Throttle pot action on MoTeC RV8

[Number 7]

Was checking the action of the throttle pot for any signs of unevenness in operation, as I’ve noticed that the engine response at small throttle openings can be a little jerky. Visually, at the point at which the throttle spindle/butterfly is at half its total travel, the indication from the TP to the laptop via the ECU is 75% throttle, i.e. significantly over reading the actual throttle position. The TP registers correctly at zero and WOT. It’s also noticeable that, for the first 2/3 or so of throttle travel, the TP scale is such that even small throttle movements result in a relatively large increase in the % seen by the TP, whereas towards the end of the throttle travel, the same small movements in throttle travel result in small increases in the % seen by the TP. The TP does not seem to produce a linear signal. Is this normal behaviour for TP’s? If not, I will need to replace it, but won’t a TP with a different scaling (more linear) mean that the part-throttle mapping will need adjusting?
TIA

7.

[stevieturbo]

Number 7 said:

Was checking the action of the throttle pot for any signs of unevenness in operation, as I’ve noticed that the engine response at small throttle openings can be a little jerky. Visually, at the point at which the throttle spindle/butterfly is at half its total travel, the indication from the TP to the laptop via the ECU is 75% throttle, i.e. significantly over reading the actual throttle position. The TP registers correctly at zero and WOT. It’s also noticeable that, for the first 2/3 or so of throttle travel, the TP scale is such that even small throttle movements result in a relatively large increase in the % seen by the TP, whereas towards the end of the throttle travel, the same small movements in throttle travel result in small increases in the % seen by the TP. The TP does not seem to produce a linear signal. Is this normal behaviour for TP’s? If not, I will need to replace it, but won’t a TP with a different scaling (more linear) mean that the part-throttle mapping will need adjusting?
TIA

7.

1. If it is reading 0-75% then whoever set the car up, hasnt done so properly.

2. How the actual scale appears under foot, as to being linear or not, is totally irrelevent.
If indeed the car is mapped TPS vs RPM, then all that matters is how the load sites are actually set out, and how the car is mapped.
Back to 1...if the load sites reflect a min of 0%, and a max of 75%TPS, then the scaling doesnt really matter, although naturally its nice for it to reflect a 0-100% scale.

In short...what you appear to be seeing doesnt matter IMO.

[Number 7]

It's definitely reading to 100% throttle, it's just the non-linear operation that's concerning me - does it indicate wear? Having relatively large jumps in TP% vs. actual butterfly movement, partcularly at small openings, makes a smooth progression more difficult. Maybe I just have to mash the pedal

7.

[stevieturbo]

How are the load sites set out ?

Is there an actual problem as to how the car is driving ?

Ive only ever set up one car with TPS as main load....and TBH, I never even looked at whether it was linear or not.

Every other car has been MAP as main load, although TPS does play a part with accel enrichment.

Unless there is a specific problem, I wouldnt be worried about it. Airflow through thte TB certainly wont be linear with opening.

[CrashTD]

Number 7 said:

Having relatively large jumps in TP% vs. actual butterfly movement, partcularly at small openings, makes a smooth progression more difficult.

Throttle Pedal Movement Vs Throttle Position wont effect the car/ECU. The ECU thinks in Throttle Position not pedal travel.

If you say there is a difference between Throttle Position and what the ECU picks up then I suspect there could be play between the spindle and the TPS. I had this when setting up my ECU where the first dozen or so degrees wouldn't be picked up. As a result there would be no reading on the TPS then straight up to 20%, giving this non-linear correlation you describe.

As stevieturbo says - a linear relationship is not essential providing the Throttle Position is correct to the Load-Site on the map.

I don't know how your car operates the throttle from the cabin but some general advice from my own car which runs a cable- The TPS converts the physical movement of the Butterfly into an electronic signal. As a human you feel a relationship between pedal movement and power.

The closer you can get these two the better, one of the main reasons for Fly-by-wire. I changed a throttle cable, all the rubbers etc from my pedal to my TB and it made a world of difference. The slightest breath on the pedal used to open the butterfly.

Didn't make the car go any faster but made it alot more drivable around town especially with a paddle clutch

Edited by CrashTD on Sunday 25th January 20:30

[trackcar]

What you are describing is entirely typical and indeed desirable characteristic.
The greatest rate of change of airflow per degree of butterfly movement is off idle, then the rate of change per degree gets progressively less the more the throttle is opened, this means that to get good mapping at low throttle sites where the airflow increases rapidly as the throttle is opened you need to stack the throttle sites closer together, and gradually space them further apart towards fully open throttle.

[Ultra Violent]

Have you set the input curve for the TPS sensor correctly. If it is not linear then you will obviously need to define the TPS input voltage as none linear.

[Pumaracing]

trackcar said:

What you are describing is entirely typical and indeed desirable characteristic.
The greatest rate of change of airflow per degree of butterfly movement is off idle, then the rate of change per degree gets progressively less the more the throttle is opened, this means that to get good mapping at low throttle sites where the airflow increases rapidly as the throttle is opened you need to stack the throttle sites closer together, and gradually space them further apart towards fully open throttle.

All well and good except neither the TP or the ECU has a clue what the airflow is at each throttle position or even that the TP is connected to a butterfly with non linear flow characteristics rather than just a spindle on a CNC machine. The TP itself should produce a perfectly linear signal with respect to motion. If it didn't it would be almost impossible to control anything properly with it. That can be checked with a voltmeter.

Assuming the TP isn't faulty the only reason for one showing a greater % movement in the software than at the spindle must lie in the software. Check to see if there's a menu option for TP position encoding. If so it may have been set to exponential rather than linear. If you change anything the mapping will be lost of course.

[trackcar]

What he is reading is via the mapping software, not the raw signal off the throttle pot from what i read of the OP.

[GreenV8S]

Pumaracing said:

The TP itself should produce a perfectly linear signal with respect to motion. If it didn't it would be almost impossible to control anything properly with it. That can be checked with a voltmeter.

The only throttle pots I've got any experience have been linear. But it is common in other applications where a potentiometer is used as a position sensor, to use a logarithmic potentiometer rather than a linear one in applications where more sensitivity is needed at one end of the scale. Give that the throttle response of the engine is highly non linear, I would have thought it makes sense to use a non-linear pot so that the pot is most sensitive in the region where the engine is most sensitive. I don't see why that would cause any difficulty to the ECU as long as the behaviour of the pot is consistent.

[Pumaracing]

GreenV8S said:

Pumaracing said:

The TP itself should produce a perfectly linear signal with respect to motion. If it didn't it would be almost impossible to control anything properly with it. That can be checked with a voltmeter.

The only throttle pots I've got any experience have been linear. But it is common in other applications where a potentiometer is used as a position sensor, to use a logarithmic potentiometer rather than a linear one in applications where more sensitivity is needed at one end of the scale. Give that the throttle response of the engine is highly non linear, I would have thought it makes sense to use a non-linear pot so that the pot is most sensitive in the region where the engine is most sensitive. I don't see why that would cause any difficulty to the ECU as long as the behaviour of the pot is consistent.

Log pots are more expensive than linear pots and there's no way a randomly chosen log scale could properly approximate a specific use such as a butterfly flow potential. This is stuff you do in software not hardware which is why car pots are linear and mapping software takes care of the rest. That isn't even taking into account the linearity of the throttle mechanism which in many cars (most cars) has some non linearity built in so that a given pedal travel leads to less butterfly movement at the start of the travel and more at the end of it. That's how you achieve sensitivity of throttle opening in most cases and trying to match that with the TP linearity would be pointless.

As someone who has designed and sold Throttle Body systems and the throttle pots to go with them I've never seen a non linear car TP and can see no reason why anyone would use one.

[GreenV8S]

Pumaracing said:

As someone who has designed and sold Throttle Body systems and the throttle pots to go with them I've never seen a non linear car TP and can see no reason why anyone would use one.

Most likely they don't exist then. I've certainly never seen one. But the extra sensitivity could be useful sometimes. For example, I know that somebody with oversize throttle bodies using TPS for load, who simply can't map the low load / low rpm region of the map because the throttle pot is not sensitive enough. Opening the throttle far enough to pick the revs up by 500 rpm at idle only changes the TPS A/D output by 1. That's not enough to give any meaningful load signal and as a result the fuelling is all over the place in that region of the map. There's nothing you can do about this in the ECU if the sensor you're working with simply isn't sensitive enough. (My suggestion was to go to MAP for load, and maybe that's what they'll end up doing.)