Google Science Fair Project

[Ahmed]

If you remember that Analog EFI thread I started about a year ago, then feast your eyes on this. I'm doing a Google Science Fair project on it. The schematics are here in a .ZIP file and there are some EagleCAD files in there too. BTW, can you make a throttle body out of PVC, if you can, then how? Thanks!

 

[Poboy kartman]

Hmmmmm- throttle body out of PVC.. ..race you the patent office sucka......

Don't see why not. Already smooth...Capital idea ol' chap.....

This could be good....

BTW- I spend more time throttling people than engines and find two bare hands sufficient. ...

 

[Ahmed]

This could be good....

This IS going to be good.

 

[exenos]

What are you meaning by analogue? No chip?

 

[itsid]

Cool idea!!!!

If that works, I'm really impressed.. I just took a quick look at the schematics
(oh boy, what a mess.. too much information you should drop a layer or two there)
Same goes for the baord layout.. looks like the auto-path feature...
I think you can do better and shorten at least 1/3 of the paths

I see a timer circuit, two amplifiers and a receiver...
yes trimmers diodes etc.
indeed no logic chip.
All adjustments made by trimmpots as far as I can see.

Impressive!

I'd really like to see that setup tuned on an industrial engine like ours (gx/clone)

What type of O2 sensor are you intending to use?
How accurately do you expect this thing to be under working conditions ?
This is really exciting..
Have you tested that thing yet?
can we see a video?

'sid

 

[exenos]

Cool idea!!!!


All adjustments made by trimmpots as far as I can see.

If this is true then this could be amazing and extremely versatile.

 

[Ahmed]

It's only a schematic for now, but I'm planning on making the board, ordering injectors and parts and testing it. It's designed to be extremely versatile and may be able used on larger, more powerful engines. Just hope someone doesn't see those schematics and patent it.

 

[Ahmed]

I have updated and cleaned up schematics, check first post!

 

[Ahmed]

I have updated shematics yet again. If you can, then please make and try out the system. Remember to use the latest copies of the board! I will be making my board soon! Ask questions, do not hesitate!

 

[exenos]

I will definitely be building one of these in the summer if it works.

 

[Ahmed]

Thanks! By the way, are my schematics sound?

 

[itsid]

Thanks! By the way, are my schematics sound?

I wish I could tell...
I see you added an air pressure sensor to the board.

That's where you lost me...

I had a rough idea of how this might work (at least I think I did)
But the air pressure sensor just doesn't make any sense on the board to me.

I understand that you need (or at least WANT) to know the air pressure in the intake, but the surrounding air? *shrugs*
So the previous MAP input made sense.. the onboard sensor doesn't... (to me!)

I really have no idea what you're doing with that information.

Yesterday I slightly modified your first schematics (basically just swapped pins) to have an easier task routing that thing
Didn't made a board layout yet or anything.
(just my way of understanding the details a little better)

So allow me to ask, you have a trigger input fed to the timer..
Why do you short discharge and threshold?
that's like skipping the timer altogether, or am I missing something?

'sid

 

[Ahmed]

I took the idea of shorting the "threshold" and "discharge" pins from this schematic here:

I put that on-board pressure sensor as an MAP sensor, a hose is run from the throttle body or retrofitted carburettor to the sensor input.

I have also updated the schematics again. Don't hesitate with the questions!

 

[itsid]

Ah.. a pressure hose... I see.
(I'd prefere wires over a hose.. but that's just me Nevermind)

Hmm.. well what should I say... revisited the harris' datasheet, and in at least one example they indeed short 6 and 7. *shrugs*
I still don't understand the reason why, but I'm okay with that

It's been too long since I last had to deal with electronics on a more regular basis,
so please excuse, if I say or ask something stupid

like this:
Would you please explain D6? (first version)
I think it's redundant if you reroute Pin2 of R11 and R8 to D7 and D8, or am I missing something again?

'sid

 

[fowler]

U need a MAF
Mass air flow

This will govern fuel delivery over varying amounts of air flow
Ie though the rev range

A TPI is ok but also not really. Needed on a petrol Efi
As the throttle plate only varys airflow

A MAP isn't really needed either
Unless u turbo it or plan to climb mountains

Think of the old carby
Just gotta match it's abilitys
Be able to vary fuel over rev range and supply it in a steady well mixed fashion

Is PVC pipe going to be able to disrupt the air flow enough to mix the fuel air

And u might wanna invest in a wide band
Might save the engine

 

[itsid]

U need a MAF
Mass air flow

......

A MAP isn't really needed either
Unless u turbo it or plan to climb mountains
...

I think this is purely about terminology here;

An air flow sensor in an older car IS a pressure sensor in a known diameter tube.
knowing the diameter and the pressure it calculates the flow rate.
(today it's indeed a hot wire anemometer, because of the higher accuracy)

most come with a air temp sensor bundled, but that's basically all the difference.
If you stick a pressure sensor in a tube, you in fact have a low res flow sensor (if you do the calculations yourself)

so MAP is sufficient since there are NO calculations just trimmings
(I still doubt it's a good idea to have the sensor on board though )

Inside the throttle body it returns the same signals as a simple flow sensor would.
All the additional precision of an anemometer will be lost at the next trimpot anyways


'sid

 

[fowler]

although an automotive map sensor is only measuring manifold pressure

it shouldn't change much if at all
just atom presure or if turboed then boost pressure

even a turbo car will run with out a map (sorta)
but runs **** with out a maf

unless u put it post throttle plate and measure vacuum

who knows
maybe i just dont under stand your explanation
ill stick to the diesels
no sensors needed there

 

[Ahmed]

But won't an MAF sensor be hard to implement? Besides, an analog EFI can't really do much with an MAF sensor's output, so I'm sticking with the MAP sensor. Also, if you want to, you can take off the on-board sensor in the schematic and replace it with the old terminal block and install your own sensor if you want. The project is kind of open-source for now, you can modify the schematics to your own needs.

 

[itsid]

although an automotive map sensor is only measuring manifold pressure

it shouldn't change much if at all
just atom presure or if turboed then boost pressure

even a turbo car will run with out a map (sorta)
but runs **** with out a maf

unless u put it post throttle plate and measure vacuum

who knows
maybe i just dont under stand your explanation
ill stick to the diesels
no sensors needed there

Well forget ANYTHING about automotive injection systems, they all (even ancient ones) make calculations based on the values send by sensors.
With this one that's not the case!

The sensors are basically all just there to controll the timer which then triggers the injector.

A MAF only sends a voltage (waveform) back to the ECU,
that calculates the amount of fuel to be injected based on said voltage. (mixed lookup table and correction values based on other sensors)
Again, this EFI cannot do that at all.

A MAP will send the exact same waveform to the EFI board,
less accurate -yes-; but still useable.

And since the board mainly consists of trimpots
those will blur out any additional accuracy anyways.

that's why -if you ask me- a MAF would be overkill for this scenario (analog efi) a MAP is more than sufficient.

get rid of the idea that you need any serious pressure to recognize a difference in the MAP signal.
if the sensor is placed inside the throttle body the output will vary based on the air flow inside the throttle body.
(I'm not sure about the connecting hose and the onboard sensor... but it should too, maybe just inversed)

'sid

 

[Ahmed]

EDIT: I got the engine but was not able to get pictures of it, nor build and use the fuel injection system due to time constraints and I had to give it away. However, I did finish the schematics and and put them up my Google Drive. Sorry for the delay.