Fuel Pump, another potential fix

Discussion in 'Crazy-Awesome almost Dakar racers (950/990cc)' started by men8ifr, Jan 7, 2011.

  1. men8ifr

    men8ifr Been here awhile

    Joined:
    Jan 28, 2008
    Oddometer:
    383
    Location:
    Midlands UK
    Bought some new contacts for my worn out fuel pump and read about several upgrades that can be done to the points to make them last longer.

    1st - Is fitting a FET to replace the points in carrying the main current to the coil, this way your points should just about last forever but requires quite a bit of reworking, some hard to get specialised parts and I'm unsure how easy it would be to reverse the procedure if the FET failed in the middle of no where.

    2nd - fitting a diode across the contacts

    3rd - fitting a capacitor across the points

    I like options 2 and 3 because they are simple and easy plus if there is a problem with a part you could quickly remove it with sidecutters and the pump is back to the original spec (handy if you were in the middle of no-where)

    My old contacts made quite a spark when the pump was ran so I used these to test 2 and 3 since you would be able to see if there is any difference, less sparking would mean less wear.

    With the diode I tried it both ways round and did not notice any difference either way except one way it got hot.

    With the capacitor 100uf there was a big difference and the sparking was greatly reduced, I would say to about 1/5 of the original amount. With a 220uf capacitor I noticed no difference (bit strange but there we go)

    I took a video of the capacitor working, when I push my finger down it makes the contact so the capacitor is connected.

    The actual capacitor is reasonably small so very easy to fit in the case, one leg can be held by a screw already there so only really needs one leg to be soldered on - pretty easy to do.

    <object height="505" width="853">


    <embed src="http://www.youtube.com/v/gExG6FrX9cQ?fs=1&hl=en_GB&rel=0&color1=0x234900&color2=0x4e9e00" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" height="505" width="853"></object>
    #1
  2. sonoran

    sonoran friend of P

    Joined:
    Sep 28, 2009
    Oddometer:
    26,525
    Location:
    Feenicks
    Seems like a simple and inexpensive way to lengthen the life of the stock pump. Thanks for sharing!
    #2
  3. men8ifr

    men8ifr Been here awhile

    Joined:
    Jan 28, 2008
    Oddometer:
    383
    Location:
    Midlands UK
    No problem - that's why I like it it's simple.

    The larger capacitor was 2200uf not 220uf - I had a look again today.
    #3
  4. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    men8ifr:
    good to see this examined. Some comments tho.
    There is already a diode across the coil. You said across the points- not sure what that will do but it must have been a big one to just not pop one direction withpoints open. The cap idea has been considered and it should do about the same thing as the internal diode. Ought to put an o-scope on it but I would expect to just see a momentary 1v (appx) pulse for the voltage drop across the internal diode when the points open. That 1v will resonate at some frequency dependent on size but regardless it's only going to charge to the diodes bias.
    Not to say that all possibilities have been realized but we (me included) are just "polishing shit" (sorry). The pump is just crap. Yes it can be modified (FET or my still available SS circuit) or just repaired frequently (points), and last fine. BTW the FET can be bypassed fine in the field.
    Easiest way leaves a little current going where it dosent belong but it's nothing. I got into the modification as an academic challange - (simple things for simple minds perhaps) and I have resistance to the alternatives. My bike still has the generation 2 kit and lots of miles.
    Perhaps there's an alternative but to get philosophic the design is at least 50yrs old, and poorly done at that.
    Always good to keep an open mind. thanks
    #4
  5. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    Wrong -there would be a 12v potential charge on the cap. NTL it would be useful to see an o-scope read on the cap. With the internal diode and the very low resistance of the coil not sure what one would see.

    further thought: addressing: why not a capacitor across the points? Does look like the right thing, but it also looks like a huge cap is needed to effect desired levels of protection. Even a 2.2mf charges pretty fast at the starting 6amps that exists in the circuit. How much is really required to extend the point's life in application? I have no idea.

    Guess we all will just have to wait and find out from you.
    #5
  6. men8ifr

    men8ifr Been here awhile

    Joined:
    Jan 28, 2008
    Oddometer:
    383
    Location:
    Midlands UK
    I was wondering why this works and couldn't quite figure it out, then a guy at work mentioned this was like a condenser on an ignition points system, here's what I found out about a condenser:

    When you remove the distributor cap from the top of the distributor, you will see the points and condenser. The condenser is a simple capacitor that can store a small amount of current. When the points begin to open, the current flowing through the points looks for an alternative path to ground. If the condenser were not there, it would try to jump across the gap of the points as they begin to open. If this were allowed to happen, the points would quickly burn up and you would hear heavy static on the car radio. To prevent this, the condenser acts like a path to ground. It really is not, but by the time the condenser is saturated, the points are too far apart for the small amount of voltage to jump across the wide point gap. Since the arcing across the opening points is eliminated, the points last longer and there is no static on the radio from point arcing.

    from http://www.familycar.com/Classroom/ignition.htm

    So it would seem that this is more of a bone-fide correção than I realised.

    Why on earth mitsubishi didn't do this in the first place I don't know, capacitors cost a few pence!
    #6
  7. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    Engine ignitions actually "ring" it does what is said in milliseconds and reverses. If you examine the size of the condenser (capacitor) and note the current is about 6 amps that cap is saturated so fast that the break still poses a problem. (BTW ringing creates a longer duration spark and is a good thing there) The other choice which doesnt "ring" is the diode suppressed concept that mimics the capacitor concept by letting the solinoid coil discharge to itself. And this doesn't saturate. This diode suppression should fix the problem but it doesn't. It is used on other solinoids like starter solinoids, glowplug relay solinoids.... I do agree Mitzu should know better- point fuel pumps had this problem 50 yr ago.
    So will the cap AND the diode - I believe so. Is it the best fix? IMO no. I'd go to the crankcase driven "vacuum" pump before just relying on the cap and diode. IMO the "easiest to greatly improve" are the two fixes I've discussed - the SS and the FET. The mechanical relay might be good too but capitilizes on the fact that it's a minute fix when it goes.
    #7
  8. zeegman

    zeegman Been here awhile

    Joined:
    Sep 14, 2007
    Oddometer:
    399
    Location:
    San Jose, Ca
    Hi RideWestKTM,

    I did the FET mod and it has been working fine thru water crossings, and heat, and a big offroad crash for a few thousand miles now. What is the FET mod field bypass you mentioned and does it require less equipment/tools/space versus taking a spare pump?:ear

    Thanks,
    Mike Z
    #8
  9. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    My point was the FET wasn't irreversable - the points are still there. If one wanted to do it wrenchless I bet it's possible by bringing out wires but what I was thinking during that entry was remove the pumps plastic cap, cut out the FET and reconnect the point wire. The little resistor could be removed or not- wouldn't matter.
    The idea of carrying a 2nd pump is IMO just silly when there are limp home/repair options so easy. The EFI not so but for the 950s - I'm just not worried.
    #9
  10. ATAK

    ATAK Been here awhile

    Joined:
    Nov 9, 2006
    Oddometer:
    212
    Location:
    arnhem, netherlands
    #10
  11. crazybrit

    crazybrit Long timer

    Joined:
    Mar 16, 2006
    Oddometer:
    8,556
    Location:
    Portland, OR
    Dunno, whats a "print". Translation seems a bit off to me. For technical stuff like this, that makes it more worrying.
    #11
  12. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    A picture is worth a 1000 words regardless of language.
    Ya ne ponnimayoo
    Ich verstea nicth
    no capiche
    no comprendea
    #12
  13. crazybrit

    crazybrit Long timer

    Joined:
    Mar 16, 2006
    Oddometer:
    8,556
    Location:
    Portland, OR
    Not sure who this is directed to, whether you're saying it has pictures and thus needs no words, or it needs pictures? I failed to follow the links last time, had I done so I'd have got to this, which has pictures:

    http://forum.lc8.nl/viewtopic.php?t=3072&postdays=0&postorder=asc&start=0

    Interesting Facet mounting pic on page 3, never seen anyone mount behind the airbox like that.

    Page 4 has a post showing a failed unit that went diving. Translation isn't easy to understand (for me)
    well, the last line is easy :D
    #13
  14. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    picture: thanks

    One look at the pictures and I knew - optical - this is the right way to do it. Gefer bench toped this method and I ran off and did it my way (2 ways) I saw this optical somewhere but when I did it didn't seem ready for sale. This looks good. #1 IMO.
    #14
  15. crazybrit

    crazybrit Long timer

    Joined:
    Mar 16, 2006
    Oddometer:
    8,556
    Location:
    Portland, OR
    atak: Is there a way to contact the guy without subscribing to that website?

    I still think the Google dutch translation isn't too good .... is there any English version of the PDF self-assembly instructions.
    #15
  16. gefr

    gefr Life is a trip

    Joined:
    Nov 26, 2005
    Oddometer:
    4,228
    Location:
    East Med, Greece
    I do not understand how the device with the infrared sensor moving is going to be fitted on the fp.
    I thought of purging the moving contact device alltogether and using only the vibrating rod.
    However the moving point is attached to a device that builds in a time lug. This is needed to give time to the solenoid to pull the membrane of the fp to fill with fuel .

    men8ifr I watched your video and noticed sparks coming from the contact of the capacitor to the bridge when you do not press it with your finger. That could give you misleading results when you do not press it, as increasing sparks???
    #16
  17. Werldcup

    Werldcup Capitan

    Joined:
    Jul 11, 2009
    Oddometer:
    12
    Location:
    San Diego, CA
    With the small amount of current getting discharged at the upper spark, I'd think it would actually decrease the spark at the points. Bend it out of the way and the points spark would be that much larger.
    #17
  18. ridewestKTM

    ridewestKTM Long timer

    Joined:
    Jun 7, 2007
    Oddometer:
    1,724
    Location:
    now in Katy TX
    I didn't look at the details just the fact that it used the optical trigger. Gefr makes two good points. There should be no sparking if done right. The optical switch should open a gate on a transistor with just a few milliamps just like my timer or the FET and points so no sparks. The other point is the lag. Without a "dead band" it could switch so fast that it wouldn't stroke. The "toggle" in the points do this for the stock set-up and gefr's set up used the points not the shaft. But because the draw-back is a near fixed duration (because regardles of fuel flow rate, draw into the pump is unchanging stroke to stroke) - a time "on" could be established even after breaking the light beam to ensure full stroke. So IMO this still CAN be a very good system. I haven't taken apart a facet or whatever, but it seems more like they are continous pulsating "on" like my timed circuit. IMO these pumps are less desireable not due to method but because -for some reason they do not match the same parameters that the mitzu does (by accident or design).
    #18
  19. ATAK

    ATAK Been here awhile

    Joined:
    Nov 9, 2006
    Oddometer:
    212
    Location:
    arnhem, netherlands
    #19
  20. crazybrit

    crazybrit Long timer

    Joined:
    Mar 16, 2006
    Oddometer:
    8,556
    Location:
    Portland, OR
    Yes I knew of that document. No way to translate pdf in place that I know of. You can only cut the text out and translate separately.

    This is the result. Translation isn't great (especially the technical/bike specific terms). I expect if you've already replaced your points or installed FET/etc you could make enough sense of it. If you've never taken the pump apart at all, I imagine a better translation would be very helpful. Someone Dutch, who speaks good English and has familiarity with the pump could probably clean this up quickly based on Google translation starting point and the Dutch pdf.

    ------

    Installation instructions DR_Bean petrol pump modification
    Print
    To print you need to assemble:
    - This guide!
    - Soldering iron (fair bit, so no soldering gun thick)
    - Piece of tin with flux (electronics solder)
    - Cutting pliers
    - Phillips
    - Allen 2.5
    - 7 key (preferably screw-cap)
    - Drop of loctite
    - At least one right hand and at least 1 hour time if you have the pump loose it.
    Read the whole story, and then follow the instructions step-by-step: this is really the easiest order! If you do not understand something, parts or tools have not, do NOT get to work but consult the LC8 forum, and get everything ready before you begin.
    1. Remove the pump: See LC8 workshop manual (forum). Bridging the service outputs with a short tubing, or make sure the pump is quite low, otherwise the gasoline are likely to damage the assemble or otherwise dangerous. For clarity, the rubber mount to the pump here removed, but this is not strictly necessary.
    2. Loosen Lid:
    - Screw the black cap of the pump with a crosshead.
    - Slide the cap over the wire (with outer cover) to the outlet so you have room to work.
    - Cut the blue outer sheath to the cable about 4 inches in:
    This gives your workspace; damage the wires.
    - Place the Gasket also as far as possible toward plug (this will be easier if you grease the threads here is to what WD40).
    3. Remove the entire gearshift (with worn spots) go the two Phillips-head screws. Also remove the paper below.
    4. Remove the soldering iron with the black wire that comes from the pump regardless of the switch contact (red arrow). The switching mechanism can be stored as a reserve.
    5. Cut (!) Now with a sharp cutter the blue-black wire by, 6-10 mm outside the black insulating sleeve (+ / - 25 mm from pump): see yellow arrow and circle.
    Now we build:
    6. Fitting the spacers for print:
    - NOTE: The spacers are required in the two holes in line with the axis, where the original screws were (the third hole for the screw the cap).
    - Under one bus is a spring washer and ground cable.
    - Connect the ground cable on the side of the third screw hole.
    - Among the other is a spring washer and a regular ring, see photo!
    - Put some Loctite on the threads of the spacers, and tighten HAND-fixed (5 Nm) with such a dopschroevendraaier-7. ATTENTION, The pump is just look, so do not turn too hard!
    7. Insert the blue-black wire (the cable with plug) in print:
    Turn the end of the wire neatly (no fraying) and insert hole where BAT stands. Insert the wire until the insulation in the hole!
    8. Solder the wire from the other side with some electronics solder.
    9. Mounting the board on the spacers
    - The large semicircular recess in the PCB is the side of the bolt hole for the cap: This is soon the black screw cap (see here how the cap comes out!).
    - Insert the short blue-black wire from the pump in the PCB hole (S +) next to the blue-black wire of the other hand, you've just soldered. (Red arrow-1)
    - Put the thinner black wire from the pump in the printgat smaller (S-) on the other side of the screw (red Arrow-2)
    - Turn the board using the two allen screws: Imbus 2.5. Under each screw a snap ring (see photo)
    - Quiet Solder the 2 wires from the pump onto the print. Let the solder flow through!
    Test: All connections are now on the board. Now you can just connect the pump (ignition off!) To the original plug on the engine. Make sure the pump and no conclusion can make prints!
    Switch on the ignition of the engine and the pump would be 5 seconds to make a quick popping sound: does not, then you have done something wrong! NOTE: If you have done something wrong, may be now fuse for the fuel pump in the zekeringbox inflated. Remove the pump, replace if necessary. the fuse and begin this step by step guide to walk again.
    Installation:
    10. Cap assembly:
    - The cap is now back on board and the wires: That fits all a bit tight, so do not use force, it can neatly fit as well fitted (possibly make. the threads and O-ring is very slightly greasy with WD40, so not too rough running during installation).
    - Push the cap back on wires up to the rubber. Press From inside (with a blunt screwdriver or similar) the rubber in the slot (Well ... not my tie Design: effe fiddling!). If the rubber is not good at session is, adjust the PCB is not soon! You can rubber on the outside look good when sitting on his location is: see picture
    - Now stop the pump with print in the cap, and pull the wire slowly. The semi-circular notch in the PCB to get where the screw cap gets stuck, see photo!
    - The ground wire, along the board, this is sufficient space, but beware that you do not pinch.
    - Put the O-ring into the rim of the cap
    - Press the cap completely and quietly on the pump. It should clearly fall into place, and the screw you the third hole in the pump to see. (Possibly cap slightly clockwise)
    - Screw the long screw the cap back in, hand!
    READY! Test the pump a little to the motor, as previously described and mount the pump in the engine.
    Finally: You can use sealant: In my opinion this is not necessary if the O-ring and Gasket well in their seats. But in any case DO NOT use silicone acid (smells like vinegar)! This affects electronics fast.
    #20