engine oil temp sensor

[dlaing]

Issues I thought were important are:

that Guzzi probably added the sensor housing because plugging into the head directly is probably too hot.

Guzzi went from a Brass sensor housing to an insulated housing POSSIBLY because too much heat was getting to the sensors body, rather than the sensor tip. This was a good idea accept there is a delay as the air has to heat up and transmit to the sensor.

I think a little solder would make an ideal conduit to the tip, or maybe a spring???

I suppose a glob of anti-seize is not a bad idea as it is pretty well sealed.

If you have a brass sensor, insulating between the sensor threads and housing with teflon tape might be a good idea.

[rologuzzi]

I'm a bit late in replying to this but I believe its an important thread. Whilst the engine will still run, a poorly or non-functioning CHT sensor will do nothing to enhance running & response of the EFI system.

 

I've had no specific symptoms of poor CHT sensor operation (my V11 has only done 7000km from new & is running very well) but after reading this thread & thinking about it for a while, it made sense to me that the ideal would be to have perfect thermal contact between the base (tip) of the sensor & the (metal) base of the sensor housing. So I removed them (luckily without damage) & after thorough cleaning, added a dollop of silver oxide thermal grease to the sensor tip & base of the housing & reinstalled. On startup, the CHT temp (as measured using Technoresearch VDSTS software) rose very quickly, certainly much faster than without the thermal grease. It was obviously having the desired effect. I shut down the engine after a few minutes, by which time the temp has risen to > 90 degress Celsius. The engine was running fine, just as before (ie. nice idle & throttle response) & no adverse effects were apparent. This was later confirmed on a couple of runs - it hasn't affected the way the bike runs. But after rethinking the issue in the light of the elevated temp readings as well as previous comments by others, I'm now not so sure this is, in fact, the ideal solution & may even have some undesirable outcomes.

 

My reasoning for this is that the tip of the sensor & the base of the housing are shaped such that there is a very small air gap between them when the sensor is screwed right in to the housing. Given that the housing is insulated (plastic) this effectively means that the sensor isn't, at any point, in direct thermal contact with the head. I assume this has been done to avoid cooking the sensor, as the head can get quite hot under normal operating conditions, possibly hotter than what is healthy for the sensor. So by adding anti-seize, thermal grease or any object which brings the sensor into direct thermal contact with the head, it may be running the risk of frying the sensor, or at best, shortening its service life.

 

It now occurs to me that perhaps the sensor actually NEEDS to be at least partially insulated & that the CHT temp information provided by a properly installed & functioning sensor is what the ECU has been calibrated to. By placing the sensor in direct thermal contact with the head, temp data may be significantly elevated & therefore adversely affect the fuel map at various points, in addition to potentially reducing the life of the sensor (or cooking it completely).

 

I haven't as yet measured the resistance of the sensor after addition of thermal grease & a couple of runs to full operating temp but will do shortly & post. This will soon tell me if I've cooked it. Resistance (cold) for CHT (before addition of thermal grease) & the air temp sensor (mounted on the airbox) was exactly as per factory manual specs.

 

Apologies for the lengthy post. I'd be interested in comments on any of the above from those who undoubtedly know more about this than myself.

 

Cheers, Tony

[Guest ratchethack]

Thanks for the informative, well thought out, and well-written post, Tony. Will be looking for your temp measurement results.

[badmotogoozer]

Tony good thinking - but don't worry about "cooking" the sensor. It's just a thermistor.

 

cheers,

 

Rj

[Skeeve]

...Guzzi went from a Brass sensor housing to an insulated housing because too much heat was getting to the sensors body, rather than the sensor tip. This was a good idea accept there is a delay as the air has to heat up and transmit to the sensor....

 

I find it far more likely that Guzzi (Aprilia) switched to the plastic part for pecuniary reasons, given that brass is about 10x-100x more expensive than an equivalent volume of plastic, and must be machined [using either a CAM turning ctr's. time or an employee's time on a screw machine] vs. running up a set of injection molds & bonking them out by the gross-tillion. [Or buying them pre-bonked, if it's some standard fitting for some other use...]

 

I find this even more likely to be the root cause given the solution for most reports of "runs rough when cold & I'm getting cr@p mileage" seems to be "swap out the plastic for the brass head temp sensor holder" & the followup always seems to be "thanks, that fixed it!"

 

I anticipate having some fun whittling my own out of some spare brass on my mini-lathe one of these days RSN, even tho' I've had no problem whatsoever from the plastic holder: I'm just of the opinion that it's better to replace a known failure-prone part w/ the superior one as soon as is practicable.

 

WRT the air gap, this would only seem to prolong the rich running of a nominally "cold" engine; I'm thinking that the only reason for a gap is so to allow for the differential expansion of the various metal parts & keep things from cracking/breaking. Anti-sieze is pretty incompressible, so unless a pinhole is added for pressure relief, it could be a failure point. Hmmm.

 

JMNSHO,

 

[rologuzzi]

skeeve's musings on this make a lot of sense to me, particulary about the "cheapskate" approach to mounting the sensor as well as the delay in transmitting correct head temp readings due to the air gap. Its difficult to know for sure whether the stock maps have been built using CHT readings obtained by direct thermal contact or the somewhat cheap & nasty setup we have to deal with on production bikes. But I think its fairly logical to assume the latter was used.

 

In any event, my bike is continuing to run well with the improved sensor-head contact. Cold starts & idle haven't been affected (they have always been very good) & I haven't done enough distance as yet to see any effect on fuel consumption. Apart from a check of the sensor resistance (to see whether the higher temps have fried it) I probably won't mess with it again until something breaks. But before then I'll do what many others already have & that is to make or buy a brass sensor holder. Its only a matter of time before the plastic unit becomes brittle & falls apart.

 

My new Direct Link hardware/software has just arrived this morning so I'm off now to explore the wide & wonderful world of fuel & ignition maps.

 

Thanks to all for your words of wisdom.

 

Tony

[dlaing]

I find it far more likely that Guzzi (Aprilia) switched to the plastic part for pecuniary reasons, given that brass is about 10x-100x more expensive than an equivalent volume of plastic, and must be machined [using either a CAM turning ctr's. time or an employee's time on a screw machine] vs. running up a set of injection molds & bonking them out by the gross-tillion. [Or buying them pre-bonked, if it's some standard fitting for some other use...]

I should have written that Guzzi POSSIBLY did it to reduce the temperature.

To my mind it is a lot more expensive to manufacture a piece that is two parts brass and one part plastic, than one that is one part brass.

But yah, if Luigi makes it in Mandello, it is going to cost a lot more than if it is mass production item.

Still I believe that the sensor tip temperature is a more accurate indicator than temperature at the threads and that insulating the sensor body will result in a more accurate reading of the cylinder head.

But I am not such a stickler that I would trade for plastic one

I am content with having wrapped the threads with teflon, but I might consider a conductor for the sensor tip, or removal of the teflon if the sensor is not heating up enough.

I need to log some temperature sensor readings

But hopefully RolloGuzzi will do it first, so I won't have to.

[Ryland3210]

I find it far more likely that Guzzi (Aprilia) switched to the plastic part for pecuniary reasons, given that brass is about 10x-100x more expensive than an equivalent volume of plastic, and must be machined [using either a CAM turning ctr's. time or an employee's time on a screw machine] vs. running up a set of injection molds & bonking them out by the gross-tillion. [Or buying them pre-bonked, if it's some standard fitting for some other use...]

 

Having seen more structural plastic parts fail than I care to count, every time it happens, I wish the part had been zinc die cast. The tooling cost is virtually identical, as are the savings in machining time. Zinc costs less than brass, and yes, more than the cheapest thermo-plastic, but when the cost of premature failure is added to the plastic part, there is no contest. So-called engineered plastics, which cost several times more than ordinary thermoplastic have their place in applications such as V-8 intake manifolds, where weight savings are key and temperatures moderate.

 

In my opinion, small parts which have to sustain mechanical stress or deal with elevated temperatures are better done with die cast zinc. It also makes one heck of a lot better conductor.

[docc]

To complicate matters, the manual specifies the torque value with Loktite 601.

 

Isn't that the red stuff?

[Ryland3210]

To complicate matters, the manual specifies the torque value with Loktite 601.

 

Isn't that the red stuff?

 

I'm not sure whether that's the permanent or "removable" version. It's probably the latter. Even so, loctite on plastic is really begging for destroying it when trying to remove it. Another reason for metal.

 

 

After just playing with one to find out its not the problem The "engine oil temp sensor" which is acutally the head temp sensor and has little to !@#$ all to do with the oil temp. What makes me say that I have a bike that has an additional temp sensor in the oil cooler and the reading out of that are radically different to the head temp sensor. Is a NTC (negative temperture coefficent) themocouple with a range of -40 to 170 ish degrees celcius. NTC means as the temp goes up the resitance goes down resitance at 10 degrees is around 3.7K ohm and at 80 degrees is 0.37kohm. Buy rights you should be able to bung a 10cent resitor in there and run home on that if it fails. The sensor does not earth through the cases oil preasure switch style but has a return line to the ECU. Hope this helps NB all temps are celcius/metric.

 

Anyone know what the specs are on the low fuel level thermistor for my '04 Cafe Sport which was just replaced under warrantee? If there is another failure, I'd like to be able to test it myself, without going through another two week diagnostic and repair lost use of the bike by the local dealer. If I'm lucky, it will just be a bulb if it happens again, which I think I can handle.

[Ryland3210]

After just playing with one to find out its not the problem The "engine oil temp sensor" which is acutally the head temp sensor and has little to !@#$ all to do with the oil temp. What makes me say that I have a bike that has an additional temp sensor in the oil cooler and the reading out of that are radically different to the head temp sensor. Is a NTC (negative temperture coefficent) themocouple with a range of -40 to 170 ish degrees celcius. NTC means as the temp goes up the resitance goes down resitance at 10 degrees is around 3.7K ohm and at 80 degrees is 0.37kohm. Buy rights you should be able to bung a 10cent resitor in there and run home on that if it fails. The sensor does not earth through the cases oil preasure switch style but has a return line to the ECU. Hope this helps NB all temps are celcius/metric.

 

Cliff's later message with links shows 3750 ohms at 20 C, rather than 3.7K at 10.

[docc]

Loctite tells me 601 is a euro spec compound most similar to 609 here in the States.

 

609 is green and specified for bushings and sleeves in housings for shafts of fractional horsepower motors.

 

Still sounds like it will contribute to destruction of the plastic housing with any removal attempt.

[dlaing]

Loctite tells me 601 is a euro spec compound most similar to 609 here in the States.

 

609 is green and specified for bushings and sleeves in housings for shafts of fractional horsepower motors.

 

Still sounds like it will contribute to destruction of the plastic housing with any removal attempt.

You betcha.

I would not use loctite on the plastic thang.

I would guess that when my sensor adaptor broke I had only put about 2 or 3 foot pounds into tightening it. (It was loose, and not seated home when I had removed it)

So I would guess it should be torqued to about 1 foot pound or less and IMHO I would use anti-seize.

 

I wonder why Guzzi used an adapter housing?

I figured it had to do with the sensor getting too hot.

If the sensor screwed right into the cylinder head it would be measuring more of the cylinder head temperature and less of the ambient air speeding past.

I figure at high speed the sensor is well cooled by passing air and the reading is low.

Likewise at low wind speed, like city driving, the sensor reading is higher, more accurately reflecting the cylinder head temperature.

So, in city driving the engine is leaned out more.

I wonder if this could have something to do with it passing emissions testing???

I have noticed that when coming to a stop after a high wind speed run the idle is low (and sometimes stalls)

But in stop and go city driving the idle is higher.

I have experienced more of this variation since switching to the brass sensor adapter housing. (but it could be coincidental as I did other things at the same time, like service the air filter.)

My plastic sensor adator housing has been on order since 7-15-06...Get your act together Moto Guzzi!!!!

But the brass one works well enough, so I have not pushed the dealer to pester Guzzi/Piaggio.

I think adding thermal grease, solder or a spring would help transmit the temperature better.

[rologuzzi]

After a few sad posts on this thread about sensor holders breaking whilst removing them, I was extremely careful when removing both the sensor from the holder & the holder from the head. On my bike they were both held in place with a weak varnish-like material (?? thread locking compound), almost clear in colour, which was baked onto the threads. It came off readily with some gentle proding from a scriber tip & thorough wash in solvent. I reassembled applying only silver oxide thermal grease to the sensor tip & inner metal base of the holder, as well as a small amount to the threads. No Loctite, as I believe the holder is a tight enough fit in the head with only a very modest amount of torque applied to it. The sensor is held in the holder with a fairly large diameter coarse thread (didn't measure the exact size) & what looks like an alloy crushable washer. It wasn't crushed much (the plastic holder wouldn't take the necessary torque) but formed a good seal, again with only minimal torque applied to the sensor. Hopefully without Loctite, I'll be able to remove them both again if necessary with a slightly decreased likelihood of damage.

 

As has been stated by others, the plastic holder does not feel like it would tolerate anything other than kid-glove treatment. I used a very well fitting 19mm open end spanner extremely carefully when things got too tight for fingers. A metal sensor holder would be a great improvement, IMO.

 

My bike continues to run extremely well with the closer thermal contact between the sensor & head, which is consistent with dlaing's (& possibly others?) experience. I haven't as yet checked the sensor for signs of life by measuring its resistance - this requires removal of the tank to do properly & I'm simply being lazy - but will do so when sufficiently motivated. But I'd suggest its likely working as MG intended.

[Guest ratchethack]

My bike continues to run extremely well with the closer thermal contact between the sensor & head. . .

Good post, Rolo. But when you say, "continues to run extremely well", I can't tell for sure if you're saying there was any change from having added silver oxide thermal grease.

 

If so, how would you describe the effects -- before vs. after?

 

TIA