CHT and EGT Monitoring for CV CARB Tuning

[Steve]

... The [aircraft] engine is surprising important when you aren't on the ground...

Yes, and the only purpose of the propeller is to keep the pilot cool.



You think I'm kidding?




Stop the propeller while the plane is in the air and watch him start sweating.

.

[posplayr]

I did not want to mess with the magnets not worry so I found these on ebay for pretty cheap. Will need to make a mount. I also got the LM-2 and LMA-3 in the mail yesterday.

I won a pair of these. They are magnetic sensors that generate a change in output based on coming close to a peice of metal. They are also active in that they have preamplifiers that square up the signal inside so that the there is more noise immunity and the device will work up to 6mm away from the metal it is detecting. So I should be able to detect each of 5 bolt on my sprocket carrier by mounting off my rear shock mount.

If this is non threaded, then you make a clamp queezes down like a triple tube clamp.

Rockwell Long Range Inductive Proximity Sensor NEW 2CT!

Manufacturer: Rockwell Aut.
Model: 872C-M6NP12-D4
Description: Rockwell Long Range Inductive Proximity Sensor NEW 2CT
Condition: New
Packaging Condition Details: Good Condition

New
Quantity: 2CT
Item: Long Range Inductive Proximity Sensors
Bulletin: 872C
Input: 10-30VDC
Load-200mA Max
Output: NO
PNP Quasi
Sens. Dist-6mm
Items are new in original manufacturer sealed packages. Packages are in good condition.

[tejasmud]

An inductive proximity sensor would be pretty slick. Make it much easier to set-up as far as sensor placement.

Literature sheet on the sensor you posted.

http://samplecode.rockwellautomation.com/idc/groups/literature/documents/pp/872c-pp002_-en-p.pdf

E

[posplayr]

An inductive proximity sensor would be pretty slick. Make it much easier to set-up as far as sensor placement.

Literature sheet on the sensor you posted.

http://samplecode.rockwellautomation.com/idc/groups/literature/documents/pp/872c-pp002_-en-p.pdf

E

Yea I'm hoping this works well. After thinking about it for a while the acceleration outputs of the LMA-3 are going to be severly limited since any full power pulls in 1-2 will result in wheelies. Also 3rd gear pulls to redline will be pretty quick.

What I like about this is the signal conditioning of the inductive pickup that is built right in. Hopefully it gives a speed which is differentiable in the software using the "slope" function.
The 6mm should provide plenty of clearance even though I'm planning on using the sprocket carrier bolts.

[posplayr]

So after you are doing spending all this money/time on a 30 year old bike, you could have bought a new fuel injected bike that puts down nearly double the HP at roughly the same displacement lol

Mmmm, the problem with using TC's on an air-cooled motorcycle engine (unlike an aircraft engine, which constantly has air flow over it) is that a motorcycle engine is designed to get HOT. So the temps you see sitting at a red light, won't be the same as what you see driving down the road at 35mph... And maybe not what you see going 70mph. I'm not sure that using TC's on an air-cooled street engine is the best way. The engine bay of an aircraft that uses a reciprocating engine is designed to keep a steady and constant air flow over the jugs (cylinder heads).

I still think the best and most logical way is to use a wideband O2 sensor, mounted a few inches from the exhaust valve.

Also, the main reason that we use CHT's on aircraft is to keep from melting the engines. The engine is surprising important when you aren't on the ground... EGT's are used for mixture setting. But if you are going to go through the trouble of mounting EGT's in your exhaust pipes, then just use a fricken' wideband o2 sensor and you can get a perfect tune with that alone. Nevermind what your temps are, if your AFR is right, your temps will be right.

Flyingsteve,
Finally getting a chance to respond to your post. My primary objective are to take some of the mystery out of the carb jetting process. My ED is starting to be pretty far away from a standard pods and 4:1 setup with Stage 3 Dynojet. I'm also looking to get a set of GSXR CV carbs to replace my 1150 carbs. So the expense of the Innovate "tuner" setup is partly to have some fun but also to save some time, hopefully not run the engine to lean and maybe learn something. I'm suspecting I'm already at 130 hp at the crank and so I'm not sure that is easily doubled by any stock bikes .

What you say about cooling and it's effect on the TC temperatures makes sense which is part of the reason I'm questioning the use of individual TCs. The inboard cylinders on the 1150 are usually hotter and they were set at the factor at 2.5 above the outer cylinders. I guess it would be nice to confirm this is a good thing to do. Of course it is going to be relative to the load and cooling effects as a function of speed.

I do now have a O2 sensor but was really not motivated to install bungs on each exhaust port for a EGT and much less for an O2-sensor. The 02 sensor will go into the mid pipe unless I can convince myself that running a copper pipe up the tail pipe gives reasonable results. I would try to isolate the cylinder is possible by pushing further up the 4:2:1 divides.

So I guess this begs the question : Why did the factory go richer on the inside cylinders?

To reduce the temperature even though the centers run a little richer? (i.e. to get more even temp despite the richness)

Or run more gas through to maintain consistent AFR because the higher temp is leaning it out.

It seems like with higher heat in the center you can never get the cylinder to be the same as the outside anyway. So at best you are trying to reduce the differences which is a compromise between temp and AFR matching.

So bottom line is, given the mixture averaging of using a mid pipe bung or tail pipe mounted O2 sensor I'll never be able to isolate individual cylinders. So is uniformity across the bank an indicator and useful. The main issue will likely be 1-4 to 2-3 cyclinder temp comparisions. I already have a 2 channel Fluke 50D that can do this. So the only benefit to monitor all pipes across all operating RPM's is if they are going to change and I need to correlate it with RPM and O-2 as I ride. It doesnt sound like I need to. I can just compare on the dual display Fluke 50D. This is probably what i will try first before putting any more money into the setup

Jim

[posplayr]

That's right. The channels could be set-up to run other 5v signal sensors, but there is only one thermo channel.

If you were to want exhaust temps EGT probes would be the most accurate. From my experience with watching CHT it is hard to establish any good metric. It is basically how 'FlyingSteve' interprets what the temps will do. The thermocouples clamped to the down tubes would have a similar affect. One thing you will notice if measuring CHT is the temps will go up or down in regards to jetting changes. Rich AFR it will be lower, Lean will be higher naturally. Never tried 4 CHT thermocouples at the same time, but my guess would be 1 and 4 would be close but different than 2 and 3, which would be similar.

Based on that description I would want to tune for AFR using O2 at the mid pipe (rich side of AFR 14.7) bung then depending on the temperature differential between ave of 1-4 and ave of 2-3 go up the inner cylinder to cool them a little at the expense of a bit more richness. The Fluke 50D will do that for me.

So if you had to pic one TC and where to locate it (the LMA-3 TC input) where woudl you put it? Spark plug of a center exhaust?

On the Speed pick-up there is some split collar set-ups for around $80 in the 3rd Party Products off the Innovate site. I have one. It is just a hall sensor J-B welded to a plastic screw, a resistor, and the collar has 2 magnets 180° apart. I actually pulled the sensor off the end of the screw and J-B welded it to a fabricated aluminum bracket to place it where I needed on another bike. Had to shim the collar to fit the wheel. If you find another set-up that will work please share.

I'll see if the hall sensor has a number on it, and check the resistor. If they are common parts that could be had from a electric supply it might be simple enough to make.


E

As mentioned I got a inductive unit and will have to see if it has mounting nuts or will need a split collar setup or if all else fails JB weld.

[tejasmud]

Based on that description I would want to tune for AFR using O2 at the mid pipe (rich side of AFR 14.7) bung then depending on the temperature differential between ave of 1-4 and ave of 2-3 go up the inner cylinder to cool them a little at the expense of a bit more richness. The Fluke 50D will do that for me.

In opinion you might go crazy trying to tweak out the temps to be even. It could possibly be done in the pilot circuit, but beyond that needles and the jets should be constants. The temps across the cylinders will trend naturally in regards to AFR and load. At WOT the AFR is most likely going to be in the 13:1 to 12:1 range, where the best pull is observed by the engine. The pilot circuit transition into the needle circuit can be tweaked to run leaner for better fuel economy at cruise. I would stay richer than 14.7:1 for sure. You will begin to see what is going on once you are able to read the AFR, and the feel of how the engine pulls should come intuitively.

So if you had to pic one TC and where to locate it (the LMA-3 TC input) where would you put it? Spark plug of a center exhaust?

Try it on either 2 or 3 then try it on 1 or 4 to see if there is any offset. There should be. I would just place it on 2 or 3 and be done with it.

As mentioned I got a inductive unit and will have to see if it has mounting nuts or will need a split collar setup or if all else fails JB weld.

lol, JB weld has it's uses. It appears from the literature the sensors you picked up have a threaded barrel. May have to buy the nuts unless it comes with.

Yea I'm hoping this works well. After thinking about it for a while the acceleration outputs of the LMA-3 are going to be severely limited since any full power pulls in 1-2 will result in wheelies. Also 3rd gear pulls to redline will be pretty quick.

I'm pretty sure the acceleration will be logged in one of the two axis’s, X or Y, but not both. Last time I programmed the LMA3 there is a selection for either axis. Maybe I missed something. The data for acceleration may have some error in lower gears if the wheel comes off the ground, but not sure. Being the change in horizontal plane would be greater it makes sense it will be skewed.

It is a similar issue with the speed sensor. Mounted to the rear, break the tire loose and the MPH will go up. If mounted to the front pull the wheel up MPH goes down. Just have to find the most reliable data.

I've always done my pulls in a higher gear similar to how it would be done on a dyno at WOT. Rolling start, quickly getting into either 4th or 5th then from about 2500-3000 RPM hitting WOT and letting it climb until over red line once the main jets are correct.

Once hooked up you will begin to get the feel of it. It is really nice to see the measurement. I usually just tuck the LM1 in the clear map pocket of my tank bag. Being able to see the display of the unit is what makes tuning the AFR so simple. A quick burst at WOT, turn around, back to the garage, adjust mains if needed, and then repeat until the AFR is close. Once close find the designated "plug chop run" one uses, and make a pass at WOT and log the run. Head back to the garage, evaluate, and tweak if needed.

On the other throttle positions I've usually just held them constant by watching the tape on the throttle housing. Tape that is marked, but a pin with a little ball pushed into the grip for an indicator. Then try to remember the log sessions when I get back to the laptop.

I think it will be a little different with CVs, might need to get the MAP to become consistant during the log session. I've only run with mechanical slide carbs. Now I will not have to remember the throttle position given I added a string pot for TPS. I'm excited to see how this works out.

I bet you will wonder why you have not been running the DAQ until now. You will learn, but it will be fun too. At least for me it is fun. Once you are comfortable with measuring the variables you have now, the thought may cross your mind, "What else can I measure and log?" Gear position? Or, suspension travel? It goes on and on.... Next thing you know you have $5k worth of stuff monitoring this and that..

One thing you might concider is a LC1 with a DB gauge set-up. Once you have all the tuning done with the LM2 and tired of carrying it with, it is nice to see the AFR.

E

[posplayr]

In opinion you might go crazy trying to tweak out the temps to be even. It could possibly be done in the pilot circuit, but beyond that needles and the jets should be constants. The temps across the cylinders will trend naturally in regards to AFR and load. At WOT the AFR is most likely going to be in the 13:1 to 12:1 range, where the best pull is observed by the engine. The pilot circuit transition into the needle circuit can be tweaked to run leaner for better fuel economy at cruise. I would stay richer than 14.7:1 for sure. You will begin to see what is going on once you are able to read the AFR, and the feel of how the engine pulls should come intuitively.

That all sounds good. Just need to get the unit installed.

I'm pretty sure the acceleration will be logged in one of the two axis’s, X or Y, but not both. Last time I programmed the LMA3 there is a selection for either axis. Maybe I missed something. The data for acceleration may have some error in lower gears if the wheel comes off the ground, but not sure. Being the change in horizontal plane would be greater it makes sense it will be skewed.

Measuring acceleration on a car in the X and Y direction makes sense as the car is not as likely to wheelie. Also the car does not bank in a turn.

On a motocycle the bike leans in the turns and it is not obvious that it is at a roll angle consistent with the turn radius. If you remain in the saddle and dont lean out it is probably pretty close however. All that means is the force is straight down in the body frame (i.e. the gravity and centriptal forces are balanced to keep the bike at a constant roll angle).

For a motorcycle getting both axis in the vertical plane (or the X-Z plane of the biek) woudl be best.

It is a similar issue with the speed sensor. Mounted to the rear, break the tire loose and the MPH will go up. If mounted to the front pull the wheel up MPH goes down. Just have to find the most reliable data.

I've always done my pulls in a higher gear similar to how it would be done on a dyno at WOT. Rolling start, quickly getting into either 4th or 5th then from about 2500-3000 RPM hitting WOT and letting it climb until over red line once the main jets are correct.

I'm figuring the rear wheel is less likely to spin than the front wheel coming off the ground.

I'll not be going to redline in 4th anywhere near my house and much less likely in 5th. Redline in 3rd is all I'll venture on the 101 FWY

Once hooked up you will begin to get the feel of it. It is really nice to see the measurement. I usually just tuck the LM1 in the clear map pocket of my tank bag. Being able to see the display of the unit is what makes tuning the AFR so simple. A quick burst at WOT, turn around, back to the garage, adjust mains if needed, and then repeat until the AFR is close. Once close find the designated "plug chop run" one uses, and make a pass at WOT and log the run. Head back to the garage, evaluate, and tweak if needed.

so you still do a plug chop santy check? At what throttle positions?

On the other throttle positions I've usually just held them constant by watching the tape on the throttle housing. Tape that is marked, but a pin with a little ball pushed into the grip for an indicator. Then try to remember the log sessions when I get back to the laptop.

Was figuring on that, but also with MAP I should be able to tell what the throttle setting was that goes with different data sets.

I think it will be a little different with CVs, might need to get the MAP to become consistant during the log session. I've only run with mechanical slide carbs. Now I will not have to remember the throttle position given I added a string pot for TPS. I'm excited to see how this works out.

I bet you will wonder why you have not been running the DAQ until now. You will learn, but it will be fun too. At least for me it is fun. Once you are comfortable with measuring the variables you have now, the thought may cross your mind, "What else can I measure and log?" Gear position? Or, suspension travel? It goes on and on.... Next thing you know you have $5k worth of stuff monitoring this and that..

One thing you might concider is a LC1 with a DB gauge set-up. Once you have all the tuning done with the LM2 and tired of carrying it with, it is nice to see the AFR.

E

I have studied instrumented flight data for years so the fascination of instrumented Moto tuning will become familiar quickly I'm sure. However my primary goal is improved and reliable tuning and not gadgetry.

Suspension tuning would probably be done best with track lap times v.s. trying to do trajectory reconstruction so I'm not sure how much value there is in that.

Thanks for you detailed response. I'll let you know how the prox gauge works.

[tejasmud]

so you still do a plug chop santy check? At what throttle positions?

Yes. I normally check the plugs at idle, and also check them a time or two at WOT. Have found two points helps establish the AFR is reading accurately and then the rest usually falls into place. I'll do fresh air calibration on the O2 sensor periodically depending how far off any reading appears to be. Just to be positive.

A good example would be where I was getting 20:1 at the LM at idle on the initial start-up a few weeks ago. I could smell it was too rich from the exhaust fumes. Quick enough it was discovered the Bosch O2 sensor was not tightened down. It was hand tight. Checking the plugs confirmed the circuit was rich. Tightening it down solved the problem with the reading.

This example also gives a good indicator of how sensitive the O2 sensor is to any air leaks. My guess is this is a similar issue Katman was experiencing with the lean reading using the end clamp in the pipe. Any leakage at the midpipe connection or pulse of the exhaust where it can suck air back in to skew the sensor will normally show lean readings or a large variation in readings. Noise. Plug chops and sensor calibration helps give confirmation.

To be honest just measuring the AFR alone is priceless. Especially when working with aftermarket/non-stock configuration carburetors, and not having any base to work from on the jetting. I plan on tuning in the VM29s, then later play with the VM33s, and I have a few slingshot sets lying around, although I'm not too keen of CVs.

Any info I can help with just ask. Not a problem. If I have had any experience with the question I don't mind relaying my observations. If I don't know the answer I'll say so, but that usually has me try to look it up to seek the answer. Any info I can find, will post the source. The Innovate forum doesn't have a ton of info on bikes, but still there is a ton of relevent info. I weed through there every once in a while.

I hear what you are saying on the "gadgetry." Once in tune, I pull the equipment off and put it back in the boxes. I'll run a LC1 w/gauge to keep tabs on the AFR, but that's about it afterwards.



E

[posplayr]

Here is a quick update of what I have been up to. Still working out the bugs but have some data collected. See below.

Problems with a pdf to show the setup; later

[posplayr]

Interim Final results of getting the Innovate stuff to work and then using it to tune my carbs. It has been a learning experience.

http://www.thegsresources.com/_forum/showpost.php?p=1303763&postcount=12

my trials and tribulations getting the LMA-3 SPEED and RPM to work properly. They are almost flawless now.

http://www.innovatemotorsports.com/forums/showthread.php?t=12410