mileage computer for a carbureted engine
 

Is there a need for a DoItYourself flow meter on an engine using a carburetor? if so, i can tell you how one i had 30 years ago was built.

Using a circular track between two plates, with a marble or ball going around counted by an electric eye.

And a magnet on the drive shaft counted by a sensor for speed.

These two signal into a computer for MPG

Sorry if I seem not to bright but what moved the ball or how was it plumbed that the flow moved the ball bearing?

Why did you use speed instead of engine rpm?

How was it calibrated?

My ride is a carb version and I would love a flow reading

Thinking about it I think I've come up with a flow meter design...

and of course the speed is for the distance part of the equation, duh.

so the big unknown for me is the computer impute part of it.

There's already a product out there for non-OBD and/or carburated vehicles. If you go to a boating supply store/website you can get gps gauges that also have a transponder that you put in-line with your fuel line. These gauges will give very accurate speed, fuel flow, and typically have MPG readouts. Unfortunately I have yet to find one even close to the cost of a ScanGauge. :mad:

Let's set a price goal of $30 for this flow meter, that should filter the ideas.

we are talking about non OBD engines.
the marble or epoxy ball is pushed around a circular track by the gasoline.
an electronics guy can tell you to make a light tach to count the ball passes(fuel volume)
the distance is pulses off magnet on the axle or drive shaft u-joint with hall effect or other sensor, a junk yard crank or cam sensor.
the calibration is like any MPG meter, count the pulses per gallon, and the pulses per mile. there are other threads on this site that tell how to use an old lap top to do the computation. but since this is just pulses and you don't have to decipher pulse width, the electronic minded can tell you how to use a divider chip (terminology) and display.

Below a very rough sketch. Visualize this stuff stacked up together and clamped with 4 bolts. The holes thur the outside ring are small enough so the ball passes over. So does some fuel, but it doesn't matter just takes more pulses to make a gallon. Next try to think of materails that are gasoline and alcohol proof to make it out of. A 1/2"x3" aluminum plate and ball mill would be nice, but lets try to think of something even more common and less expensive. Like maybe cut a 5/16" length off a piece of 3" seamless tubing for the outside ring. And 2 1/2" for the inside ring.

I guess my question is about this kind of meter is what keeps the ball moving between the "out" to the "in" ? to get back into the start place again.

And if it is a steel ball the angle of the vehicle could throw off the readings, because there can't be a real seal between the ball and the wall. so you will get some flow around and if the angle increases or decreases that would I'm guessing cause differing readouts.

You had experence with one of these years ago how good did it work for you?

personx, tell us about your idea.

Would the pulses of an electric (solenoid type with internal pressure sensitive shut off) fuel pump work?

The one i had in the 70's worked very well. it was two piece die cast, so each side had one half of a round groove. (your better seal) so i'm thinking we'll need to fill the corners in with epoxy, if we can't figure out an inexpensive way to cast or machine the groove. The ball was a plastic or rubber which was light weight. (your mass and acceleration concern). This may not be as much of a concern as we think because although the ball races around one side it would go slower around the other side. The old meter failed because the change gasoline formulation in the 70's melted the ball. so we need something round 1/4 to 5/16 in. diameter (approx. fuel line size) and immune to the chemicals in gasoline. So looking in the back of the Grainger's catalog chemical resistance chart, we see aluminum, kynar, teflon, polyacetal, nylon, tyton, carbon, ceramic, viton, epoxy. and i would add glass to the possible candidates, although on the heavy side. We are looking for a round ball made from one these material. Googling aluminum ball i find a lot of suppliers. 1000ea 5/16" weighs 1.6 Lbs. I am leaning towards aluminum right now any suggestions?

there is gasoline bypassing all the time that the engine is using gas, to take the ball pass the inlet port.

this pic of that 70's mpg computer, which still works by the way. Just no flow meter.

what I was thinking about was something that I remember from fish tanks. they had a little spinning thing to show that the water was flowing, that the pump was really working. it was like the vane type air or hydraulic pumps. 6 (or more) vanes from a central hub all sealed in a clear plastic housing.

it would be easier to machine, just cut a round hole (1" to 2") in however thick stock you want and sandwich the sides as with the other one. drill inlets and outlets holes. you could mount a magnet on each side of the hub for balance or do a flat round plate that covers the vanes and color it to set off an optical sensor.

this is along the lines that I'm thinking now, but I not really married to any ideal right now. I had not even thought of doing this until you brought it up and now I think it's a great ideal.

I'm kind of a slow thinker and had some problems visualizing how that ball would get around the backside that is why I was asking about how well it worked for you. nothing like experience with something to see the way.

anyway I'm still building it in my head and have not come up with what I think is a good design so far.


ps where at in OH are you, I grew up in Franklin (between Cincinnati and Dayton) many many years ago.

the impeller would be gasoline proof, the bearing material would be what?
lets define the pressure to be 7 psi.
on the ball in a track design, an aluminum ball would be good, but all kinds of glass beads are available now. think glass would be to heavy? know any arts and crafters, a 6mm black glass bead would be good.
let's estimate the flow per min. @60 MPG @60 MPH would 1 gal,/ hr. or about 2 oz./ min.
@30MPG would 4 oz./min.
so 2 to 4 oz/ min would be a range to test over.

I have a need to get fuel consumption feed back on my bike. But without a fuel pump to keep the marble moving I might have to look elsewhere.

Hmm, fuel pump..., what about replacing the fuel pump with a 12v positive displacement pump and small accumulator that shuts off at a couple psi? Then keep track of the revolutions the pump turns and the distance traveled?

dcb,
your fuel pressure is what 0.5 psi maybe, first i thought, graduated cylinder with a stop cock on the bottom. Then i thought an IV drip in the fuel line counted by an electric eye. and magnet on a spoke counted by a sensor.

we are going to need an electronics wiz to jump in here sooner or later and tell us how to divide one signal by another and display the result. that does not give us mpg but it would be a relative number. for comparing driving styles or whatever.

how about a solenoid style fuel pump, they pulse on and off
http://www.facet-purolator.com/solidstate.asp

I really like the drip idea :) 'cept I'd need like 15000 drips/hour on average.

I just realized, you have a positive displacement pump in your diesel. You can probably make a good guess at fuel consumption based on the throttle position and rpm alone, as long as you leave the smoke screw alone :)

D_J
Where would I find an optical sensor for this kind of project? Is one used on a auto somewere or do you have something else in mind?

I just realized, you have a positive displacement pump in your diesel. You can probably make a good guess at fuel consumption based on the throttle position and rpm alone, as long as you leave the smoke screw alone :)[/QUOTE]

A couple years ago, on the diesel, i mounted a thottle positon sensor on injector pump linkage. so at a given conditions, i can read the resistance change and get a relative number to comparing mods. i test both directions on the same road at 55mph.

for the bike a flow meter could be sized for the flow with smaller bead and track with closer tolerances.

or maybe a small solenoid pump and regulator to 0.5 psi

for sensing an object breaking a white light beam or infared, radio shack for example has several types of light sensing components and emitters.
this one was listed under compoents at the shack
http://www.radioshack.com/product/in...rt&tab=summary

it takes a small handful of these 1 dollar parts to do the whole project from two ac signals in to a voltage out we can read with a multimeter, or with more effort to its own display.

to get just fuel flow per unit time, it would take half the processing. in fact these displays on treadmills and bicycles would do that once we get the right shape signal to input to them. The flow meter is the show stopped IMHO.

i am not knowledgeable enough in electronic to know all the lastest short cuts, there is probably a chip that incorporates all these components in to one piece

There is a rather bulky shortcut to the electronics problem, but it turns it into a programming problem. If you can live with a laptop computer for a display then check out the diy mpg gauge (for fuel injected vehicles). It would be good for prototyping if nothing else and could theoretically be taught to read drips, marbles, throttle position, vehicle speed, whatever.

http://www.gassavers.org/showthread.php?t=4250

Or you could use a supermid ($170) and just figure out how to get the marble signal to look like injector pulses. The RPM displayed would be way off (and possibly the duty cycle), and it is only in metric units currently, but it would be a lot smaller than a laptop.

http://www.greenhybrid.com/discuss/f...-1-t-1-a-3015/

I THINK that there was/ is such a device available from JC Whitney. No Idea how/ if it works.
I'm guessing it is a vac gauge . . .. . .


Edit: Never mind it's just a Vac gauge with some #'s on it
http://www.jcwhitney.com/autoparts/P...002121/c-10101

i just found the patent on the flow sensor i had back in the 80's

http://www.freepatentsonline.com/451...o&stemming=off
it failed because the light burned out and the ball dissolved from the unleaded gas. in the drawing it appears more complex than what i remember.
first pic the ZEMCO computer(80's)
i also have the PDF file that i can e-mail.


for any non-ODB MFI, i have a KEL on my truck, from Poland, most $200 in it, just buying one. there is no credit link between here and Poland.
Second pic the KEL economy and trip computer. I have the manual in a word .doc.

speaking of pulse, would the supermid read the pulse from this fuel pump proportional to fuel volume. this type turns on below 2 psi and turns off at 4 psi. duration wouldn't change but frequency would increase and decrease with fuel usage.

http://www.facet-purolator.com/solidstate.asp


personx, what psi do need for your carburetor?

I'm having an epiphany in slow motion here.

I don't know if that pump will squirt out the exact same amount of fuel with each stroke/revolution, not like a gerotor pump would, but it might be close enough.

But if a pump (any pump) can turn on and off by itself within a pressure range as needed, then perhaps it is simpler and also "close enough" to just keep track of how long the pump has actually run. I.e. an 12 volt analog clock in parallel with the pump for proof of concept. It wouldn't be terribly instantaneous, but would be good for a trip. I think I'll have to try that, maybe a vacuum guage can close the gap on the lack of instant mpg.

therefore we have: miles traveled/(total seconds pump was on ON * average gallons per second the pump moves when installed)

Just have to log the start time, and the start miles and see how close it is at the next fillup to figure out gallons per second the pump is moving.

this bike computer is less than 20 bucks and it can count magnetic pulses per hour. so you would have "live" fuel volume per hr.

http://www.cateye.com/sites/cateye/u...s/en/VL500.pdf

so i borrowed the little woman's speedometer off her bike and placed it next to an operating fuel pump, and the display says i going 175 km/hr.

on the bike there is a lower limit on speed, below which, the display goes to zero. this occurs at 2 Hz more testing is needed to determine if the pump will trigger the counter at vary low volumes. or what the flow rate is at 2 Hz. i am thinking make the volume of the pipe after the pump very non-compressible so that the accumulated pressure drops rapidly, there by
causing the pump to cycle on and off (short cycle) for a small volume of fuel.

this bike computer has better spec's, if they are true. $22 (topeak 130)
http://www.topeak.com/sub/documents/M-TPC3-GB-5-05.pdf

That's awesome :)

Hey, does that pump shut off when the pressure is reached or does it just keep running?

this version of the pump i have does not shut off, it runs continuosly when i dead head it with a needle valve. so i am looking around for an oil pressure switch or something i can throw in line with it. then i can check different flows vs. time.

Wow, you've got it all set up there. Course we are up to $80 with pump/bike computer/pressure switch.

I really like the bike computer idea, they are like $10 at my local store. Even if it is just fuel consumption and not mpg it can be real useful on a regular route. It could be adapted to optical operation easily enough.

Perhaps, with a "marble", or turbine, or paddlewheel, or?? optical arrangement in the fuel flow, and some fudging with the wheel size adjustment on the bike computer, it could be possible to come up with a useable fuel consumed number (or factor of 10 number) in the distance travelled display (or the current consumption rate in the mph display).

Google check time: the bike computer idea is tried, with a $60 sensor. The second link has a schematic:
http://11109.rapidforum.com/topic=11...6&search=sigma
http://diverse.freepage.de/cgi-bin/f.../fuelmeter.htm
http://www.v-stromforum.de/files/ver...nzeige_117.jpg

Note: it specifies .025L/min on the low end of the sensitivity range on that sensor. It won't even notice my 250 idling.

so lots of options on a display but the inexpensive flow meter prob. remains.

the pressure sw idea is proof of concept with the stuff i have. the on/off pumps are in the $40 range. i am just interested in how much fuel would need to be bleed off to lower the line pressure from say 4 to 2. that pump needs to restart a couple times a second for the frequency to be high enough to trigger the bike computers.
the close tolerances on the small flow meters keeps the cost up.

Ok, here is a thought, kind of plays on the drip idea, basically use a bike computer or whatever to keep track of metered units of gas. No real tight tolerances required except the solenoid valves need to work. Now just an electrical problem Again! :) Will be near-real time depending on capacity of the resivoir and flow, I'm fine with a 10 second delay at full throttle myself if the valves will last 10 times longer as a result.

http://ecomodder.com/forum/attachmen...1&d=1204683170

Assuming circuitry to make it all work (the controller), theory of operation:

start state: valve B is closed, valve A is open, fuel is flowing into the chamber and raising float C.

1. opto-coupler d1,d2 is interrupted by the float and the controller closes valve A and opens Valve B. The float begins to decend.

2. opto-coupler e1,e2 is interrupted by the float and the controller closes valve B and opens Valve A and sends a trigger to the counting device (i.e. bike computer).

3. goto 1

Still need to sort out venting the resivoir when the float is supposed to be moving...

reminds me a little bit of the old style gas pump, where they pump how much you want up in the glass cylinder on top to measure it, and then let it drain down into your tank. don't see why it wouldn't work, gets around the low flow problems, and some of the sloshing problems.

i was sort of kidding when i said we could count the drops like an IV dripper, but the way you got those photo sensors mounted it would count the drops, just like my planter counts every seed it plants with an infrared sensor and IR light. might have to put a little dye in the fuel to make it show up better.

on those bicycle computers, do they just have a reed switch inside or a mag. pickup? because a reed sw might be vibration sensitive.

dcb, how did you draw that?

well i tried in vain tonight to get a bike computer to read the magnetic pulses from 12v solid state fuel pump. the digital odometer would not follow the increasing and decreasing frequency of the pump consistantly when i varied the flow with a needle valve. i had the pump cycling on and off between 2 and 4 psi pressure. but when it turned on each time it had to make 5 or 6 thumps to catch back up to 4 psi. and the bike computer counted those 5 or 6 pulses as 100 or 120 MPH instead of the actually fuel being delivered which was 20 or 25 MPH. the other style pumps that are longer and cylindrical might pump more volume in one stroke an work better, also the pressure switch i was using had a 2 psi delta that i couldn' easliy narrow. with a 1 psi or less delta and a higher volume per thump and it might work.

I thought the fuel delivered would have been in the distance travelled, don't know how usefull mph will be without the flow sensor. But if it is still off by a factor of four on distance travelled, then maybe see if you can tell the bike computer that the "wheel" is 4 times smaller in circumference.

The bike computer I had, the sensor would click when I brought the magnet near itm so assuming it was a reed type.


p.s. picture was done in paintbrush, saved to .gif

distance times a constant would be gallons but i never got that far, because MPH wouldn't track GPH. MPH times the fudge factor is GPH. the concept has not been dispproven.

who, had a pic of the Zemco flow meter. The part on the left is the recirculating ball flow meter.

thanks, AndrewJ I'll watch it. John

it sold for $25plus 6 shipping with a few seconds to go.

to summarize so far we have found that some Zemco flow meters are still on ebay and selling for $25 to 30. the output is frequency proportional to fuel flow.

dcb
has found a circuit diagram to convert pulses from a flow meter to a bike computer for measuring mileage on a carb'd engine.
http://diverse.freepage.de/cgi-bin/f.../fuelmeter.htm

also a site showing a bike computer displaying mileage.
http://11109.rapidforum.com/topic=11...6&search=sigma

This makes it possible to measure to mileage of a carb'd engine for about $100 to $120. which is still above our price goal.

parts dissaembly of recirculating ball flow meter
 

I found the old flow meter and it looks like it wouldn't be to hard to fix.
Also found old feeding metering pump.

pic 1 both halves of meter showing hole for the light source on left half, mount for photo resistor on right half.

pic2 input side showing flow damper and metering jet

pic 3 also found a precision flow up to 295ml/hr. (tube stretched over rollers measures the fluid.)

pic 4 drop detector assembly for the medical flow meter

Cool, what is the diaphram about? Is inlet pressure applied to the non-spring side?

I just realized looking at this that the "marble" is probably the same "density" as gasoline. A heavy marble might not be as accurate, it would have too much inertia, and tend to sink to a low spot, and a light one would tend to float.

yes, the spring pushes against inlet pressure.

The diaphram is an accumulator to damp the fuel pump pulses. Providing a more constant pressure to the restricting nozzle. Which is just brass jet.

The correct ball was made of hard plastic or rubber and was the same density as gasoline. It melted either from heat or fuel. the ball needs to be opaque. The ball in the photo is for illustration purposes. It is from a straight pin and the wrong material. I am looking for a black 0.250" glass or plastic ball.

the light source failed and I am going to replace it with a LED. But even with a ball bearing in inside and my flashlight, I can get the display to count fuel by blowing thur it.

I am going to put it on my mechanical injected diesel.