Welcome to Netrider ... Connecting Riders!

Interested in talking motorbikes with a terrific community of riders?
Signup (it's quick and free) to join the discussions and access the full suite of tools and information that Netrider has to offer.

Solved Temperature gauge over-reading?

Discussion in 'Technical and Troubleshooting Torque' started by jack_1313, Apr 5, 2016.

  1. Hello :)

    My Hornet 900 has recently begun showing extreme temperatures on the analogue gauge. By this, I mean that while I ride around town, the gauge will climb into the red, then beyond the red, and then all the way across to the stopper for the speedo needle.

    I don’t think the gauge is reading accurately for the following reasons:

    Firstly, the bike is not popping the radiator cap and loosing/boiling coolant. I’m not sure exactly when that should occur, but I imagine there’s no way the needle should be able to climb past the red without the cap giving way.

    Secondly, the bike isn’t showing other signs of overheating and the cooling system components all appear to be working fine. In other words, I’ve already inspected/tested the impeller, thermostat, fan, and coolant (ECT) temperature sensor.

    Thirdly, there are signs of electrical issues. The needle spikes a centimeter or so when the fuel injection primes (engine off) and when the engine is running. It also spikes once the fan turns on. The blinkers also cause the needle to jump up and down a little bit. However, periphery devices (auxiliary lights, heated grips) aren't causing the gauge to spike.

    The fan is switching on just as the needle climbs into the red (at which point the needle suddenly spikes).

    It looks to me like what is happening here is that the ECT sensor is working correctly and the fan is probably turning on at the right time, but other electrical systems on the bike are interfering with the signal before it gets the to gauge itself. I’ve already looked for loose ground connections and damaged wires and couldn’t find anything.

    Has anyone come across an issue like this before? Any ideas where I should be looking?

    (I haven’t opened up the instrument cluster itself yet.)
  2. is the fan turned on by the ECU or by a separate temp switch? have wiring diagram? (what year?)
  3. Sounds a bit like an earthing issue for the dash cluster.
    • Agree Agree x 1
  4. #4 jack_1313, Apr 15, 2016
    Last edited: Apr 15, 2016
    Thanks for your responses – sorry I went radio silent. I had to put the temperature issue on hold after my exhaust fell off :S

    The fan is turned on by the ECU. I do have a wiring diagram, although unfortunately I’m not very good at reading it.

    What I can see is this:

    The ECT sensor has three wires running to it. One is pink and white and runs directly from ECU – I guess this is the live wire? Another is green and orange, presumably a ground wire, and runs back to the ECU, with several other sensors and the fuel injectors splicing into it along the way. The other wire is green and blue and runs to the gauge.

    It looks like the gauge gets power from some universal source in the cluster (there’s only one wire coming out of it) and then grounds through the ECT sensor, which provides variable resistance. (Less resistance = higher reading.)

    So if the ECT sensor is good - as it seems to be - and assuming that the fan is actually coming on at the right time (it's coming on just as the gauge enters the red, which makes sense given that the gauge is reading high), then I think there must be an issue either along the gauge-ECT sensor wire or in the instrument cluster itself.

    I already bypassed the pin connector that this wire runs through in the headlight, which had no effect.

    I don’t think the issue could be in the cluster because the gauge is spiking a lot when the fan comes on, and no wiring related to the fan runs through the cluster AFAIK.

    I think the next step will be to run my own wire from the gauge to the ECT sensor – that ought to tell me if there’s some issue along the gauge-ECT sensor wire with buried in the harness.
  5. Just as a peace of mind thing it would be worth measuring the battery voltage at rest and when the engine is running, and when the engine is at cruise revs. Instrument spiking can be due to low voltages in your charge circuit so that motors starting etc have a greater effect. In addition to providing start power the battery acts as a smoothing device to hold voltages and deal with short term loads. The alternator deals with the averages and tops up the battery but is not good at short term loads. Instrumentation also can be out of calibration if voltages are out of specification.

    How old is the battery?
    Does the bike turn over and start quickly and reliably?
  6. And give the battery terminals a quick clean, just in case the contact is a bit dodgy
    • Agree Agree x 1
  7. yup Green/Orange is ground wire, as it goes to IAT as well as MAP and TPS
    Yellow/Red from the TPS and MAP is their +5V feed from the ECU (and their other wire is their output back to ECU). the Blue/White wire next to it in ECU connector is also power.

    then there's a gap in ECU connector.. the next group is low level inputs, ie MAP/TPS/cam position
    the Pink/White (pin 13) is the last of those, so is unlikely it is a power wire. more likely it is ECU coolant temp output wire.

    so the green/blue is output to gauge, Pink/White is output to ECU (for fuel/fan control) and green/orange is probably ground (can check that easily with multimeter).
    by output, I mean voltage is provided from gauge and ECU and the ECT has (dual) variable resistance.

    gauge cluster likely gets power from the Red/Green wire going from fuse to up near the speedo

    so if speed and tacho are not going weird, the power to all the gauges is probably ok, and as you think, something between the temp gauge and the sensor.
    sensor probably has 2 internal temp sensors, one goes to gauge, one to ECU, so the ECU side could be ok, and the gauge side a bit weird... sensors don't die often...

    with all the other observations when higher current devices turn on, kind of leaning to an earthing issue closer to the battery itself, or battery to frame?somewhere
    • Like Like x 1
    • Agree Agree x 1
  8. #8 jack_1313, Apr 19, 2016
    Last edited: Apr 19, 2016
    Hey guys, time for a quick update. Thanks very much for your comments.

    Regarding the battery voltage, I checked the other day. I was getting 12v with the bike turned off and 13.3v with the bike running at idle. That seemed a little low to me, but there are two things to consider: 1) I'd been messing around with the bike in the garage and might have drained the battery a bit, and 2) I was measuring from my auxiliary power bus bars/board, not the battery itself. The bus bars are wired to the battery (with a relay along the way) with some pretty thick wire, but there still might be a voltage drop.

    However, the real breakthrough came tonight as I was checking the various ground connections.

    There is a thick cable connecting the underside of the petrol tank to the frame of the bike. EDIT: the cable had snapped. I thought the function of this cable was only to make sure the tank doesn't fall off while you've got it propped up to work on the engine (an idea supported by the fact that Honda calls it the "CABLE, FUEL TANK STOPPER"). But today I had the bright idea that it might actually serve some electrical purpose grounding the tank, perhaps a purpose related to the fuel tank sensor on the other end of the tank? It's a bit counter-intuitive to me (Surely the tank itself - filled with explosive petrol - couldn't be part of a live circuit?), but I am electrical dumbass. It's probably perfectly clear on the wiring diagram.


    In any case, I let the bike heat up and watched what happened when I manually rejoined this wire with my fingers. Firstly, a spark! My multi meter is telling me that there is about 1 volt flowing through that connection. Secondly, sure enough, the temperature gauge dropped! It still spiked very high when the fan switched on, but I probably might not have been creating the best connection.

    So the next step is to properly replace this cable connection. Will it fix the issue entirely or only partially? I guess I'll find out in a few days.

    There's a good chance the gauge itself is rooted now though, given the amount of time it has spent pinned against the top stopper. I think it might be time to install an auxiliary gauge - one that actually gives me a number read-out of the temperature so there's no guesswork needed.
    • Like Like x 2
  9. Good on you for diagnosing this. One little hole in your electrical theory which is no criticism but it helps to get these things right.

    Voltage is a Difference in potential, you can think of it as a difference in height. Standing on a box is a low voltage standing on a cliff is a high voltage. It provides the push, current is the thing that flows, it is electrons scooting along the wire. Resistance is the stopping of this current from travelling. The three are related in a rule called Ohms law. Voltage = Current x Resistance

    So what the one volt reading across this strap is telling you is that that strap is a high resistance and it is limiting current flow. It should be very low resistance so you should get low voltage across it. You need to make sure it is bonded well to both ends.

    Well diagnosed.

    Ebay has this option as another Water Temp display: Top Gauge Digital Motorcycle MotoBike Water temp Gauge Meter RED LED Waterproof
    • Like Like x 1
  10. Unfortunately I have no real background in electronics, as you can probably tell. I understand some basic practical aspects but have a very weak conceptual understanding. So thanks for the explanation :)

    I just reread my post and realised I left out a key sentence! The cable in question had snapped. (I had been aware that it had snapped, but, as I mentioned, I didn't realise it served any electrical purpose.) So that's why I was talking about joining it back together with my fingers :)
    • Informative Informative x 1
  11. I wouldn't think the pump grounds through the tank (it should have two wire connector)
    same for the fuel level sensor. if single wire, then it does ground through tank, but if 2 wire, it doesn't.

    the grounding strap from tank to frame may just be to prevent static buildup (and spark) in the tank.

    have you measured voltage and resistance from the battery negative terminal (the battery post itself) to the frame?
    should be nearly no resistance, and should not have voltage drop.
    you may still have other grounding issues. look for the ground wire at the dash, and see what the resistance or voltage drop is to the battery (with +ve battery post disconnected)

    the spark you saw when rejoining may have been tank static.

    just for kicks, here is a table for the voltage required to jump an air gap
    Measuring high voltages

    that's 4700V for 1mm jump, and 30,800volts for 10mm jump!! (hence when spark plug sparks can jump around 10-15mm usually)
    even static sparks you see from clothing (like when taking off synthetic clothes in the dark) is thousands of volts!
    • Like Like x 1
  12. Time for an update.

    First, thanks both for your help :)

    I have no idea when I'll next have chance to get to a dealership to pick up the broken cable part, so I went ahead and fashioned a nice thick wire to serve its purpose for the time being.

    The good news is that my temperature gauge is no longer telling me that my engine is a puddle of molten metal on the ground beneath my bike. The gauge now gives a reasonable reading and doesn't spike when the fuel pump primes, when the engine is on, or when the blinkers are on. (I imagine the headlight etc. was also interfering with it.)

    The gauge still spikes a little when the fan turns on. In retrospect, I think the bike has always behaved this way during the ~100,000k that I've had it. I'd previously noticed the needle jump when the fan comes on but mistaken it for a non-uniform temperature occurring in the cooling system when the bike transitions from moving to stationary. In other words, I got the cause->effect relationship the wrong way around - I had thought the fan was turning on because the temperature in the coolant passing the sensor suddenly jumped, whereas the gauge was actually jumping because the fan turned on. A minor over-reading while the fan is running also explains why my bike always seemed to struggle to stay out of the red when stuck in traffic on really hot (40+) days.

    Now, whether or not this is normal behavior, I'm not sure. It might be that the gauge is now working as designed, or it might be that there is some residual issue from before my time with the bike.

    I went hunting for other bad ground connections and didn't find any. No resistance from the (negative) battery terminal to the frame, measured at the main grounding point. No real resistance through the ground wires I checked.

    Regarding to electricity coming from the tank, it's still a bit of a mystery. There's definitely a constant current running across that broken cable I replaced, which explains why the original cable was shielded. There are three wires going into the fuel pump, one of which is green. My guess is that one is for the pump itself, one is for the fuel level sensor, and the green one is a shared ground wire. So I don't think it's the pump grounding through the tank. (I guess it could be the sensor? There is actually some strong evidence for this, but I won't go into.)

    As oldcorollas said, it might be a static electricity thing. (Maybe what Honda really means by "tank stopper cable" is "the cable that stops your tank from exploding"?)

    Whatever the source, my theory is that without the "tank stopper cable", the electricity finds some other passage to ground and in the process messes with the electrical system, adding resistance in places that it shouldn't. This makes sense in my head, but - as I said before - I'm a complete moron when it comes to electronics.

    In any case, the problem appears to be basically fixed now. At the very least, the bike is functioning the same as it did before the excessive gauge readings started. A numerical auxiliary gauge is still on the table but not really a priority for me anymore. It would involve cutting a coolant hose and there's really only one place under the tank that I could do that - a place that is already a bit crowded because the wires for my auxiliary HID projectors are running through that space :)

    Thanks again!
    • Like Like x 1
    • Winner Winner x 1