51 replies to this topic

[Wuzak]

Fly by Wire Throttle vs Mechanical Linkages - what is more responsive to driver inputs.

 

What sort of response time would a driver notice?

 

Regarding the current turbo hybrids, would the time to convert the input from the driver to power output at the wheels with a blend of ICE and MGUK be noticeable?

[gruntguru]

Great question. Typically FBW will be slower - there will always be some lag between movement of the pedal and movement of the throttles. The current F1 PU's are not typical however. There is scope for the response to be faster, even faster than a mechanical linkage.

 

 - Current F1 cars run very lean (lambda >> 1.0) under all conditions, even WOT. If the trailing throttle mixture through a corner was say lambda 1.4 (ER = 0.71), the power could be increased almost instantly by approximately 30%, by increasing the mixture strength to the "best power" value (approx lambda 0.9, ER =1.11 ) but of course only up to the 100 kg/hr limit. Within a very short time the throttles will have opened and the air flow will increase to restore the "best efficiency" mixture for normal operation.

 

 - FBW can be tuned to "anticipate" movement of the pedal by introducing a "derivative" component into the control loop. For example if the pedal is advancing very rapidly, the control loop will anticipate that the pedal is heading towards a much higher value than the current position and advance both the throttles and the injector duration beyond the usual static value for the current pedal position. This anticipation can be used to some extent to "tune out" delays in the throttle actuator, the intake airflow and the engine itself.

[Canuck]

Interesting.  I've had a reasonable amount (>10,000 km) of experience in 3 different FBW vehicles from two different brands and find them all slightly different, equally frustrating and continue to miss the responsiveness of my old BMW. The Lexus hybrid is entertainingly quick off the line but still stops to think about launching between the pedal-mashing and the actual go. I boosted someone on the weekend who'd left their interior light on for two days (and nights) and had a very dead battery at -23C, using my Expedition. Trying to hold the engine speed at 2000 or 2500 RPM (to get the alternator output up) was a farce. It was either racing towards 3500 or dropping to idle with the tiniest of tiny pedal movements. Which is the inverse of it's action on the road where you try to give it a little nudge and it ignores the input until you go from a nudge to a boot - I have a very hard time staying off the brakes in rush-hour traffic as it either doesn't move off idle or surges forward (where the old BMW was a piece of cake to hold at almost any speed).

 

I'm not certain what the benefits are (for a civilian vehicle), beyond perhaps traction control - which drives me nuts 95% of the time and is welcome the rest.

[gruntguru]

Very annoying when the OEM gets the no-load mapping wrong. Same applies for slow-speed, light-load operation in traffic jams etc.

 

I am sure Greg can chip in with more, but some of the benefits of FBW throttle are:

 - idle speed control

 - stability/traction control

 - cruise control

 - 3D mapping of pedal/throttle/rpm

[Greg Locock]

Throttle progression has caused many a bar fight. In the 1990s Mercedes were wedded to a very progressive pedal, you really had to mash it to get full throttle, whereas other companies preferred a very rapid response to tip-in, so the last half of the pedal travel didn't do much except change the shift points of the auto. Even with a direct mechanical linkage the plot of MAP vs  pedal vs rpm was complicated, because of course at 50% redline rpm you only need the throttle butterfly to be half way open to get 0 MAP. Designing the snail cam to give the right throttle feel was quite an art.

[mariner]

I think that the ability of FBW to anticipate driver intention to improve response , per Gruntguru, is only possible with traction control. FBW helps enable that too so it ought to allow better overall response if programmed correctly, or with performance in mind.

 

IIRC one trick of the old Jag XJR saloon with a big supercharged V-8 was that when you pressed " resume" on the cruise control it used a very big throttle opening to return to set speed fast - quite cute.

 

 

The real oldies here might remember vacuum secondaries such as Pontiac GTO tri-power. As you opened the throttle, which was attached mechanicaly to only the centre carb, vacuum depression sensing triggered the other two carbs wide open rather suddenly  - not good if you bought one to drive round wet English roundabouts.

Edited by mariner, 12 January 2016 - 09:41.

[Wuzak]

Great question. Typically FBW will be slower - there will always be some lag between movement of the pedal and movement of the throttles. 

 

How much time?

 

The question is not about the throttles, but the response of the engine to input.

 

I would think that having FBW enables parameters such as ignition and injection can be modified at the same time, or even ahead of, the throttles moving.

 

Also, with direct injection, variable length intakes and an electrically controlled supercharger/turbocharger might it actually be possible to do without the throttle?

[Wuzak]

Interesting.  I've had a reasonable amount (>10,000 km) of experience in 3 different FBW vehicles from two different brands and find them all slightly different, equally frustrating and continue to miss the responsiveness of my old BMW. 

 

I think in road cars the throttle mapping has to be on the conservative side, as most customers are not Lewis Hamilton.

 

And is the responsiveness down to the FBW system entirely or, at least in part, to the engine itself?

[MatsNorway]

In computer gaming (yes that is a million dollar industry now) Mouse acceleration is 95% of the time turned off (genre dependant), as it does not give a predictable acceleration as it too depends on how fast you move your hand. Is this solved in FBW?

 

Main issue is that you need X amount of movement to get the desired result. And that movement needs to be muscle memory.

 

Variable acceleration would be like having a variable steering output depending on how fast you rotate the steering wheel.

 

Reality: everyone wants X amount of steering angle, Regardless of how fast they do it.

 

But.. i guess you could use FBW to compensate for engine inefficiencies of mechanical nature.

[Greg Locock]

Well, if you borrow your mates car and floor the throttle in neutral, how long does it take to red line? It isn't ten seconds, and it isn't 0.1 seconds.

[Fat Boy]

Throttle progression has caused many a bar fight. In the 1990s Mercedes were wedded to a very progressive pedal, you really had to mash it to get full throttle, whereas other companies preferred a very rapid response to tip-in, so the last half of the pedal travel didn't do much except change the shift points of the auto. Even with a direct mechanical linkage the plot of MAP vs  pedal vs rpm was complicated, because of course at 50% redline rpm you only need the throttle butterfly to be half way open to get 0 MAP. Designing the snail cam to give the right throttle feel was quite an art.

 

You know what my answer is going to be, don't you? It's a matter of horses for courses.

 

On a 'street course' car with plenty of HP where wheelspin is an issue, I prefer to have a soft tip-in. It just helps the driver put the power to the ground in a more controlled manner. If you're on an oval that is 'flat' or 'almost flat' (i.e. few throttle lifts), then I prefer to have a strong tip-in (helps prevent pit stalls) and any lifts have only a minor effect on power. A road course is somewhere in the middle and more linear.

[Lee Nicolle]

FBW is a pain if for no other reason you cannot rev the engine from under the bonnet. Did two last week and just bringing up RPM to circulate water or check charge rate is a pain.

[gruntguru]

Mats. I wouldn't expect the "mouse acceleration" component of F1 FBW to do more than compensate for some of the lags in the system. The ideal system would be one where torque at the wheels correlates exactly (and instantly) to each pedal position.

 

Wuz'. I think the output of an F1 engine would begin to change within one crank rotation (approx. 6 ms). The next injection pulse is never more than 720*/6 = 120* away, compression TDC perhaps 90* later and the increased firing pressure will become evident 90* after TDC (or less) for a total delay of 300*. Of course it will take longer for torque to fully catch up with demand if increased airflow is needed.

 

Forgot to mention in my original post. To improve response, it would be possible to briefly increase fuel flow beyond the 100 kg/hr limit. This would depend on the time constant (averaging) of the fuel flow measurement system but should allow at least a few cycles of significant overfuelling without breaking the rules.

[Kelpiecross]

On the subject of MAP and engine load control versus throttle (or accelerator) position - for example you can get full MAP/maximum cylinder filling  (and thus full engine load)  at (say) 2000rpm with the throttle position maybe only one quarter open etc.  However if you were controlling the engine load (amount of cylinder filling) by varying the amount of late closing of the inlet valve - the amount of  cylinder filling is directly proportional to the amount of LIVC used the MAP remaining at a maximum.  Thus with the engine example above - one quarter accelerator depression would mean 1/4 load - not full load. 

 Just how the response to the accelerator position would feel in a car with LIVC engine load control is a bit hard to judge - you would probably get used to it - or maybe the car maker would arrange the amount of LIVC (by FBW possibly) to give the same "feel" as a conventional throttle plate.    

Edited by Kelpiecross, 13 January 2016 - 11:51.

[MatsNorway]

The ideal system would be one where torque at the wheels correlates exactly (and instantly) to each pedal position.

 

Woa woa. Hold it. Could one go really crazy and use the engine to take out oscillation in the drive system? (flex in driveshafts etc??

 

Number of benefits if thats possible. More controll over the launch, better throttle feel at not full throttle etc.

Edited by MatsNorway, 13 January 2016 - 19:13.

[Fat Boy]

Woa woa. Hold it. Could one go really crazy and use the engine to take out oscillation in the drive system? (flex in driveshafts etc??

 

Number of benefits if thats possible. More controll over the launch, better throttle feel at not full throttle etc.

 

Did you enjoy your time working on that Nissan LMP1 car?

[gruntguru]

 

Woa woa. Hold it. Could one go really crazy and use the engine to take out oscillation in the drive system? (flex in driveshafts etc??

I suppose so, for oscillations of significantly lower frequency than the engine firing.

[kikiturbo2]

Very annoying when the OEM gets the no-load mapping wrong. Same applies for slow-speed, light-load operation in traffic jams etc.

 

I am sure Greg can chip in with more, but some of the benefits of FBW throttle are:

 - idle speed control

 - stability/traction control

 - cruise control

 - 3D mapping of pedal/throttle/rpm

 

also, on turbocharged cars you can cut the lag after throttle release by keeping throttle open and minimising the engine output by other means.. Also you can kick the throttle for rev matching etc...

Having said that, I hate FBW throttle, and see no point in linear throttle angle vs TQ relation, but that is just me..

[MatsNorway]

Did you enjoy your time working on that Nissan LMP1 car?

What?

[Fat Boy]

Woa woa. Hold it. Could one go really crazy and use the engine to take out oscillation in the drive system? (flex in driveshafts etc??

 

 

 

What?

 

Humor, and a call-back to a different thread. This is the type of utter bullshit they spent entirely too much time on.

[MatsNorway]

I see. I also take it as a jab at them not focusing entirely on the right thing. For instance perhaps they shoulda started with a more traditional car... to.. idk.. gather some info on KERS etc.

Edited by MatsNorway, 14 January 2016 - 19:57.

[Lee Nicolle]

You know what my answer is going to be, don't you? It's a matter of horses for courses.

 

On a 'street course' car with plenty of HP where wheelspin is an issue, I prefer to have a soft tip-in. It just helps the driver put the power to the ground in a more controlled manner. If you're on an oval that is 'flat' or 'almost flat' (i.e. few throttle lifts), then I prefer to have a strong tip-in (helps prevent pit stalls) and any lifts have only a minor effect on power. A road course is somewhere in the middle and more linear.

Over the decades I have used throttle travel as a driver aid. Too the point of having a couple of positions on the linkages. To speed or slow the action.

How you would achieve that with an electronic throttle though is a mystery to me. Can it be done?

 

I do know that in some forms of motorsport that FBW is illegal. Though in the past so were cables.

[Fat Boy]

I see. I also take it as a jab at them not focusing entirely on the right thing. For instance perhaps they shoulda started with a more traditional car... to.. idk.. gather some info on KERS etc.

 

As usual, you're off-line, off-pace and out of touch.

 

Mostly the jab was at you, but the comparison was because _clearly_ they had problems. SMH...You're even annoying to ridicule.

[Fat Boy]

Over the decades I have used throttle travel as a driver aid. Too the point of having a couple of positions on the linkages. To speed or slow the action.

How you would achieve that with an electronic throttle though is a mystery to me. Can it be done?

 

I do know that in some forms of motorsport that FBW is illegal. Though in the past so were cables.

 

Pedal travel has some relation to throttle blade angle. It really doesn't matter what connects the two, you can manipulate them with the same basic approach. I think the biggest problem with FBW is a perceptible of a time lag between the two. In most situations, I think that's been eliminated or, at the very least, minimized.

[MatsNorway]

Why you start being a dick for no reason. Problem with your personal life so you try to take it out on someone else?  How sad is it that you remember better than me what i have said in some other thread. You should get a temp ban for being poisonous.

Edited by MatsNorway, 14 January 2016 - 23:31.

[Lee Nicolle]

Pedal travel has some relation to throttle blade angle. It really doesn't matter what connects the two, you can manipulate them with the same basic approach. I think the biggest problem with FBW is a perceptible of a time lag between the two. In most situations, I think that's been eliminated or, at the very least, minimized.

I have never noticed any lag. Though that on std road cars.

Compared with turbos I doubt it is a real issue.

[Greg Locock]

I hadn't come across the lag criticism of FBW before, is that a common one? Any particular makes? There's no inherent reason why it should be so, although of course the calibrators may deliberately damp or slow the butterfly motion for improved emissions, in particular. I'm just wondering if there is an issue with lean mixture on tip in, which of course back in the day was 'solved' by using an accelerator pump.  I know that 90s era Falcons used to rap (det) for 3 cycles if you tipped in aggressively, that was quite deliberate, and perfectly safe, as a fuel saving measure. "The sound of economy" as the calibrators called it. it was safe because one of the dyno tests was to run the engine in detonation for hours at a time.

[Fat Boy]

This sounds made up...

 

I have not experienced many FBW systems, but one was on a Ferrari FXX. It was awful. I don't know if that's common or not, but it was not good. There might have been any or all sorts of partial throttle engine management problems. Full throttle was sorted....that's for sure.

[Fat Boy]

Why you start being a dick for no reason. Problem with your personal life so you try to take it out on someone else?  How sad is it that you remember better than me what i have said in some other thread. You should get a temp ban for being poisonous.

 

I'm being a dick because you continually come up with really dumb stuff. Say fewer dumb things and I'll seem like much less of a dick. Having said that, I'm sorry I hurt your feelings.

 

The other thread I referenced was the Nissan LMP1 thread, and it had nothing to do with you. Please don't take this as an invitation to go ruin it.

[Wuzak]

I hadn't come across the lag criticism of FBW before, is that a common one? Any particular makes? There's no inherent reason why it should be so, although of course the calibrators may deliberately damp or slow the butterfly motion for improved emissions, in particular. I'm just wondering if there is an issue with lean mixture on tip in, which of course back in the day was 'solved' by using an accelerator pump.  I know that 90s era Falcons used to rap (det) for 3 cycles if you tipped in aggressively, that was quite deliberate, and perfectly safe, as a fuel saving measure. "The sound of economy" as the calibrators called it. it was safe because one of the dyno tests was to run the engine in detonation for hours at a time.

 

My question is the "perceptible lag" a function of the FW system or the engine itself?

[Wuzak]

Over the decades I have used throttle travel as a driver aid. Too the point of having a couple of positions on the linkages. To speed or slow the action.

How you would achieve that with an electronic throttle though is a mystery to me. Can it be done?

 

I'm guessing the way the spring and damper system on the electronic throttle could be set up to give similar action.

 

The FBW system itself can be set up to give all sorts of effects with relation to pedal travel.

[MatsNorway]

Blabla im a grumby ****er

Just shut up in my direction. I get good answers from others. Go marvel at sylinder walls made of something else than steel. Thats a simple question too and you went apeshit there as well.

Edited by MatsNorway, 15 January 2016 - 01:25.

[Canuck]

I was thinking about this after making my comments above. I've only driven the two brands - Ford and Lexus - and found them both to act not as I expect, which is a remarkably vague description. Perhaps I expect to much giddy-up out of the Expedition - the thing is shockingly fat after all and asking it to move like it's svelte is perhaps unfair. I have a handful of sensors I should be able to rig together to generate a plot of pedal position, throttle blade position and acceleration. Maybe. It would be interesting to see if the lag between pedal input and throttle blade position is as marked as I think, or all in my head.

[Fat Boy]

Just shut up in my direction. I get good answers from others. Go marvel at sylinder walls made of something else than steel. Thats a simple question too and you went apeshit there as well.

 

Ugh, you even suck at trolling...

 

Do you want a Mulligan? Fine by me. Take some time and try again.

[Kelpiecross]

What's a "Mulligan"? 

[Lee Nicolle]

I'm guessing the way the spring and damper system on the electronic throttle could be set up to give similar action.

 

The FBW system itself can be set up to give all sorts of effects with relation to pedal travel.

It is a still a mechanical thing. Purely a solenoid driving the throttle body.

[Lee Nicolle]

I hadn't come across the lag criticism of FBW before, is that a common one? Any particular makes? There's no inherent reason why it should be so, although of course the calibrators may deliberately damp or slow the butterfly motion for improved emissions, in particular. I'm just wondering if there is an issue with lean mixture on tip in, which of course back in the day was 'solved' by using an accelerator pump.  I know that 90s era Falcons used to rap (det) for 3 cycles if you tipped in aggressively, that was quite deliberate, and perfectly safe, as a fuel saving measure. "The sound of economy" as the calibrators called it. it was safe because one of the dyno tests was to run the engine in detonation for hours at a time.

Greg, 90s Falcons use cables. Yes they use electronics that can at times be a pain in the arse.  Like all manufacturers since the mid 80s.

[kikiturbo2]

I hadn't come across the lag criticism of FBW before, is that a common one? Any particular makes? There's no inherent reason why it should be so, although of course the calibrators may deliberately damp or slow the butterfly motion for improved emissions, in particular. I'm just wondering if there is an issue with lean mixture on tip in, which of course back in the day was 'solved' by using an accelerator pump.  I know that 90s era Falcons used to rap (det) for 3 cycles if you tipped in aggressively, that was quite deliberate, and perfectly safe, as a fuel saving measure. "The sound of economy" as the calibrators called it. it was safe because one of the dyno tests was to run the engine in detonation for hours at a time.

 

I have driven a lot of FBW throttles and there is no real lag.. the problem is in the throttle mapping. This is especially the case where companies introduce "sport" buttons which, among other things, alter throttle mapping, making it "faster" to get to OT which gives the driver the feeling of the engine having more power. Quite often this introduces unwanted problems with throttle controll, not only because you have to adapt to diffetent throttle strategies, but also because the really quick throttle opening tends to give you traction problems in the wet. BMW 2M being the worst of the lot... really bad.

 

Personally I dont like forced torque mapping on the throttle, esp on turbo cars.. I quite like the traditional turbo response where I give it a fixed 25% throttle and then ride a wave of rising torque as the turbo spools up... lazy, I know.

[Greg Locock]

I know the 90s Falcons were cable throttle, one of the many bits on EF I developed the throttle cable abutment fitting, a collapsible rubber plug. The idea was to reduce the amount of vibration in the throttle pedal and it worked perfectly - a particular bee in the bonnet of a chief engineer. They later cost reduced it out (in order to maintain throttle pedal feel it had a non linear spring characteristic, hence was quite complex). The replacement part was not as thoroughly tested as mine, saved them 30c per car, and caused a recall because it sometimes jammed the throttle wide open. Well did I laugh my head off?

[Lee Nicolle]

I know the 90s Falcons were cable throttle, one of the many bits on EF I developed the throttle cable abutment fitting, a collapsible rubber plug. The idea was to reduce the amount of vibration in the throttle pedal and it worked perfectly - a particular bee in the bonnet of a chief engineer. They later cost reduced it out (in order to maintain throttle pedal feel it had a non linear spring characteristic, hence was quite complex). The replacement part was not as thoroughly tested as mine, saved them 30c per car, and caused a recall because it sometimes jammed the throttle wide open. Well did I laugh my head off?

And those bloody cables still break the outer cables. Multi or single point!

Though I have adapted a complete EB throttle body cable and pedal to my XE. Works ok with the Holley. And yes I do have two  lever points on the carb. 

[Kelpiecross]

I can't quite picture this - 6 or V8?  Got a photo of the engine? 

[Greg Locock]

I6, i never worked on the V8.As I remember the cause was that the throttle body was on the manifold obviously and for some reason the bracket for the outer was on the rocker cover. The dimensional stack across all those interfaces was enormous, and the vibration of the manifold relative to the rocker cover was significant. I think we tried a brace between the two but tooling would have been excessive. Quite why we didn't move the outer to the throttle body is lost in the mists of history.

[Kelpiecross]

I understood LN's meaning to be that he used both an EB throttle body and a Holley carb.

[rachael]

I was thinking about this after making my comments above. I've only driven the two brands - Ford and Lexus - and found them both to act not as I expect, which is a remarkably vague description. Perhaps I expect to much giddy-up out of the Expedition - the thing is shockingly fat after all and asking it to move like it's svelte is perhaps unfair. I have a handful of sensors I should be able to rig together to generate a plot of pedal position, throttle blade position and acceleration. Maybe. It would be interesting to see if the lag between pedal input and throttle blade position is as marked as I think, or all in my head.

 

 

You don't need any sensors to do this, just a cable to connect to the OBD port and some logging software to compare throttle pedal position and throttle plate angle.

 

Rachael.

[Canuck]

Indeed - for the two FBW cars that would do, but the old BMW which serves as the comparative reference / baseline is pre-OBD (Motronic 1.0) and lacks even a throttle position pot in favour of a pair of switches to indicate idle or WOT.

I would guess that there's an accelerometer in the later cars, in the stability control hardware at the very least. Is that data available on the buss as well?

BTW - nice to see you again.

[Canuck]

I suspect my knock-off ELM Bluetooth adapter and a phone would provide a rudimentary log those three things.

Edited by Canuck, 17 January 2016 - 15:27.

[rachael]

Indeed - for the two FBW cars that would do, but the old BMW which serves as the comparative reference / baseline is pre-OBD (Motronic 1.0) and lacks even a throttle position pot in favour of a pair of switches to indicate idle or WOT.

I would guess that there's an accelerometer in the later cars, in the stability control hardware at the very least. Is that data available on the buss as well?

BTW - nice to see you again.

 

 

Accelerometer data will definitely be available.  I'm not sure it will be a fair test with the old BMW though as they cared less about emissions then so could squirt a load of extra fuel in which probably acts to make the car feel more lively.

 

Rachael.

[gruntguru]

Having said that, I . . . . see no point in linear throttle angle vs TQ relation, but that is just me..

I agree. However the relationship between pedal and torque is important and FBW enables this relationship to be uniquely defined for each rpm band.

[gruntguru]

I understood LN's meaning to be that he used both an EB throttle body and a Holley carb.

I'm glad somebody understood something there!

[Lee Nicolle]

I can't quite picture this - 6 or V8?  Got a photo of the engine? 

I am just using the EB throttle body  pedal and cable to work a Holley. The cable is the right length, better than the genuine XE Ford V8 cable actually.

No EFI was contemplated in this operation! Yet alone FBW . I am making far more power with a carburettor  for a few hundred than several thousands worth of efi. And the car is lighter too without miles of sensors,computers wiring harness.

And no. PLEASE do not tell me to use a factory efi.,, apart from the fact it would make less power there is no such thing for a Cleveland.

A proper efi using a 4bbl  style throttle body and 8 injectors in the manifold works well, and costs a LOT more too. Or better 8 throttle bodies which costs a lot more.

Edited by Lee Nicolle, 18 January 2016 - 11:37.