Dreaded SH__46 Error

Yamaha FJR Motorcycle Forum

Help Support Yamaha FJR Motorcycle Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Reading up on page 5-11.

2. Adjust

  • Gear position sensor angle
a. Connect the gear position sensor coupler to the gear position sensor.
b. Connect the digital circuit tester to the gear position sensor.

(that's just a multimeter)

And then they want the multimeter connected to probe 1 and 2 (outer two) again to measure the voltage.

Now, aside from doing what yamafitter did of stripping the wires, how do I do this with the gps hooked up? I seem to be missing something here.

 
Reading up on page 5-11.
2. Adjust

  • Gear position sensor angle

a. Connect the gear position sensor coupler to the gear position sensor.

b. Connect the digital circuit tester to the gear position sensor.

(that's just a multimeter)

And then they want the multimeter connected to probe 1 and 2 (outer two) again to measure the voltage.

Now, aside from doing what yamafitter did of stripping the wires, how do I do this with the gps hooked up? I seem to be missing something here.
What I've done in the past is to get some multi-stranded wire, strip off the insulation from one end for a suitable length, then cut off all but one of the strands. Now push the remaining strand into the connection socket, and re-connect the plug, so jamming the single strand into the connector. You can take voltage measurements form the other end of the wire.

Pictures from when I was measuring fuel gauge voltages:

(Click on image for larger view)



With this connector I had to feed the strand through the body of the connector, this would depend on its design. Obviously I have one for each of the two connections I wanted to probe. Don't let them short to anything they shouldn't, I taped mine carefully to ensure no shorts.



 
Sometimes depending on the probe you have on the meter leads you can stick the probes along side the wire at the connector and get a proper connection. I sometimes have luck doing this but it is a hit or miss proposition. My Fluke multimeter also has a set of alligator clip leads which are the ones I used in the photo I posted.

Having spent 34 years in the Instrumentation and Industrial Electrical Trades I'm OK with stripping the insulation on the wires I need to take a reading. The ideal way to do this would be to build a jumper that would plug into the connector and have the required tap offs for you to get your readings but if you don't have an electrical background it may be difficult for you to source the required parts to build one.

If you go with stripping the wires please make sure that the power is off while stripping the wires and make sure that nothing is shorted before turning on the power. Afterwards use a good quality electrical tape to tape the wiring back up. Try to do this in an area that is not exposed to a lot of heat since the heat from the motor can have a negative effect on the electrical tape and reduce its insulation value.

When you have everything adjusted correctly put your display into Diagnostic Mode as described on Page 8-69 in the manual and then go to Diagnostic Code 21 (d:21 - see page 8-74) and check the operation of the gear position switch as a final check.

 
Last edited by a moderator:
One other option is to stick a (small diameter) pin through the wire you are interested in & clip your alligator clips to that. Usually easier to repair the small diameter hole left in the insulation..........................

 
Thanks for the tips guys, I'll figure out how to roll forward on this. I just wonder how the shop does it? Of course, from what I've read, the majority of the shops don't even know how to work on the AE, so there's that.

 
Well, I'm not putting this one to bed yet. I put the sensor back on, didn't do the deal about calibrating it for voltage but ran through the SH__64 recalibration (I figure it needs to re-sense the position each time it's removed and put back on) and it's been a week with no error.

I do wonder though. When I pulled it, the sensor isn't a sealed unit and doesn't seal to the bike. There's a seal on the bike around the shaft coming out to keep dirt and debris out of the internals of the bike but nothing really for the sensor. There was some small pieces of grass and general debris between the sensor and the bike that I cleaned up. I wonder if there might have been something gotten inside of the sensor that got knocked free or something by my process of taking it off and testing it? Who knows.

I'll report back if it returns. I've been wary of posting in case i jinx myself.
sad.png


 
Looks like I made it a month. Came back again this morning.

I can see why the shops hate working on these AE's. Not enough of them out there that have problems to get experience with them for starters. Second, the service manual sucks. It basically goes like this "If you have this error, check this on this part, if it fails, replace the part." There's no steps, it's one step. And if that step doesn't work there's no additional information or test to try. Plus, the error isn't repeatable either. I can't go out and make the error come up, so it'd be like they'd fix something and give it back to the customer and the customer would be back a month later bitching because they can't seem to fix their bike.

This weekend is slated for tearing the forks apart as I've got a leaky seal, so I'm doing bushings and all while I'm in there so there won't be much time for looking into this one.

That said, I'm contacting two other posters who had this issue but never posted if it was resolved. In addition, here in a few weeks I may be pulling the clutch back apart for an inspection as that seems the next place to go. It doesn't seem likely though as they have issues in 5th while opening the throttle heavily, and mine is happening while steady cruising. In fact, I've put it in 5th and opened the throttle heavily and can't get it to happen. I also know that while it's flipping between 4th and 5th on the readout that I'm NOT accelerating and am just cruising.

 
Please go back and check the voltage reading as discussed earlier. You may be just at the outer edge of the 2.4 to 2.6 VDC range which would explain the intermittent nature of the problem. It took some time for me to get the reading close to 2.50 volts and make sure you take a reading after you have everything tightened up. I would be very interested in the "as found" reading.

One word of caution. Accuracy in this case matters and I can trust what my multimeter is telling me since I still have access to a instrument calibration lab and have checked my meter. Some of the really cheap meters are not all that accurate but even a cheap meter should be accurate enough to do this job. Just make sure to use the lowest range available that will give a proper reading. In this case I think you will need to use the 0 - 20 VDC range since the meter will probably over-range on the 2 volt setting.

 
Well, this pretty much went away for a good number of months (but of course for other reasons it didn't get ridden for portion of those months) except for one or two odd times. This past week it's been back with a vengeance. At least once every ride now. However, it seems to be worse than before and that might just be good.

Previously, it'd start flipping between 4 and 5 on the gear indicator. Too many times of that will pop the amber light and the code and then it's restart time. Still getting that, though sometimes it's more severe. It'll jump even lower than 4 and on Friday I got a video...been too lazy to edit...and saw with my own eyes it go all the way down to 1 before jumping back up. That's pretty crazy.

Anyway, the bike didn't respond any differently, so if it were the clutch, there should have been something noticeable going on besides just the gear indicator. I'm thinking since the manual points to the sensor, and with it going that wonky, it's got to be an intermittent issue with the sensor. $240 is a lot of cash for a little potentiometer but more and more I'm thinking that's the issue. I'm guessing that since it works most of the time on the bike it makes sense that it'd work most of the time in my garage while I've tested it, but that it goes crazy every once in a while. Anybody see any reason to not jump down that path? I got no problem paying the $$$ to fix it but I and my wife will not be happy campers if it ends up being something else.

 
When the TPS started failing on my FJR (before it was known to die an early death) it was intermittent, like your suspect gear position sensor. I purchased a DMM from Harbor Freight that was essentially 'disposable' at just $6, I Velcroed it to the handlebars where I could read it while riding and ran wires down to the TPS so I could monitor the output voltage from the sensor. As soon as my bike misfired it was clear that it would misfire at two very narrow rpm ranges and only there, regardless of the gear and it was immediately known that the TPS was bad. I used blue painter's tape to hold the wires in place while I was riding.

If you can cobble a meter onto your gear indicator you need to check for two things 1) put the probes on ground and the output wire of the sensor and make sure that when the gear indicator display changes gear (without a gear shift) that the voltage does indeed change on the output wire 2) if the output voltage does change (as we suspect it does) then put the probes between ground and the +5 volt reference voltage from the ECU and make sure that it remains steady and doesn't fluctuate when you get a false gear change. If the +5 volts fluctuates, that's the cause of your problem. If the +5 volts is stable but the output voltage changes the sensor is bad. It is possible for the sensor to be intermittent and it can also be heat sensitive.

The things that make up the gear sensor circuit are:

The 5 volt reference circuit from the ECU, you should use the ECU ground wire at the sensor to measure this voltage to ensure both gnd and +5 volts are stable

The output voltage from the sensor

It is possible for the voltage output from the sensor to be good but the connector or the ECU pin has corrosion or contact problems. If +5 volts is stable and the output of the sensor is stable and correct the problem is in the wire or connections to the ECU or the (highly unlikely) possibility that the ECU has a bad input circuit.

The picture is a clickable link.



 
Last edited by a moderator:
I like it. I actually was thinking about this and wondering if there was a way to do it. There's an HF on the way home from work. Wonder if it'll fit in my X Mount.

Bad weather this week will give me time to fiddle with it too.

 
Last edited by a moderator:
...Anybody see any reason to not jump down that path?...
Certainly looks like the sensor or its connections. I would suggest checking its wires and any connectors (maybe only the connector into the MCU, can't see reference to one into the sensor).

I'd unplug the connector into the MCU, check for any signs of corrosion or bad wire to pin crimp. Measure the resistance of the sensor from these pins - between blue and black. The "wiper" connection is green with yellow marker, this should read a lower resistance to each of the blue and black wires (and should add up to the blue-to-black resistance.

Maybe use the manual method to change gear

(Click on image for larger view)



See that the wiper to black resistance changes consistently in each gear.

This is a poor-man's version of the "SH_ _65" diagnostic.

If it shows a fault, follow the cable back to wherever it dives into the transmission, if there's another connector, check that for corrosion or whatever.

Only after this proves the sensor is bad would I go down the replacement path. The only other option after that would be a bad MCU, which would be considerably more expensive.

 
Last edited by a moderator:
Thanks guys, looks like I'll have time to fiddle in the garage perhaps tomorrow night. I'll report back. I appreciate everyone's assistance. Thanks!

 
Well, been busy as hell. Family came into town for ten days, work has had me slammed, I haven't been on the forum and the bike just sits. Finally decided to run some wires and try some tests. Not sure what I did but my potentiometer seems now to be a resistor. It only gets one reading no matter the position. Sigh. Well, now I have the motivation to order a new one, hope that solves it. Oddly, my local stealership was cheaper than any place online, so I sit and wait.

 
Last edited by a moderator:
One more thought my friend:

It is possible that the sensor is affected by heat. It might give you all the right readings while it is at ambient temperature but once up to operating temperature, things go awry.

If I overlooked something in the (very helpful) posts to that effect, I apologize.

I want to also echo Yamafitter's advice about a quality Multi-Meter. The meters I use at work are calibrated once a year. I sometimes bet my life on the accuracy of those meters. After using good stuff all these years it is hard to trust anything else.

 
Well, my last ride, that was unrideable, it was acting up within a few minutes of the ride starting.

I'd ride it over to your place for testing...if you were closer.
sad.png


 
So, progress and a brick wall.

Got the new sensor in, finally got some time to fiddle with the bike.

Sensor installed per 5-11 in the manual. Instant SH__21 error. No worries, that means the output voltage is wrong, which I expected because I hadn't calibrated it yet. Followed the procedure on 5-11 along with Mcatrophy's example of feeding wires into the coupling to get it to a very careful 0.81V, which is right in the middle of the 0.71-0.91 necessary. I wanted to be in the middle in case of any error in my testing methods or equipment.

I proceed to check errors, the SH__21 error already cleared, I have none. Good! All is going well. Time to proceed into the SH__65 sequence on page 8-178 to program the eprom with the values. All goes well until I pull the brake handle and see the dreaded flashing light.

Ugh.

Well, maybe the battery is low from sitting for a while. I plug up the tender and it's red. Stays red for a while, but eventually overnight it goes to solid green. I let it sit the rest of the day because of that silly thing called work and come home and run through SH__65 again. Flashing green light. Sigh. I've done this before when I tested out the old sensor and it worked like a charm but not now.

For grins I decide to try starting the bike in neutral. It just spins but doesn't sound like it's getting any fuel at all. Now, I don't know if because the sensor isn't programmed right it's cutting the fuel out or what it is, but I can hear the starter running hard but none of the sound of it trying to start even a little bit.

Checked connections for the main ECU as well as the YCCS ECU and all looks good there.

Unfortunately, SH__65 on page 8-178 ends with "If the data is not written successfully, the hand shift indicator light will flash." There's no instructions or troubleshooting that I can find for that failure. I may be at the point of pulling it onto the Kendon and dropping it off at the dealer for a resolution. Sigh.

 
Not in my near future.
sad.png
I'll likely have to pay to have it fixed. Oh well. Maybe they'll see how distraught I am and offer me an amazing deal in trade on an amazing deal on a 2014 model that I can't walk away from.

 
I hate to say this but its your clutch :( I had the same problem an finally replaced the clutch and

works fine now. The problem is the motor is running faster then the wheels are turning and the

computer wants you to shift. (clutch slipping).

 
Top