i used to "not like" DBW...then i started building and driving swapped cars every day. DBW is clean to wire, simple to setup, fully adjustable and adding a/c and cruise controls is just a few clicks. but the best part is being able to dampen throttle response at slow speeds like driving over speed bumps and dips in the road without the jerking of cable setups when your foot moves relative to the throttle pedal. if you're a DBW "hater" then you havent driven a proper-setup one yet. 👍
What you said about throttle cables sticking reminded me of seeing an unlimited hydroplane that had a cable the driver could pull to disconnect the cable from the throttle.
What are you fitting DBW to or tuning it in? 🔵 Building a fast car? Get $400 OFF the all-inclusive VIP online course package deal: hpcdmy.co/offery161 🦵Kickstart your EFI Tuning knowledge. Get 50% OFF your first online course: hpcdmy.co/offery161
Great video Andre and HPA team. I recently saw another RUclipsr Dyno tune a Kawasaki H2 supercharged motorcycle back when it was new. He develops performance parts for these types of bikes. While on the dyno, you can see the throttle body smoothly modulate to 100% open even tho he demanded 100% to start the run. Interestingly, once the bike started getting to the higher RPMs, the bike started closing the throttle, killing power, before it was anywhere near the redline. They flashed the ECU to give 100% throttle in the areas where it was closing and the bike picked up like 80whp. Crazy computers😆 - Pantyeatr1
Interesting. Traction control or anti-wheelie tables in the background or just the usual 'lazy' DBW factory control that makes us all generally underestimate them? - Taz.
@@hpa101 it's hard at say, but you can clearly see what the bike was doing on the dyno screen. The bike was backing out way before redline and cutting the HP by at least 30hp.
I’ve had several turbo engines that made dramatically more power at part throttle at lower RPMs than full throttle. Would be great to have the ability to prove that out with DBW.
Great video Andre! One correction though, you mentioned in cable throttle scenario there's nothing ecu can do to remedy the stuck blade scenario, this is not true as if theres both tps and pedal sensors employed the ecu will know if the cross-check fails and there are ofcourse other methods of torque reduction in that scenario :)
At that point, why not use DBW if you are going to install everything required for it? The only legit reason I can think off for cable throttle is for the feel, kinda like how cable clutch has a unique feel.
You can actually get really good PID figures using one of matlab's fairly recent applets. It takes in the step responce (so from no power to all the power, see what happens) and works its magic to get you 99% there. I believe they also do a fair bit of SIL stuff that can automate the final tuning process and get you the exact responce you want, so you don't have to manually fiddle with it. Not sure how that would perform accounting for adverse conditions though 🤔
Can you tell me more about this? Having issues with my dbw not staying on target and would like to have some guidance as to what PID settings should be
@mkgxcommanderx It's an add-on app for a package called matlab (sold by mathworks) called "PID Tuner". I've seen it demoed and it seems to work pretty well. Thing is, it needs good quality raw data to work with. Do you get access to those? If a step responce is not something you can get out of the ECU, you may have to manually measure it and use a proper DAQ on the position sensor. The step responce is simply applying full voltage to the input all at once. Then also taking it away all at once. So it can be measured in both directions, in case it differs. Another way to examine this is to use an oscilloscope (a deep memory DSO in particular can be very handy), again on the position sensor, and fine-tuning the PID constants until it responds exactly the way you like. All of this equipment is quite expensive though (including matlab, so even with raw data from the ECU it's not cheap). If you don't have ready access to it, you should consider having your PID loop tuned by a professional. If it's a one-off, it'll be a considerable saving. If you do have the raw data (especially if they are in graphical form), you can have a go at adjusting the numbers yourself, but be warned, it's a massive time sink. You may also consider a book (or five) on control engineering at that point 😁
Is there setting for recommend frequency with say a gm gold blade throttle body I know with the aem infinity there are setting and presets for different throttle body's I had always considered what was the default setting to be to slow and the setting from the presets you could actually hear the dc motor hum with the engine off ignition on
Hello, is it ok to splice in app signal to both the original ecu and a haltech ecu? I need the original throttle position wiring to the original ecu to keep some features the car has. Thank you for your time.
Only one way to find out 😅 Can't really give you a straight answer on that as it's not something we have done, it might depend a bit on your specific ECU (OEM) and what that ECU is still controlling vs your Haltech - Taz.
I would imagine an ITB engine would likely make more low rpm torque at partial opening due to the big throttle area and potential of better swirl/tumble/ and port velocity all those things are a more important at low RPM .. Hey Andre why haven't you updated this ECU to run the NSP firmware and software ??
The NSP software appears to be missing the APP to TPS Calibration curve option. Looks like the only way to do it is to manually change numbers in the Pedal to Throttle map. Anybody know of a way to access the APP to TPS Calibration curve?
Having issues can't set idle control on dbw lol when I put to closed loop idle is static at 900 doesn't move but when it's in open loop it starts at 1200 slowly lowers to 1000 not sure what I've got wrong
Thanks for reply I'm only a bronze member lol I'll upgrade if there are people that know the haltech inside out Do you guys need a log or just screen shots
You should be able to get 1 month for free if you haven't already claimed it. I think it pops up with you try to watch a webinar but don't have Gold Membership: www.hpacademy.com/previous-webinars/ If you want 100% Haltech feedback from 100% Haltech people, the Haltech forums are your best bet as while Andre has used the platform for years and knows Haltech well I can't guarantee it will just look at your log and point out your issue in 2 seconds. Sadly it doesn't always work like that with these things. That said, sometimes that is also exactly how it works 😂 A log and your map file would be best. The more info you can give people with things like this the easier it is to actually help rather than just take guesses - Taz.
I have a k24 swapped Scion frs with a haltech elite 1500 ecu. While cruising if I am at very little throttle pedal input around 10-14 percent or so maybe lower and the car will buck as if the throttle body blade is snapping shut. I have the stock FA20 throttle body wires to the ecu and this is the only problem I have with the DBW and I can’t track down which setting might be causing this.
Datalog tps when the condition is happening. If it's not moving then I could guess it's hitting engine protection or a severely rich portion of the map
@@ReubenHorner I don’t have any engine protection set up on the ecu and I’ve data logged it before and the AFR is perfect. That’s why it’s been such a problem to track down
5 and a half minutes in and I'm already disappointed. First of all, 60ms is NOT a short time when you're in a situation of heightened awareness and focus like when you're driving. In a blink of an eye, you might miss so much information that could determine whether you live or die, an amount of information that you are fast enough to process. Also, the speed of a blink of an eye varies a lot. When you're driving or riding, your eyes open and close multiple times faster than when you're relaxed. With that out of the way, I'll use some numbers for reference. A cable throttle reacts at very close to the speed of sound in steel because the cable is made of steel and mechanical waves and disturbances propagate at the sound of speed or close to it respectively. You might say, and you'd be correct, that electric signals travel at 90% the speed of light in a wire but that's irrelevant and I'll explain why. Now, in a throttle cable of a motorcycle which is around 81cm, the pull of the cable as you twist the throttle will reach the throttle actuator at a speed of ~5000m/s and at that point the actuator will immediately start moving. Do you know how long it takes for the actuator to start moving from the moment you start twisting your wrist with a throttle cable of 81cm? It takes only 0.164 milliseconds. If you can twist your grip fully in 20ms, do you know how long it will take the actuator to open the butterfly valve from 0% to 100%? 20.164ms. That's the time it took you to fully twist the grip+the time it takes the tension stress to propagate from your wrist to the throttle body and open it. The human brain is able to see and recognize an image of a single frame in a 1000fps screen and the more you train it the more info it will be able to extract from that single image in 1ms of time. If you can extract meaningful information with your eyes from an image in 1ms, you can definitely feel the difference in tactile feedback between 0.164ms and 60+ms. Those 60+ ms will determine if you'll blip the throttle smoothly or not for instance, then again you'll say we have autoblippers for that. The only real usefulness of fly by wire is electronic aids and electronic aids numb your skills as a driver. They do make you faster because a computer will process all the info faster and more efficiently than a human and it will operate everything in unison on its own with the human's input as a suggestion rather than a command and optimize output, but it will do so in expense of your skills because the car/bike is essentially driving itself. When you want the maximum torque applicable, you will not modulate the throttle properly, you will just go wide open and the ECU will do the modulation for you. What operators had to do manually in the 80s and 90 now the computer does for you. With a cable you'd know that at x RPM and y Lean angle, the optimal throttle would be z% and open it that much, now you just whack it open and the ECU calculates x, y and z and does it for you. Where's the skill in that? Now, why is the speed of electric signals that is massively faster than mechanical waves irrelevant? Because there's the middleman, the ECU, AND the delay of the servo. Although, both of these can be negated by using faster CPUs and and more expensive servo units that can work with shorter delays than hundreds of a millisecond. Do CPUs that fast exist? Yes, even the most basic embedded 1GHz CPU is fast enough to process wheel speed, TPS signals, Pedal/twist grip position signals, RPM, O2, Mass airflow and all the other parameters fast enough and send the signals in a few thousandths of a millisecond but I don't know if there are servo motors that can react that fast in starting and ending an operation once a signal is received. Now, do manufacturers, both OE and aftermarket, consider the cost of more decent CPUs and fast servo motors worth it to make their systems react better than a mechanical system? From what I've seen so far, no, they do not. Or they simply do not care, or servo motors that can keep up with fast CPUs and exceed the speed of a mechanical solution aren't yet available. Now I'll watch the rest of the video out of curiosity to see what's next. EDIT: At 8 minutes... If a cable throttle body jams open, there's the return cable that forces the throttle body closed and what do you mean the ECU can't do anything about it? Every injected vehicle, cable or not, has a TPS. There's nothing stopping a manufacturer from putting a grip/pedal position sensor in a mechanical system and monitor the discrepancy between the grip/pedal position and the TPS, if there is any, and cut fuel and spark if any discrepancy is observed. I'd assume those already are made like that, it would be absurdly stupid if billion dollar manufacturers couldn't do that. Also, a properly maintained cable throttle system will not have these issues and sensors would be just a redundancy, there's no way, under normal operating conditions, that a properly maintained cable would just snap or a throttle body would just jam open, dirt, grime and gunk do not magically just spawn out of thin air and jam equipment. I don't remember the last time I manually cleaned my motorcycle's throttle body or lubed the cables and there wasn't even a hint of hesitation or resistance from the throttle grip and the bike had seen quite a bit of bad conditions before I wrecked it.
@@hpa101 I never claimed to be a top notch rider or even mildly talented. I simply stated some simple things. 60ms is noticeable for a human brain, 0.164 isn't. And now that you mention it, is it possible to convert a ride by wire to a cable throttle? I think any motorcycle newer than 2010 is ride by wire...
The simple point there was unless you have data and experience to show it's a problem for your application, it's not really a problem 😎 There are other factors beyond response time alone too, some of which are also discussed in the video here. At the end of the day you can do whatever you like with your own setup to suit you and your application 🤘 - Taz.
Are you stating as fact that a cable throttle has never jammed or failed and can never, ever, EVER, jam or fail? That's a strong statement. Not sure I'd say that about DBW OR cable throttles personally, but DBW is hardly 'sketchy' 😎 - Taz. PS: You might enjoy (or hate, but hopefully enjoy!) this: www.hpacademy.com/blog/drive-by-wire-dbw-systems-are-trying-to-kill-you-but-not-really/? - Taz.
i used to "not like" DBW...then i started building and driving swapped cars every day. DBW is clean to wire, simple to setup, fully adjustable and adding a/c and cruise controls is just a few clicks. but the best part is being able to dampen throttle response at slow speeds like driving over speed bumps and dips in the road without the jerking of cable setups when your foot moves relative to the throttle pedal. if you're a DBW "hater" then you havent driven a proper-setup one yet. 👍
What you said about throttle cables sticking reminded me of seeing an unlimited hydroplane that had a cable the driver could pull to disconnect the cable from the throttle.
Nothing is infallible and many things are always finding new ways to catch us out 😂- Taz.
Nice and useful video.
Haltech DBW calibration wizard looks great.
What are you fitting DBW to or tuning it in?
🔵 Building a fast car? Get $400 OFF the all-inclusive VIP online course package deal: hpcdmy.co/offery161
🦵Kickstart your EFI Tuning knowledge. Get 50% OFF your first online course: hpcdmy.co/offery161
Great video Andre and HPA team. I recently saw another RUclipsr Dyno tune a Kawasaki H2 supercharged motorcycle back when it was new. He develops performance parts for these types of bikes. While on the dyno, you can see the throttle body smoothly modulate to 100% open even tho he demanded 100% to start the run. Interestingly, once the bike started getting to the higher RPMs, the bike started closing the throttle, killing power, before it was anywhere near the redline. They flashed the ECU to give 100% throttle in the areas where it was closing and the bike picked up like 80whp. Crazy computers😆 - Pantyeatr1
Interesting. Traction control or anti-wheelie tables in the background or just the usual 'lazy' DBW factory control that makes us all generally underestimate them? - Taz.
@@hpa101 it's hard at say, but you can clearly see what the bike was doing on the dyno screen. The bike was backing out way before redline and cutting the HP by at least 30hp.
All about emissions restrictions. Power,noise you name Government bodies require different restrictions per countries.
Great video, such a solid channel. Thanks guys 💪🏻
Cheers mate, appreciate that! - Taz.
I never realised just how powerful the Haltech elite is. The customer axis tabling is Motorsport level stuff! I want one now...
Many other brands do it mate, it's not that special now
@@sexyfacenation any examples? I am hunting for a new ECU but the haltech is not cheap...
@@ApolloBSG Link definitely do it.
I’ve had several turbo engines that made dramatically more power at part throttle at lower RPMs than full throttle. Would be great to have the ability to prove that out with DBW.
Informational as always
Great video Andre! One correction though, you mentioned in cable throttle scenario there's nothing ecu can do to remedy the stuck blade scenario, this is not true as if theres both tps and pedal sensors employed the ecu will know if the cross-check fails and there are ofcourse other methods of torque reduction in that scenario :)
At that point, why not use DBW if you are going to install everything required for it? The only legit reason I can think off for cable throttle is for the feel, kinda like how cable clutch has a unique feel.
@@patx35 personally I would go for dbw for sure, but you might be limited by class rules to run cable for example.
You can actually get really good PID figures using one of matlab's fairly recent applets. It takes in the step responce (so from no power to all the power, see what happens) and works its magic to get you 99% there. I believe they also do a fair bit of SIL stuff that can automate the final tuning process and get you the exact responce you want, so you don't have to manually fiddle with it.
Not sure how that would perform accounting for adverse conditions though 🤔
Can you tell me more about this? Having issues with my dbw not staying on target and would like to have some guidance as to what PID settings should be
@mkgxcommanderx It's an add-on app for a package called matlab (sold by mathworks) called "PID Tuner". I've seen it demoed and it seems to work pretty well. Thing is, it needs good quality raw data to work with. Do you get access to those?
If a step responce is not something you can get out of the ECU, you may have to manually measure it and use a proper DAQ on the position sensor. The step responce is simply applying full voltage to the input all at once. Then also taking it away all at once. So it can be measured in both directions, in case it differs.
Another way to examine this is to use an oscilloscope (a deep memory DSO in particular can be very handy), again on the position sensor, and fine-tuning the PID constants until it responds exactly the way you like.
All of this equipment is quite expensive though (including matlab, so even with raw data from the ECU it's not cheap). If you don't have ready access to it, you should consider having your PID loop tuned by a professional. If it's a one-off, it'll be a considerable saving.
If you do have the raw data (especially if they are in graphical form), you can have a go at adjusting the numbers yourself, but be warned, it's a massive time sink. You may also consider a book (or five) on control engineering at that point 😁
Is there setting for recommend frequency with say a gm gold blade throttle body I know with the aem infinity there are setting and presets for different throttle body's I had always considered what was the default setting to be to slow and the setting from the presets you could actually hear the dc motor hum with the engine off ignition on
Hello, is it ok to splice in app signal to both the original ecu and a haltech ecu? I need the original throttle position wiring to the original ecu to keep some features the car has. Thank you for your time.
Only one way to find out 😅
Can't really give you a straight answer on that as it's not something we have done, it might depend a bit on your specific ECU (OEM) and what that ECU is still controlling vs your Haltech - Taz.
What compute architecture do you guys use in your ECUs?
works great on bmws if you know hat you are doing 🙃😊
Why did you have to mention blinking!!!! All I could see after that was you blinking 🤣🤣🤣🤣
Quick question. Do you happen to know where in tunerpro you can change from dbw to dbc? I been looking and can't seem to find it anywhere. Thanks
I would imagine an ITB engine would likely make more low rpm torque at partial opening due to the big throttle area and potential of better swirl/tumble/ and port velocity all those things are a more important at low RPM .. Hey Andre why haven't you updated this ECU to run the NSP firmware and software ??
The NSP software appears to be missing the APP to TPS Calibration curve option. Looks like the only way to do it is to manually change numbers in the Pedal to Throttle map. Anybody know of a way to access the APP to TPS Calibration curve?
After doing a burn out my car will sometimes stall. Would increasing the close damping reduce the chance of stalling after doing a burnout?
Limp mode should be illegal.
Having issues can't set idle control on dbw lol when I put to closed loop idle is static at 900 doesn't move but when it's in open loop it starts at 1200 slowly lowers to 1000 not sure what I've got wrong
Why I'm here now lol but doesn't go into that much detail
Are you an HPA member? Post this up on a forum with a bit more information: www.hpacademy.com/forum - Taz.
Thanks for reply I'm only a bronze member lol I'll upgrade if there are people that know the haltech inside out
Do you guys need a log or just screen shots
You should be able to get 1 month for free if you haven't already claimed it. I think it pops up with you try to watch a webinar but don't have Gold Membership: www.hpacademy.com/previous-webinars/
If you want 100% Haltech feedback from 100% Haltech people, the Haltech forums are your best bet as while Andre has used the platform for years and knows Haltech well I can't guarantee it will just look at your log and point out your issue in 2 seconds. Sadly it doesn't always work like that with these things.
That said, sometimes that is also exactly how it works 😂
A log and your map file would be best. The more info you can give people with things like this the easier it is to actually help rather than just take guesses - Taz.
I have a k24 swapped Scion frs with a haltech elite 1500 ecu. While cruising if I am at very little throttle pedal input around 10-14 percent or so maybe lower and the car will buck as if the throttle body blade is snapping shut. I have the stock FA20 throttle body wires to the ecu and this is the only problem I have with the DBW and I can’t track down which setting might be causing this.
Datalog tps when the condition is happening. If it's not moving then I could guess it's hitting engine protection or a severely rich portion of the map
@@ReubenHorner I don’t have any engine protection set up on the ecu and I’ve data logged it before and the AFR is perfect. That’s why it’s been such a problem to track down
Fuel or spark cuts?
Of coarse it wouldn't be science otherwise.
5 and a half minutes in and I'm already disappointed. First of all, 60ms is NOT a short time when you're in a situation of heightened awareness and focus like when you're driving. In a blink of an eye, you might miss so much information that could determine whether you live or die, an amount of information that you are fast enough to process. Also, the speed of a blink of an eye varies a lot. When you're driving or riding, your eyes open and close multiple times faster than when you're relaxed. With that out of the way, I'll use some numbers for reference. A cable throttle reacts at very close to the speed of sound in steel because the cable is made of steel and mechanical waves and disturbances propagate at the sound of speed or close to it respectively. You might say, and you'd be correct, that electric signals travel at 90% the speed of light in a wire but that's irrelevant and I'll explain why. Now, in a throttle cable of a motorcycle which is around 81cm, the pull of the cable as you twist the throttle will reach the throttle actuator at a speed of ~5000m/s and at that point the actuator will immediately start moving. Do you know how long it takes for the actuator to start moving from the moment you start twisting your wrist with a throttle cable of 81cm? It takes only 0.164 milliseconds. If you can twist your grip fully in 20ms, do you know how long it will take the actuator to open the butterfly valve from 0% to 100%? 20.164ms. That's the time it took you to fully twist the grip+the time it takes the tension stress to propagate from your wrist to the throttle body and open it.
The human brain is able to see and recognize an image of a single frame in a 1000fps screen and the more you train it the more info it will be able to extract from that single image in 1ms of time. If you can extract meaningful information with your eyes from an image in 1ms, you can definitely feel the difference in tactile feedback between 0.164ms and 60+ms. Those 60+ ms will determine if you'll blip the throttle smoothly or not for instance, then again you'll say we have autoblippers for that. The only real usefulness of fly by wire is electronic aids and electronic aids numb your skills as a driver.
They do make you faster because a computer will process all the info faster and more efficiently than a human and it will operate everything in unison on its own with the human's input as a suggestion rather than a command and optimize output, but it will do so in expense of your skills because the car/bike is essentially driving itself. When you want the maximum torque applicable, you will not modulate the throttle properly, you will just go wide open and the ECU will do the modulation for you. What operators had to do manually in the 80s and 90 now the computer does for you. With a cable you'd know that at x RPM and y Lean angle, the optimal throttle would be z% and open it that much, now you just whack it open and the ECU calculates x, y and z and does it for you. Where's the skill in that?
Now, why is the speed of electric signals that is massively faster than mechanical waves irrelevant? Because there's the middleman, the ECU, AND the delay of the servo. Although, both of these can be negated by using faster CPUs and and more expensive servo units that can work with shorter delays than hundreds of a millisecond. Do CPUs that fast exist? Yes, even the most basic embedded 1GHz CPU is fast enough to process wheel speed, TPS signals, Pedal/twist grip position signals, RPM, O2, Mass airflow and all the other parameters fast enough and send the signals in a few thousandths of a millisecond but I don't know if there are servo motors that can react that fast in starting and ending an operation once a signal is received. Now, do manufacturers, both OE and aftermarket, consider the cost of more decent CPUs and fast servo motors worth it to make their systems react better than a mechanical system? From what I've seen so far, no, they do not. Or they simply do not care, or servo motors that can keep up with fast CPUs and exceed the speed of a mechanical solution aren't yet available.
Now I'll watch the rest of the video out of curiosity to see what's next.
EDIT: At 8 minutes... If a cable throttle body jams open, there's the return cable that forces the throttle body closed and what do you mean the ECU can't do anything about it? Every injected vehicle, cable or not, has a TPS. There's nothing stopping a manufacturer from putting a grip/pedal position sensor in a mechanical system and monitor the discrepancy between the grip/pedal position and the TPS, if there is any, and cut fuel and spark if any discrepancy is observed. I'd assume those already are made like that, it would be absurdly stupid if billion dollar manufacturers couldn't do that. Also, a properly maintained cable throttle system will not have these issues and sensors would be just a redundancy, there's no way, under normal operating conditions, that a properly maintained cable would just snap or a throttle body would just jam open, dirt, grime and gunk do not magically just spawn out of thin air and jam equipment. I don't remember the last time I manually cleaned my motorcycle's throttle body or lubed the cables and there wasn't even a hint of hesitation or resistance from the throttle grip and the bike had seen quite a bit of bad conditions before I wrecked it.
You can run a cable throttle on your own build if think something like a.60ms response is whats holding you back 🤘 - Taz.
@@hpa101 I never claimed to be a top notch rider or even mildly talented. I simply stated some simple things. 60ms is noticeable for a human brain, 0.164 isn't. And now that you mention it, is it possible to convert a ride by wire to a cable throttle? I think any motorcycle newer than 2010 is ride by wire...
The simple point there was unless you have data and experience to show it's a problem for your application, it's not really a problem 😎
There are other factors beyond response time alone too, some of which are also discussed in the video here.
At the end of the day you can do whatever you like with your own setup to suit you and your application 🤘 - Taz.
Drive by wire is sketchy and will never be as reliable as a simple mechanical setup.
Are you stating as fact that a cable throttle has never jammed or failed and can never, ever, EVER, jam or fail? That's a strong statement.
Not sure I'd say that about DBW OR cable throttles personally, but DBW is hardly 'sketchy' 😎 - Taz.
PS: You might enjoy (or hate, but hopefully enjoy!) this: www.hpacademy.com/blog/drive-by-wire-dbw-systems-are-trying-to-kill-you-but-not-really/? - Taz.
cable can never be as good as DBW when you know how to set it up on an elite series
@@adrianask9519Define "good".
It takes more skill to effectively handle a cable throttle, but it pays dividends.