Bench testing the B041440 demand sensor

[lucian]

Since some of these sensor have become problematic, I thought I would attempt to bench test one I got from NCDan which was repeatedly throwing the urgent service warning, and the po 155 fault code.  Being a contact less design, it cannot be ohm tested like a traditional "wiper" style potentiometer. It needs to be powered up in order to test the output signals from the four variable tracks. Here's what I came up with.
  First you will need a few things;
 





  A multimeter set to dcv
 A reliable 5v dc power supply ( this one a charger for a cheap rechargeable screwdriver)
 A couple test leads
 Cold beer optional
 
Here  is the sensor , which is actually two identical magnetic potentiometers mounted opposing each other, with a central actuating spool  the throttle cables connect to
The plastic cover pops off the center section to give acsess to the spool which you can manually turn while bench testing to watch voltage values rise and fall , hopefully in a linear fashion









note the blue (master) and white (slave) sides marked with dots
also note the pin connectors ,six on each side
here is a guide to which pin does what, three per track a positive/ negative and signal



 I found the pins too tight for three alligator clips so I improvised with short pieces of light gauge wire with the casing expanded so i could slide them over the pins and then connect those to my test leads
connect the positive lead of the meter to the signal pin
connect the positive from the power source to the positive pin
connect the negative of the meter to the negative from the power source and to the negative pin
and your off









here are the sample values from the service manual on top, and the results from Dans faulty sensor on bottom



note track c-d specific to the po155 code
clearly track d is out of spec with greater that 1400Mv deviation   She's junk!
The good news is only the white side , if another turns up with a faulty blue side we can peice together a good 500 dollar sensor

[DougG]

Yay Dave!!!

Thank you for that most helpful and detailed post.  It really gets to the heart of the "Urgent Service" fault code.  I know it helps me.

Be well,
DougG

[lucian]

Of course these same track values can be tested with the sensor in the bike with Guzzi diag.

[Ncdan]

Ain’t it great to have a buddy like Lucian( Dave) !!!!!
Good work my friend👍

[lucian]

Thank,s Dan  I'm gonna hold on to this bumm sensor for you. Some day when your to old for full power mode we can put it back in and you can tool around in limp mode.

[Ncdan]

Thank,s Dan  I'm gonna hold on to this bumm sensor for you. Some day when your to old for full power mode we can put it back in and you can tool around in limp mode.

Lol, always looking out for you old NC buddy:)

[DougG]

"Of course these same track values can be tested with the sensor in the bike with Guzzi diag."

You are right (of course).  But whether you get a new TGPS...$450+, or a used one from ebay for a lot less..it is reasonable to test it before installation.  You have showed us the way

Thanks again.

Be well,
DougG

[LowRyter]

just curious, can these things be swapped out on the side of the road?  If so, a spare along with the tire plug kit might be warranted.

[lucian]

just curious, can these things be swapped out on the side of the road?  If so, a spare along with the tire plug kit might be warranted.

They could in theory , you have to pull the tank. You will also need the cable and a laptop or tablet with the software to re calibrate the handle and throttle learners. Actually, the bad sensors seem to reveal themselves very early on, the majority of them have been reliable and are used in many brands and models. It also seems that the defective ones I have heard about all seem to fail on track c-d and throw the same po155 code.
  I am certain that both the blue and white sensors are identical as they both have the same Marrelli numbers and when testing the white side mv readings at full throttle are identical to the blue side readings at released throttle. that is because they sit inverted,opposed to each other , I would speculate that the ECU interprets the signals as a percentage of throttle opening, thus at 50% throttle opening the Mv signals  reach equilibrium .  this inversion would eliminate the possibility of the ECU misinterpreting o% throttle for 100% throttle.

[kballowe]

Great post.  Good information to have - for us 1400 owners

 

[Kiwi_Roy]

It seems like a terribly complicated system, I assume the sensors are Hall Effect?
So you still have a cable from the twist grip to this device, what actually moves the throttle plate in the throttle body?
I always assumed the sensor was right in the twist grip

Since its two identical pots and only one goes bad if owners send you all the bad demand sensors you should be able to make one good one from 2 faulty ones no?

[pete roper]

It is 'Complicated' in a way but it has to be. While if the four signals may not cause issues if they fail to correspond to what the ECU is programmed to do the fact is that if you don't have a 'Fail Safe' bad things can happen.

Remember all the hullabaloo when RBW and DBW were first mooted? Vehicles were going to go berserk because suddenly the control of the throttle plate was no longer a 'Physical' connection. Did it happen? No, it didn't! And why? Because of the failsafes!

For a year I had a Mazda CX3, I halted it, but not because it was DBW. My current Ducato van is also DBW, it's faultless, (Well not quite. If I re-engage cruise control after having to slow down it will actually accelerate faster on re-engaged 'Cruise' than it will if I just floor the pedal! Weird, but it's a van. Like I give a shit!). Thing is not the fact it's a RBW system on the Cali's, it's just a shitty, poorly manufactured component.

RBW allows much finer control of fueling and emissions than a 'Conventional' system, (With the caveat that it has to be done right.) so it is the way of the future for IC engines. Get used to it. It's a dying technology anyway so who cares?

Pete

[lucian]

Throttle cables rotate the spool on the demand sensor  ,4 input signals are cross checked /interpreted by the ECU and 4 output signals from ECU sent to throttle body which is operated via a 2 track potentiometer and electric  motor.  The two tracks from this are also checked /interpreted by the ecu and variables added from input from lamdas, air pressure sensor, temp sensor ect.
My whole propose in this exercise  was to determine if, in fact, the two potentiometers of the demand sensor are in fact the same. It appears to be the case, and It should not matter weather it is a blue or white marked side , only that the proper connector is plugged into the proper side due to the opposing rotations which causes one signal to go from 1v to 4v and the other from 4v to 1 v from 0 to full throttle.  I would be interested in any failed sensors out there to continue this experiment. dave

[Kiwi_Roy]

Pete / Dave
Thanks for the explanation of the interaction between Twist Grip and Throttle Plate.
Yes I do remember when Audis went through the sudden acceleration crisis, I worked for Volkswagen at the time and for some reason the management thought it necessary to explain to us how it wasn't possible.

With 1400cc I can see why the traction control system needs to run interference between the rider and the wheels LOL

[pete roper]

I took the 1400 Stelvio out today. I was just about to shitcan the project but Michael talked me out of it. The map still needs work, principally at the very top end and just off idle. It pops and farts on a light throttle between 2,000 and 3,000rpm and goes flat at WFO especially above 6,000. Can't work out whether it's going rich or lean at the moment. Mark will have to log that because when you go WFO at >6000 on a  big port, single spark 1400, especially in the lower gears, shit gets real very, very quickly due to the stupid amount of torque.

I've only had the TC cut in once so far though when I gave it a bit too much welly coming out of a roundabout and the back end let go, the Guzzi W5AM TC system is such shit though that it hooks up so crudely it is more likely to toss you over the high side than save your arse! The Stelvio though is such a wide barred, long wheelbase, pig that even I was able to catch it and wrestle it back into some sort of semi-controlled direction pointing!

Pete

[gthorpe]

I was dealing with the frustrating P0155 fault code on my Eldorado for a while.

The only dealer in the Phoenix, AZ area that carried Moto Guzzi dropped the brand, so no warranty work was available.

Fortunately I got good advice from Lucien and Pete Roper on this forum and was able to confirm the demand sensor was bad and I replaced it myself.

My Eldorado is running great again, hopefully the new demand sensor doesn't give any more trouble.

[Ncdan]

I took the 1400 Stelvio out today. I was just about to shitcan the project but Michael talked me out of it. The map still needs work, principally at the very top end and just off idle. It pops and farts on a light throttle between 2,000 and 3,000rpm and goes flat at WFO especially above 6,000. Can't work out whether it's going rich or lean at the moment. Mark will have to log that because when you go WFO at >6000 on a  big port, single spark 1400, especially in the lower gears, shit gets real very, very quickly due to the stupid amount of torque.

I've only had the TC cut in once so far though when I gave it a bit too much welly coming out of a roundabout and the back end let go, the Guzzi W5AM TC system is such shit though that it hooks up so crudely it is more likely to toss you over the high side than save your arse! The Stelvio though is such a wide barred, long wheelbase, pig that even I was able to catch it and wrestle it back into some sort of semi-controlled direction pointing!

Pete

That’s interesting Pete. Do you think we may be better off to turn the TC off? Heck, I made it just fine for 50 years riding like a wild man without TC.

[pete roper]

The Stelvio TC is, as said, very crude. Really it just retards the spark grossly thus limiting torque to the rear wheel. The 7SM system used on the 1400's and one assumes the V85 is far more sophisticated and actually manipulates the throttle plate and does it so fast it's impossible to dump the clutch and spin the wheel out. Back when they first came out I was at the importer and they tried to get it to step out on a polished concrete floor. Winding it up to 6.5K and dropping the clutch just resulted in little 'Ticks' on the floor as it only took the merest of discrepancies between the front and rear wheel sensor inputs to kick the TC off and it would slam the throttle shut. The only reason the wheel spun even fractionally was the energy stored in the flywheel. Quite impressive.

Pete

[lucian]

I agree the TC on modern bikes is a wonderful thing. I never was much on tech wizzardry but without tc , esspecially on a lot of modern , high power sport bikes.  a lot of us would be in over our heads real fast.  On the Cali I just set it on one and forget it. It's saved my bacon plenty on wet or dirty conditions. Keep it turned on Dan.

[Ncdan]

Roger that Dave and Pete. I’ll check and see what mine is set on. I think it’s in 2 but will have to check after while. Thanks guys!