fly by wire throttle

[Fred]

I bet you think the ford pinto engine is a good engine . name me a good engine from that era

I wouldn't touch a non 4/5 valve per cylinder engine with a barge pole with one exception. That exception is not the dirty old gutless pinto.

FE-DOHC, RB20DE(T), RB25DE(T), RB26DETT, 1UZFE, 1GZFE, 3SG(T)E, 4G63, VG30DET, B16A, B18C, 4AGE, M50, BPT/BPD/BP-ZE, VH40, VH45, etc etc.

All from within that era. All cast iron blocked. All built like brick shit houses :-)

the reason why you need to do major work to get better than std is because std is so good.

I don't know about you John, but I'd prefer to not have to change the rods out to push 12psi in. That's a major job. The rods were fine, but built to provide that 200hp, and not 300, and if you try, they will fatigue and fail.

certainly not with the minimilsitic approach to sensors to are trying to do.

This proves that you haven't grasped the concepts involved here yet. Given that a particular user has 10 sensors he can probably use most of them, however a user that has only 6 will be stuffed if the code RELIES on having 10. If it relies on 6 and can make use of 10, it is a much better situation for everyone. The guy with 10 gets maximum benefit, as does the guy with 6. Both are happy. You would have it such that only the guy with 10 was happy, however he would be absolutely no happier than he would be if the guy with 6 was also happy. Do you see my point?

when you get it to run as well and economical as std ecu --then you know you have got to grips with thing and have an ecu system that works .

This is the default, many OEM ecus leave a LOT to be desired. It is not hard to extract better drivability and economy and emissions etc from them if you so choose. MegaSquirt is a solid enough system to do that easily. It isn't really suited to 200 - 500hp / litre engines though due to its batch injection. Below that they are fine for what they are.

car makers are getting more power now from std engines and doing it with more economy --that is by anyones judgement a good measure of effiency--

True, how though? By lightening components and cutting margins of safety present in older less efficient engines. This directly correlates to an engines ability to handle power. Modern ones can't handle power (ah la forced induction) because they are built lightly to be efficient. They are great stock, and poor for boost in most cases without expensive things like rods etc etc.

I am old enough to remember when engines only lasted 40-80k-now modern engines just don,t wear out if kept serviced correctly .

This is true when they are making OEM power levels, yes.

I think this is where i stop posting cos you are still living in the dark ages --

That's a shame. Obviously you have some experience using ECU's, however equally obviously you aren't familiar with what goes on inside them. I sincerely hope you stop back and read this as I'm sure there is a lot we can both learn from each other. I've tried my best over multiple posts to explain how it is, and Jean has also stepped in on the 6 sensor thread and explained how not requiring an wideband doesn't mean that you will get any less benefit from one when you do choose to use one etc. We are not living in the dark ages, we are being practical and pragmatic. There is a lot of variety in engine hardware and supporting all of it cleanly is quite a challenge. One that most companies like GEMS do not even attempt. They only support common stuff as listed on their site. DIY is different to that and the range of installations will be large at some stage in the future. We are doing our best to plan for that.

Fred.

[KW1252]

Back on the topic about FBW throttles, here are few ideas about safeguards against various mechanical, electonic and electric malfunctions:

As the Autospeed article mentioned, the devices use dual potentiometers. There are also accelerator pedals with opposing potentiometers. The beauty of it the summed voltage should always be Vin, minus voltage losses. If it's not, there's a problem.

A watchdog is a good start.

For DC motor:
Prohibit 0 and 100% duty cycle; use the signal edge to drive a hardware time-out circuit.

For stepper motor:
Step sequence detection.
PWM modulation for falling edge detection as earlier.

Allowed rate: If the throttle signal is changing too rapidly or too slowly (possibly except at 0%), it is garbage. Power off.
Same for the PWM duty/step signal frequency to throttle plate.

And as Autospeed mentioned, sanity check for throttle output against engine conditions.

Mechanical failures should be handled by the controller itself. Software failure (code errors, circuit malfunction) should be handled by an external watch module. Most of these could be arranged by discrete components (0% change of coding or software execution errors) but it'll take a lot of components and board space. A microcontroller would do all this for much less parts, but have the problems of coding and execution errors; however, the routines are fairly simple to evaluate and debug, and the microcontroller can be really basic (like PIC12 series) so using several for redundancy is feasible.

[Fred]

If you're using several pics for redundancy, they should be independently designed and implemented such that no flaws are common.

0 and 100% are valid outputs to a DC throttle motor (I've implemented this in person with excellent results).

As is instant rate change. If I want to literally stamp the pedal, the ADC will go from 0 to full in a single sample (depending upon sample frequency) and you should be able to converge upon that value as quickly as possible with full 0 or 100 DC.

Again, implemented all that for a fast acting and accurate response, but I still wouldn't want one on my car, the indirection isn't nice, there is no feel, you can't notice physical issues with the throttle by feel, etc etc.

Fred.

[jharvey]

In my drive by wire Jetta, the throttle is wide open pretty much all that time. I guess it might close some under certain conditions during idle, but I understand the throttle is more of a safety feature than anything else. Basically if the ECU notices the injectors are dumping in extra fuel, the ECU has a separate mechanism to limit intake air, such that it can limit power. Other than that the throttle is wide open, and the intake air pressure is regulated with the turbo. I guess that's why the 1.8T is known for breathing well.

There is no foot feel for issues with the throttle, so if it were to back fire, I'd have no idea it was doing that. Perhaps a handy feature for FreeEMS would be to use the MAP to catch such spikes and toss up an indicator light or something.

[Fred]

Either your Jetta is diesel, has two throttles, or you're wrong. Are there any other options?

Diesel regulates torque via fuel supply with full air supply always.
Two throttles can be used, one electronic one mech for traction and/or safety control of other kinds.
A gasoline/petrol engine with a wide open throttle will just destroy itself in one way or another.

If you are talking about the 20 valve 1.8 turbo engine I think you are, then there is no way its wide open all the time, no way at all.

Fred.

[jharvey]

I'm honestly working off a very small source of data. I posted the manual snippet below which is just about all I know about this TB. The manual has a couple paragraphs about how it uses the TB at idle, and it has a couple paragraphs about how it regulates the pressure via turbo waste gate and BOV, then it only has a small section about the TB. Click the snippet below to enlarge.

So it limits the RPM to 1200 if there is a problem. I read this as TB closed. It also notes the TB may be much more open than the TP is commanding. WOT or not, I don't know, but it could be saying that. The vague nature of the manual, could imply they didn't want you to know it's WOT, as that can raise false alarms. Who knows. If anyone wants to hang off the front of the car and check, while I put the engine under full load, I'm willing, but a waver will be required first

[Fred]

That makes sense :-) You could do nice things to the drive-ability of it by taking the throttle input as a torque percentage request and acting accordingly. It won't be WOT when light load is required, though. It just says more throttle when you request less, so it probably opens the wastegate for less boost and opens the throttle for more efficiency (lower pumping losses) and gives you the torque you want, but without boost instead of with boost and a restriction :-)

Fred.

[MotoFab]

Either your Jetta is diesel, has two throttles, or you're wrong. Are there any other options?

Diesel regulates torque via fuel supply with full air supply always.
Two throttles can be used, one electronic one mech for traction and/or safety control of other kinds.
A gasoline/petrol engine with a wide open throttle will just destroy itself in one way or another.

If you are talking about the 20 valve 1.8 turbo engine I think you are, then there is no way its wide open all the time, no way at all.

There's also the BMW Valvetronic system which controls intake volume by directly controlling intake valve lift, that in lieu of a throttle plate.

Diesels also use throttle bodies. A closed throttle eliminates pumping losses.

Jim

[Fred]

There's also the BMW Valvetronic system which controls intake volume by directly controlling intake valve lift, that in lieu of a throttle plate.

Good point, but the old 1.8 20v engines don't have that.

Diesels also use throttle bodies. A closed throttle eliminates pumping losses.

Wait a second, pumping losses are from pulling against vacuum on the intake! Not from compressing and then decompressing the gas pulled in, right? If I'm wrong about this, educate me. It's 1am and it's been a long day and isn't over yet...

Fred.

[MotoFab]

You get back the energy from drawing the vacuum against the throttle plate, because that vacuum helps lift the piston on the compression stroke.

Pumping losses are from pumping. With a blocked intake there is very little air being pumped. If you consider a completely closed intake and exhaust, there are no pumping losses. There's only friction from the motor mechanicals.

Block the hose on a shop vac and the motor speeds up. Block the intake port on an electric-powered piston or diaphragm vacuum pump and it uses less current.

While traveling at a medium rpm in second or maybe third gear, switch off the ignition and take your foot off the gas. While coasting down in gear, open the throttle and notice the increase in deceleration. That is really noticeable on a high power to weight ratio vehicle, like a motorcycle for example.

Jim