DIY_EFI Digest Sunday, 8 December 1996 Volume 01 : Number 375 In this issue: Re: Engine braking???? Re: Flywheel/Encoder patterns Re: Injector duty cycles Re: Flywheel/Encoder patterns Re: Flywheel/Encoder patterns Re: Flywheel/Encoder patterns See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- From: Mazda Ebrahimi Date: Sat, 07 Dec 1996 11:14:21 -0500 Subject: Re: Engine braking???? talltom wrote: > > A friend of mine has a 96 chev 3/4t 4x4(with a 350) and just got back > from > a hunting trip. This guy was a GM line mechanic for 20 years and > has owned his own shop for over 15 years, and he insists that his > new truck does something so that once the brake is tapped going downhill > it will not under any condition gain speed. Neither he or the other > mech can figuire out why, but both swear to it. > My question is is it possible that the guys at ac-delco have rigged > things to retard timing to the point that the compression stroke > would be infringed up on as necessary for deceleration??? Makes me > wonder if some wise ass has done a little abs system interfacing here???? > > I realize that some of the company guys have non-disclosure agreements > to deal with, but I'd appreciate as much insight into this as is > pheasible! Hi Talltom, That is strange. My company installs aftermarket alternative fuel systems on vehicles, and the Chevy P/U is a popular vehicle. We have a fair amount of experience with that truck, since we certified the aftermarket system for EPA on the 1996 MY. I have logged about 3,000 miles behind the wheel of about a dozen of these trucks with 4.3L, 5.0L, and 5.7L engines (mostly C-1500 and a couple of C-2500s), and haven't noticed anything special about deceleration. In fact, two things stick out in my memory that may or may not help you: At speeds lower than approx. 45 MPH, briefly after entering decel, engine RPM drops to about 1100 RPM, so there can't be a whole lot of engine braking going on. The second thing is that if the engine is warmed up, and you are in closed loop, the ECM conducts EGR system tests, in which it cycles the EGR valve and looks for changes in MAP to verify its operation (required by OBD-II). This goes on at other engine operating points also, but its fairly strong at decel close to idle. Hope this helps, Mazda ------------------------------ From: "Jeffrey Engel" Date: Sat, 7 Dec 1996 11:01:02 +600 Subject: Re: Flywheel/Encoder patterns Mazda Ebrahimi Jeffrey Engel wrote: > > > > > Subject: Re: Flywheel/Encoder patterns > > > From: Kalle Pihlajasaari > > > > > Hi There, > > > > An encoder giving absolute crank position would implicitly supply an > > interrupt when the position changes value. Reading crank position directly > > means you wouldn't have to fool with timers or counters to derive it. You'd > > also know exactly what the position is upon start-up without having to store > > the last known position info upon shutdown. No way around the cam sensor tho > > that I can see. > > The only real world advatage I see to the encoder is quicker starting with lower > emissions. Updating the degree count takes very little time in the processor. We still > have to employ a timer to calculate RPM, right? > Actually, the above is far too sage to have been written by me. I got into this discussion by remarking that Grumpy Jenkins found enough crank flex (at 650+ HP) to make sub-degree measurements useless. Of course he had to deal with crank AND cam flex because the distributor is at the back of the engine. Someone a while back made the point that burning is a function of time, not crank angle. Would it be practical to find TDC of the cylinders and use a time offset? je jengel@xxx.net ------------------------------ From: Mazda Ebrahimi Date: Sat, 07 Dec 1996 11:47:34 -0500 Subject: Re: Injector duty cycles Roger Feingold wrote: > > I'm working on sizing injectors for a '88 2.8 V6 MPFI. Does anyone have > experience with the max duty cycle when at WOT and redline? My computations > show stock 16 lb/hr injectors will be maxed out by the time the engine hits > 5000 rpm and 12.6:1 A/F. > Also, has anyone correlated the O2 sensor voltage vs. A/F ratio? > > Thanks, > > Roger.I had always been told for proper injector sizing, you don't want to exceed 80% duty cycle at WOT, max RPM. But someone on the mailing list said their injectors on a 3.8L turbo GN go upto 99% at WOT. If you calculated your fuel reuiqrement based RPM and displacement, did you consider maximum Volumetric Efficiency for that engine is less than 100%? I think that engine makes about 145 HP at 4600 RPM (these numbers may be off a little, but lets use them for sake of argument). at 145 HP, the engine should be consuming approximately 15 lbs of air per minute (I got 15 by dividing 145/10 and rounding up. It's an approximation) (or 900 lb/hr). Given 12.6 A/F, you need 71.4 lb/hr of fuel, which divided by six cylinders results in 12 lb/hr injectors. Given you have 16 lb/hr injectors, they shouldn't see much more that 75% at normal conditions (maybe higher when ambient air temp is cooler), which comply with the 80% rule of thumb. Do you know what the actual duty cycle for those injectors is? Best Regards, Mazda Best Regards, Mazda ------------------------------ From: Mazda Ebrahimi Date: Sat, 07 Dec 1996 11:56:23 -0500 Subject: Re: Flywheel/Encoder patterns Jeffrey Engel wrote: > Actually, the above is far too sage to have been written by me. I > got into this discussion by remarking that Grumpy Jenkins found > enough crank flex (at 650+ HP) to make sub-degree measurements > useless. Of course he had to deal with crank AND cam flex because > the distributor is at the back of the engine. > > Someone a while back made the point that burning is a function of > time, not crank angle. Would it be practical to find TDC of the > cylinders and use a time offset? > > jeI'm happy to get everything right at one degree or less error. I think its proven that the method you are saying works well, but my motive for going to 180 teeth wasn't just better accuracy. I did it to simplify the required programming. by having one pulse per degree, you no longer have to estimate when the next event is comming up by calculating the time required at a given engine speed to reach the next coil or injector firing. When the degree count is advanced to the right number, you turn your device on. Best Regards, Mazda ------------------------------ From: Todd Knighton Date: Sat, 07 Dec 1996 08:57:39 -0800 Subject: Re: Flywheel/Encoder patterns Mazda Ebrahimi wrote: > > As soon as I get a little more involved in actual programming, I'll come up with a lot > of questions for you. Regarding the length of the interrupts, this is the way I'm > planning things right now: I'll be running a foreground loop with no time driven tasks, > which constantly calculates fuel and spark values based on latest readings. The result > of the most recent completed calculations are four numbers: inj_delay, inj_ontime, > ign_delay, ign_ontime (dwell). When an interrupt is generated, the following happens: > -advance degree count, > -check for reference count (i.e. at this degree count, do I have to load any > counters? This only happens 8 times a rotation (v8)). > -if reference is reached, load decoder with external counter value; > -return > this is in addition to calculating RPM. > > I'm running a 16 bit 32 MHz processor. Sounds like you should be fast enough, don't know if your flywheel sensor will have high enough resolution to work at any high speeds, a hall effect would work better. How big are your counters, 16, 26, 32-bit. Running at 32Mhz you have to make sure your counters won't overflow at low speed, this is why Porsche stayed with 6Mhz with 16-bit counters, any faster and you'd be overflowing the counters all the time waiting for the next tooth. To get things going you could just setup some constants initially, Coil On tooth, Coil Off Tooth, Injector On Tooth, Injector Off Tooth with some basic injector times and ignition timing just to get the thing fired to work out your hardware. Then start adding in functionality as you go. Todd Knighton Protomotive Engineering ------------------------------ From: Mazda Ebrahimi Date: Sat, 07 Dec 1996 14:37:33 -0500 Subject: Re: Flywheel/Encoder patterns Todd Knighton wrote: > How big are your counters, 16, 26, 32-bit. Running at 32Mhz you have > to make sure your counters won't overflow at low speed, this is why > Porsche stayed with 6Mhz with 16-bit counters, any faster and you'd be > overflowing the counters all the time waiting for the next tooth. All my counters are sixteen bit. But remember, the counters used for driving the injectors and coils are not processor clock driven. They are driven by the crank pulse, so their count would never exceed 719. > To get things going you could just setup some constants initially, Coil > On tooth, Coil Off Tooth, Injector On Tooth, Injector Off Tooth with > some basic injector times and ignition timing just to get the thing > fired to work out your hardware. Then start adding in functionality as > you go. Sounds like a good plan to me. Thanks! Best Regards, Mazda Ebrahimi ------------------------------ End of DIY_EFI Digest V1 #375 ***************************** To subscribe to DIY_EFI-Digest, send the command: subscribe diy_efi-digest in the body of a message to "Majordomo@xxx. A non-digest (direct mail) version of this list is also available; to subscribe to that instead, replace "diy_efi-digest" in the command above with "diy_efi".