TPS break points at low throttle openings

[audiomick]

Finally got around to furthering the Quest to set the TPS base value.

After a first attempt several weeks ago, in which I had neglected to screw out the idle screw and make sure the "choke" wasn't holding the throttle valve open, and I measured 500 mV, I did it right today.

The TPS had apparently not been readjusted since the bike left the factory about 65,000 km. ago. There was still yellow and red sealing paint on both screws. What I found was 250 mV.

I reset it, aiming for the 157 mV that Meinolf recommends. With the movement that happens when it is all screwed up tight, I landed at between 159 and 160 mV. Given that the workshop manual states a tolerance of +/- 15 mV, I reckon that is ok.

After fiddling around with the "choke" a bit to see how it might be best set up, I tried starting it. The starter button funny business reared its ugly head again, but she started after about 4 or 5 depressing clicks. First time since about last June.

Idle was ok to be going on with. I hope to be able to spend some more time on it tomorrow and finish off the "decent tune up".

[docc]

10 minutes ago, audiomick said:

Finally got around to furthering the Quest to set the TPS base value.

After a first attempt several weeks ago, in which I had neglected to screw out the idle screw and make sure the "choke" wasn't holding the throttle valve open, and I measured 500 mV, I did it right today.

The TPS had apparently not been readjusted since the bike left the factory about 65,000 km. ago. There was still yellow and red sealing paint on both screws. What I found was 250 mV.

I reset it, aiming for the 157 mV that Meinolf recommends. With the movement that happens when it is all screwed up tight, I landed at between 159 and 160 mV. Given that the workshop manual states a tolerance of +/- 15 mV, I reckon that is ok.

After fiddling around with the "choke" a bit to see how it might be best set up, I tried starting it. The starter button funny business reared its ugly head again, but she started after about 4 or 5 depressing clicks. First time since about last June.

Idle was ok to be going on with. I hope to be able to spend some more time on it tomorrow and finish off the "decent tune up".

Taking the TPS from 250mV to 160 should give you some outstanding results across the map!

I would still rather strongly suggest that the +/- 15 mV in that addendum of the Workshop Manual is highly inadvisable if you have time to make it right.  Based upon reported findings from the Look-up Tables, 157mV +/- 3.5mV is most accurate and desirable.

Even with that tight tolerance, your 160mV checks out!

[audiomick]

2 hours ago, docc said:

...the +/- 15 mV in that addendum of the Workshop Manual is highly inadvisable if you have time to make it right. 

Of course. The more accurate, the better.

As it happens, I am quite concerned with accuracy. Some might even say "annoyingly pedantic". Looking at the "official" tolerance serves more to help me sleep well at night than to excuse my unwillingness to get it right. My difficulty is more in the direction of recognising when it is well and truly good enough.

[audiomick]

3 hours ago, docc said:

 Based upon reported findings from the Look-up Tables, 157mV +/- 3.5mV is most accurate and desirable.

Even with that tight tolerance, your 160mV checks out!

Another thought regarding fixing the throttle valve down with cable ties, or pushing it shut to do the "base level" measurement:

I observed today that pushing the throttle shut changed the measured voltage by about 4 mV. On the basis of that, I reckon there is no point in taking the extra effort to tie  or push it shut. Make sure it is clean and unhindered, let it snap shut a couple of times, and do the measurement. Any further measures just wont make a significant difference.

[audiomick]

Another another thought:

I just did some sums. As I mentioned further up, I'm quite pedantic by nature, so don't take this as a "near enough is good enough" thing.

I observed today that the "full range of travel" of my TPS was from the set 157 mV at fully closed to just over 4.8 V at fully open. As far as I know, the TPS / ECU combination has a resolution of 256 steps. So...

4,800 / 256 = 18.75

which means that the +/- 15 mV in the workshop book is less than one "step" in the resolution of the system, and 4 or 5mV variance is a relatively small value in the context.

[docc]

22 minutes ago, audiomick said:

Another another thought:

I just did some sums. As I mentioned further up, I'm quite pedantic by nature, so don't take this as a "near enough is good enough" thing.

I observed today that the "full range of travel" of my TPS was from the set 157 mV at fully closed to just over 4.8 V at fully open. As far as I know, the TPS / ECU combination has a resolution of 256 steps. So...

4,800 / 256 = 18.75

which means that the +/- 15 mV in the workshop book is less than one "step" in the resolution of the system, and 4 or 5mV variance is a relatively small value in the context.

This is very interesting stuff to the pedantic amongst us! (We may have to take this discussion to a dedicated thread as these "catch-all threads" just swallow our discussion up.)

Seeing your pedantic, and raising you an eyebrow:

One must consider the range is 4800 minus the 157 (4643). With 256 steps, one step is 18.1 mV (or +/- 9).  I thought I read the 7mV steps from Meinolf, but will have to try and find that reference . . .

I still maintain the Workshop Manual addendum of +/- 15 mV is suspect.

[Lucky Phil]

1 hour ago, audiomick said:

Another thought regarding fixing the throttle valve down with cable ties, or pushing it shut to do the "base level" measurement:

I observed today that pushing the throttle shut changed the measured voltage by about 4 mV. On the basis of that, I reckon there is no point in taking the extra effort to tie  or push it shut. Make sure it is clean and unhindered, let it snap shut a couple of times, and do the measurement. Any further measures just wont make a significant difference.

It's not about the accuracy predominantly Mick it's about the consistency and the fact that with the throttle blades gently cable tied shut you have one less thing to think about and check and also the blades don't try and follow the TPS as you adjust it.

 

Phil

[Lucky Phil]

1 minute ago, docc said:

This is very interesting stuff to the pedantic amongst us! (We may have to take this discussion to a dedicated thread as these "catch-all threads" just swallow our discussion up.

Seeing your pedantic and raising you an eyebrow:

One must consider the range as 4800 minus the 157 (4643). Assuming the 256 steps is correct (?), one step is 18.1 mV (or +/- 9).  I thought I read the 7mV steps from Meinolf, but will have to try and find that reference . . .

I still maintain the Workshop Manual addendum of +/- 15 mV is suspect.

Depends on where you are in the TPS range from memory docc and what TPS you have the linier or non linier. 

Also there are 256 break points but they are not distributed evenly. From 1000 rpm to 3000rpm and 2.4 degrees to 12.7 the there are more break points then the rest of the range so you have finer throttle control at smaller throttle angles.

 

Phil

[docc]

6 minutes ago, Lucky Phil said:

Depends on where you are in the TPS range from memory docc and what TPS you have the linier or non linier. 

Also there are 256 break points but they are not distributed evenly. From 1000 rpm to 3000rpm and 2.4 degrees to 12.7 the there are more break points then the rest of the range so you have finer throttle control at smaller throttle angles.

 

Phil

Ah, yes the "non-linear" thing.

So, how many break points from 2.4º to 12.7 ?

[audiomick]

10 minutes ago, docc said:

 I thought I read the 7mV steps from Meinolf, but will have to try and find that reference . . .

Quote from Meinolf (via DeepL.com) from the German forum:

"Furthermore, the 8-bit ADC used in the 15M/RC has a very low resolution, it can only calculate in steps of 5V / 256"

Regarding 8-bit resolution:

https://en.wikipedia.org/wiki/8-bit_computing#Details

256 steps.

Regarding what Phil wrote about the non-linear TPS response, yes I have read that elsewhere, and consider that to be accurate.

@docc if you think this is worth splitting off, go for it.

[audiomick]

A further quote from Meinolf from the same post, which confirms what Phil wrote about the non-linear response of some TPSs, I believe the earlier ones, which includes the V11:

"The difference may seem small, but a meticulous adjustment of this value is absolutely necessary. The interpolation points are so close together in the low load range that a shift of one row can quickly occur due to normal wear and tear in conjunction with the function of the ADC. And this is immediately noticeable in the low load range."

Meinolf was writing regarding a California Vintage, but I believe the ECU and throttle bodies were more or less the same as the ones in the V11.

[docc]

4 minutes ago, audiomick said:

A further quote from Meinolf from the same post, which confirms what Phil wrote about the non-linear response of some TPSs, I believe the earlier ones:

"The difference may seem small, but a meticulous adjustment of this value is absolutely necessary. The interpolation points are so close together in the low load range that a shift of one row can quickly occur due to normal wear and tear in conjunction with the function of the ADC. And this is immediately noticeable in the low load range."

Meinolf was writing regarding a California Vintage, but I believe the ECU and throttle bodies were more or less the same as the ones in the V11.

This ^^^

If the "average" break points through the 157-> 4800 mV range is 18.1 Mv (+/- 9), can the closer break points for the lower throttle angles be derived?

[Lucky Phil]

32 minutes ago, docc said:

Ah, yes the "non-linear" thing.

So, how many break points from 2.4º to 12.7 ?

2-11.9 is 9 break points. In that TPS range to 3200 rpm there is 72 cells.

The map goes to 8500 rpm and 84.8 degrees of throttle angle.

Sorry about the break point and degrees in previous posts, those are for my bike.

 

Phil

[audiomick]

1 hour ago, docc said:

 I thought I read the 7mV steps from Meinolf

No doubt that was referring to the lower range of the non-linear TPS. I saw a graph somewhere. The non-linearity is not a continuous curve, but rather a steep straight line, a bend, and another less steep straight line. 

1 hour ago, docc said:

I still maintain the Workshop Manual addendum of +/- 15 mV is suspect.

Yes, I was also very suprised at how large that range is. Thinking about the numbers involved, I can see it as a pragmatic compromise within the context, but really accurate is mostly better. And lets me sleep better. 

[Lucky Phil]

58 minutes ago, audiomick said:

No doubt that was referring to the lower range of the non-linear TPS. I saw a graph somewhere. The non-linearity is not a continuous curve, but rather a steep straight line, a bend, and another less steep straight line. 

Yes, I was also very suprised at how large that range is. Thinking about the numbers involved, I can see it as a pragmatic compromise within the context, but really accurate is mostly better. And lets me sleep better. 

https://www.bikeboy.org/ducatitps.html

 

Phil