electronic throttle

You forget the main culprit - LEAD FREE SOLDER.

Bean counters. And greenies. Lead free solder and bean counters on the
same project can definitely cause problems.

But bean counters and engineers on the same project ALWAYS cause
problems - whether mechanical or electronic - which IS why I qualified
both conditions - electronic and mechanical - as properly designed and
implemented.

You caught that.

ANd all told, Toyota has had a lot less of those problems, up 'till
now, over the long haul, than virtually ANY American manufacturer.
Toyota electronics over the years have been WAY above average.
NipponDenso components in particular.

 

The people who are pushing this issue have no idea about complex
software systems. You can not test for all of the potential
combinations of events.

Backing off on the throtle while getting airborn at the crest of a
hill with the brakes gently applied and the A/C on? I can come up
with hundreds and hundreds more. And everyone her can probbly match
those numbers without a repeat.

The EE who hacked a failure is an academic, not a real world person.
You have to apply the 20-80 rule. 20% of the problems will generte
80% of the failures. You are more likely to grind a brake pad to
metal than to have the concotted failure he caused. Much more.
--

- dillon I am not invalid

You know, I can't think of nothing finer than a fine
naked woman holding a gun. And you got all kinds of
fine going on.

Frankie Figs

 

at very close proximity. there is not a "good relationship" between
proximity and output for distance. thus, if you read you link again,
you'll see that those devices are used basically just as a pulse
counters, or on-off switch sensors, not proximity distance gauges. hall
effect potentiometers use this principle i think you'll find - they
count the number of pulses from origin, then electronically integrate,
not measure distance.

this is an example of an appropriate integrator:

http://www.analog.com/static/imported-files/data_sheets/AD5220.pdf

 

yes indeed.

some perspective for those that don't understand - a lot of good quality
electronic components are often rated for 50,000 hours use. higher spec
is available rated for 100,000 hours or more. a typical car driving
300,000 miles will on average only be operational for 10,000 hours.

 

even lead solder fatigues under thermal cycling. the mitsuba main relay
in older honda civics is a classic example.

 

dude, the numbers of inputs for a car is not only finite, it's actually
quite small. it's also very well mapped. the "combinations of events
is" therefore very testable.

he shorted two wires. any car would react in some way to that.

 

except for price. i had a potentiometer tps fail on a civic at about
300k a while back. an integrating, pulse counting, hall effect device
might have lasted longer, but at that mileage, i really doubt it's worth
the 10x price tag.

besides, let's get back to reality for a second: wtf is the logic of a
hall effect potentiometer with an analog output being used as an analog
input for a a/d converter???

 

Nope - they are used direcly as rotary controls like potentiometers,
with on-chip amplifiers built in..Generally they are single ended - in
other words using only N or only S magnetic polarity so the voltage is
only positive or only negative - not using rail to rail dual power
supplies. They are LINEAR, and actually extremely simple in concepr
and application.
The schmitt trigger type is as you say, strictly on and off, with
designed in hysteresis. The fact that there IS hysteresis supports the
fact that the output of the Hall "cell" itself IS linear, or at the
very least variable.

 

That is true - but lead-free lets a lot MORE gremlins out of the
cage!!!
And actually, in the vast majority of situations the fail IS "safe" -
with the result of a failure being non-operation - not full.

 

The actual Hall effect sensor chip is worth something like $2.39 each
in quantity. They are NOT integrating pulse counting devices. They are
a simple monolythic 3 terminal chip that, for all intents and
purposes, is a direct drop-in replacement for a resistive
potentiometer. A moving magnet varies the flux seen by the sensor,
varying the voltage output on the "wiper" terminal from 1 to 4 volts
with 5 volts applied to the "ends" if it were a pot.

I bought complete hall effect throttles for e-bikes for less than
$6.00 canadian direct from the manufacturer in HongKong about a year
ago.. That's handgrip, bracket, hall-effect and magnet - completely
assembled and wired, and functional and shipped to my door in canada -
quantity 6.

I've used them on Curtis E-bike controllers and 200 amp Navitas golf
cart controllers (Curtis 1206X clone).
Wish that stuff was available back when I built the Electric Fiat.

 

i think a better input device would be the kind of precision counter
system you find on digital micrometers and inkjet printers. direct
digital signal, totally clean. spending 10x on an electronic version of
the same kind of analog device that was in use 20+ years ago doesn't
make a lot of sense to me.

 

ok, i guess i'm out of date.

 

Don't know where you get 10X, Jim. The devices are pretty cheap - and
VERY reliable. Optical encoders, or even magnetic encoders, are FAR
more likely to cause problems in a noisy and dirty automotive
environment.

 

Actually, in Ford-speak anyway, it is called CAN (controller area
network)... Not to be confused with other computer (module communications)
networks on the modern automobile (we have had many networks and multiple
networks on the same car... as many as three separate networks on one
vehicle in the past).

Today, the Fords in the showroom have only two networks... high speed CAN
and medium speed CAN.... And it isn't a bus... it is a BUSS.

The BUSS is the two wire "common" that joins all the HS CAN modules through
the HS CAN+ and the HS CAN - or the MS CAN+ buss and the MS CAN- buss that
joins all of the MS CAN modules. There will be at least one module connected
to both networks and sometimes there might be two (I haven't seen any
networks that have three "sharing" modules). There will be one "gateway"
module... this is the module that good scan tools will use to access other
modules no m atter what network they are part of...

Now... the reason for two networks should be somewhat obvious. Some module
functions (engine control, braking, shift control and some other critical
functions) need to happen quickly and need to have high input scan rates.
Other functions (HVAC, entertainment/navigation modules - "creature comfort"
modules and "electronic toy" modules) can live with lower scan rates and
"slower" functionality - cheaper to build....

It shouldn't be much longer before we progress to the next level regarding
automotive communications networks and anything I can say will, overnight,
become ancient history. Baud rates will increase, module functionality will
increase, the modules will have improved self diagnosis.

Like the computer on your desk - if you have a car in your driveway, the
module communications network will be already obsolete... They are now
perfecting the next generation of electronic miracles.

 

check online prices. resistive potentiometers are about 10x cheaper
than hall effect.

magnets don't like it hot, especially not over time, and attract some
types of crud. optical can be pretty good on both counts.

but my main point though is still "why bother"? if we want a reliable
sensor, we should be looking at something with a direct digital output,
not this weak signal -> amplifier -> analog -> digital nonsense.

 

Not really - bus and buss are alternate and acceptable spellings - just
like gauge and gage. Many FSM's spell it bus. Bus is in fact a way
more common spelling than buss in industry, including the military.
Buss is almost obsolete. Usage kind of wins out over time. Perhaps
the Bussmann Fuse company (now Cooper Bussmann) spelling is confusing you?

 

Yes they would. If you want to know why just google Prius big hand
syndrome. It happened because the NHW11 (2001-2003) used a dual
potentiometer assembly and one or both pots got noisy. See:
http://techno-fandom.org/~hobbit/cars/DAS_Accel_pedal_renovation.pdf

BHS didn't get much press coverage IMO because the failure caused the car to
lose power not go out of control hence it wasn't newsworthy.

 

great post daniel - thank you. i did something like this for a civic
whose potentiometer had failed after ~300k miles. but it was an old one
with mechanical throttle and the results were just stuttering, not
failure. the guts of the civic's potentiometer were single track with a
~10 fingered wiper.

 

This electronic throttle thing is great. If you believe everything that
could be invented has already been been invented, do this: Replace
something already invented with something else already invented
and call it a new invention.

This has been successfully done in automobiles by throwing away the
familiar throttle cable and replacing it with a whole new system of
electromechanical parts; Hall-effect position sensors, electric wire
cables, electrical connectors, power transistor amplifiers, servo motors,
and an expanded computer program to control it all.

The brilliant scientists who created this new invention promise:

"significant improvement in mileage and pollution
control and significantly better control for things
like cruise control and automatic transmissions too"

They have revealed no numbers about how significant the improvements
will be, but they have pointed out that the new system will bring a big
improvement in safety. The wire inside the old throttle cable might bind
or snap, whereas the Hall-effect/position sensors/electric wire
cables/electrical connectors/power transistor amplifiers/servo motors
and expanded computer program use electronic parts reliable for millions
of operations. There is no way the new system can fail.

It's only a matter of time before these scientific savants come up with
other applications, such as eliminating the ripcord in parachutes.
Instead of tugging on a ripcord, a skydiver could just touch the button on a
Hall-effect sensor on his chest, and an electrical cable draped over his
shoulder would transmit a signal to the parachute's computer telling it
to let the power amplifiers energize the servomotors to release the latch
on the chute. Instead of a conventional ripcord that could bind or
break, a reliable electronically controlled ripcord actuation system
would finally take all of the fear out of parachute jumping.

The possibilities are endless. Are you tired of pulling a cord to raise
your venetian blinds? Do it with a servo control system just like the
electronic throttle. How about your lawnmower starter cable? Your
church bell rope? Your light fixture pull chain? Your tampon string?

As our lives are impacted by their leadership, the wisdom of the throttle
cable replacement scientists will be demonstrated again and again.

Best regards to all,

Rodan.

 

Is it not true that the drive-by-wire systems have a cable connecting
the accelerator pedal to the first electronic part? If so, a chain is
only as strong as its weakest link - literally in this case. If that's
the case, they'd be foolish to say that one benefit of the hi-tech
solution is the elimination of the cable. I can believe some of the
claims of better control of engines systems for power and emissions and
possibly enhanced safety if it's done right, but they should leave out
the part about eliminating the mechanical linkage.