What Are Durable Alternatives to Potentiometers for Electronic Throttle Systems?

[RexRoentgen]

Im currently designing a system to convert my cable pull throttle to an electronic one as part of a bigger project. Initially I was looking at using a potentiometer to indicate how far the acvelrator pedal had been pressed, but am wary of such a system due to the potential wearout. Can someone recommend a similar piece of electronics that will give me similar operation (voltage change as a function of rotation or perhaps a digital output) with high cycles to fatigue? Ideally this sensor would last 5-10 years under normal operation.

 

[RexRoentgen]

Should have done more research before posting my question. I ended up locating a hall effect based sensor thatll fit the bill.

 

[berkeman]

Glad to help!

Welcome to the PF, BTW.

 

[meBigGuy]

linear hall effect sensor may be a bit pricey. If so, you might consider optical rotary encoders, although that takes some additional electronics.

 

[Ryoko]

You may want to take a look at this new toy from TI (danger! PDF file)

http://www.ti.com/lit/ml/slyb212/slyb212.pdf

 

[dlgoff]

You may want to take a look at this new toy from TI (danger! PDF file)

http://www.ti.com/lit/ml/slyb212/slyb212.pdf

Heck, for $29 I might get one of their evaluation modules. I can think of lots of fun applications.

I though this image from the referenced EDN article was interesting from the selling point of view:

 

[meBigGuy]

Sorry to hijack the thread (but the OP seems to be gone) Could this be the basis for a metal detector? Is 16 bit impedance and 24 bit inductance capability enough to compete with good metal detectors? Or are they much more sensitive?

 

[HowlerMonkey]

He's running small block chevy and all of his questions can be fixed by using a factory ecu from a GM car from mid 80s to 1995.

You can connect to the obd port and read throttle position, timing advance, and a ton of other sensor inputs and engine parameters.

You can do this as far back as the early 1980s GM cars.

It's simple, connect GM hei distributor for "reference" to the ecu and whatever sensors you want to "see" when you connect through the obd port.

They make plenty of cables for pre OBDII cars regardless of manufacturer.

Get the right cable/software package and you can glean all you need.

 

[RexRoentgen]

While I am using a small block chevy, it is in a 73 vette and its a 68 block. So I am not sure how the ecu would help.

 

[Highspeed]

Perhaps you could throw out some more information about your project and more possibilities would come to light. A newer vehicle pedal with a throttle sensor built-in may be another option.

If you're planning to drive this vehicle with throttle-by-wire, please build in a redundent back-up. One failure could be catastrophic!

 

[HowlerMonkey]

Well...you could install tuned port injection as found in the later vettes which makes pretty good power and has the option of an ecu that shows throttle position and ton of other data.

You can pick the entire setup off of an iroc z Camaro at the "upull" yards for surprisingly cheap.

 

[RexRoentgen]

Perhaps you could throw out some more information about your project and more possibilities would come to light. A newer vehicle pedal with a throttle sensor built-in may be another option.

If you're planning to drive this vehicle with throttle-by-wire, please build in a redundent back-up. One failure could be catastrophic!

The project I'm working on has larger ambitions but I'm currently focused on implementing a rev matching system in my C3 corvette. It seems like the right combination of all the things I'm interested in. In this thread I sort miss-titled it because I'm not actually looking for a throttle-position sensor per say, but rather a rotational sensor with a high cycle life. I used throttle position sensor since I figure its very similar to my intended application of determining how much the accelerator has been depressed.

I've considered purchasing an accelerator pedal and haven't completely ruled it out but am not sure how well it will fit with the other pedals in terms of geometry. When I'm home this weekend I'll take some measurements to get an idea on the feasibility for it.

I definitely agree with you that redundant backup is definitely necessary, and one thing my job has impressed upon me is coming up with numerous fail safes. I never thought I'd do a failure analysis for a hobby project, but when your life is on the line it has to be done. I'll definitely be making sure there are redundant checks and provisions in place before this project sees the road.

As far as getting an ECU or converting the system over to EFI which can give me most of that data, in my opinion it sort of takes away from some of the fun. I'm enjoying the challenges presented by having to figure this entire system out.

 

[meBigGuy]

a rotational sensor with a high cycle life.

That would be an optical rotary encoder, optical quadrature encoder, or optical shaft encoder. (plus their magnetic equivalents).

http://www.quantumdev.com/products/optical_encoders/index.html

http://info.dynapar.com/magnetic-optical-wp-optical-rotary-technology-cta

Here is a sample circuit (page 5) if you want to convert rotation to resistance. There must be some product out there that does this nicely.
http://m.eet.com/media/1142647/18018-30603di.pdf

Depends on exactly how you want to interface things.

 

[HowlerMonkey]

All of the functionality you desire already exists within many ecu controlled systems which also includes the redundancy that makes them safe.

The "throttle" sensors on new pedals are dual sensors for that redundancy.

What you get with an ecu based system is the drive by wire and precise spark control as well as the ability to change your parameters through the datalink connector.

You can simply delete the injectors and put the carburetor and intake of your choice on the car but you won't be getting both horsepower and driveability of a full on injection system like we get from our corvette that does 250mph in the standing mile and then idles like a kitten on the drive home from cape canaveral.

 

[RexRoentgen]

I agree that I can go out and purchase all that I am looking for, but that's not what I am interested in doing. I want to make the components and do it myself.

 

[RexRoentgen]

To clarify my intentions with this project further. The end goal isn't to bolt on horsepower or add any cool functionality. Its ultimately to learn about engine control and design. So yes I can go and purchase an ecu to do all that I want it to do, but that's gettig away from the point of why I want to do it.

Ill take a look at using an optical encoder in the system. What advantage does it have over a hall effect sensor?

 

[HowlerMonkey]

In the average drive by wire car, the throttle pedal sensor is two potentiometers being fed a 5 volt reference.

The signal the sensors return are usually the opposite of each other.

Most drive by wire throttle plate position data is also 5 volt potentiometer.

I'm not sure hall effect or optical sensing is really necessary.

 

[meBigGuy]

Optical vs Hall is mostly a matter of what is available, what it costs, and what you are trying to do. One of the articles I linked to probably outlines it well.

The problem with hall or optical sensors is that you genberally need interface logic to convert pulses to resistance, voltage, or whatever.

Obviously Howler knows a lot more than I regarding engine controls. I've not done any engine control projects.