roper plate revisited

[pete roper]

I think the choice of the type of temperature sensor is dead easy. It's a unit that is used in trillions of other crappy vehicles. Why not go for ecconomy of scale and fit it to a crappy Guzzi as well? Like it or not we're talking about mass production. Even if with Guzzi 'Mass' has an entirely different meaning!

 

My guess is that you'll find exactly the same sensor in most if not all Aprilia Scooters and just about any other Italian motorbike and tons of Italian, (And probably other European.) cars The big difference being that most if not all of them are water cooled. The cost of developing a 'Stand alone' product for a small production vehicle like a Guzzi big block would make cost prohibitive. So we end up getting the rough end of the pineapple as usual and we have to work out ways of getting a marginal system to be the best it can be.

 

Don't like it? Simple! Buy an 'Onda. Like it or not if we insist on buying these irracible, under-developed bits of shit we're going to have to do a fair amount of development work ourselves. One of the good things is that most of the people who can do this are barking mad enthusiasts who don't mind sharing what they learn.

 

Cost for the above information is $750US. Please deposit it into my Nigerian bank account

 

Pete

[macguzzi]

[dlaing]

The chart on the sensor goes like this( I was lazy so there are missing rows at ten degree increments)

C° Kohm

-40 100950

0 9750

40 1598

80 0.377

125 0.102

 

Now I have not run tests, but assuming those are the limits of what it can read, I am prettty darn sure I have had the bike closer to the upper limit than the lower.

Cooling fins, plastic insulation, and airgap might actually make sense for this non water cooled engine.

[raz]

The chart on the sensor goes like this( I was lazy so there are missing rows at ten degree increments)

C° Kohm

-40 100950

0 9750

40 1598

80 0.377

125 0.102

 

Now I have not run tests, but assuming those are the limits of what it can read, I am prettty darn sure I have had the bike closer to the upper limit than the lower.

Cooling fins, plastic insulation, and airgap might actually make sense for this non water cooled engine.

The only purpose of that sensor is to tell the ECU what temperature (ballpark) the engine has so the ECU can figure out if it needs to boost the fuel (like a good old choke). I'm not sure what you mean, would you like to target the center of that table? Why?

I'm pretty sure the exact output (like 84°C or 92°C) is not important at all, but speed is. If the engine is warm the ECU should know about it. The ECU is NOT interested in having the reading severly lowered by rain and cold wind. That will result in unnecessary fuel boost.

[dlaing]

The only purpose of that sensor is to tell the ECU what temperature (ballpark) the engine has so the ECU can figure out if it needs to boost the fuel (like a good old choke). I'm not sure what you mean, would you like to target the center of that table? Why?

I'm pretty sure the exact output (like 84°C or 92°C) is not important at all, but speed is. If the engine is warm the ECU should know about it. The ECU is NOT interested in having the reading severly lowered by rain and cold wind. That will result in unnecessary fuel boost.

If people are moving to the brass fitting, packing it with conductive, and wrapping it all in insulating tape, I would be concerned about going off the chart.

I have not road logged the sensor to see if I am hitting 125°C, but if I am, than others will be off the charts.

I don't think that is a good idea.

Even if they are just bumping the temperature up 20° on the chart, they are running the bike leaner.

This may be fine for PCIII'd bikes, but for people running the stock ECU with open airbox and aftermarket mufflers, this could just be more of a bad thing.

Since I have Tuneboy and can change the numbers to whatever I want, I suppose I could optimize it by packing the sensor, but while insulating the threads with teflon tape to keep temperature down. We want the sensor tip to get hot as quickly as possilbe, not the sensor. We want the sensor body to stay cool, but not get overcooled by rain....Perhaps we should be putting a rain shield around it!

The hotter you make the sensor with conductive goo, the more the rain will cool it down, creating a bigger change at the ECU.

There is a lot going on here with a solutions that don't appear to be well thought out.

To my mind the ideal solution would be a silver probe deep into the heart of the cylinder with an insulated sensor housing. I think Luigi was on to something with the plastic housing. If combined with something conducting the heat to the sensor tip, I think we would have a near ideal situation.

The insulated plastic also insulates away the effect of cool wind and rain.

Need more data!

[pete roper]

Just to get back to the original subject I had a post from Todd at MPH saying that there were people asking about availability of the plates. Well I'll be posting 10 of them off on Monday so they should be there in about a week but until then get off the cripple's case or he'll insert his prosthesis up your date

 

Pete

[raz]

If people are moving to the brass fitting, packing it with conductive, and wrapping it all in insulating tape, I would be concerned about going off the chart.

I have not road logged the sensor to see if I am hitting 125°C, but if I am, than others will be off the charts.

I don't think that is a good idea.

Even if they are just bumping the temperature up 20° on the chart, they are running the bike leaner.

This may be fine for PCIII'd bikes, but for people running the stock ECU with open airbox and aftermarket mufflers, this could just be more of a bad thing.

Since I have Tuneboy and can change the numbers to whatever I want, I suppose I could optimize it by packing the sensor, but while insulating the threads with teflon tape to keep temperature down. We want the sensor tip to get hot as quickly as possilbe, not the sensor. We want the sensor body to stay cool, but not get overcooled by rain....Perhaps we should be putting a rain shield around it!

The hotter you make the sensor with conductive goo, the more the rain will cool it down, creating a bigger change at the ECU.

There is a lot going on here with a solutions that don't appear to be well thought out.

To my mind the ideal solution would be a silver probe deep into the heart of the cylinder with an insulated sensor housing. I think Luigi was on to something with the plastic housing. If combined with something conducting the heat to the sensor tip, I think we would have a near ideal situation.

The insulated plastic also insulates away the effect of cool wind and rain.

Need more data!

As long as you don't turn this supposed oil temp sensor into a cylinder head temp sensor I can't see how you would overheat it. If you experience 150°C oil temp you might want to check your oil cooler thermostat. My guess is only a series of dyno runs would get you there. The operating range (and the table) probably just means that the reading above 125°C is less predictable. It will still be lower resistance at higher temp and I would expect the oil to suffer from overheat long before the sensor does.

 

I'm not sure about WM1.5 but a stock WM1.6 will do the same correction at 150°C (or higher) as it will at 113°C. At temps over 65°C there is a wierd negative correction.

FWIW Cliff's MyECU stock table will stop correcting at 50°C, it never goes negative but just zero.

 

You are right about the conductive goo will help cooling too. The 1100 Sport sensor is probably better. It sits in the LH valve cover and the bare NTC gets splashed with oil all the time.

 

Are we hijacking this thread? Maybe we should move.

[pete roper]

The earlier sensor in the rocker cover is much, much worse and even more inaccurate.

 

Pete

[raz]

The earlier sensor in the rocker cover is much, much worse and even more inaccurate.

 

Pete

Not mine anyway. I have a dipstick thermometer in the sump and it usually reads pretty much the same as the temp sensor.

[pete roper]

Not mine anyway. I have a dipstick thermometer in the sump and it usually reads pretty much the same as the temp sensor.

 

 

Bad choice of words on my part. The sensor may be perfectly accurate but its location lends itself to cooling so it will give an accurate reading to the ECU that is far too cool and the ECU compensates by over fueling.

 

Pete

[raz]

Bad choice of words on my part. The sensor may be perfectly accurate but its location lends itself to cooling so it will give an accurate reading to the ECU that is far too cool and the ECU compensates by over fueling.

No, that was what I meant. But on the other hand, I was wrong .

Before engaging in a fight with you I looked at a datalog from a ride in some 10°C and the oil temp from sensor actually dropped from 80°C to 60°C pretty quick. As soon as I stopped it started climbing again. I usually don't ride when it's that cold and I had the impression a reading taken while riding (I can see the sensor temp with Cliff's Optimiser) would match the dipstick. But apparently it can be real bad!

[dlaing]

As long as you don't turn this supposed oil temp sensor into a cylinder head temp sensor I can't see how you would overheat it.

I don't think oil is getting pumped by the sensor adapter.

I recall it being oil free.

I think it is simply measuring the temperature of the cylinder.

I could be wrong. Anybody know for sure?

Is there a market for a tiny windshield to help maintain a more accurate temperature?

FWIW I found this on an MV Agusta site.

http://www.mvagusta.net/forum/showthread.p...2228&page=3

The Ohms match our bike almost identically. The voltage readings should be close enough to test the sensor, roughly....I could not find a photo of an MV Agusta sensor, but the Ducati 748, 916, 996, etc., sensors look like they MIGHT work in our bikes. I checked a Ducati manual and the numbers correspond correctly, although they only measure up to 80C and not 125C, presumably because the Ducs are water cooled.

I could not find a manual for an air cooled Ducati...

Should we move these posts to a new thread?

[raz]

I don't think oil is getting pumped by the sensor adapter.

I recall it being oil free.

I think it is simply measuring the temperature of the cylinder.

I could be wrong. Anybody know for sure?

It is called "oil temperature sensor" in WHBs (like the one for V11sport) and other publications but that may be for historical reasons. I was thinking about putting a sensor in the sump but that could show too warm instead. Imagine a real hot bike being parked in cold rain and wind for a while. Oil in sump will still be pretty warm but cylinder heads pretty cool. So we want cylinder head temperature of some sort but a "real" CHT sensor (like a ring replacing the spark plug gasket) would probably be overkill, and anyway it is not a plug in replacement.

 

This thread (1st reply) show an interesting alternative: http://guzzitech.com/PHPBB2/viewtopic.php?t=922 Apparently the commercial item is very expensive but you could DIY by just epoxying the NTC to your blanking plug.

[pete roper]

Couple of things here.

 

1.) It's a temperature sensor. I don't think it matters what you call it! I also belive that once the engine is HOT there is no need for any real variation in the mixture. Why should there be? There never was on carbureted engines and they worked well enough for well over half a century!

 

2.) You really, really don't want to stick the temperature sensor in the sump! Why? Because as soon as you're riding in the rain the oil temperature plummets! My SP has both a pressure and temperature guage, both are *reasonably* accurate. I can sit in traffic in Canberra in 40* heat and the oil temp will hit 150*C! Get moving and it will rapidly drop to back to 115-120*C. Ride into a storm, (As you often get in the afternoons at this time of year.) and suddenly, even though air ambient will still be 30*C +, the oil temperature wll plummet to roughly air ambient in a matter of a couple of minutes, if that! Evaporative cooling and simple conduction to the water running of the sump means that the oil gets quite cool very quickly. In Winter the effect is even more extreme.

 

I have to say I was surprised when I first observed this, I hadn't realized exactly how radical the effect of water on the sump would be but it IS a fact. You can do a simple test with your hand. Go for a flog in the wet and then put your hand on the sump. If it's more than luke warm I'd be surprised, even on one of the lean running modern bikes.

 

Pete

[raz]

Yeah that is the gist of it. I think one of the Right Things you could do with this is what Cliff does. Treat anything above 50°C as "we do not need a choke". Still, I'm interested in whatever quick'n'dirty fix you could do to make the OEM happy. The potentiometer solution, while fool-proof, is down right ugly.