| Author | Message | ||
     Gregoxb |
And so it begins again. The hesitation when I roll on the throttle is beginning again... It is very gradually starting to become more prominent..... Sigh... | ||
| Greg_e |
Did you send the original to me? If so I probably shouldn't cut it open. Could one or more of the wires be damaged where they go into the connector? | ||
| Greg_e |
Is there room to lay it flat over next to the fuse box, then go ride it? That would put a different flex on the wires and might tell you what you need to know. | ||
| Gregoxb |
Didn't send it yet. I'll try it tomorrow. | ||
| Reepicheep |
Interesting that it starts slow and gets worse over time. Makes me think there is a bad sensor input that the ECM is trying to "fix". I'd work through the ECMSpy data and see if anything doesn't make sense, for example head temperature sensors. Unfortunately the CPS changes too fast to really be "debuggable" from the ECM output, you need an oscilloscope. Probably the same with the O2 sensor. I want to do something, but don't have the time. I'd like to go through the ECM inputs with a paperclip and a cheap scope and document what each input should look like on a properly running bike, with a short procedure on how to set up the scope and the running conditions to demonstrate it. It would be a really nice troubleshooting resource. | ||
| Greg_e |
Yes, that would be a good resource. And yes, I agree that even the o2 sensor is difficult to troubleshoot from ecmspy. That said, most of the time that an o2 goes bad, it shows a regular low or high reading which messes up the mixture. I've had a few go low on cars as they started failing. Put a new one on my XB when I did the top end build because I just didn't want to bother with changing it later, and so many other things were new that I thought it would give me the best success. The most valuable would probably be an early CPS and a late CPS. Do you remember if those are really 5 volt signals? I have a decent datalogger that can handle up to 5 volts and records into a computer, screen shots or even the actual data (software is free) can then be analyzed. | ||
| Reepicheep |
Yup, it's 5v, and you can tap it with a paperclip from under the seat. You don't even have to take anything apart. I have screen shots on a thread somewhere showing the signals on a $90 hand held digital scope. They aren't that fast, just hard for the ECM to spot. | ||
| Greg_e |
If I find some extra time on my hands, or my CPS starts to fail, I'll definitely grab some with the data logger. Would be nice to have a 5 volt RPM trap too, might have to probe the gauge to see if I can at least get the raw pulses. | ||
| Reepicheep |
The signal you get from the CPS is a multiple of the RPM, as there are many teeth in the cup. It would be cool to see if you could measure the acceleration and deceleration of the crank by careful monitoring of the CPS signal. I don't know why, but it would be. | ||
| Greg_e |
The datalogger has a time scale in milliseconds, so it should be visible. What would be cool would be to see the difference in time between the spark and the trigger to see how accurate the advance really is. Considering most of my work will be on the Electra Glide, maybe I'll check and see what it's CPS is doing, think it is still pulses and not data stream, but there is a ton of data stream stuff going on in those machines (yet no oil coolers or cooling fans or o2 sensors???). | ||
| Reepicheep |
That would be interesting Greg, if you do it, let me know what you see. I've wondered about using a hall effect sensor in there, it would be pretty easy to retrofit it to something optical, and doing so seems intuitively like it would be a lot more precise, but I don't know for a fact how consistent and precise a hall effect switch really is (especially one like this one, where the magnet and sensor are a single unit measuring a slot in a ferrous cup, as opposed to measuring a magnet going past). | ||
| Reepicheep |
(found one web reference that indicates a particular brand of hall effect sensor remains linear at .5% to 1.5%, and that's not total error, that's uncorrected nonlinearity error, so I'm guessing that a hall effect sensor is well under 1% error, which is probably "fine" for an internal combustion engine, which is probably why everyone is using them for that application.  ) | ||
| Greg_e |
Many IC engines use a hall effect, including most of the points to electronics conversion kits out there. Pretty sure they are generally fast enough. | ||
| Reepicheep |
Yup. I've had one fail on a Saab. | ||
| Gregoxb |
Still trying to iron out this issue. Bike is running, but intermittently, the engine runs like crap. Extra vibration at riding speeds, decel popping, and power fade. My buddy rode it and said the engine felt "hard." I am wondering, can a bad head temp sensor be the culprit? Perhaps a frayed wire or just a bad unit? Thoughts? (Message edited by gregoxb on October 12, 2015) | ||
| Sharkytattoo |
Bad Throttle Position Sensor | ||
| Gregoxb |
Now when I plug in my Buelltooth, it does not power up. Can that all be related to the ECM cables? Could the pressure of the seat have damaged the cables? If so, how would I go about checking/fixing if indeed there is a problem with the ECM cables? | ||
| Greg_e |
The data connection gets power from the ecm, so yes it could be related. I would put a multimeter in resistance mode on the pin for the data power ecm side (ecm disconnected) and on the power pin for the data port, and then flex that harness back and fourth and watch the meter. Guessing you see a change for low resistance to high resistance or open. Fixing it properly may require a new harness. | ||
| Gregoxb |
Whew, this gremlin is killing me. I'll put a multimeter on it and see what happens. | ||
| Tq_freak |
Greg is this similar to what yours is doing? https://youtu.be/h2S2NEABecs | ||
| Gregoxb |
No. That looks like bad intake gaskets. (Message edited by gregoxb on October 29, 2015) | ||
| Gregoxb |
I'm fairly certain I am dealing with an electric gremlin in the form of a chaffed wire or a poorly contacting wire that was dislodged or damaged by the pressure that my seat was exerting onto the ECM cables. | ||
| Greg_e |
Looking at how the harness was pressed down, I have to agree. If there was extra wire, I'd suggest buying the connectors and cut and crimp new ones on, but I'm sure there isn't that much slack. It would need to be cut before the bend point so you'd lose 2 to 3 inches. About the only thing I could suggest is finding connectors and make a short inline extension. Then you could cut the harness short enough to remove the bad spot. But you might have better luck find a good used harness, than finding mating connectors and the crimp tool. | ||
| Desotonorsocal |
Hey all, So it sounds like I have the same problem... A month and a half of chasing down electrical "problems" (which was me being convinced I fixed it everytime I found a shady looking wire) only to find out the seat was pressing and rubbing those ECM wire bundles BADLY. I cut a hole out of the seat, bent the wires up slightly, and rode it for a few weeks without issue. Now ORIGINALLY, the symptoms were more like a bad temp sensor, super rough at any RPM, CEL coming on/off, etc. NOW, however, as of yesterday, I was riding home on the freeway and she started dying on me (with a chick on the back of the bike no less, ouch) and I had to half-clutch and keep on the throttle HARD and constantly blip it to get us off the freeway. I limped it another 5 miles in town like that. So today I re-checked the wires and was about to head out (she starts fine, but after 3k RPM it's what was described above no power, rough all around etc) so she's toast. I put her on ECMSpy and basically at idle it keeping 'dipping', and also I can't seem to get idle to stay. No codes thrown by ECMSpy btw. Here's what it's doing: https://youtu.be/1pwkk9_uVw4 Someone mentioned bad TPS? | ||
| Desotonorsocal |
Also, if you look at the video (or don't, lol) ECMSpy goes from Closed Loop to Open Loop any time the idle 'dips'. I don't even know what open or closed loop means, it just seems something is 'wrong', heh. Any comments are appreciated! Thanks! | ||
| Greg_e |
Closed loop - sampling the o2 and adjusting the afv. Open loop - just running the programmed settings. If the rpm drops below a certain point, it will fall out of closed loop. Since you've had a wiring issue, that's the first place I would go for a fix, probably another wire has gone bad. Chasing anything else is kind of a waste of time, the wiring will always be in question until it is finally fixed completely. Wish I had suggestions, but I have an R not an S so I haven't seen it first hand. | ||
| Desotonorsocal |
Alright, so I looked through the wires and reset the TPS to no avail. The bike seems *slightly* smoother but same problems that sound like above: - Rough idle - Dipping on idle - Holding at any RPM I can feel 'surging' - Like posted above, the engine feels 'hevy' across the board. Definitely sluggish - Idle sometimes drop til the motor dies if I've run it long enough - seems to smell a bit more rich than normal Anyway, there you have it, my newest 'issue', heh. I can't justify $650+ for a complete wiring harness plus a weekend or more worth of work. :/ I've spent too much on the bike thus far. | ||
| Greg_e |
Should be able to get a salvage harness for $100. Check the for sale section, thought one guy had a couple he was parting out. I really should get myself a new set of forks from him. | ||
| Desotonorsocal |
@Greg_e, I hear ya and I'll take a look on this site, but I've looked around (ebay, etc) and have been reticent about purchasing such because frankly, the xb's have all had problems so I'm concerned of old wires, ya know? So it sounds like either cut and splice or new harness. Again, I'll take a look now though! | ||
| Greg_e |
But... If you have a second harness out of the bike, it will be easier to make repairs on some of the soldered connections. That's one normal area of failure, the other is where is snakes around the steering neck and get's flexed a lot. If you really wanted to go the whole way... You can buy new crimp pins for all the connectors and string new wires. That's the longest route, but also should be the most certain way of fixing the issues once and for all. |
