TPS adjust/TB balance/air bleed screws

[dlaing]

Micha's method compared to the How To method:

Step one - Set your valves to world settings

Step two - Set your bleed screws to open 1 full turn

Step three - Synch throttle bodies at just off idle (around 1800 rpm)

Step four - Set idle to 1100 using left idle screw adjuster only

Step five - Set TPS to 3.6 degrees (I forget what that translates to in mv..someone here will know)

Step six - Ensure idle trim is set to zero

 

One important difference is that the How To method leaves nothing to chance, whereas the Micha's method takes shortcuts by making assumptions.

 

For example, it assumes the left idle screw controls both throttles and that the right idle screw and fast idle is not in contact.

Synchronizing the throttle bodies at just off idle should eliminate interference from the two idle screws so that the connecting rod is in control, just as in the How To method, except that any variation in the flow rate delivered by the two bleed screws will require an offset of the two butterflies in order to compensate. The assumption here is that the difference is negligible. The How To method avoids this interaction.

The How To method quotes 521 mv as the middle of the published ranges as a guideline. That corresponds to 3.5 degrees, almost identical to Micha's 3.6.

 

The other important difference is that Micha's method apparently disregards the 150 mv at throttle closed baseline calibration in favor of forcing the TPS to be adjusted to 3.6 degrees (532mv) at 1 full turn of bypass opening and 1100 RPM. In the case of the How To method, the TPS is calibrated at throttle closed, and if air bypasses are not opened at least 1/2 turn to achieve 1100 to 1200 RPM, then the idle screw is loosened by closing the throttles in steps (which will reduce TPS voltage) until they are.

 

Let's try an example using Micha's method. Suppose the TPS is calibrated correctly, 150mv at throttle closed.

Air bypasses are opened 1 full turn and left there. At 1100 RPM, it so happens that the TPS indicates 3.0 degrees. Micha says adjust the TPS to read 3.6 degrees anyway. The result is a richer mixture throughout the throttle range.

 

In my opinion, so long as the assumptions the Micha procedure makes are reliable, it has merit provided one is willing to accept the fact that it will likely result in an uncalibrated relationship between TPS voltage and actual throttle opening. For some riders and bikes this may be entirely satisfactory. However, it is important to understand that maps provided by others which one may wish to use as is, rely upon a TPS calibrated to the throttle closed baseline. The How To method should be used when setting up the bike for dyno tuning or using a map provided for a particular set of modifications or special purposes as is.

 

A lot of this has been gone over in previous posts. I just thought I would summarize it from my point of view.

 

regards, John

I knew we could count on you for a good analysis.

Since some will resort to Micha's method, maybe we should improve it a little:

 

Step one - Set your valves to world settings Intake .15mm, Exhaust .20mm

Step two - Set your bleed screws to open 1 full turn

Step two point one - Back the right idle screw and fast idle off so that they will not make contact with the motion of the right throttle body butterfly valve.

Step three - Synch throttle bodies at just off idle (around 1800 rpm)

Step four - Set idle to 1100 using left idle screw adjuster only Note: Veglia tachometer is notoriously inaccurate. Use a digital tachometer, or computer running Power Commander, VDSTS, Tuneboy, Axeone, or whatever.

Step four point one - balance at idle using bleed screws (not sure if Micha would approve, but I think it would help)

Step five - Set TPS to 3.6 degrees 524mV or 532mV depending on your source. Actually it is a range of millivolts.

Step six - Ensure idle trim is set to zero Or if you have access to a CO meter, adjust trim to ideal setting (not sure Micha would approve, maybe it is best to keep at zero)

 

Still after those (IMHO) improvements, ignoring the 150mV baseline has potentially negative implications, but apparently neither method works perfectly for all bikes.

(I can imagine Micha reading this rolling his eyes )

[v11cafe]

I knew we could count on you for a good analysis.

Since some will resort to Micha's method, maybe we should improve it a little:

 

Step one - Set your valves to world settings Intake .15mm, Exhaust .20mm

Step two - Set your bleed screws to open 1 full turn

Step two point one - Back the right idle screw and fast idle off so that they will not make contact with the motion of the right throttle body butterfly valve.

Step three - Synch throttle bodies at just off idle (around 1800 rpm)

Step four - Set idle to 1100 using left idle screw adjuster only Note: Veglia tachometer is notoriously inaccurate. Use a digital tachometer, or computer running Power Commander, VDSTS, Tuneboy, Axeone, or whatever.

Step four point one - balance at idle using bleed screws (not sure if Micha would approve, but I think it would help)

Step five - Set TPS to 3.6 degrees 524mV or 532mV depending on your source. Actually it is a range of millivolts.

Step six - Ensure idle trim is set to zero Or if you have access to a CO meter, adjust trim to ideal setting (not sure Micha would approve, maybe it is best to keep at zero)

 

Still after those (IMHO) improvements, ignoring the 150mV baseline has potentially negative implications, but apparently neither method works perfectly for all bikes.

(I can imagine Micha reading this rolling his eyes )

Please,....not again.....

[dlaing]

Please,....not again.....

But it will only take four maybe nine more pages....

[v11cafe]

But it will only take four maybe nine more pages....

Ok then,... go ahead!!!!

[raz]

The problem is the inaccuracy of the voltmeter when trying to differentiate 4.860V from 4.850V, meanwhile differentiating 150mV from 160mV or even 151mV is not a problem.

So, the value of reading at WOT is to eliminate the problem that Tom experienced...we just have a technical inaccuracy issue.

You can measure with the same resolution at WOT. Just measure between the wiper and +5V instead of wiper to ground. That way the voltage will be below 2 volts (and in fact at a very similar range as the base voltage referenced to ground) so you'll get three decimal digits on virtually any multimeter. There is no downside, you will just get a tenfold increase in accuracy resolution.

[Mr. Bean]

I knew we could count on you for a good analysis.

Since some will resort to Micha's method, maybe we should improve it a little:

 

Step one - Set your valves to world settings Intake .15mm, Exhaust .20mm

Step two - Set your bleed screws to open 1 full turn

Step two point one - Back the right idle screw and fast idle off so that they will not make contact with the motion of the right throttle body butterfly valve.

Step three - Synch throttle bodies at just off idle (around 1800 rpm)

Step four - Set idle to 1100 using left idle screw adjuster only Note: Veglia tachometer is notoriously inaccurate. Use a digital tachometer, or computer running Power Commander, VDSTS, Tuneboy, Axeone, or whatever.

Step four point one - balance at idle using bleed screws (not sure if Micha would approve, but I think it would help)

Step five - Set TPS to 3.6 degrees 524mV or 532mV depending on your source. Actually it is a range of millivolts.

Step six - Ensure idle trim is set to zero Or if you have access to a CO meter, adjust trim to ideal setting (not sure Micha would approve, maybe it is best to keep at zero)

 

Still after those (IMHO) improvements, ignoring the 150mV baseline has potentially negative implications, but apparently neither method works perfectly for all bikes.

(I can imagine Micha reading this rolling his eyes )

 

Thanks for adding in the details Dave.

 

At the risk of adding to the nine pages my only quible with the above would be your added step 4 point 1. By balancing the TB's under step 3 at 1800 rpm and only 1800 rpm with the bleed screws open, you are already taking them into account with the balancing. If you adjust the bleed screws at this later point, would you not be offsetting the balance you have already done?

 

R

[Ryland3210]

Thanks for adding in the details Dave.

 

At the risk of adding to the nine pages my only quible with the above would be your added step 4 point 1. By balancing the TB's under step 3 at 1800 rpm and only 1800 rpm with the bleed screws open, you are already taking them into account with the balancing. If you adjust the bleed screws at this later point, would you not be offsetting the balance you have already done?

 

R

 

Of course it is true to say you would be offsetting the balance previously created. I think Micha's intent is to keep the procedure simple. His method balances at 1800 RPM as a priority over balancing at idle. I agree with that priority, and would do without the final balancing at idle. Those who want to take more care can use the iterative balancing procedure in the How To section.

 

Let's try another example using Micha's method. Suppose the TPS is calibrated correctly, 150mv at throttle closed.

Air bypasses are opened 1 full turn and left there. At 1100 RPM, it so happens that the TPS indicates 3.9 degrees. Micha says adjust the TPS to read 3.6 degrees anyway. The result is a leaner mixture throughout the throttle range.

 

Measurements of TPS voltage at WOT are subject to the accuracy of the 5 volt reference voltage. What if the reference voltage is off by 1%, which is not unreasonable to expect for the typical reference chip? Then the 5 volts can be anywhere from 4.95 to 5.05, a range of 100 mv. Measuring the difference between the 5 volt reference and WOT is better. However, a 1 degree unbalance at WOT has little effect, whereas a 1 degree unbalance at off idle and cruise has a big effect. What if the ratio of linear cable motion to butterfly rotation varies between throttle bodies? For any and all of these reasons, I strongly suggest using the lower end of TPS range for calibration.

[dlaing]

You can measure with the same resolution at WOT. Just measure between the wiper and +5V instead of wiper to ground. That way the voltage will be below 2 volts (and in fact at a very similar range as the base voltage referenced to ground) so you'll get three decimal digits on virtually any multimeter. There is no downside, you will just get a tenfold increase in accuracy resolution.

You are THE MAN!

I have some experimenting to do now....

[dlaing]

Thanks for adding in the details Dave.

 

At the risk of adding to the nine pages my only quible with the above would be your added step 4 point 1. By balancing the TB's under step 3 at 1800 rpm and only 1800 rpm with the bleed screws open, you are already taking them into account with the balancing. If you adjust the bleed screws at this later point, would you not be offsetting the balance you have already done?

 

R

Good point, we should add a few repeat steps....

IMHO idle balance is especially important for those of us that spend hours per year waiting at the Red for the Green, and for any of us prone to popping while decelerating.

[dlaing]

snip

 

Measurements of TPS voltage at WOT are subject to the accuracy of the 5 volt reference voltage. What if the reference voltage is off by 1%, which is not unreasonable to expect for the typical reference chip? Then the 5 volts can be anywhere from 4.95 to 5.05, a range of 100 mv. Measuring the difference between the 5 volt reference and WOT is better.

Yes, thanks to Raz's suggestion we should use the 5volt reference, so we have no problem there.

 

However, a 1 degree unbalance at WOT has little effect, whereas a 1 degree unbalance at off idle and cruise has a big effect. What if the ratio of linear cable motion to butterfly rotation varies between throttle bodies? For any and all of these reasons, I strongly suggest using the lower end of TPS range for calibration.

Yes, but it could be less of a problem than Tom M.'s being off by 100mV.

Something is wrong with the 150mV baseline on some bikes.

Maybe it is carbon built up.

 

Maybe it is a variation of ratio of linear cable motion to butterfly motion, between throttle bodies.

Maybe it is a variation of ratio of TPS motion to butterfly motion, between throttle bodies of different bikes.

Maybe it is bad tolerances.

 

Using WOT as the baseline does not preclude using the lower end of the TPS range for calibration. We would still be targeting 3.5 or maybe even 3.6 degrees at the lower end of the TPS range specifically at idle, but unlike Micha, we would not be tweaking the TPS to match the idle, but rather the idle would be tweaked to match the TPS, as our method does.

 

What is physically happening when the butterfly valve fully closes?

Is the edge of the valve digging into the bore of TB?

What is happening when the butterfly valve fully opens?

Is there some sort of throttle stop at WOT?

 

Right now I can't strongly recommend our method, although I think it is better than anything else including Micha's, and I certainly can't recommend the WOT method until I test it, and maybe get some guinea pigs to test it.

 

Another solution to the dilemma that Mr. Bean and Tom M. experienced is a compromise

Do our method, but rather than bumping the idle up 100mV to reach 521mV and getting too high of an idle, I suspect a variation of more than oh, maybe 25mV to obtain the proper idle should require resorting to tweaking the TPS to match the idle rather than tweaking the set screw to the point that Tom did and getting an idle TPS reading of 440mV or 480mV depending on bypass.

 

And still another method would be to start with the 150mV base, and if you can't get 1150RPM at 521mV with 1/2 to one full turn of bypass, then you should check the WOT TPS voltage and it should read 150mV between sweep and +5V. If it does not, then we grab a slide ruler and use the soon to be famous compromise algorithm. Perhaps too complicated for some, but hopefully few will run into the problem that these guys did.

 

Or we can steer them to the Micha method and wish them best of luck.

Any other ideas?

Sorry if I come across as argumentative, it is not my intent. I simply think we need a better all encompassing solution. And yes, I see some of the holes in every solution I have put on the table.

[Skeeve]

Good point, we should add a few repeat steps....

IMHO idle balance is especially important for those of us that spend hours per year waiting at the Red for the Green, and for any of us prone to popping while decelerating.

 

Hmm. Never much cared if the balance is off a tad at idle. Always balanced at idle ['cause it's easy to do as a 1st step], then rolled slowly on the throttle to bring the engine up to 3k rpm ['cause above this rpm the mercury tubes are bouncing around too much and at full throttle there's little difference between 100% open & 95% open...] and balance there as closely as possible. Go back to idle and see how far things are off & halve the discrepancy. Double check at 3k rpm, & if within reason, Robert's your mother's favorite brother. [Or in my case, my dad's younger, only, brother... ]

 

Whatever. The key is that the carbs [on bikes w/ those] or throttle bodies are balanced & pulling the same vacuum in the rpm range where the engine spends the most time/fuel! Obviously, few of us spend all of our time idling, & it will take many hours of idling to equal 5 minutes of runtime at full throttle.

Because an engine with unequal vacuum on the cylinders is quite literally working against itself.

 

As an example, on my old Yamaha XS Eleven [former VP of the XSives & have the patch to prove it! ], I would religiously keep track of mpg. It would avg. 32 for a month [big, inefficient engine on a bike w/ lard that makes our v11s look svelt!] and then wham!, drop to 28. Rebalance the carbs [a 30 minute operation, mostly because you had to remove the tank & hook up a remote fuel supply] and mileage immediately returned to normal.

 

Yes, I could have spent time fiddling w/ the carbs to get better mileage, but gas was only $1/gal back then, so why bother? The XS still got double the mileage of my truck & outdid most cars, so I wasn't crying about it...

 

Anyway, balancing at idle only resulted in a minor improvement to 30mpg on that bike [i try to avoid routes that result in a lot of stop'n'go idle time: I'd rather turn right on red & keep rolling on an alt. route than sit around waiting for a traffic sensor [b]negligently adjusted not to recognize motorcycles[/b] & hoping for a car to come up behind me while praying the BDC driving wouldn't ram me from behind... ]

 

That's my experience. Balance at 3k rpm is more important than idle, but you may need to sacrifice a little at the higher rpm to get an even idle. They're both important, basically.

[Guest ratchethack]

That's my experience. Balance at 3k rpm is more important than idle.

:!:

[raz]

What is physically happening when the butterfly valve fully closes?

Is the edge of the valve digging into the bore of TB?

What is happening when the butterfly valve fully opens?

Is there some sort of throttle stop at WOT?

I'm pretty sure you can easily ruin the TB's by snapping the plates too hard against the bore, too many times. Some instructions literally say you should do that (not using the word 'too' of course) but I don't like it.

At WOT there are tabs providing throttle stops similar to the idle stops, except there is no screw, it's permanent.

 

If it does not, then we grab a slide ruler and use the soon to be famous compromise algorithm.

Slide ruler

 

Sorry if I come across as argumentative, it is not my intent. I simply think we need a better all encompassing solution. And yes, I see some of the holes in every solution I have put on the table.

+1, I'm just enjoying the discussion. For people not interested in theoretical ping pong I'd just say try the pinned How-to and if it don't work out good, try Micha's method instead. Or the other way round if you wish. Probably one of them (or both) will produce a nice running bike.

[Ryland3210]

I'm pretty sure you can easily ruin the TB's by snapping the plates too hard against the bore, too many times. Some instructions literally say you should do that (not using the word 'too' of course) but I don't like it.

At WOT there are tabs providing throttle stops similar to the idle stops, except there is no screw, it's permanent.

 

 

Slide ruler

 

 

+1, I'm just enjoying the discussion. For people not interested in theoretical ping pong I'd just say try the pinned How-to and if it don't work out good, try Micha's method instead. Or the other way round if you wish. Probably one of them (or both) will produce a nice running bike.

 

I agree that snapping the plates closed is risky. Gentle pressure is enough.

 

It's slide rule, not ruler. The latter is used for scribing lines.

 

I agree completely with your last comment. No point in arguing with success, only with ignorance and errors!

[dlaing]

Balance at 3k rpm is more important than idle

But I am not going to recommend the Micha method that ignores the idle balance. (I am pretty sure Micha would agree and some of the steps in his methodology probably just got lost as it made it this forum)

 

I'd just say try the pinned How-to and if it don't work out good, try Micha's method instead.

The combination of the two in that order seems to be the best instructions that we have.

I thought we could do better.

Maybe we still can, but the instructions will lose their simplicity.

 

No point in arguing with success, only with ignorance and errors!

The problem is all the methods can result in "success" with errors, and that is still worth arguing over.

The rider does not know something is amiss unless the bike exhibits symptoms.

Spotting rich running symptoms is usually more difficult than spotting lean running symptoms.

So often people say their bike runs perfectly. Sorry, but perfection of this sort is unobtainable. The engine advance is always a little off perfect, and the mixture is always a little off perfection. But we can make an engine seem to run perfectly, if we get no popping, missing, pinging, surging, etc.

But we should be content if get no symptoms.