is it possible simple pulse width modulation

Topic: is it possible simple pulse width modulation (Read 2459 times) previous topic - next topic

is it possible simple pulse width modulation

December 07, 2006, 03:25:05 AM

ive been reseaching some fabrication sites out there relating to TBI and CFI systems and it dawned on me an idea i have.

if its true that the EEC controls the pulse width modulation of a given fuel injector , wouldnt you be able to duplicate this by trial and error using an R/C time constant?

you already have the resistor (TPS) which varies depending on throttle position so a known value capacitor would be added to complete the circuit.

with this circuit, couldnt you vary the "TIME" inwhich the injector is open and closes depending of throttle position. It seems to me this would be too simple for words.

is it possible simple pulse width modulation

Reply #1 – December 07, 2006, 09:19:09 AM

And so megasquirt was born....

is it possible simple pulse width modulation

Reply #2 – December 07, 2006, 10:03:35 AM

The only thing I can see that might be difficult would be to try to keep the frequency constant while changing the pulse width (which is what ECM's do). Changing a resistor in an R-C circuit will change the pulse width but it'll also change the frequency. I'm sure there's a way to do it with a 555 timer though. I've got a program that lets you enter resistor and capacitor values to change frequency/pulse width in a 555 timer - I've been trying to attach it to this post but my internet connection keeps dropping. I'll try again later

is it possible simple pulse width modulation

Reply #3 – December 07, 2006, 12:59:52 PM

so..........for the sake of conversation, frequency is defined as the opposite of time. You already have your fequency and amplitude. The frequency is ''how long does the injector stay on and stay off with the same input voltage. The input voltage to an injector never changes from what i know so the only thing left is to regulate how long it stays on. IMHO, a timer circuit is a basic RC time constant and seems redundant.

therefore you have your frequency built into your engine by refering the math towards the rotating crankshaft.

What we know........
1 there are 60sec in one minute
2 the crank makes two complete revolutions to create work...
down-intake of fuel/air
up- compress air and fuel
down-spark pushed the piston down
up-exhaust gas is let out
cycle repeats here
3 a typical engine idle rpm is about 800

the drawings below are intended for a typical sefi setup because all the math would be different for a CFI application but still work the same, just different timing of the injector pulses since there are only two injectors.

Essentially , an injector on an sefi application would pulse every other complete rotation of the crank as described above.

the 800rpm and 4000rpm drawings are based on the same math,,,

@800rpm
800rpm/2=400rpm (this narrows down our time so the injector pulses every other crank rotation)
60sec/400rpm=.15sec or 150ms

@4000rpm
4000rpm/2=2000rpm (this narrows down our time so the injector pulses every other crank rotation)
60sec/2000rpm=.03sec or 30ms

Your fequency is the crank rotation ,all i think is needed would be a variable capacitor perse' to experiment with and how to take the output. An independant 555timer would work to but i dont see why you would need both.

*the only problem i see is making it very linear,, it should by nature but i think i need to review my ELI the ICE man stuff.

is it possible simple pulse width modulation

Reply #4 – December 07, 2006, 01:07:00 PM

I read it,I re-read it and thought about it,and I still don't understand it.

is it possible simple pulse width modulation

Reply #5 – December 07, 2006, 01:21:30 PM

How do you discharge the capacitor?
With a capacitor in series with the TPS resistor you get one pulse as the capacitor charges.

http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=6235

is it possible simple pulse width modulation

Reply #6 – December 07, 2006, 02:01:22 PM

I know my drawings depict an impossible situation,, i only did them to depict a known number of how often an injector would need to pulse,, not to make a car run. Ill correct myself here,, i think i understand now what you mean Thunderchicken on the frequency control,,, I should have actually siad "HOW LONG THE INJECTOR STAYS OFF or HOW LONG THE INJECTOR STAYS ON, not both.

vinnietbird
im just trying to think of a way to run a set of injectors (home grown) without the use of an eec. All that would be required next to make the car run would be the old time points dizzy.

softtouch
im just kicking around ideas,, getting the cobwebs cleaned out of the brain. I had to laugh when i saw what you wrote, that very thing was going on in the back of my mind.

is it possible simple pulse width modulation

Reply #7 – December 07, 2006, 03:47:20 PM

Both of those square wave drawings you show have a 50% duty cycle (which is the term used to measure pulse width). You can have a differnt duty cycle with the same frequency - imagine holding your engine at 3000 RPM with no load, then imagine holding it at 3000 RPM going up a hill towing a trailer. The frequency will be the same, but the injectors are obviously going to be open longer to allow more fuel. *EDIT* Ok, I'm back and full

So as this image shows:

The RED example shows a 50% duty cycle (the shaded area shows the "on" time)
The BLUE example shows a 25% duty cycle
The GREEN area shows a 75% duty cycle

All three examples are the same frequency.

As you stated, Scott, the ECM does not really govern the frequency, the tach signal does. The ECM also doesn't really control the amplitiude either - system voltage (12V) is the amplitude, although some more sophisticated engine management systems may do so (Ford EEC does not). The ECM simply controls (or modulates) how long the injector stays open during each "firing" (the length of injector pulse, or "pulse width modulation"). It does so based on several parameters it receives from sensors on and in the engine, including MAP, TPS, ECT, IAT, VSS, O2 and of course in mass air cars, the MAF. It takes readings from those sensors, adjusts the injector pulse width according to a table in the ECM programming, reads the O2 sensors to see how good a job it did, and fine tunes the pulse width from there. And it does this every time an injector fires.

Because so many variables decide how much fuel is required it'd be impossible to make a rudimentary fuel system that injects fuel based solely on engine RPM. Even a carburetor constantly adjusts fuel flow based on engine requirements. You would need inputs from at least four sensors: TACH, TPS, ECT, and MAP. TPS and TACH alone would not do the trick. MAP is required to determine engine load (higher MAP = higher load = more fuel, lower MAP = lower load = less fuel, even though TPS and TACH may remain constant) and ECT would be required for cold running enrichment (and warm engine leaning). All other things equal it takes more fuel to run a cold engine than a warm one.

You might be able to set an engine on a stand and make it run with TPS/TACH alone, but the moment the load on the engine changes the mixture would go lean because the simple RC circuit could not compensate. If you base your injector pulse width on a warm engine it would be very difficult to start and keep running when cold. Similarly, if you built the circuit to make the engine run while cold it would be too rich when it warmed up.

You're a thinker, Scott - I like that :D

is it possible simple pulse width modulation

Reply #8 – December 07, 2006, 06:22:17 PM

Carmen you sure you didn't help build the megasquirt? You have just educated me greatly.

is it possible simple pulse width modulation

Reply #9 – December 07, 2006, 06:54:44 PM

Are you wanting to build this system just to see if you can,or can't you buy a system that will work,like the new CFI syatems with the stand alone harness and computer?Just curious.I have noticed you seem to be a little creative.I can build a car,change every component,modify most anything and even set up a good audio system in the car.I've done all of that,but,when it comes to things like you're doing,WAAAAAAAAY out of my teritory.I am intrigued though.I'll be following this story.Keep the info coming and the results when you're finished.We get a lot of questions asked (myself included),but we seldom relay the final chapter,the results.I'll get my bowkl of popcorn and sit back and watch.

is it possible simple pulse width modulation

Reply #10 – December 11, 2006, 12:14:32 PM

I think i have a solution to adding in the missing link of "PULSE WIDTH". Manipulating an injector to fire based on "knowns" was easy but making the injector stay on longer and maintain the timing of when it does so was the hard part.

First off, leme get to the nittiy gritty

1 the drawings show a dizzy ***WITHOUT*** the cover so it is easier to see whats going on in there.
2 There is no TFI cause i dont think you need one.
3 there is no TPS cause i dont think you need one.
4 This is a points dizzy cause i think you need one.
5 there is no EEC cause i dont think you'll need one

The drawings,,,,,
PARTS LIST
-one calibrated spring
-one flat surface with a carbon layer that is durable and conducts
-the flat surface here is triangle shaped cause thats the way it turned out as i drew
-the flat surface has a pivit point such as a cap screw to allow motion
-the flat surface has a spring attached that is calibrated
-the outter rim of the dizzy holds a contact ring with 8 contacts
-the contact ring simply slips over the outter edge of the dizzy
-this contact ring is designed to allow the dizzy cap to mount on top of it

How it works
KOEO, the flat surface makes no contact with the contactor ring

KOER
-upon starting, the flat surface will make contact with the contactor ring by centrifugal force due to the dizzy spinning (see first drawing)
-when contact is made between spring loaded flat surface and the contactor ring, a ground is supplied to the already powered up injector resulting in a pulse of fuel that is proportional to the surface area of the contactor ring and the flat surface. This is where the *PULSE* is modulated.
-When you step on the gas (see second drawing), the flat surface is extended outward even more due to centrifugal force. This allows the injector to be supplied with a ground for a longer period of time due to the shape of the flat surface resulting in a longer *PULSE WIDTH* or,, simply said, the injector stays on longer sending more fuel.
-the timing of when the injector fires with relation to other injectors did not change, only the amount of time each injector stays on depending on the centrifigal force dominating the contactor ring to flat surface.

This is just a basic prinl drawing and i am sure that there is a design waiting to be tweeked. I am only illustrating a prinl, not a final product. If the prinl is working, the design can be made im sure.

I think this is about as simple as it gets ,, i love uncomplicating simple stuff.

now it needs a name,,,hummm,, how about Centrifugal Fuel Injected Distributor,,,,CFID,, has a ring to it i think : )

is it possible simple pulse width modulation

Reply #11 – December 11, 2006, 01:13:10 PM

So you've figured out how to make the injector pulse width respond to engine RPM, but there's still the matter of load. You need a way to fatten up the pulse width for higher engine loads. At any given RPM the load can vary considerably.

An easy way to do this, I suppose, would be to make a vacuum-operated fuel pressure regulator that has a wide range of operation. At idle (high vacuum/low manifold pressure) the fuel pressure could be low, but at WOT (low vacuum, high manifold pressure) the pressure could raise. This would vary the fuel being delivered even though frequency and pulse width would remain the same.

is it possible simple pulse width modulation

Reply #12 – December 11, 2006, 01:32:48 PM

hummm.........i think you nailed. I never accounted for that although i thought i did.

so a vac operated FPR and its design would remain the same with no changes and deliver the fat needed. ? Your saying that two pulses could be the identical width but one requires more fuel presure while the other could require less depending on the load?.

I think i see only one problem with my idea here. Its when you step on the gas. See, if you step on the gas pedal with my design, you are only inserting more AIR into the intake at the first moment in time. I would assume that when more air is introduced to an idle rpm, there would be a slight,,slight increase in rpm. This is my hope because with that said, this would cause the initial centrifugal force required to increase the fuel delivery (increasing pulse width). If it worked, it would be a ramping cycle perse'

do you see what im saying,,? I imagine only a throttle plate system on the intake that regulates incoming air.

I actually considered using a vac advance dizzy to create this pulse width thing and have it actually work off vac presure instead of centrifugal force but it sort of hurt my brain to think that hard,,lol.

is it possible simple pulse width modulation

Reply #13 – December 11, 2006, 04:44:43 PM

Exactly - imagine an engine on fast idle at 1500 RPM - this is a high vacuum/low pressure condition that would require very little fuel.

Now imagine that same engine in a car travelling at 70MPH in OD. IT's still gonna be turning about 1500 RPM, but the throttle will be open more and the vacuum will be slightly less.

Same car again, starts going up a long, slow grade. Not enough to activate the kickdown, but enough that you've got to lean on the engine a bit to keep your speed up. Still the same 1500 RPM, but the throttle will be open even more and the vacuum will be even less.

In all three scenarios the frequency (RPM) and pulse width (based on your centrifugal idea) would be the same. Adjusting fuel pressure by vacuum would help deliver enough fuel for each condition, but it would still be very innacurate. You would still need some kind of TPS to let your system know the desired engine output. You'd also need something to momentarily richen the mixture during quick throttle application to prevent the sudden rush of air into the intake from temprarily leaning the mixture and causing a bog. Think of the accelerator pumps in a carburetor. The normal fuel circuts in a carb do not react fast enough to sudden throttle openings, so the acc pump squirts a little extra in.

In your case what you could do is figure out a way to really fatten the pulse width momentarily, or add a second injector for throttle enrichment (but you'd have to make sure it only did so momentarily - slamming your foot down is one thing, but keeping it down requires a different strategy altogether), or even by adding a mechanical squirt pump like a carb.

is it possible simple pulse width modulation

Reply #14 – December 11, 2006, 07:06:16 PM

So the answer to your question.. there is no simple pulse width modulation. :D