Linear Throttle Position Sensor setting notes for all models - Written 08/08
Summary: a discussion of the linear TPS output, official and non official set up procedures, specific issues for Euro 999S/999 ‘05 and USA 749 models, 2V models with idle control valves and fitting Ducati Performance ECU.
Note: In this report I reference the Mathesis diagnostic tester, which is what I was using when the report was written. It’s many years since I used the Mathesis. The common diagnostic tools now are as follows:
DDS the second generation Ducati electronic diagnostic tool stand alone.
Technoresearch’s Centurion S PC based. Formally the MDST and VDSTS. This was also the official MV Agusta tool, but it also works on many different brands and can be purchased in a few different capability specs with prices that vary accordingly.
Texa Navigator (supercedes the Axone) stand alone. Nominally the official Aprilia electronic diagnostic tool, but available as a generic all brands tool as well. As I worked at an official Aprilia and Moto Guzzi dealer we had an Axone that for general workshop convenience was upgraded to the “multi-brand” level.
There are more of these electronic diagnostic tools becoming available all the time.
The linear TPS introduced with the 800 and 1000 2V models and 749 and 999 Testastretta models of 2003 works differently to the earlier non linear style TPS. As the name says, the output of the TPS is linear. The “non linear” name of the earlier TPS implies a curve of sorts, but in reality it’s a two stage linear output. The graph below shows this. The dark blue line is “Non Linear”, red is “Linear”.
The main difference in use is that the baseline setting procedure is totally different. The old method of “150mV at fully closed” no longer applies. The new setting procedure is electronic and activated though the diagnostic tools.
What happens when you initiate the procedure through the diagnostic tools (Ducati Mathesis or DDS, Technoresearch VDSTS, Ultimap UMD diagnostics, Axone, etc) is that the ECU looks at the voltage coming out of the TPS at that point (throttle twist grip released) and stores that voltage as being equal to the nominated idle setting as stored in the ECU software file.
This nominated idle setting is specified as degrees of throttle opening. For example, the 749 is 1.3 degrees, 999 is 2.3 degrees, 800 is 2.7 degrees, open loop 1000 is 3.0 degrees, and closed loop 1000 is 3.4 degrees. This figure is not adjustable except at ECU software level, which pretty much means not at all with the 5.9M and 5AM ECU.
The assumption here is that the throttle bodies will be set correctly with the idle stop holding the throttle blade at the specified angle. This is not always the case. Nor is the relationship between output voltage and throttle angle based on any “zero point” figure that you can check.
The official line is that you never touch the idle stop screws. If you do need to replace a TPS or an ECU you carry out the TPS resetting procedure via the electronic diagnostic tools once the new TPS is fitted and away you go. We also reset the TPS as part of the first service procedure and then every 20,000km. This is more to ensure the setting is actually checked and correct than anything else.
If there was a problem from new it was usually apparent we had one 800M years ago that obviously hadn’t been done at the factory, and it ran quite poorly at low RPM. Checking the idle throttle opening showed 0 degrees at idle from memory, so we ran the reset procedure and away it went happily.
One issue that can come up is that if you adjust the idle stops for whatever reason, there’s no official baseline procedure to set them back to where they were like there was for the older Non Linear TPS. And if you do an electronic reset procedure after adjusting the idle stop, there’s then no way to know how much you’ve varied them.
At least on the Ducati models the TPS is on the same throttle body as the idle stop. The Moto Guzzi system uses mostly the same hardware, but the idle stop is on one throttle body, with the TPS on the other and connected to the idle stop through the balance adjustment mechanism. So if you adjust the balance, you have to reset the TPS. I really don’t understand why they do it that way.
Anyway, back to the story and on to 749 and 999. When we first got Linear TPS bikes I debated whether or not we should be trying some sort of baseline procedure, but due to the inevitable time constraints of servicing we didn’t bother. There didn’t appear to be any issues with most models anyway they all seemed to run quite well. Then with the first of the 999S we delivered we had an issue that I thought may have been TPS related, and some info from Jason Etter at Section8Superbike (now at Department of Suspension) about what they were doing provoked me into it.
What Section8 was doing was carrying out baseline TPS set ups on the 749 and 999 models by backing all the stops off so that the throttles were fully closed, carrying out a TPS reset via the Mathesis, winding the idle stop in until the throttle opening displayed through the diagnostic data was double what the reset procedure had given, then carrying out another TPS reset via the Mathesis.
Which basically meant, on a 999 for example, you backed everything off, reset the TPS (which gave 2.3 degrees via the diagnostic data), wound the idle stop in until you had 4.6 degrees, then the second reset gave the true 2.3 degree opening.
The issue we had was with the 999S models, being a flat spot just as you tried to take off. Combined with the sintered clutch plates they were still running in them at that point, they were nigh impossible to get off the line cleanly, especially in the wet. So I got one of them in to have a bit of a play. At this point I should also add that getting the top off the 999 air box is a real exercise in frustration, so Section8 were drilling holes in them to access the idle stop screw. It’s what we do too.
With this particular problem bike I performed a TPS reset without touching the bike (we’d done it already, but just to be sure) then wound the idle stop out. At fully shut, the diagnostic data was still showing 0.5 or 0.6 degrees of throttle opening. I returned the idle stop to where it was before and removed the throttle bodies. I then hooked then up to the Mathesis, which has a function where you can bench test the TPS the Mathesis provides the 5V power source and reads the output voltage. As delivered, the TPS was putting out (from memory) 430mV on the idle stop. With the idle stop wound all the way out this dropped to around 365mV, meaning the opening was 65mV. I hooked the throttle bodies back into the bike loom and did a TPS with the idle stop all out and then wound in another 2.3 degrees of throttle opening. Reconnected to the Mathesis TPS function again now showed 465mV.
Which I took to mean that 2.3 degrees of throttle opening was 100mV from fully shut. We had another set of throttle bodies kicking around from a 749, so I did the same thing to them. The 749 throttle dropped 65 or so mV from as delivered to fully shut, and doing the “double the reset degrees” thing took me pretty much back to where I had started. So it seemed 1.3 degrees throttle opening corresponded to 65mV from fully shut, and 2.3 degrees corresponded to 100mV from fully shut.
I emailed Jason with this info, and he said they were finding it the other way around the USA spec 999 seemed to be right, but the 749 were wrong. And that their issues were with the 749 models. At that point I checked the parts catalogue for variations there were only two throttle bodies listed for 749 and 999 models being 749 and 999. There was no Euro / USA distinction like there often is. So the difference could only be the idle stop “degrees of throttle opening” setting, and somehow that 749 / 999 variation must have been transposed into Euro / USA. Don’t ask me why, I have no idea.
You will note here I have not worked off the voltage coming out of the TPS that’s because it’s irrelevant. The 749 and 999 TPS (like many later ones) is part of the throttle bodies it can’t be moved and as such the voltage is what it is. It’s the change of voltage that is important here, not the actual voltage value.
So I set up the throttle bodies of this 999S with the idle stop giving 100mV of throttle opening from fully closed. Then I refitted them to the bike, reassembled it all and did a TPS reset so the ECU was reading 2.3 degrees throttle opening on the idle stop. I then finished the set up by resetting the idle speed, running balance, idle mixture and went for a ride. Result problem solved. It was as simple and clear cut as that, and my arse was much relieved.
Initially I wrote a procedure from this result that had us reaching into the air box through the RH air filter opening and hooking directly into the TPS with the Mathesis. But the resulting contortion and damage to hands and arms and general agro (our man Danno was not very impressed with my procedure at all) led to us doing what Jason had described using the TPS reset function to show the required throttle opening.
The procedure we adopted is laid out below. As a general rule, the value of doing this on our Euro spec 749 and 749S models was debateable, but on 999 models, especially all 999S and ’05 onward 999 (with the S cams) it often made a noticeable difference to how easy they were to get moving from a standing start.
749 / 999 TPS setting procedure
I think there’d be a few dealers in Australia familiar with this, as I passed it on to the tech support man at NFI after it’d worked for us, but it’s still not an official procedure of any sort.
When I first played with the 749 and 999 throttle bodies I graphed the results on graph paper to see what it looked like, as I assumed there had to be a mathematical relationship to it. Below is a scan of the page the graph lines have faded, but the important bits are there.
You can see my gradient calculations (the angle of the line, or rise over run) that give a figure of approximately 36mV per angle of throttle opening. The 18mV offset at 0 degrees I find a bit odd, and don’t really understand, but it appeared to be consistant. What this means is that, if you need to baseline your Linear TPS throttle bodies after you or someone else has played with them, you can use the following equation to calculate how many mV you need to open the throttle for the correct idle stop setting:
Opening (mV) = [Req’d throttle opening (degrees) X 36] + 18
For example, a 749 would be [1.3 X 36] + 18 = 64.8 and a 999 would be [2.3 X 36} + 18 = 100.8
You do need to know the required throttle opening here, and the best way to find that is to perform a TPS reset using the electronic diagnostic tool and see what it says. You tend to get variations of +/- 0.1 degree in the number displayed, but if you’ve got a problem due to an incorrect TPS output versus actual throttle opening relationship then that variation will not make any difference.
There is one thing that I can’t explain though, and that’s the variation in gradient of the above graph and the initial graph shown at the top of this report. On that initial graph, the gradient is more like 46.7mV per degree of throttle opening. I’m choosing to ignore that though, on the basis that 46.7 doesn’t fit with what I require. Actually, the data I read out of the Mathesis and used for the initial graph tends to give 40mV per degree up to 12 degrees of throttle opening, and then 46.7mV per degree from 15 degrees of throttle opening up.
Moving on to the 2V models. If you look at the above plotted graph, you can see “1000 SC” toward the right hand end of the horizontal scale. This is because I had an issue with a 1000 Sport that I simply couldn’t explain so I thought I’d do a TPS baseline check.
The issue with this bike (not sold by us) was that, at around 800km from new, the automated fast idle function as regulated by the idle control valve stopped working. This made the bike fairly hard to start and it wouldn’t idle when cold. At around 1,000km it has its first service completed by someone else (not the selling dealer) which made no difference. It came to us with around 1,700km on the clock to see if we could fix the issue.
I did a few tests and everything checked out ok, including the idle control valve. So I thought I’d check the TPS baseline to see where it was. I checked the mV setting of the TPS as delivered, so I knew where to go back to if I needed to, then did the reset like the 749/999 procedure above. The problem was the change in mV coming out of the TPS wasn’t as much as it should have been for 3.4 degrees of throttle opening. I carried out the procedure again with the same result, so was convinced the error wasn’t me. I’m not sure why the procedure didn’t work, as I can see no reason why it shouldn’t have, but that’s what happened.
We had another Sport 1000 in bits for a crash repair at the time, so I checked it too. On it the mV change from as delivered to fully closed was much more like what I expected based on my plotted graph. So I then redid the problem bike with the idle stop set 140mV above fully closed. Performed the TPS reset procedure using the electronic diagnostic tool and then put it back together. We had to wait until the next morning to test the cold start, when it appeared to be back to normal. The customer confirmed this when he’d taken the bike back.
I’m not sure why this problem showed up after 800km of being ok, or what the problem actually was. If the throttle opening the ECU is reading is over a certain number of degrees the idle control logic appears to not operate the valve, but on this bike it was the expected 3.4 degrees. For example we had an S4Rs we fitted a DP full system to that wouldn’t start after we’d fitted the ECU prior to the TPS reset procedure using the electronic diagnostic tool being performed. When we hooked up the Mathesis to do the procedure we first checked the throttle opening at idle and it was around 10 degrees typically cruise at 100km/h or so. Carrying out the reset procedure took it back to the correct 2.3 or 2.4 degrees (I forget now) and on the next start the fast idle function worked correctly.
Which brings us to the next point fitting a new genuine or Ducati Performance ECU
When you fit a new standard or Ducati Performance ECU the ECU will look at the voltage coming out of the TPS with the throttle closed and assign it a value of degrees of throttle opening. How it does this I really don’t know, as we have seen anything from zero to 10 degrees on the diagnostic data before the TPS reset procedure using the electronic diagnostic tool is carried out. I have also seen the number appear to be correct, but then the value doesn’t change until the throttle has been opened a few degrees.
So the first thing to do when you fit a new standard or Ducati Performance ECU is to carry out a TPS reset procedure using the electronic diagnostic tool. If the bike does start and run ok, as many people say their bikes do, it’s luck and luck alone. There is no reason why the setting should be correct.
The next thing to do on an open loop ECU (all DP ECU are open loop) is to set the idle mixture. If the idle mixture is right out of the box then you’re doubly lucky.