep in mind that I had previously had to dremel out the metal plate under the original rivets, so as to gain access to their ends. This previous job made screwing a breeze this time around. :evil

    After installing everything back together, I put the bean can on the bike and she fired right up, without any hiccups. If anything, while adjusting the timing advance (only by ear for now), I felt like I could get a smoother idle than I ever could before. The video is in Romanian, but the airhead speaks crisp Bikesperanto: :lol3



    I put a few thousand kilometres on the girl since then. Even managed to crash it a little. :fpalm But the sensor has performed flawlessly so far. :gyro

    So there we go: $8,45 sensor with $3,05 shipping. Here is the link to the product page again. Remember: select the HKZ121E version when ordering.

    Thanks again for all the help and take care! I'll let you know if something comes up.
    #71
    aptbldr, Square1 and Warin like this.
  12. boxerboy81

    boxerboy81 Stay Horizontal

    Joined:
    Jul 22, 2009
    Oddometer:
    4,069
    Location:
    Melbourne, Oz
    Good job. That's the same sensor available here in Oz from Jaycar.
    #72