DIYEFI.org Forum

http://www.physicsforums.com/showthread.php?t=182862

An interesting discussion of AFR with relation to different fuels etc.

Fred.

I had some thoughts about the fuel / injection. I'm wondering if the injection pulse is intended to be synced with the tach? Such that if you pointed the injector right at the valve, you could inject directly in the cyl, when the valve is open, reducing the need for XTau or equivalent. I think that if we intend for this, it will make many other parts much simpler and more robust.

So assuming the valve is open for half of one cycle, you would have about 5 ms to inject directly in the cyl. This would reduce heating the fuel, decrease the response, ect.

If you size this way, would you have a chance of making it idle? I think that it's likely, should be able to maintain a good 1ms on time for about 1k RPM, but would the error brackets get to large to handle nicely. Not really sure, but gut feel tells me theres a good chance it would work all around.

Hmmm any how just a thought, hope the above made some sense.

A low power engine might idle, but I'd need something like 4000cc/min injectors to make my power and do it that way.

A big part of the accel issues in ms are due to lack of timed injection. Whether it falls on the open or closed valve is less important than how often the updates are and how consistent the mixture is between cyls due to timing differences with batch injection. Timed injection is right at the top of the list of things to do. This thread is only about how much to inject, not when :-)

Start a new one about sequential injection and/or accleration/transient/delta enrichment if you like.

Fred.

this just came up again on msextra...

How do you all feel about referencing the map to barometer and having an altitude correction based on barometer reading? Have we already hashed this out?

From an old Nines article (Saab owners mag):
http://web.archive.org/web/200512141818 ... tm#physics

Old school 'how to deal with altitude'. It appears that if we use the previous eqns, but reference the barometer (map/barometer) for the equations and add a correction graph to lean out the mixture as oxygen decreases?

Gearhead

As shown in Table 1, “dry air” is a mixture of nitrogen, oxygen, argon, carbon
dioxide, etc., which are all, practically speaking, perfect gases, and so obey (4). The
composition of dry air is nearly homogeneous below 20 km. Except for water vapor and
ozone, whose concentrations vary greatly, the concentrations of the other principal
constituents of the atmosphere, i.e., N2 , O2, Ar, CO2 , Ne, He, Kr, H2 , CH4 , and N2 O,
are nearly homogeneous up to an altitude of about 80 km.

http://kiwi.atmos.colostate.edu/group/d ... nOfAir.pdf

I'd take that article you linked with a grain (or ocean) of salt.

There are already multiple options for it in the code, some finished, some not, but it will be taken care of. As usual, it is modular enough to make it easy to change the way we go about it or indeed add new ways.

Fred.

I do not know about grain of salt, but it is more 'home remedy' than pure science.

One thing I did note was that Saab deals with this on their EMS by adaptation. It has smart adaptation which is persistent. As altitude goes up, it leans it out to still achieve stoich. It reverses this as it goes back down. No equations, just feedback controlled adaptation. Just as it does with timing based on knock as octane changes.

Gearhead

I think it's important to have the option of an open loop solution regardless.

everything you ever wanted to know about humidity:
http://www.sensirion.com/en/pdf/product ... dity_E.pdf

Thanks, I skimmed it now and will have a more in depth read when I get a chance :-)