Throttle Position Sensor

[Mr. Bean]

. Maybe Mr Bean has read the 485mV TPS @ 3.6 degrees but from a zero start setting and without the 150mV pre-dialled in?

 

GJ

 

That is exactly what I did based on what MI told me. They don't use the 150mv starting point. They had me set all the engine parameters to what they said. Bleeds to 1 turn open, idle at 1050 +- 50, balanced at 1800 to 2000rpm and THEN set the TPS to 3.6 deg. Which he said would be around 485mv. Mine was spot on this when I tested it. (Well actually it read 483mv...but very close anyway.)

 

On my bike I could never get the "stock factory settings to work at all. If I set my TPS to 150mv with everything disconnected, then hooked everything up and tried to set the idle to 1050 and the TPS to 525mv...my bleed screws would have to be all the way closed! I tried it a number of times changing and tweaking things to get all the "suggested" factory parameters to work. Nothing worked well for me until I did it Micha's way. Finally I have a bike with a stable idle speed regardless of temperature and FINALLY it doesn't ping anymore under hard acceleration. It is also much smoother taking off from a standing start than before.

 

Given that...I'll keep the 485mv setting! But of course your mileage may vary. I may have a wonky TPS for all I know.

 

Randy

[dlaing]

So, are we finally in agreement then?

 

If there is a need for a reference point to start setting up a 15M ecu equipped bike, including the V11Sport, and you don't have access to whatever brand of software, then 525mV @ 1060 +/- 50 rpm idle is a good point to aim for. Higher or lower readings may be be reached as a final tweaking but if you start at 485mV and then apply your +/- .5 degree (approx 100mV) then it could drop way too low. Maybe Mr Bean has read the 485mV TPS @ 3.6 degrees but from a zero start setting and without the 150mV pre-dialled in?

 

GJ

Jack, I was never disagreeing with you.

But actually, I am not at all sure we are in agreement, because now I really don't understand why Mr. Bean got different numbers.

What is a "zero start setting"?

Does that mean that the ECU assumes that when you turn the key on you are at idle and it marks the map there at 3.6°?

If so, that makes setting it to 3.6°±0.5 kind of useless, and doing it by millivolts would make more sense.

Also, I still don't understand why one even bothers with an at idle TPS reading.

If you set TPS to 150±15mV and then set the idle to 1000-1200RPM with bypass screws out about 1/2 turn, you should be good to go.

Why bother with 3.6°±0.5 or even 525mV±100mV at idle?

It should lie within that range anyway, assuming you set TPS to 150±15mV and then set the idle to 1000-1200RPM with bypass screws out about 1/2 turn and started the ECU with throttle backed off. Right?

But then in order to get bike to run right, Mr. Bean HAD to do it to 3.6°, so this leaves me very confused.

Feel free to keep disagreeing...

[guzzijack]

By 'zero start setting' I was surmising that Mr Bean hadn't set his TPS to 150mV with the butterfly closed i.e. he had 0mV at that point - from his message above it seems that I guessed correctly.

 

Without the start setting of 150mV his TPS would of course read lower than the expected 525mV at a rotation of 3.6 degrees. The fact that his reading was 485mV would suggest to me that something is indeed amiss (maybe the TPS itself) as he should be getting 525 - 150 = 375mV. That's probably why the 'normal' method has never worked for him before. Of course I defer to the guys at MI as they see far more bikes than I do so maybe their method is something they have refined to deal with a problem they are aware of and I'm not .

 

Why are we getting so wound up about the relationship between degrees of throttle opening and millivolts read at the TPS? Well, I guess that we need to establish the link because the factory literature, which is in the main aimed at dealers, assumes that some form of software will be used to set the bike up and therefore all references are to degrees of throttle opening. Many (most?) home mechanics will only be armed with a multimeter so datum points in mV such as those we have discussed are much more important to them.

 

IMHO in the majority of cases on any of the bikes with EFI the most important one is the 150mV baseline setting - okay so there is sometimes a need to even play with that on some bikes - followed by the 'recommended' at idle point to aim for i.e. 525mV on 15M ECU bikes. As we know that figure is not set in stone and it doesn't work for all bikes but it is a good starting point.

 

GJ

[Ryland3210]

By 'zero start setting' I was surmising that Mr Bean hadn't set his TPS to 150mV with the butterfly closed i.e. he had 0mV at that point - from his message above it seems that I guessed correctly.

 

Without the start setting of 150mV his TPS would of course read lower than the expected 525mV at a rotation of 3.6 degrees. The fact that his reading was 485mV would suggest to me that something is indeed amiss (maybe the TPS itself) as he should be getting 525 - 150 = 375mV. That's probably why the 'normal' method has never worked for him before. Of course I defer to the guys at MI as they see far more bikes than I do so maybe their method is something they have refined to deal with a problem they are aware of and I'm not .

 

Why are we getting so wound up about the relationship between degrees of throttle opening and millivolts read at the TPS? Well, I guess that we need to establish the link because the factory literature, which is in the main aimed at dealers, assumes that some form of software will be used to set the bike up and therefore all references are to degrees of throttle opening. Many (most?) home mechanics will only be armed with a multimeter so datum points in mV such as those we have discussed are much more important to them.

 

IMHO in the majority of cases on any of the bikes with EFI the most important one is the 150mV baseline setting - okay so there is sometimes a need to even play with that on some bikes - followed by the 'recommended' at idle point to aim for i.e. 525mV on 15M ECU bikes. As we know that figure is not set in stone and it doesn't work for all bikes but it is a good starting point.

 

GJ

 

Here's my take on Mr. Bean's experience:

 

I think it is very unlikely that his throttle closed TPS voltage was zero. It makes sense that it was near 150 mv. When he increased the throttle opening until the TPS was at 525 mv he was opening it farther than it would be at 485, therefore letting in more air. That explains why he had to close down his air bypasses to get the idle RPM right. In order words, holding the idle RPM as a constant, there is an inverse relationship between how far the throttle is opened and how far the bypasses are open. There is no mystery in this.

 

Once the throttle plates are balanced, the more they are opened, the less open the air bypasses will end up to achieve the target RPM. If the throttle plates are opened too much, the air bypasses will have to be closed altogether. So the 485 mv setting happens to be low enough that the bypasses must be opened to reach 1060 or 1100 RPM. The advantage of this is then the bypasses can be used to conveniently balance the vacuums and fine tune idle RPM.

 

Theoretically, although I haven't tried it, the same idle RPM could be reached by keeping the bypasses closed, and adjusting the idle adjusting screw. However, this means that the ability to balance using the connecting rod adjustment at cruise RPM and independently balance at idle using the air bypasses is lost.

 

So here is a proposed summarized procedure to append to the draft procedure for those seeking perfection:

 

A When checking balance at cruise RPM, make any fine correction needed using the connecting rod adjustment.

B. Check balance at idle RPM. If OK, done, if not, rebalance at idle using the air bypass screws, and go back to step A.

[dlaing]

snip

Hmmmmm....I wonder what your base TPS is set to now.

And I wonder if the ECU flags the zero start to always be 3.6°, regardless of mV????

If I get a round tuit, I mess with my currently near perfect settings, just to find out

[dlaing]

Without the start setting of 150mV his TPS would of course read lower than the expected 525mV at a rotation of 3.6 degrees.

I don't understand why you would be certain of that conclusion.

I can understand why a start setting lower than 150mV would of course read lower than the 525mV at 1050 RPMs with airbypass screws out 1 full turn

Until I read Mr. Bean's post I always assumed 3.6 degrees would equal ~525mV, regardless of whether TPS was initially set to 100 or 200mV. I think if was set to 0mV it would run like crap.

Finding out that 3.6 degrees does not always correspond to ~525mV has surely thrown my world of understanding upside down.

MUST VERIFY!

Why are we getting so wound up about the relationship between degrees of throttle opening and millivolts read at the TPS? Well, I guess that we need to establish the link because the factory literature, which is in the main aimed at dealers, assumes that some form of software will be used to set the bike up and therefore all references are to degrees of throttle opening. Many (most?) home mechanics will only be armed with a multimeter so datum points in mV such as those we have discussed are much more important to them.

I agree.

I only got wrapped around it because I thought Mr. Bean's 485mV/3.6° correlation indicated that the range of acceptable mV at idle was lower than what I had been led to believe from MPH's charts.

Now, I am more perplexed than ever.

I thought we were on route to a simple, clear, and definitive set of tune-up instructions.

Still I think Ryland's instructions may still be very close to ideal, and just needs some polishing as I suspect the degrees open can safely and effectively be ignored by all but the cheese heads with computers hooked up to their bikes.

IMHO in the majority of cases on any of the bikes with EFI the most important one is the 150mV baseline setting - okay so there is sometimes a need to even play with that on some bikes - followed by the 'recommended' at idle point to aim for i.e. 525mV on 15M ECU bikes. As we know that figure is not set in stone and it doesn't work for all bikes but it is a good starting point.

GJ

I agree.

[dlaing]

Here's my take on Mr. Bean's experience:

snip

Sounds good to me.

FWIW mine is set to just under 150mV, I seem to recall 148mV. At idle I think it is just under 1200RPM, I should verify with computer, and mV at idle is 527. Left air screw is one half turn and right is nearly one full turn.

It could be running better, so I am NOT suggesting these settings are a panacea.

But this is the closest I have ever gotten to 150mV, so I don't want to mess with it.

[guzzijack]

I've been thinking about this after posting yesterday and realised a glaring error in what I was thinking. Of course the ECU doesn't 'see' degrees of throttle opening, it merely responds to the TPS signal in mVs and converts that to degrees of opening as observed via the software programme(s).

 

Why there should be a measurable difference (485 v 525) between the two methods discussed at an indicated 3.6 degrees is indeed a strange one .

 

Looking at the previously available information on how to adjust/setup the EFI it starts with the EFI manual for the P7/P8/16M bikes. The method of TPS adjustment in that was laughable - twist it and allow it to return under it's own spring pressure'!!!

 

Then came the baseline method of 150mV with throttle fully closed and the 'target' voltages at idle for the different bikes.

 

Following that, and presumably coinciding with the more widespread availability of tuning software, came the specifications for setting at degrees of throttle opening at idle rpm with no mention of baseline or target mV readings - why would there be, the literature is aimed at those (dealers in the main) with the tuning software.

 

I'm guessing, but it could be that the difference between the two methods of setting up (degrees v mV) has arisen because of that changeover. It would be interesting is someone (Mr Bean maybe?) could measure TPS output with the throttle butterfly fully closed and disconnected after setting at idle using the MI method. After all, there will still be many riders who can only use the mV method.

 

GJ

[Mr. Bean]

I am swamped for the next couple weekends...but I gotta admit this has my curiosity peaked. I will try and get a few moments in the next couple weeks to disconnect everything and close the butterfly fully and see what it shows. I remember when I was doing the procedure I thought about doing it then just to see, but never got around to it.

 

Does anyone know what the effect the TPS has on ignition timing? I assume as we back it off it retards the timing and advancing the TPS in turn advances the timing? One of the items I always fought on my bike was pinging...especially in the 3500 to 5000 rpm range. I've tried to resolve it with a PCIII even going as high as +35 or 40 on the map in those areas to no avail. I began to believe that my timing was more the cause than mixture. Before when my idle would show around 525mv I could not get the bike to stop pinging in that rpm range. Now with the current setting it doesn't ping anymore...I assume because the timing is now retarded from where it was before?

 

Randy

[Ryland3210]

I've been thinking about this after posting yesterday and realised a glaring error in what I was thinking. Of course the ECU doesn't 'see' degrees of throttle opening, it merely responds to the TPS signal in mVs and converts that to degrees of opening as observed via the software programme(s).

 

Why there should be a measurable difference (485 v 525) between the two methods discussed at an indicated 3.6 degrees is indeed a strange one .

 

Looking at the previously available information on how to adjust/setup the EFI it starts with the EFI manual for the P7/P8/16M bikes. The method of TPS adjustment in that was laughable - twist it and allow it to return under it's own spring pressure'!!!

 

Then came the baseline method of 150mV with throttle fully closed and the 'target' voltages at idle for the different bikes.

 

Following that, and presumably coinciding with the more widespread availability of tuning software, came the specifications for setting at degrees of throttle opening at idle rpm with no mention of baseline or target mV readings - why would there be, the literature is aimed at those (dealers in the main) with the tuning software.

 

I'm guessing, but it could be that the difference between the two methods of setting up (degrees v mV) has arisen because of that changeover. It would be interesting is someone (Mr Bean maybe?) could measure TPS output with the throttle butterfly fully closed and disconnected after setting at idle using the MI method. After all, there will still be many riders who can only use the mV method.

 

GJ

 

It appears to the me that the difference between Mr. Bean's 485 mv at 3.6 degrees and MPH's chart showing 524 mv at 3.6 degrees is explained by Mr. Bean's settings and measurements made without reference to the throttle closed voltage. If his was actually 111 mv instead of 150, that would explain his 485 mv.

 

Therefore, I agree that Mr. Bean's 485 is not a valid number and another datum is needed. Let's look at the MPH chart: Within a small range of throttle opening, the change in mv/degree should be fairly constant.

Here's what the math says:

Degrees mv mv/degree

3.4 501

3.6 524 115 from 3.4

3.9 567 143 from 3.6, and 132 from 3.4

4.1 579 60 from 3.9, and 111 from 3.4

10.4 1258 108.14 from 3.4

 

I find it extremely unlikely all these numbers are correct. Assuming 108 is the closest, since it is taken over the largest range, then 150 + 3.4 X 108 = 517 (a difference of 16), and 150 + 3.6 X 108=539 (a difference of 15)

 

These descrepancies suggest that the chart was constructed by measuring voltages, rather than based on the engineering theory, and that the actual TPS voltage was 134-135 mv during the test. If so, the most reliable is the 108.14 number. That gives us a revised table as follows:

Degrees mv

0 150

3.2 496

3.4 518

3.6 539

As I indicated previously, it seems that it makes little difference whether one uses anything from 3.2 to 3.6, because the air bypasses will end up compensating for that when the idle RPM is adjusted to the 1000 to 1200 range. On top of that, apparently my tuning, just completed at 485 mv, corresponds to 3.1 degrees, and seems to run just fine. It also explains why my air bypasses had to be opened more than 1 turn.

As dlaing pointed out, the MPH chart shows no difference in fuel delivery between 900 and 1300 RPM for the throttle settings in the range of interest. Therefore, it does not seem to matter which of the above TPS voltages one targets for. The air bypasses will simply be opened to whatever it takes to get to the target idle RPM. Since in that range the fuel delivery is constant, using a smaller throttle position will not result in saving fuel, as I had earlier believed when I thought fuel delivery would increase with throttle position (within that range). It still remains that by offsetting the TPS to "trick" the ECU could be used to richen or lean the mixture at higher throttle openings and RPM's above 1300.

So where this leaves me is to change the target TPS to 518 at 3.4 degrees. 539 and 3.6 degrees would probably work also, but the lower number provides more margin for error in TPS calibration, to make sure the air bypasses are open a reasonable amount to achieve the idle RPM and make it easy to balance.

I propose to set the idle RPM target at 1100 to 1200, and to leave out target air bypass settings.

 

Before I cast this in stone, any further comments?

[dlaing]

So where this leaves me is to change the target TPS to 518 at 3.4 degrees. 539 and 3.6 degrees would probably work also, but the lower number provides more margin for error in TPS calibration, to make sure the air bypasses are open a reasonable amount to achieve the idle RPM and make it easy to balance.

I propose to set the idle RPM target at 1100 to 1200, and to leave out target air bypass settings.

 

Before I cast this in stone, any further comments?

The recommended idle and TPS sound good to me.

Please cast it in plaster, for now

But....

Air bypass screw adjustment effects a few things:

It effects the mV to RPM relationship at idle.

I believe it has a very slight effect on mixture at idle, but I don't think it has as much effect as a similar screw does with carburetors. I believe one should not worry about its effect on mixture.

The effect on mixture when decelerating is more dependent on idle RPM setting than air bypass or TPS settings.

Air bypass effects balance at idle and can and should be used to balance at idle. (Some disagree. Do we need to sway them? or need to be swayed by them???)

Balancing with bypass screws is more forgiving at about 3/4 turns out than at 1/4 turn out.

I would favor a minimum air bypass screw setting of 1/2 turn and the lagging cylinder should have air bypass increased to what ever is need to balance at idle, after cruising balance has been achieved.

I'd like to comment more about the MPH chart on mV to degrees relationship, but I need more time for that, including time in garage with TuneBoy

[dlaing]

Does anyone know what the effect the TPS has on ignition timing?

The first map should represent most of our bikes. the second map is the Titanium muffler kit ECU map.

[guzzijack]

It appears to the me that the difference between Mr. Bean's 485 mv at 3.6 degrees and MPH's chart showing 524 mv at 3.6 degrees is explained by Mr. Bean's settings and measurements made without reference to the throttle closed voltage. If his was actually 111 mv instead of 150, that would explain his 485 mv.

 

Before I cast this in stone, any further comments?

 

I don't think that the fact that Mr Bean didn't set the throttle closed voltage will directly influence the disparity between 485 and 524 in the way you have described it. The ECU and whatever brand of software used to access it are only responding to the TPS output in millivolts and converting that to degrees of throttle opening. I initially thought that the lack of a 150mV baseline would jumble the figures but it shouldn't have any influence at all on the outcome - in essence the two methods of setting up (degrees versus mV) are two completely different ways of looking at the same thing but there should be a constant relationship between the two.

 

The fact that Mr Bean's software sees 3.6 degrees as equating to 485mV and Dave's software gives 524mV the same value is where the conundrum arises - throw into the mix the different methods of setting up that each have followed might provide an answer (maybe not!).

 

Also when trying to equate degrees to mV you might want to consider that the output from the TPS is not a straight linear progression. It is possible that it has an advancing (or retarding) curve to it. I believe I have seen the spec. for the PF3C kicking around the web somewhere but this appears to be confirmed by the FIM website. Here's the relevant quote;

 

"The computer measures the butterfly position using the Throttle Position Sensor (TPS). This sensor is precisely aligned on the butterfly shaft and afftects not only fuel delivery but ignition advance as well. Many owners are tempted to move this sensor on the shaft, as you can get more fuel delivery from the ECU in this way. But there are several goods reasons not to do this:

 

The Weber TPS sensors are NOT LINEAR. If you change the position from the factory setting, not only do you change the fuel delivery but you change the ignition advance. This means that the bike will have too much advance at partial throttle, leading to detonation (pinging) problems which were not there before.

The factory used a specific setting for it's mapping, and we use the same setting for our mapping. So if you want to achieve optimum results with our chips, then you should set the TPS to the factory settings. Then your engine will operate as close as possible to the engine we used for testing."

 

GJ

[Ryland3210]

I don't think that the fact that Mr Bean didn't set the throttle closed voltage will directly influence the disparity between 485 and 524 in the way you have described it. The ECU and whatever brand of software used to access it are only responding to the TPS output in millivolts and converting that to degrees of throttle opening. I initially thought that the lack of a 150mV baseline would jumble the figures but it shouldn't have any influence at all on the outcome - in essence the two methods of setting up (degrees versus mV) are two completely different ways of looking at the same thing but there should be a constant relationship between the two.

 

The fact that Mr Bean's software sees 3.6 degrees as equating to 485mV and Dave's software gives 524mV the same value is where the conundrum arises - throw into the mix the different methods of setting up that each have followed might provide an answer (maybe not!).

 

Also when trying to equate degrees to mV you might want to consider that the output from the TPS is not a straight linear progression. It is possible that it has an advancing (or retarding) curve to it. I believe I have seen the spec. for the PF3C kicking around the web somewhere but this appears to be confirmed by the FIM website. Here's the relevant quote;

 

"The computer measures the butterfly position using the Throttle Position Sensor (TPS). This sensor is precisely aligned on the butterfly shaft and afftects not only fuel delivery but ignition advance as well. Many owners are tempted to move this sensor on the shaft, as you can get more fuel delivery from the ECU in this way. But there are several goods reasons not to do this:

 

The Weber TPS sensors are NOT LINEAR. If you change the position from the factory setting, not only do you change the fuel delivery but you change the ignition advance. This means that the bike will have too much advance at partial throttle, leading to detonation (pinging) problems which were not there before.

The factory used a specific setting for it's mapping, and we use the same setting for our mapping. So if you want to achieve optimum results with our chips, then you should set the TPS to the factory settings. Then your engine will operate as close as possible to the engine we used for testing."

 

GJ

 

 

OK, let's review the fundamentals:

The ECU has no way to directly know the actual throttle position. That's the purpose of the TPS. If its voltage output is incorrect at zero degrees open, it must necessarily be incorrect at all throttle positions. Therefore it will definitely affect the"outcome".

 

Regarding linearity, my analysis of mv/degree on the MPH table showed that not only is the TPS linear over a small range of rotation (which is true for linear potentiometers and for small changes in rotation for non-linear potentiometers) up to 4.1 degrees, but it is nearly perfectly linear all the way to 10.4 degrees. I discount the 3.9 entry of 567 mv as a typo, because if you insert 557 in the table in it's place, the linearity works out fine, at 111 mv/degree in that region. Perhaps it is non-linear above that range, but I think that's unlikely. Why? because non-linear pots are more expensive than linear ones, and any non-linear function desired can readily been created in the table, which is clearly non-linear as it is shown. I believe that is what is probably meant by the Weber/non-linear comment.

 

So we get right back to the importance of calibrating the TPS sensor so its output is 150 mv with throttle closed so that the ECU inputs the actual physical position of the throttle. That way, the tables for fuel and ignition advance are accurate. As your quote points out, offsetting of the TPS will affect mixture, but it will also affect timing. Therefore, it is a better idea to modify the table than to offset the TPS, unless one studies the table and understands that both factors will change together.

 

Now to review how the idle TPS setting might affect economy in a small way. Let's assume that the ECU interpolates the fuel delivery between the entries on the table (ANYONE KNOW ANY DIFFERENT?). For example, there is a change towards richer, from 1.99 to 4.18 degrees on the table in the range of 900 to 1300 RPM. Therefore, the larger the throttle opening (and with it, TPS voltage) the richer. Conversely, adjusting the idle stop screw for a lower TPS and opening the air bypass screws to obtain the target RPM will result in less fuel consumption at idle. How much? I think the difference will be so small, only a perfectionist would be concerned. I have been accused of that, to which I confess. Since I have already set mine up at 485 mv, I'll leave it there and save a little fuel.

 

A side comment on the two ignition advance tables. I noticed the non-standard table advances the timing more at idle. In my experience, based on the numbers in the table, that will increase efficiency at idle, and require less bypass opening to obtain the target RPM. If I get a tune boy, one of the things I would do is advance the timing at idle from the standard 9 degrees at idle, and also as it shows in the green area. I would expect more power and less fuel consumption for the amount of increased advance shown. Of course, then it is wise to watch for pinging.

[Ryland3210]

Please review my draft procedure, which has been updated to what I think is a final version.

:!:

Thanks,

John