Faulty Fuel Level Strips

Discussion in 'GS Boxers' started by ARG, Jan 22, 2012.

  1. Dorian

    Dorian huge carbon footprint

    Joined:
    Jul 7, 2004
    Oddometer:
    770
    Location:
    WA State
    Thanks again for the info Joel. I get the gist of your answer but lack the equipment and knowledge required to perform the repair myself (although based on your instructions I could probably work that part out). Maybe you could offer a fuel strip repair service?

    The fuel gauge on my '08 12GS seems to work (for now) but I get misleading readings for the first 40mi or so after I fill-up, then it settles down. I use the trip odo anyway - but I'm not keen on an always-on low fuel warning light.

    Cheers, Dorian
    #61
  2. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    We could do this. Fooling the system into thinking there were a half tank would be hard as the resistance of the strip rises as the part of the strip thats above the fuel level heats, but I se no reason we couldn't fool the system into thinking and always displaying the tank as full.

    All I would need is for someone with a working strip to drop by my house for an hour. I know the resistance of the sensor and heater circuits as it's burned into my brain from checking so many, but I need to know how much current flows on the heater circuit to size that resistor.

    1: 1/8th watt 68 ohm resistor for the sensor. (if memory serves, measure while bike is here)

    1: X watt (probably 1 watt) X ohm resistor, with thermal properties that raise it's resistance a bit as it heats up for the heater circuit.

    Shape the ends of the resistor leads to line up with the bikes plug.

    Pott all but the end of the leads in a little Epoxy

    Include 2 high quality black zip ties, one to keep the thing plugged into the plug on the bike, the other to zip the whole thing to the fuel pump harness right next to it.

    Installation would require removing 3 screws. No need to even open the tank.

    Put it on Ebay for $5 and free shipping and I could move up to eating Ramen brand noodles instead of Nissin :)

    I never hung with the R1200GS crowd here, All my friends have F8's or G650's or S1000's. Whisper to someone local in St. Louis with an R12 to come by and if really desired, I could mock something up.


    As to fixing the old sensors. I don't want to get within 1000 yards of the anger surrounding that when I only have a test pool of 3 bikes that my fix worked on, and personally I think BMW should just man up and build a conversion. The new floats aren't compatible with the current firmware, but BMW could tweak the firmware and do the conversion.
    #62
  3. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    37,869
    Location:
    OAK
    Joel, are your sure that the sensor isn't capacitive?
    #63
  4. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    100% certain, but capacitive is what I had assumed till the first time I looked at one. It's just resistive, thermally modified during the heater cycle.

    LOWWWW technology, easy to fool.
    #64
  5. bboshart

    bboshart Been here awhile

    Joined:
    Jan 31, 2010
    Oddometer:
    144
    Location:
    Atwood, Ontario
    So, if you could make a device that receives the signal from the new float and converts the signal to simulate the old fuel strip you could make some serious sales? :deal
    #65
  6. drmajor

    drmajor Been here awhile

    Joined:
    Feb 1, 2002
    Oddometer:
    564
    Location:
    Charleston, SC
    Joel.

    Looking into the Fluke 1520. Discontinued. Replacement is the 1507 $498.00..

    Ouch...

    Can you think of another, cheaper way to shock this thing?

    What was the voltage & amperage output of the 1520?

    Can this be replicated with something else???

    Thanks for your input!!!
    #66
  7. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Tell me about it! My 1520 recently broke so going to have to buy a 1507 soon as there are many applications for it in my field.

    Very low current, high voltage. I can think of many sources of that kind of voltage around the average house, but none with that low of current. Look in electronics supply catalogs. You want 1000 volts 1 - 20 mA max current, then a 40 pF 1000 volt capacitor.

    I don't know the exact ideal specs because I never researched it as a repair technique. With Volvo it was a curiosity but they weren't interested in repairing them as they wanted a fuel level measurement device that didn't break in the first place so ultimately rejected using strips all together.

    My FSE with BMW was fascinated, but that poor guy already had a zillion things on his plate and has since been promoted to North America parts boss.

    In any case, can't blame BMW for not mocking up a procedure that required having a megger in dealer service departments. Master techs would try using them in places they shouldn't and blow up bike computers lol.

    OOO, perhaps a real small compact fluorescent light, like one that claims to put out the same light as a 15 watt incandescent bulb.

    If you cozy up to the right dealership, it's against the rules to give them to you but they will have dozens of strips in boxes saved for warranty dept should they ever call for them to be shipped back which they won't.

    Strips fail so often they actually litter BMW service departments trash cans and counters.

    Dumpster dive? Just ask the service manager or warranty clerk the right night to do it and their trash will be filled with strips and other defective parts, also technically against the rules as you are supposed to destroy all old parts but nobody does.
    #67
  8. drmajor

    drmajor Been here awhile

    Joined:
    Feb 1, 2002
    Oddometer:
    564
    Location:
    Charleston, SC
    #68
  9. grpweld

    grpweld Been here awhile

    Joined:
    Apr 19, 2010
    Oddometer:
    324
    Location:
    Fallbrook, Ca. San Diego
    My gauge doesnt drop until I've gone about 110 miles, & when the tank is empty it still shows 3-4 bars, Is this a faulty fuel strip also?
    #69
  10. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    It's hard to say. That wouldn't be the common way for a fuel strip to go but can happen.

    Just as likely tha adaptation is off, or perhaps the strip came off its hanger and is setting on the bottom of your tank.

    Without hands on, meter on diagnosis, I don't know. Sorry.
    #70
  11. qman8

    qman8 GS09

    Joined:
    Jan 24, 2012
    Oddometer:
    550
    Location:
    new philadelphia, ohio
    Dude ... Be careful 20mA is 20 watts at 1000v . There is no way that fluke is putting out that kinda power... That's why the cap. Is needed so the voltage stays higher for a little longer ... A megohm meter is going to have very high internal impedance and will provide 1kv only to the highest of resistive loads...and the strip is not!... At the very least u need to put a several Meg resistor in line BEFORE the parallel cap .
    #71
  12. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    The flukes 1520's short circuit current is 2 mA, it will be less when it's anywhere near 1000v and yeah, the cap is in parallel to add current for a brief period of time + the megger has some capacitance itself.

    You want a short circuit current no greater then 20 mA at maximum.

    But yes, just a few mA will stop your heart if it goes through it just right, and 1000v can certainly get it there, don't kill yourself :)
    #72
  13. qman8

    qman8 GS09

    Joined:
    Jan 24, 2012
    Oddometer:
    550
    Location:
    new philadelphia, ohio

    Well yeah..I wasnt really thinking of HIS safety when i wrote that...:huh...but it looks to me that if he connects that source up he will fry the strip for sure, and probably light the fuel too! The 20mA figure is misleading. Thats probably a stated maximum rating, for very short duration on the Fluke...and in reality probably never approaches .01 of that...but im guessing.
    #73
  14. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    20 mA, as in 0.020 amps, is the MAXIMUM short circuit current you would want to send through the fuel strip sensing circuit.

    Voltage is immaterial to melting or burning things up as long as your power source is limited to no more then 20 mA

    You are not trying to heat the circuit trace with current, you are trying to create an arc across where the circuit trace is broken between the laminated layers of plastic.

    The strip sensing circuit, when working, is around 2,700 ohms. A power supply that is "current limited" to a maximum of 20 mA will fall to 54 volts once the broken part of the circuit is bridged. (ohms law for voltage I*R=E (I=intensity of current, or amps. R=resistance, or ohms. E=electro motive force, or voltage))

    So we now know from ohms law that a voltage source that is limited to a MAXIMUM of 20mA will in fact supply 54 volts once the higher voltage arcs across the break and creates a carbon trace or whatever it does that fixes them.

    To find Power, which it the case of this resistive circuit will also be heat, ohms law says (I*E=P (P is for power, otherwise known as watts)) 54*0.02=1.08 watts.

    1 watt is not going to melt the strip or the sensing circuit in my opinion.


    Now, if the strip were delaminated, or there was so much capacitive current that it melted through the strip, you bet, thats a good ignition source for anything flammable! So is the fuel pump which has carbon brushes that arc the whole time the bike is running, as are static discharges that occur between the body of the tank and metal parts such as the pump all the time.

    Why don't gasoline tanks blow up all the time??? Simple, fuel vapors inside a tank are normally not in a flammable state because they lack oxygen. In other words that might seem more understandable to mechanics, the fuel mixture in the tank is normally to rich to ignite.

    The lower flammability limit of gasoline is 1.4%. Uncompressed, anything less then a 1.4% gasoline to air mixture won't ignite.

    There is also an upper limit. Anything more then 7.6% fuel to air and it also won't ignite.

    Atmospheric pressure and many other variables have an effect, but only small ones in regard to conditions a fuel tank will be in.

    On a cool spring day, gasoline mixtures in a fuel tank will be upwards of 50%

    Now, if the tank is nearly empty and you have been blowing air in the open cap for a while, you might get a boom.

    If it is negative 50 degrees F out and the cap is open, you probably will get ignition because this is so cold that the fuel is not evaporating, but at any sane temperature, gasoline rapidly evaporates and turns any sealed container into a vessel that is WAY to fuel rich to ignite let alone explode.


    This principal is why as a child you could make molotov cocktails with gasoline, light them and let them burn all day without a boom till you tossed them and broke the glass. This is why when you poured too much alcohol in your spud gun, it wouldn't light. This is also why when you drilled a hole near the bottom of a 55 gallon drum, dumped a bunch of trash, wood and gasoline in, and lit it from the hole, you lost all the hair on your arm and burning logs and trash rained down in a 75 foot diameter. The top of the vessel was open so you DID get a flammable, or in my case, explosive mixture.

    A note on flight 800 who's center tank exploded in NY as I recall. Jumbo jets run on kerosine. Kerosine does not start evaporating in mass till around 130 F, so jet plane tanks are normally too lean to ignite but when nearly empty and heated by all the things that heated that tank climb into the flammable range. This would have never happened if planes ran on gasoline.

    Lastly, a note on all the cars that blow up on TV. That is TV, it is not real and not one single person from Hollywood would know reality even if it crawled up their ass and died. Those cars explode because special effects guys hook remote detonated explosives to their fuel tanks. The boom is from the bomb, the fire afterward is from the fuel that was blown around after the bomb flattened the fuel tank, which is the same thing 60 minutes did when they and Ralph Nader were trying as usual to destroy American car manufactures on behalf of the Asian manufactures they are paid lobbyist for, but I'm getting far afield now

    At normal temperatures in a sealed gas tank (cap closed), gasoline will not ever burn or explode.

    I have been wrong before, so if you think this may be one of those times and you don't wish to become a screaming alpha, remove the strip from the tank before working on it, font create any sparks are have any static discharges as you do cause once the cap and fuel pump access is removed, theres plenty of air to support combustion in the tank :)
    #74
  15. ARG

    ARG Adventurer

    Joined:
    May 27, 2011
    Oddometer:
    54
    Joel,

    Will you be going to the BMW MOA rally in MO this July? Sure would like to meet you !!!!!

    ARG
    #75
  16. qman8

    qman8 GS09

    Joined:
    Jan 24, 2012
    Oddometer:
    550
    Location:
    new philadelphia, ohio
    Wow...keystroke quantity impressive...but keep in mind that current is entirely dependent on voltage. Thats where your logic fails. Point is that HIS circuit is NOT the same as the megohm meter and i want to keep the poorbastard from at the very least blowing the strip and worse injuring himself or the bike....for anyones sake please DO NOT USE the circuit shown.
    #76
  17. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Voltage does not determine current. In a DC circuit V/R, or voltage divided by resistance determines current. This is not my theory or logic, it is an accepted law of physics but was first postulated by George Ohm way back in the 1830's.

    Now, where I think we are failing to communicate is on the subject of resistance. There are two prime points of resistance in this instance. 1 is the fuel strip, which if working properly would have resistance value near 2,700 ohm.

    If an unlimited current supply at 1000 volts was hooked to the fuel strip sensing circuit, 0.37 amps would flow. This is WAY to high and would equal 370 watts! ( 1000/2700=0.37 ) ( 0.37 X 1000= 370)

    But there is a second resistance and that is the resistance of the power supply. When I say "current limited to 20mA max" or "short circuit current of 20mA max" I am using common terms to say "use a power supply that won't put out any more then 20 mA under any circumstance" Or for those not into metrics, 0.020 amps maximum.

    If the power supply can't put out more then 20 mA, then the voltage will be 54 volts because that is the voltage that will flow 20 mA through a 2,700 ohm load.

    The circuit linked in the post way above is a regulated 1000 volt supply, and without testing it or spending days working up the math, it's hard to say what it's short circuit current will be but likely higher then desired. It won't be insane because it is powered by an 800 mA 12 volt transformer. If efficiency was perfect, and it never is, this circuit is designed to output 9.6 watts (0.8 X 12 = 9.6)

    9.6 watts, less losses is it's operating power. Thats too much power and is going to have WAY more then 20 mA of short circuit current, so there we agree, though the circuit is most defiantly NOT 20 watt. Power out can not be greater then power in or we wouldn't need energy anymore.


    So to you qman8, or anyone else contemplating this project, don't proceed further until you understand what the terms "current limited" and "short circuit current" mean. I am suggesting a power source with a minimum of 1 mA of short circuit current and a MAXIMUM of 20 mA and an open circuit voltage around 1000 volts.

    Also make sure the circuit does not have greater then 40pF of capacitance. Think of a capacitor as a battery that can instantly discharge all the energy it is storing. Regardless of the current available from your power source, it will eventually charge the capacitor up to the source voltage. A bigger capacitance will blow shit up.

    Is everyone with me on the whole "20mA current limited" or specified another way that means the same thing "20mA short circuit current". It is important and what I think qman8 is not understanding.
    #77
  18. qman8

    qman8 GS09

    Joined:
    Jan 24, 2012
    Oddometer:
    550
    Location:
    new philadelphia, ohio
    "Voltage does not determine current"...Dude! it DOES..thats fundamental electronics 101....Your stated formulas are correct, but resistance is normally constant. Stated another way, you can have voltage without current flow, but you cannot have current flow without a voltage differential, so again, current is completely dependent.
    First, the 2700 ohm figure is meaningless, because the fuel strip is broken, so resistance = infinity (theoretically). Now you are going to introduce enough voltage to achieve an ionizing arc, then your resistance drops dramatically . Thats when the internal impedance really comes into the picture. So, when that arc forms you want just enough current to do the job,but not enough to damage the device. With the megohm meter, the voltage will plummet when the arc starts, the cap will try to keep a constant voltage and discharge to do so, and then the system drops to a steady state current. Here is where the damage can occur. SO, I'm saying keep enough resistance in series to keep that damage from happening...what that minimum resistance can be is determined by the max current rating of the strip. If you know that, then you can estimate a substitution reisistance.
    #78
  19. drmajor

    drmajor Been here awhile

    Joined:
    Feb 1, 2002
    Oddometer:
    564
    Location:
    Charleston, SC
    Wow,
    Seems there should be a cheap-safe way to zap this thing.

    I would probably take the sensor out of the tank anyway. The Fluke is way to $$$.
    #79
  20. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Oh jeeze, Im arguing with a physicist, engineer, or someone who likes to read a lot. I'm familiar with all 3 and have argued with all 3 while teaching "electronics 101", to bring everyone up to the same playing field before diving into instructing industrial automation and other classes using real power in real classrooms. The electricians that were taking the classes for their certification or continuing education requirements were easy. The physicists, engineers, and electronics majors were a pain in the butt.

    So lets stew this down qman8

    Are you suggesting that a "current limited power supply" that has a short circuit current no greater then "20 mA" is going to fall to a steady state of dissipating more then 1.08 watts of power through this fuel level sensing strip?

    Are you suggesting that a current limited power supply with a short circuit current of no greater then 20 mA will fall to a steady state of greater then 54 volts once the arc ends after creating a carbon track and powering into a 2,700 ohm circuit?

    Are you suggesting that 20 mA at 54 volts will dissipate more then 1.08 watts into this circuit?

    Are you suggesting that 1.08 watts is more then the sensor circuit of this fuel strip can safely dissipate forever?

    Have you ever seen a resistance fuel strip?


    As to the strips normal operating envelope, I have never bothered to measure it. It is likely the same as the other 99% of sensors in the automotive world, 5 volts from a 10k ohm power source.

    What it's normal operating voltage, current, or power dissipation are is not material because we are trying to do something that is not normal and will do something that 5 volts from a 10k ohm source will not, fix the strip!

    P.S. After the whole part about an arc creating an ionized field with the impedance plummeting to essentially zero (absolutely correct though it is still in series with the 2.7k ohm of the rest of the circuit), and went on to talking about keeping enough resistance in series to prevent damage, was that you agreeing with me while studiously refusing to agree with me?

    The power supply which is "current limited" and has no greater then a "20 mA short circuit current" IS THE RESISTANCE IN SERIES TO LIMIT CURRENT.

    If you are suggesting bad things will happen if a current is steadily passed through the strip sensor circuit greater then it it normally is from the bikes computer..... Well thats probably 5 volt through 12,700 ohm,or 0.39 mA.....

    I would suggest the following. Grab a broken strip, dealerships are littered with them so shouldn't be hard, and find out.

    I suppose an EE would want to see a white paper. I have never bothered to try to find out who makes the strip BMW uses but it's probably Bosch, Siemens, or Marelli.

    A physicist is going to want to explode it. I'm down with that, take video :)

    A person that reads a lot, not sure, but look at an actual strip if you haven't already which I'm thinking you haven't. It,s a fat trace with a lot of surface area.

    So lets have it, what that I have specifically stated do you disagree with? And if I get off my duff, grab a strip, pass 20 mA through it for an hour and it doesn't melt, do I get a prize?

    What about if I run a loop of 22AWG wire into my fuel tank wrap some silicone tape around the individual conductors as they pass through the gas cap so they don't short there, shut the gas cap on the wires and pass 20 amps through the wire so that it heats white hot and burns in half. Do I get a prize if it doesn't blow up or ignite the tank?

    I passed more then 20 mA continuously through the volvo strips which look about the same and we tried everything under the sun to ignite fuel vapors inside a sealed gas tank, so i'm cheating but would still like a prize and will put it all up on youtube for a good enough one :)
    #80