Fuel injector pulse stretcher for E85 ethanol fuel conversion
rawelk
Posts: 2
A friend of mine was asking what I would do in the following circumstance, and it prompts my question.
He is thinking about converting one of his cars to use E85 (85% ethanol) fuel when it becomes available in our area. A Brazilian company makes an after-market conversion kit that will do the job, but it costs on the order of $250.
What it essentially does is intercept fuel injector firing pulses and monitors the car's ECM pulse width output, multiplies the “on” time by 1.27, and re-outputs the stretched pulses to the fuel injectors.
I mentioned using a Basic Stamp, and was envisioning something along these lines.
1.Put the ECM pulse output through a resistor divider. Tap the divider for a digital input signal to the BSD. I don't know what the maximum burden resistance would be and still provide enough loading on the ECM, but it seems fuel injectors comes in 4 ohm and 16 ohm (and perhaps other) varieties. I'm guessing a suitable brute force approach would be to use a pair of 6.8 ohm, 10 watt resistors for the divider (which would give the BSD a pulse voltage no more than ½ of the car's DC supply ... call it 7 volts nominal.
2.Monitor the ECM output for an injector firing pulse. When the leading edge of a pulse is sensed turn on a digital output (controlling a suitably sized transistor) to fire the fuel injector. Start a timing loop for that cylinder.
3.When the ECM pulse goes low note the timing loop accumulated value, and multiply by the 'stretch factor' (nominally 1.27 for E85 fuel), and save that value.
4.Continue the timing loop until the timing loop accumulated values => the calculated stretched value, then turn off the fuel injector output, reset the timing loop, and wait until the next ECM firing pulse.
There would be four pins set up for input looking at the 4 cylinder ECM output, and four pins set up for output going to the injectors.
5.I'm figuring it won't do to throw the BSD supply directly across the car battery, and we'd need a 7509 regulator or functional equivalent.
6.Haven't done much research on this, but different pulse stretchouts are required for different ratios of ethanol to gasoline – don't know if its a linear relationship or not. User 'tuning' for a particular fuel mixture could be done using five digital inputs (perhaps connected to a DIP switch) set up to provide binary weighting. For instance, “00000” is a multiplier of 1.00 (straight gasoline), “00001” is a 1.01 multiplier, “00010” is a 1.02 multiplier, etc. 5 bits is good for 31 states (so all DIP switches on is a multiplier of 1.031).
I've done a quick Google search, and have seen Basic Stamps used in various fuel injector control applications, but none doing a pulse stretchout for E85 ethanol fuel conversion similar to that described.
Was wondering if anyone on the forum has experimented in this area, or has a URL to someone who has.
Thanks,
Bob
He is thinking about converting one of his cars to use E85 (85% ethanol) fuel when it becomes available in our area. A Brazilian company makes an after-market conversion kit that will do the job, but it costs on the order of $250.
What it essentially does is intercept fuel injector firing pulses and monitors the car's ECM pulse width output, multiplies the “on” time by 1.27, and re-outputs the stretched pulses to the fuel injectors.
I mentioned using a Basic Stamp, and was envisioning something along these lines.
1.Put the ECM pulse output through a resistor divider. Tap the divider for a digital input signal to the BSD. I don't know what the maximum burden resistance would be and still provide enough loading on the ECM, but it seems fuel injectors comes in 4 ohm and 16 ohm (and perhaps other) varieties. I'm guessing a suitable brute force approach would be to use a pair of 6.8 ohm, 10 watt resistors for the divider (which would give the BSD a pulse voltage no more than ½ of the car's DC supply ... call it 7 volts nominal.
2.Monitor the ECM output for an injector firing pulse. When the leading edge of a pulse is sensed turn on a digital output (controlling a suitably sized transistor) to fire the fuel injector. Start a timing loop for that cylinder.
3.When the ECM pulse goes low note the timing loop accumulated value, and multiply by the 'stretch factor' (nominally 1.27 for E85 fuel), and save that value.
4.Continue the timing loop until the timing loop accumulated values => the calculated stretched value, then turn off the fuel injector output, reset the timing loop, and wait until the next ECM firing pulse.
There would be four pins set up for input looking at the 4 cylinder ECM output, and four pins set up for output going to the injectors.
5.I'm figuring it won't do to throw the BSD supply directly across the car battery, and we'd need a 7509 regulator or functional equivalent.
6.Haven't done much research on this, but different pulse stretchouts are required for different ratios of ethanol to gasoline – don't know if its a linear relationship or not. User 'tuning' for a particular fuel mixture could be done using five digital inputs (perhaps connected to a DIP switch) set up to provide binary weighting. For instance, “00000” is a multiplier of 1.00 (straight gasoline), “00001” is a 1.01 multiplier, “00010” is a 1.02 multiplier, etc. 5 bits is good for 31 states (so all DIP switches on is a multiplier of 1.031).
I've done a quick Google search, and have seen Basic Stamps used in various fuel injector control applications, but none doing a pulse stretchout for E85 ethanol fuel conversion similar to that described.
Was wondering if anyone on the forum has experimented in this area, or has a URL to someone who has.
Thanks,
Bob
Comments
For example - If you have 6000 RPM [noparse][[/noparse]at red line] and 8 cylinders, that would come to 48000 pulses / minute. 48000/60 would give you roughly 800 pulses per second. That would be relatively slow for a microprocessor. The accuracy of your microprocessor would increase with its clock speed. The BS24p would crunch a better number than the BS2.
The 7809 [noparse][[/noparse]a 9 volt positive regular] would help pre-regulate the power. Filter capacitors and an EMI in front of it for spikes caused by turning on and off brake lights, hitting high and low beam on the headlights, would be helpful. Some guys with more knowledge than myself might say that is really unnecessary and refer you to one chip that will go it all.
I guess you could simply have the BasicStamp detect the leading edge and immediately transmit an extended pulse of any value you will 1.27, 1.28, or so on. IF the relationship is purely linear it would merely be a matter of finding a way to compensate for any delay. I imagine that there might be a mechanical adjustment similar to advancing and delaying a traditional automotive distributor.
As you mention, a non-linear relationship begins to get a bit hairy. If the relationship is not linear, you may have to monitor your engine performance under a simulated load in order to create a curve - you may require a different curve than the original. You may also need to advance the ignition for slower burning fuel. I don't think you can depend on fudging one number or on purely textbook examples as you really are customizing an ignition system and it would be best to fit it to the engine in question via observation rather than generalize. Additional loads [noparse][[/noparse]such as power steering, power brakes, and air conditioning] may factor into the whole.
Even if you could get a simplified hack to work, would it be fuel efficient and polution friendly? Race car drivers don't have to worry about both and drag racers worry about neither. But, everyday drivers usually want to convert to save money while avoiding getting into trouble with the law. It is a more complex challenge.
Do you have an oscilloscope available? It would make the whole project much easier. You would have real data, not a bunch of guesses. It could also verify you power supply is clean. An automotive engine diagnotic computer wouldn't hurt either.
You might just forget about this 'pulse extending scheme' and just create a BasicStamp fuel injector control according to more basic principles. It would just be tuned according to the different fuel. In other words, don't get hung up on one number. It may be an advertising gimick to 'sex up' a rather generic product.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
Personally, I'd stick with a company you can go back to for warranty stuff...but that's just me!
The hobbyist in me likes to explore things like this....however, for a $250 black box that does the magic for me, and is hopefully warrantied/gauranteed etc....well, I'd hate to kill something in my $30,000 vehicle because I was trying to save a buck.
A BS2sx is $60...say $75 after tax and shipping. Add additional circuitry to protect your stamp and to 'masage' the input/output at about $30 (guesstimating).
So, you're at $100 before you've put your time in it....
You'll probably need to feed something to your ECM....and hopefully you won't need to change your O2 sensors....but a dead ECM will cost you $800 (depending on car) and an O2 sensor is a few bucks too.
If this is a beater car, go for it....but I'd caution you to not kill your daily driver!
Good luck!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Also, I believe that the stoichemetric fuel mixture for E85 is probably different than gasoline, and thus the fuel mixtures will be off, and require correction the ECM. E85 vehicles have a fuel sensor to detect the ratio of gasoline/ethanol and then can compute the appropriate PWs. I believe that ignition timing is also adjusted for E85...
Another consideration of the E85 conversion, is that the entire fuel system needs to be stainless steel/teflon, this includes the tank, lines, pump, fuel rails. Alcohol is corrosive. I bet that the conversion 'kit' doesn't include any of this!
It is generally not recommended to convert a gasoline based vehice to E85...
Rick
Post Edited (rockin_rick) : 3/1/2006 1:03:31 AM GMT
I'm pretty sure the car in question, while not a beater, isn't a high value vehicle, and the owner wouldn't be unduly upset if it was semi-trashed as a result of this experimentation. He has a good background in auto mechanics, and would be readily able to refit the car with SS tubing, et al., when the time came. I think the kit he was looking at is the one described at this website.
http://www.abcesso.com/index.html?lang=en-us&target=d26.html
Its worth reading if only for some of the unintentionally humorous mistranslations and phrasing into English.
I also read through the wiki article ( http://en.wikipedia.org/wiki/E85 ) and several others, and figure the Abcesso kit can't be doing much more that stretching out the injector pulse- the way it's wired it couldn't be doing much more.
Rick's point about exceeding the fuel injector duty cycle design rating is well taken, and something I was't considering. If they are in fact designed to work at 70 to 75% output then multiplying the pulse width by 1.27 would kick it up to the 89 to 96% range, and thats probably too close to the edge.
Bob
Seems a little too easy. They don't mention the need to convert fuel tank and supply lines and they maintain that the plugs don't need changing either. Most likely your car will run fine for a while and it won't hurt the basic functionality of the engine. But you will need to change injector plugs either more often or after you destroy your original set and go to alcohol rated ones.
It would be nice to switch back and forth if you are in an area with 'problematic fuel distribution'. But, everything would still need to be rated for the tougher use.
Sadly, I suspect the automotive manufacturers have it pretty well thought out in terms of how to make more in the after market on conversion. Tanks, plugs, ignition control, ....
The real question is how smooth you want it too run, or how cheaply you want it to run. I had a '51 Chevy PU that got 10mph and leaked a lot of oil, but it only cost $300USD and would stand tremendous abuse. Best of all, it was paid for. No job, just park it.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
I was up in Northern BC and came across a guy in an old dodge ram stuck on the side of the road in the middle of nowhere!
Poor bugger just picked up the truck for $500...came with an external fuel tank in the bed for those long trips.· Well, the fuel was sitting in the external tank too long and it either 'staled' or picked up too much water from condensation and he was basically stuck (having exhausted his main tank).
Anyhow, got him going....but if you run dual fuel vehicles (whether different fuels or just different tanks) be aware that you shouldn't drain one completely and then switch to the other.
In the gov't we had to do the PR/PC thing and get vehicles with natural gas as well as regular gasoline.· Problem was, the vehicle started on gasoline and once running warm, would switch to natural gas (if that was the last option selected).· Well, if you ran the gastank dry, you couldn't restart the car once you stopped it!· Haha....oops!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."