Questions on checking a CDI module w/multimeter

Discussion in 'Thumpers' started by Sa-tevo, Aug 30, 2004.

  1. Sa-tevo

    Sa-tevo Jetmechs' Winston Wolf

    Joined:
    Jul 30, 2002
    Oddometer:
    505
    Location:
    Near the compound of Peachtree City Georgia
    I was helping a friend this morning who has a Yamaha watercraft (Jetski?) that will rev out but has lost power. Idles and revs fine, but no muscle. In the course of poking and prodding, I checked the CDI module with a multimeter and had many resistance measurements that were out of whack with the Clymer manual we had to work with. The Clymer manual has a matrix of wire combinations to check and expected readings to have on the meter. The wire combinations that should read open when checked did, but the resistance readings where all way above the book ranges (I.E. book says 30 to 40 kilohms, read 3 megohms). Same results using another meter, and both meters checked ok with known resistors. I could see having one squirrelly reading, but all ranges were out of parameters. The stator and coils read ok, within spec values.

    Anyone with experience checking CDI's have any suggestions?
    #1
  2. cpallen

    cpallen Nearly Adventurer

    Joined:
    May 26, 2003
    Oddometer:
    1,023
    Location:
    Seattle, WA
    Are you using a digital or analog voltmeter? It's been awhile, but I think that you would need a digital volt meter with a fair amount of current available to bias the semiconductor junctions on diodes, thyristors, transistors etc that would be found in a solid state device.

    Can you compare readings with a known-good cdi? I suggest you look at the air and fuel supplies too.
    #2
  3. Sa-tevo

    Sa-tevo Jetmechs' Winston Wolf

    Joined:
    Jul 30, 2002
    Oddometer:
    505
    Location:
    Near the compound of Peachtree City Georgia
    Both meters were digital, a Simpson and a MAC automotive. The shop manual didn't give a spec for what meter to use. I'd like to check it again using my Fluke (which was at work), but my friend was eager to buy another CDI in the hopes of getting it running well again, so we will be able to compare soon. I was just reluctant to pass judgement on the CDI

    It was a sudden drop in power. The fuel filter was replaced, no help. I didn't check the inlet to the impeller for obstructions, but we did check the engine coupler and it was in good shape.

    An idea that comes to mind is that the wire terminals are corroded behind the crimps, but if it was corroded bad enough to see on a meter, the engine wouldn't run.
    #3
  4. Frank Warner

    Frank Warner Traveller

    Joined:
    Mar 14, 2002
    Oddometer:
    6,593
    Location:
    Sydney, Australia
    Warning - different models can be very different ...

    Another thing .. these tests are with the battery turned off?

    The book might say "greater than 30 kohms"? or ">30kohms" in that case 3 Mohms is fine.

    Does the book have any checks for continuity? (>5 ohms) If you have a reading of 3 mohms instead of 30-40kohms you may have a broken wire.

    ---- All the above are just ideas. I'd not be testing a CDI (or EFI) with a meter except for the supply to the unit, the sensors connected to the unit and the output devices connected to the unit. Once those are established good then I'd be looking for waveforms with a CRO ... sorry 'merican Oscilloscope...
    #4
  5. motu

    motu Loose Pre Unit

    Joined:
    Dec 28, 2001
    Oddometer:
    4,915
    Location:
    New Zealand
    I'm with Frank,but I'd use my DSO - check inputs and outputs,if the correct signals are going in and nothing is coming out I'd kinda think the CDI unit is stuffed.
    #5
  6. Sa-tevo

    Sa-tevo Jetmechs' Winston Wolf

    Joined:
    Jul 30, 2002
    Oddometer:
    505
    Location:
    Near the compound of Peachtree City Georgia
    Frank and Moto,
    Thanks for chiming in. I really liked Moto's scope shots of ignition activity from a few years ago. On this job I left my personal Tektronix scope at work, thinking it was more straightforward.

    I checked the excitation coil on the stator, the common grounds, the coil and the overtemp circuit that puts the CDI into limp mode, with no problems found for book readings, continuity or for insulation resistance to ground. The CDI is powered from the stator, like a dirt bike.

    The book gave several wire combinations that should have readings in an expected range, such as from 30 to 40 kilohms, where I read over a meg in those combinations. To tell the truth, I don't think the CDI is the cause of any problems, as my experience in the past with RZ350LC engines has been that CDI is all or nothing, since spark ignition engines rely so much on proper spark at proper time. This is really acting more like a drivetrain fault, such as a slipping clutch on a motorcycle or car.
    #6
  7. motu

    motu Loose Pre Unit

    Joined:
    Dec 28, 2001
    Oddometer:
    4,915
    Location:
    New Zealand
    I've never been able to figure out CDI units or solid state regulators by checking resistances,without knowing exactly what is embeded in the epoxy it's just guess work.From the wiring diagrams of a couple I looked at basicaly you have an earth,power supply,some sort of charge coil,this should put out a scary high voltage,then a low voltage input to trigger the capacitor and last the output to the coil.Actualy an old analoge multimeter is good for this,you don't really need to see exact voltage readings,but the activity - you will be seeing the vertical trace of the scope with the needle swing,a DMM just tumbles numbers.

    You just dunno what's inside these things - one time we were having trouble with a GSX750,ran,but not quite normal.The previous owner had fitted a new CDI unit - so it couldn't be that eh? Anyway,someone picked all the epoxy out and we found the outputs bridged - it was firing all four cyls at once.
    #7
  8. cpallen

    cpallen Nearly Adventurer

    Joined:
    May 26, 2003
    Oddometer:
    1,023
    Location:
    Seattle, WA
    This is someone's interpretation of the CDI on an early Transalp FYI.

    [​IMG]

    There is some theory of operation and o'scope waveforms here:

    CDI repair information in German

    I maintain that the input impedance of the voltmeter and the current output of the ohmmeter function of an analog multimeter (unless it's a VTVM) will be too low to provide accurate voltage and resistance measurements in a solid-state device (when measuring p-n junctions). Sure, viewing outputs with a 100 mhz digital storage scope would be preferred but many of us don't have this kind of gear in the garage (or down at the dock).

    Someone drove my brother-in-law's jetski up on the beach and sucked a moderate amount of sand into the jet pump and I think that the symptoms afterwards were very similar - no power due to scored impeller or cylinder resulting in very low output volume.

    Good luck.
    #8
  9. Frank Warner

    Frank Warner Traveller

    Joined:
    Mar 14, 2002
    Oddometer:
    6,593
    Location:
    Sydney, Australia
    And here I think is the real problem...
    #9
  10. kevbo

    kevbo Rubbery-Lip Flappin' PHI

    Joined:
    May 29, 2002
    Oddometer:
    1,478
    Location:
    Albuturkey
    I'd be looking for a spark leak on the HT side.

    Get a cheap AM radio with an earphone. Tune it to static between stations, so you can hear the spark noise. If the noise changes a lot when the problem happens, you have a spark leak.

    It takes a lot more voltage to fire the spark under heavy load. It can be fine with light load, and totally not work under heavy load.
    #10
  11. Dolly Sod

    Dolly Sod I want to do right, but not right now

    Joined:
    Aug 18, 2003
    Oddometer:
    15,250
    Location:
    Richmond, Va
    Something to consider.. On many outboards that run on a magneto ignition system, the stator has two windings with two different outputs (Not counting the charging system for the battery, making it three outputs)..

    The first output is for the idle ignition. It will put power out from idle up to about 3000 rpms.. After that the voltage coming from the idle side of the stator drops from around 400 volts to somthing like 50.. At about 3000 rpms the other side of the stator, the high speed side, begins to pick up and supply voltage to the cdi from 3000 rpms up..

    I have seen on many outboards that when the highspeed side goes, the motor will still zing up to redline with no load, but as soon as a load is put on the motor, she'll go no where above 3000 rpms or less, and have no power.. But she'll idle all day long and sound great...

    But in my experience, an ohm's test of the CDI, switchbox, and coil, were always tests to confirm the demise of one of the units. In the shop, we always tested the outputs first. If the stator is putting out 400 volts at cranking speed then it's good. If the coil is getting 400 volts at cranking speed then the switchbox is good, and probably the trigger is good as well. If there's spark at the gap tester, the coil is good. Next we would test the high speed side of the stator, which usually put out about 70 volts at cranking speed..

    If we weren't getting the correct voltages, then we would fall back on the ohm's tests to verify what we pretty much already knew..

    But of course you cannot test voltage from a stator with a regular multimeter..

    You need a peak reading adapter to read the peaks of the voltage pulses coming from the stator.. Peak reading adapters plug right in to a regular multi-meter.. Also Mercury marine sells an analog meter that has a peak reading adapter built in.. (60$)..

    Maybe this helps?
    #11
  12. Dolly Sod

    Dolly Sod I want to do right, but not right now

    Joined:
    Aug 18, 2003
    Oddometer:
    15,250
    Location:
    Richmond, Va

    Another suggestion.. I know that many cdi's have diodes inside them.. so perhaps you are trying to read ohm's the wrong way across a diode? Try switching the leads from your multi-meter.. You have to apply a current to measure the resistance, and maybe your meters send the voltage from the opposite leads..
    #12
  13. Sa-tevo

    Sa-tevo Jetmechs' Winston Wolf

    Joined:
    Jul 30, 2002
    Oddometer:
    505
    Location:
    Near the compound of Peachtree City Georgia
    Thanks for the added input. The manual has the wire pairs checked both ways, swapping leads of the meter. My friend has located a CDI for $75USD, so it could be worth it as a spare on a ten year old jetski. Another co-worker also got the similiar readings using other meters. Maybe the manual is wrong, which is known to happen.

    Will report furthur.
    #13
  14. Sa-tevo

    Sa-tevo Jetmechs' Winston Wolf

    Joined:
    Jul 30, 2002
    Oddometer:
    505
    Location:
    Near the compound of Peachtree City Georgia
    The spare CDI didn't materialize. Everything was put back together again and it ran fine.

    In my field it's referred to as "Reracked, checks ok". Who knows, maybe the connections just needed reconnection after ten years.

    Thanks all.
    #14
  15. motu

    motu Loose Pre Unit

    Joined:
    Dec 28, 2001
    Oddometer:
    4,915
    Location:
    New Zealand
    Yeah,sometimes I reckon it's just a call for attention,like kids.You give them some quality time and are rewarded with smiles - it really was you,take the credit.
    #15