mileage computer for a carbureted engine

[dcb]

Yup, what goes into the pump is a simple function of rpm, what goes out the return line is everything that didn't go into the cylinders (more or less).

[diesel_john]

oops i didn't realize in post 137 you referenced an entire other thread on the subject. Looking at all the springs inside the pump i think my take on the TPS accuracy is wrong. And since the van pump has an internal relief path, RPM probably isn't accurate either.

[dcb]

Hmm... Looks like you are correct about RPM, ah well, sounded too easy
I doubt the relief path could easily be plumbed to the return line.

Yah TPS is a little bit on the wonky side as a control sleeve indicator as well I think.

[diesel_john]

theoretical pressure curve
Actually the curve will have high frequency noise superimposed upon it. A low pass filter, move the transducer up and down a straight section until the minimum noise is observed. I don't want to say look for a node line... The ear is good frequency analyzer and the entire event can be heard with a mechanics stethoscope.
Notice the pickup has two piezo's, and they oriented opposite each other, side to side motion will be cancelled somewhat, since piezo's have polarity. but the expanding tube will produce plus or minus voltages in both at the same time.

[consaka]

Originally Posted by diesel_john

theoretical pressure curve
Actually the curve will have high frequency noise superimposed upon it. A low pass filter, move the transducer up and down a straight section until the minimum noise is observed. I don't want to say look for a node line... The ear is good frequency analyzer and the entire event can be heard with a mechanics stethoscope.
Notice the pickup has two piezo's, and they oriented opposite each other, side to side motion will be cancelled somewhat, since piezo's have polarity. but the expanding tube will produce plus or minus voltages in both at the same time.

I need a setup like that for running a timing light on my diesel. My custom engine has no real way to set timing. I built it from a combination of parts from different era engines, such that there were no timing marks for the pump. It is working well but I just set it by ear. It has been bugging me to know what the timing really is.
As for mileage of a mechanical pump the only way you are going to do that is with a flow sensor like your zemco. The return line will be a problem because not only does it take the bleedby from the injecters(most of these injectors are a machined tolerance fit(no orings)) but it also takes bypass fuel from the pump to keep it cool. If you bring your return line back and T it in front of the zemco(engine side) you will need to cool it with a transmission cooler or something. The volume of the cooler should not affect accuracy of the zemco at all as long as it doesnt have an air pocket. Even with an air pocket only the instant mpg my fluctuate but the rest would be fine.
Also if you operate in very cold climate you might need a heater in your fuel tank to keep the diesel from gelling. The return line also helps warm the tank a bit in cold weather.
To do this right you really need two flow meters similar to the zemco with a logic board that subtracts the return line pulse's from the supply pulse. Unfortunatly I am not an electronics engineer so I am not sure how to design the logic board or it would be done already.

[diesel_john]

we have been kicking the diesel angle around in this thread
Diesel instrumentation - Page 4 - Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com
In fact i put together a prototype for a timing trigger yesterday, you might need to attenuate the signal a bit with a couple Meg pot until you just have one flash for each injection pulse.
Parts: 2 used GM knock sensors, one 1/4" FPT x 1/4" FPT pipe coupling.
Tools: a drill bit the size of my injection line, a 1/4" pipe tap, hack saw or dremel.
Procedure: drill a hole in the center of the coupling, thread the coupling deeper on both ends until the knock sensors will reach in to the hole, cut one side of the coupling out until it will fit over the injection line.
Place coupling on a straight section of the injector line.
Hand tighten the knock sensors until they clamp on the injection. Wire the sensors in parallel. Attenuate signal as needed. see pic

[max_frontal_area]

love it! knock sensors, which GM cars and models can they be found in?
i have been in need to get a tach on my rabbit but swapping one if you can locate it requires jumping trough too many hoops

[diesel_john]

for a tach its much easier than that. just get a crank position sensor or cam position sensor off any late gas engine, or even an ABS sensor from the differential of a ford pickup. mount it so it detects the bolts on your crank sheave or water pump sheave, for a tach. or drive shaft bolts if you need speedo signal

[thecheese429]

Originally Posted by wyatt

...The fluid pushes up on a ball that is more dense than the fluid you are working with...

That sounds a lot like a ball flow meter like on a shielding gas regulator (for welding)(or oxygen I guess)


Like so:

[wyatt]

Exactly. Only in this case the working fluid is a liquid, but the concept is exactly the same. The only thing that needs done then is to figure out a means of measurement. I would suggest using a steel (able to be attracted by a magnet) ball in the tube, and wire coiled around the outside in several sets. The ball induces a current in the wire (like at a stop light) and the position of the ball can be determined.