DIY_EFI Digest Wednesday, 5 June 1996 Volume 01 : Number 156 In this issue: re: MAF Conversions re #3: MAF Conversions re: MAF Conversions MAF vs MAP Re: MAF Conversions Re: MAF vs MAP Re: MAF vs MAP How about no MAF or MAP? re: MAF vs MAP See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- From: SRavet@xxx.com Date: Tue, 4 Jun 96 12:40:30 CDT Subject: re: MAF Conversions RABBITT_Andrew@xxx.au Wrote: | | 'Everyone' talks about MAF conversions of existing EFI systems but can | anyone answer me a few questions about the whole exercise? | | 1- what does a MAF sensor offer that a MAP and inlet air temp sensor | don't? (or what do people think it offers?, ie what are the main | reasons for the conversion?) MAF adapts better to engine mods. SD systems use a pre-calculated volumetric efficiency for your engine, while a MAF system actually measures the air entering the system. An SD system can compensate somewhat for changes like headers or a new intake, but only when it is in closed loop mode. Someone else asked: | But you can calculate Mass Airflow from manifold temp, air density and | engine speed. Why is this so inferior? The engine speed part is the limitation. If you make an engine mod that increases your rev limit, or changes your VE at any rpm, you have to recalibrate the look-up table. A MAF based system would just measure the increased flow. Closed loop operation will see a lean condition after these mods, and adjust accordingly, but that won't help your WOT, when the lookup table is used by itself. | | 2 - are analogue (voltage) output devices used more commonly than | frequency output? | | 3 - how are they normally sampled in these conversions, synchronously | at a determined crank angle, synchronously at any angle at all, or | asynchronously? | | 4 - how do you know that the airflow you're measuring is actually | there? Do people just slap them on and hope, or is some science | applied to make sure inlet manifold resonances are not upsetting the | signal? | | 6 - what kind of manifold filling compensation do people use? I think usually they are put far enough upstream that they don't see any intake effects. Isn't that one purpose of the plenum? Some of the Ford people here had talked about modeling the intake using differential equations or something, but that seems a little complex for a DIY project. At least, it is for *my* DIY project. Can anyone here recommend books about control system theory? Books aimed at the experimenter, that is, not textbooks. I had a controls class once, I won't mention what my grade was. Let's just say that I barely didn't have to repeat it. Surely some has written an easier-to-understand book on the topic. | | Andrew Rabbitt | Orbital Engine Company | | | | - --steve Steve Ravet sravet@xxx.com Baby you're a genius when it comes to cooking up some chili sauce... ------------------------------ From: pfenske@xxx.ca (peter paul fenske) Date: Tue, 04 Jun 1996 14:24:22 -0700 Subject: re #3: MAF Conversions >>1- what does a MAF sensor offer that a MAP and inlet air temp sensor >> >>>MAF directly measures Air Mass. Fueling is simply based on AF ratio >>>desired not extensive engine mapping. In other words a change in >>>engine parms does not require a recalibration within limits. > >But you can calculate Mass Airflow from manifold temp, air density and >engine speed. Why is this so inferior? > >>N = PV / RT is the map calculation. The fact is even using lookup >>and base pulse fueling many adjustments are required. Digital systems >>are not continuous but discrete. Thus while theoretically you could >>have map work perfectly it requires a great deal of uP HP. As for >>responding to changes, speed density is based on a load percentage >>as related to manifold pressure. WOT has the highest pressure, idle >>the lowest and fueling is based on where the engine is. IE base pulse >>is modified by %load. It is this calculation that gets mucked up >>when you change engine cam. With a lower vacume ,higher pressure PW >>increases which results in too rich a mix. Thus the lookup table or >>slope and base of the %load has to be changed when engine parms change. >> >>>MAF is continuous. Just like MAT or CTC. >> >What do you mean by continuous? Most ECU's are digital and require >some form of ADC, therefore you have to decide where (time-wise) to >sample the sensor output. > >>Guess its late. Oh well. More specifically response time is the key. >>Mass airflow devices are either vanes or hot wire. These devices respond >>relatively slowly to fluctuations and will average the airflow. >> >>6 - what kind of manifold filling compensation do people use? >> >>>Usually taken care of in PE corrector. sori not ve in maf. >> >VE? Volumetric Efficiency? What sort of corrector/algorithm? > >>VE tables are part of the speed density system. The base pulse >>gets modified by a value in a lookup table indexed to RPM. >>In a MAF system there is an enrichment table. This allows some >>variance in A/F ratio with changes in RPM. > >re: airflow fluctuations, I've data from a 1.8l 4-cyl that I'm working >on that at one speed-load point has airflow varying from 20g/s to >70g/s at twice crank speed! This would not be uncommon. How have >people overcome this in MAF conversions? > >>Once again instantaneous vs average airflow. You are interested in >>the average airflow albeit over a short time period. >>As for reversion, ram tubes, runners whatever usually run from the port >>to a plenum. The throttle body feeds the plenum and the MAF feeds the >>throttle body. Thus the MAF is both far removed from the valve and >>plenum airflow tends to average individual cylinders. If you are worried >>about manifold pulsing look at the output of a MAP sensor sometime. >>The AD data from a MAP has a first or second order digital filter in >>software usually. > >Anyways if you want SAMS has a good book out on engine controls. In addition >there are many SAE papers on the subject. >>BEEN A PlEASURE: peter ------------------------------ From: Jeffrey.Giberstein@xxx.EDU (Jeffrey Giberstein) Date: 04 Jun 96 19:48:51 EDT Subject: re: MAF Conversions - --- You wrote: I had a controls class once - --- end of quoted material --- My roommate had one thrice. I'm waitng for my grade... Cheers ------------------------------ From: Michael Kent Date: Tue, 4 Jun 1996 18:03:36 -0700 (PDT) Subject: MAF vs MAP The biggest reason to go to a MAF system is simply its flexibility. One can make numerious changes to a motor, dramatically altering it's breathing abilities and still run quite well. Change the characteristic of a speed-density motor beyond it's adaptive abilities and risk loosing the motor to lean conditions etc. Talking with a calibrator from Ford, he mentioned numerious problems with the MAF system, including non-linear output from the meter, reversion (backflow), leaks down stream, and contamination. All in all though it was a much better system than the speed-density they used to run since it allowed the EEC-IV to adapt better to the widely varying motor's coming out of the factory. MK ------------------------------ From: FIScot@xxx.com Date: Tue, 4 Jun 1996 22:06:04 -0400 Subject: Re: MAF Conversions In a message dated 96-06-03 21:13:36 EDT, you write: >'Everyone' talks about MAF conversions of existing EFI systems but can >anyone answer me a few questions about the whole exercise? >1- what does a MAF sensor offer that a MAP and inlet air temp sensor >don't? (or what do people think it offers?, ie what are the main >reasons for the conversion?) I am not sure of the conversions you are talking about, but the MAF is a much more accurate way to measure airflow. The MAP sensor equipped engine relies on a VE table, temp sensors and the MAP to *calculate* airflow. It can work quite well, and the MAP sensor has the advantage of being much cheaper than a MAF sensor. But the MAF is a better way. If an engine modification alters the VE of the engine, the VE table in the ECM is now incorrect. The MAF really doesn't care about VE, it just measures airflow. >2 - are analogue (voltage) output devices used more commonly than >frequency output? I have no idea which is used more. The freq type may have some advantages. >3 - how are they normally sampled in these conversions, synchronously >at a determined crank angle, synchronously at any angle at all, or >asynchronously? Good question. The injected fuel quantity (indicated by injector PW) for a cylinder should be matched to the air that the cylinder took in, and at the desired AFR. If one measured the number of freq style MAF pulses between spark reference pulses, you would know the amount of air taken in per ref pulse, which would then be the amount of air for one cylinder.. One might also use a multi-toothed crank wheel, and measure the amount of wheel travel against MAF pulses. In using a DC type MAF, I guess a running average might give you an idea of the airflow. >4 - how do you know that the airflow you're measuring is actually >there? Do people just slap them on and hope, or is some science >applied to make sure inlet manifold resonances are not upsetting the >signal? I have no idea there. >5 - How about reverse flow? Is this any concern in the average >conversion? The MAF can measure airflow both ways. It may not be linear in the reverse direction though. This may be a problem if you use a cam with a lot of overlap that causes reversionary pulses. >6 - what kind of manifold filling compensation do people use? I guess you are refering to transients? By monitoring the TPS, you could cover those by adjusting fuel to suit the TPS delta. The MAF could also be monitored for a delta, indicating the need for more fuel. Scot Sealander FIScot@xxx.com ------------------------------ From: Jeffrey.Giberstein@xxx.EDU (Jeffrey Giberstein) Date: 04 Jun 96 23:25:13 EDT Subject: Re: MAF vs MAP - --- Michael Kent wrote: The biggest reason to go to a MAF system is simply its flexibility. One can make numerious changes to a motor, dramatically altering it's breathing abilities and still run quite well. Change the characteristic of a speed-density motor beyond it's adaptive abilities and risk loosing the motor to lean conditions etc. Talking with a calibrator from Ford, he mentioned numerious problems with the MAF system, including non-linear output from the meter, reversion (backflow), leaks down stream, and contamination. All in all though it was a much better system than the speed-density they used to run since it allowed the EEC-IV to adapt better to the widely varying motor's coming out of the factory. MK - --- end of quoted material --- Correct me if I'm wrong, but both GM and Ford "MAF" systems still run a MAP sensor to deal with situations that the MAF can't handle (transients, I presume). Jeff Giberstein Dartmouth Formula Racing ------------------------------ From: "S. McManus" Date: Tue, 4 Jun 1996 20:52:42 -0700 (PDT) Subject: Re: MAF vs MAP > (backflow), leaks down stream, and contamination. All in all though [MAF] > was a much better system than the speed-density they used to run since it > allowed the EEC-IV to adapt better to the widely varying motor's coming > out of the factory. > > MK > Maybe I'm confused or jumping to conclusions here, but is this implying that the quality control from the Blue Oval factory is so bad that the maps for a SD system would have to be individually tailored to EACH car coming from the factory to keep A/F correct? Thus, a MAF system was needed to make up for manufacturing tolerences? Ed H, any insight? Sean McManus University of Washington Formula SAE ------------------------------ From: "Tony Bryant" Date: Wed, 5 Jun 1996 17:10:24 +1200 Subject: How about no MAF or MAP? My pet project at the moment, is a EFI (& Elec. Ign.) system for my Datsun 510 club race car. I've taken one look at your average flap type mass airflow sensor (280ZX), and decided that 1) It will severly restrict air flow 2) The plumbing problems are a real hassle (same applies to hot wire types) 3) The extra plumbing makes the second intake harmonic unusable. Since I'm using twin dellortos as the throttle bodies, there is no plenum chamber to take a stable manifold vaccum reading off. So MAP is out. I've decided therefore to use a throttle position vs. engine speed calculation. A map based lookup is a problem, because it is hard to set while driving(!), and resolution is limited due to ROM space. The solution I have used is to solve two simultaneous equations to find the air flow: AirFlow = M.A.Pressure * VE(RPM) * RPM { Just like in a MAP system } & AirFlow = CriticalFlow(Throttle Pos) * compressible gas function(M.A.Pressure) I have no false hope about getting emissions quality control from this scheme, but: 1) Since New Zealand is a clean green country , we don't have such silly pollution laws (yet :-). 2) It's (mostly) a race car. 3) I want everybody to smell my Avgas. The question I have, is does anybody have a general equation for the critical flow for a given throttle position, with real co-effecients? p.s. It is now running, but without knowing the critical flow above closed throttle, the mixture at part throttle is "interesting". p.p.s Has anybody here had any experience with measuring mixtures by measuring peak spark voltages? Any comments? *********************************************************** * "Insanity is the only sane response to an insane world" * * >> bryantt@xxx.nz << * *********************************************************** ------------------------------ From: RABBITT_Andrew@xxx.au Date: Wed, 5 Jun 1996 13:44:08 est Subject: re: MAF vs MAP - --- Jeffrey Giberstein wrote: Correct me if I'm wrong, but both GM and Ford "MAF" systems still run a MAP sensor to deal with situations that the MAF can't handle (transients, I presume). Jeff Giberstein Dartmouth Formula Racing - --- end of quoted material --- MAF sensors can handle all the transients you can throw at them provided they're set up properly, ie, they're sampled correctly and you have some form of manifold filling compensation. I suspect that the additional MAP sensor is for added functionality, perhaps EGR control or maybe something else. Andrew Rabbitt Orbital Engine Company ------------------------------ End of DIY_EFI Digest V1 #156 ***************************** 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".