DIY_EFI Digest Tuesday, 3 September 1996 Volume 01 : Number 258 In this issue: EFI332 home page Re[2]: Newer VATS Vehicles Jump starting an EFI car? Re: Ancient History Re: DIY_EFI Digest V1 #256 Re: Ancient History INFO to develop a software K_knock sensor Data/Spec. Sheets Re: Ancient History, and Djet and cams ISO-9141 Re: FORTH Re: Ancient History Really long water injection post, $0.04 worth Re: knock sensors, ancient history Re: Ancient History Cams and FI Manifold/throttle bodys-ED:-) Bosch d disease Re: How tall is tall? Re: d disease Mercedes mechanical Re: FORTH Bosch to Hitachi MAF Re: K_knock sensor Re: Throttle Body Sizing Re: Bosch d disease See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- From: David Redpath <101513.357@xxx.COM> Date: 03 Sep 96 05:13:26 EDT Subject: EFI332 home page Is the EFI332 home page operational ? ------------------------------ From: dzorde@xxx.au Date: Tue, 03 Sep 96 18:08:38 Subject: Re[2]: Newer VATS Vehicles Why not just carry a spare computer, unplug the original, plug the other in and off you go. ______________________________ Reply Separator _________________________________ Subject: Re: Newer VATS Vehicles Author: diy_efi@xxx.edu at INTERNET Date: 8/30/96 12:40 PM > What about the next year, after the thieves figured out how to get around it? > After all, if guys on the list can do it, so can anyone with enough > technical knowledge. Technical knowledge is great but it doesn't necessarily give you time to defeat it. The guys here that have cracked it have done it one of two ways. Either by trying each of 15 resistor values until it unlocked (with a 3-4 minute delay between each try) or by hacking the PROM which requires removing the computer, pulling the CalPak, removing the PROM, downloading it, editing it, blowing a new PROM and putting it all back together again. A thief just doesn't have the time to do all of this. He might be able to carry a spare CalPak with a modified PROM already in it but even if the thief decides to concentrate on a Corvettes there are about 8 different CalPaks he would have to carry. It still easier to use a tow truck or steal the keys. John Faubion jfaubion@xxx.net ------------------------------ From: Jeremy Pronk Date: Tue, 3 Sep 1996 21:09:41 +1000 (EST) Subject: Jump starting an EFI car? Hi, I am mainly a lurker on this list, but I should be getting a car with aftermarket EFI to play around with soon. Anyway, this morning the next door neighbour needed a jump start, and I remembered hearing somewhere that jumping starting an car with EFI can burn out the ECU? If so, what is safe method, and how does it actually get burnt out (I would like to learn)? regards, Jeremy, Sydney AUS '69 510 4dr ------------------------------ From: Dirk Wright Date: Tue, 3 Sep 1996 08:28:13 -0400 (EDT) Subject: Re: Ancient History On Mon, 2 Sep 1996, Robert J. Harris wrote: > > Liquid fuel does not burn - period. It must be vaporized before it burns. > All forms of carboration - including fuel injection - atomize the fuel > hoping enough vapor will be formed to start combustion so that the heat and > violence of combustion will finish the job. Power and fuel efficiency are > directly related to how well that is done. Um, sorry. Gasoline vapor explodes, it does not burn. Liqiud gasoline burns. Gasoline vapor ignited in a combustion chamber will cause detonation, which is highly undesirable. The spark ignites atomized gasoline droplets as they pass through the spark plug gap during the end of the compression stroke. The transformation of liquid gasoline droplets into hot exhaust gases by the process of combustion is what pushes against the piston during the power stroke. Gasoline vapor is not what is desired. What is desired is extremely fine gasoline droplets of uniform size. This is where fuel injection is superior to carburation. The droplets are more uniform with fuel injection than they are with carburation. **************************************************************************** Dirk Wright wright@xxx.gov "I speak for myself and not my employer." 1974 Porsche 914 2.0 "A real hifi glows in the dark and has horns." 1965 Goodman House **************************************************************************** ------------------------------ From: Dirk Wright Date: Tue, 3 Sep 1996 08:37:47 -0400 (EDT) Subject: Re: DIY_EFI Digest V1 #256 On Tue, 3 Sep 1996, Tony Bryant wrote: > > > > > How about the real mechanical FI systems? Like off a '69 Mercedes coupe. > No I'm not talking about K-Jet, I talking something that looks like a > diesel fuel pump/distributor. I've only seen it in a workshop manual. > > Anybody with any experience with these, care to comment? There was a SPICA system on my old '74 Alfe Romero that looked like that. I did figure out how to adjust it, but of course, I was too unimformed at the time to buy the shop manual. As I understand it, they use very high pressure to the injectors. Higher pressure would make for finer atomization if the nozzle is designed right. Of course, the BMW2002tii used Kugel-Fischer (sp?) mechanical injection. I'm guessing, but I think most electronic FI use lower pressures because it would take too much electrical current to unseat the pintle of the injector in order to open it. **************************************************************************** Dirk Wright wright@xxx.gov "I speak for myself and not my employer." 1974 Porsche 914 2.0 "A real hifi glows in the dark and has horns." 1965 Goodman House **************************************************************************** ------------------------------ From: Dirk Wright Date: Tue, 3 Sep 1996 08:41:40 -0400 (EDT) Subject: Re: Ancient History On Tue, 3 Sep 1996, Tony Bryant wrote: > > > > What I want to know, is what happens when you have radical > cam timing? This nice puddle of fuel on the inlet valve > gets blown back up the inlet duct when the valve opens? > Makes nice flames, if you don't run an air cleaner :-) > But this surely can't make for repeatable fuel volume supplied > to the cylinder? As Ralph pointed out, the vacuum goes down at lower rpm due to valve overlap. A vacuum responsive FI would make the mixture too rich for the engine to run. I think carburators are more tolerant of this low vacuum at low rpm because they are passive devices and because they are less precise in fuel metering. **************************************************************************** Dirk Wright wright@xxx.gov "I speak for myself and not my employer." 1974 Porsche 914 2.0 "A real hifi glows in the dark and has horns." 1965 Goodman House **************************************************************************** ------------------------------ From: prebuf@xxx.uy (Pablo Rebufello - CECAL - ) Date: Tue, 3 Sep 96 09:29 GMT-3:00 Subject: INFO to develop a software Hi! I'm new in this list. I got a degree in Computer Science, and my brother works as a mechanic. I'm planning to develop a software to manage and visualize the data (I/O) to full control the EFI, and everything else that could be read from a computer. But I need information about de EFI, and any electronical sensor. I know it depens on the car, but at least I have to try with just one at the beginning. What I need is: - - protocol of comunication from/to the devices. - - how the data is transmitted from/to the device, as well as description of the data to send/receive from/to the device to be read/write from/to the computer. I know I'm asking for a lot of information, so anything will help. Source of information such as books, internet address, etc will help much. Thank you!! Pablo Rebufello prebuf@xxx.uy URUGUAY - South America ------------------------------ From: cloud@xxx.edu (tom cloud) Date: Tue, 3 Sep 1996 08:59:47 -0500 Subject: K_knock sensor I don't do the race circuit, only want to go fast some times and just drive the others. The way I tune my toy is to set the advance until I get a little rattling on long hills on hot days (we have some of both of those) and back it off a little. So, Todd, or anyone else who cares to chip in, what's wrong widdat? tom cloud ------------------------------ From: cloud@xxx.edu (tom cloud) Date: Tue, 3 Sep 1996 09:14:21 -0500 Subject: Data/Spec. Sheets There seems to be a lot of ill-informed discussion on this group regarding various sensors, etc. I'm not putting anyone down, as I do the same thing. Seems like many of us suffer from the same malady -- a lack of data on the components in our systems. Can anyone shed light on where to get data on the various transducers used in EFI systems? (esp. EGA, MAP, MAF, etc.) I would like to find data on specific auto makers' components. They obviously have the data and I cannot see any reason why they would care if individuals have it -- can't be trade secret. Their engineers have to have access to it. It's gotta be published somewhere. Maybe Ed Hernandez (where'd he go?) could comment on where to get this data from Ford (and, yes, I do have a MotorSport book on EFI and Emissions, but it only gives 'typical' data). Can anyone help ? (Todd ?) Thanks, Tom Cloud ------------------------------ From: "Robert J. Harris" Date: Tue, 3 Sep 1996 08:04:45 -0700 Subject: Re: Ancient History, and Djet and cams Remember the Reichstag - ---------- > From: RD Rick > To: diy_efi@xxx.edu > Subject: Re: Ancient History, and Djet and cams > Date: Monday, September 02, 1996 10:42 PM > > Correcting myself here: fuel squirts at beginning of intake stroke on > two cyl, and beginning of exhaust stroke on the other two. > Note: injector duration at Op Temp, WOT, 6,000 rpm, is 8.3 ms, while > the valve is open only 5 ms. > Note for the timed/untimed discussion - the injector is pushing fuel for far longer than the valve is open. At this point, the question is moot. To have enough fuel to make power in this region, it must store fuel by puddling it on the valve and waiting for the inrush of air to finish mixing it. When it stores it there is of no consequence. The closer to the cylinder the fuel is injected the less the mixture problems. Sequential port fuel injection kicks butt over throttle body primarily because it generates a higher quality mixture that is more closely tuned to the engine. It starts with an unfair advantage because it eliminates all the complications and compromises of moving a gas/liquid mixture thru plumbing and instead deals with an air only system. Enough about timing - anyway works - get the mixture right and run. Take maximum advantage of the mechanical properties that you can. Leaves less for the control system to solve. GMC roots mechanical positive displacement 6-71 superchargers make monsterous horsepower yet the whole high pressure fuel air charge comes out of the blower totally asynchronous with regard to valve openings and cylinder filling and it has worked very well for 40 years of racing and street!!!!!. ------------------------------ From: Jens Knickmeyer Date: Tue, 3 Sep 1996 17:14:00 +0200 (MET DST) Subject: ISO-9141 Hi all! Are there chips available for the ISO9141 standard? Anything would be welcome, from pure electrical interfaces to controllers which have parts of the protocol on chip. Thanks, Jens ('92 VW Polo-G40) ------------------------------ From: David Redpath <101513.357@xxx.COM> Date: 03 Sep 96 11:14:12 EDT Subject: Re: FORTH I've seen a few postings recently about FORTH, I've done a fair bit of programming in FORTH, so as my first contribution here goes :- The microcontroller we used was an Hitachi H8, the original development board was from a company called Triangle Digital Services, they are in the UK and have a distributor in America I believe. Their development system came with lots of developed code, everthing from a multi-tasking kernel, Fast Fourier Transformation code, A-D software, multiple LCD screens, PWM outputs etc. It was very well supported, and might could well be a good basis for a EFI controller. ( I guess ). Also a book I bought called 'Forth - the new model' came with a disk with a FORTH compiler/interpreter on it. which run on a PC. So for learning Forth that might be a good starting point. ( I could found out the ISBN if required. ) Hope some of that is of some relevance. ------------------------------ From: Arnaldo Echevarria Date: Tue, 3 Sep 1996 12:02:01 -0400 Subject: Re: Ancient History >On every engine - a small portion of exhaust gas is pulsed into the >intake manifold the amount depending on cam timing, load, and engine >RPM. This is then sucked back into the cylinder diluting the charge. >For you old timers, this is why you must richen a carb when putting >headers on an engine - because the reversion pulse is damped - leaving >less flow thru the carb (the first pulse backwards - followed by the normal >forward pulse.) Untimed fuel systems have a fuel air mixture that is >pushed backwards up the manifold (reverted) before being sucked back >in. As this is very hot gas with no oxygen present (e.g. no flame!!) and >lots of CO2, and this hot gas is quenched by contact with relatively cold >(far below ignition temperature) matter in the intake tract - cold mixture >cold intake valve, cold head, all that happens thru almost all ranges of >performance is charge dilution and therefore less than theoretical power >is developed. Aside - a very rich fuel mixture such as present when >fuel is "puddled" on the valve is not combustible. A lean - i.e. >minimally vaporized mixture such as from a mistuned carb or timed >injector is explosive!! > >What timed fuel injection in conjunction with proper cam timing does is >allow this reversion pulse to go up and come back out the exhaust port >before adding fuel to the incoming air. Properly done, no air pump is >needed for the cat converter as the reverted air pulse shoves oxygen >out the exhaust. The chamber gets a full, dense properly mixed charge >and makes better power, fuel efficiency and lower emissions. > Could someone elaborate more on this? Excuse my ignorance, but I'm trying to learn more about combustion theory. I've been hot rodding cars for years now and have played with different carb / efi setups, but I'm trying to learn a little more on what "really" goes on (as in good techniques vs bad techniques of getting fuel to burn using injectors or throttle bodies etc). What I'm going to build (no laughs, please) is a fuel efficient pontiac 455 (If that ever existed) using some sort of fuel injection. Right now the car runs low 13's in the quarter and gets 12mpg (city) using a Quadrajet 750cfm; I want to improve that to 15-17mpg. I think the cheapest way to go is to use a tri-power manifold and hook up three 2barrel tbi's. I'm pretty good with building microcontrollers so now I'm just trying to decide a baseline by which the fuel injectors will fire (before tdc? during? after? during exhaust cycle too? in between? how long?) Of course I cannot get specific numbers, each engine is different, but I'm just trying to get an idea on how to approach the problem. The system will be a closed loop using o2 sensors and a map sensor (small cam, good vacuum). I'm interested to learn more about when should I inject fuel for best power and when should I do so for best efficiency. It seems that a Hillborn system just dumps fuel and fuel guaranteeing full combustion but at the expense of efficiency. I have so many questions but no words to put them together; I guess I'm just looking for a little "knowledge". >In all endeavors we tend to forget what has worked and is working in our >efforts to achieve the new. We get enamored with new technology and >scorn the old. We should instead - understand the baseline from where >we want to go and use that to our advantage. Use only from the new what >is needed to solve the unsolvable with the old - progress - not perfection. > I totally agree. Arnaldo E. aec@xxx.net ------------------------------ From: "Edward C. Hernandez" Date: Tue, 03 Sep 1996 12:34:16 -0400 Subject: Really long water injection post, $0.04 worth >...an article involving water injection to reduce detonation. My >question is: What happened to this wonderful idea? I saw an ad toward >the back of the magazine in which Edelbrock that was selling their >Vari-Injection I've been running my Vari-Jection unit for 6 years without problems. > If you forget to keep the tank filled and you don't have sophisticated >controls to either tell you of this or to shut the advance curve down >to stock when there is no injection taking place, you will cause >yourself a problem.... I just include this in my list of fluid levels to check, like how much gas do I have... >Octane rating has *nothing* to do with how fast the fuel burns. An >octane rating is a measure of an engine's resistance to detonation... No, octane rating measures a FUEL's resistance to detonation. >Flame front speed is very similar for all gasolines regardless of >octane rating. That depends on what you call similar. Different octane fuels do, in fact, burn at different rates. How important that is depnds on how well calibrated the seat of your pants is or how important that last 0.01 in the 1/4 mile is. >Water does not slow the combustion process. Its primary function is to >reduce the intake charge and combustion chamber temperatures, which >help the engine avoid detonation. No, it's primary function is to asborb some of the heat of combustion, reducing peak in-cyulinder temperatures, which improves the knock limit. Driving a phase change of water(ie, liquid to steam) takes a lot of energy. That kind of energy isn't available in the intake tracts of a normally aspirated engine. There's more in a turbo. I think I agree about the flame speed not being affected. >To my knowledge, knocking is not a fast burn. The burning >process takes some time (for the flame front to travel accross the >cylinder). As this burning is happening the pressure in the cylinder >rises dramatically. The pressure is also rising (BTDC) due to the >piston compressing the mixture. If these two pressure rises combine >(due to over-advancing the timing), the resulting pressure = internal >energy of the gases may exceeded the activation energy of the >combustion reaction. This energy will be present at all points in the >unburnt gas, so all molecules of unburnt gas are likely to react, and >when the do the pressure rises very suddenly (around TDC) with shock >damage resulting to your precious lump o' steel. Excellent and accurate explanation! But... >Knock resistance is a measure of the activation >energy required to combust (react) the mixture. Slowing the >burn rate can be achieved a number of ways, but its net effect is >that the peak unburnt gas energy density is lower because the >actual peak will occur when the piston is somewhat past TDC. >Knocking is not burning, so knock resistance cannot be a measure of >burn rate. I disagree. The best analogy I can think of is an old fashioned bomb fuse burning its way to a big black bomb. Think of the bomb as the knock limit. If it goes off, you've ignited the end gas as explained above: it explodes al at once. The fuse represents the parameters that control the knock limit(pressure, temperature and TIME). Your goal is to burn the mixture completely before the fuse reaches the bomb. Anything you do to increase pressure or temperature shortens the fuse, and the converse is true. Time is defintely a variable, which means that the knock limit is related to burn rate. Therefore, another approach is to outrun the fuse, which fast burn combustion chambers can do. Anyone notice how high the compression ratios of modern 4V engines (and even some 2V) are getting? Ed Hernandez, still catching up... Ford Motor Company ehernan3@xxx.com ------------------------------ From: Todd King Date: Tue, 03 Sep 96 09:49:00 PDT Subject: Re: knock sensors, ancient history <<< From: talltom Date: Mon, 2 Sep 1996 09:36:12 -0700 (PDT) Subject: Re: Knock sensors I did a Corvette with a turbo on it and used a old buick spark retard system, and while it worked, I wasn't really very happy with it. It would retard 3 degrees for every clatter it didn't like, and hold it for 20-60 seconds. The problem was that it was inconsistent, back to back runs would vary widely. In fact one day it decided to run harder than it had previous and fried the tires hitting second. The result being the rear end stepped out to the left and the car climbed the curb to teh demise of the right front corner. My question is what kind of expieriences have others had with knock retard systems? I'd think that by now this would have been refined some. >>> The knock system is subject to false knock indications and must be carefully calibrated for the various engine combos, sensor placement, etc. However, properly executed it works very well. On the '87 Turbo Regals the ECM processing of the knock system (ESC) signals can be programmed for retard attack rate, timing recovery rate, etc. Never will we see the timing retard linger for more than a brief instant though- you may have had a false knock problem? <<< From: "Robert J. Harris" Date: Mon, 2 Sep 1996 10:06:26 -0700 Subject: Ancient History He drilled the heads so that the untimed variable flow injector sprayed directly on the intake valve from the back side of the valve pocket. What was lacking in mechanical sophistication was made up in simple physics. ... From: "Tony Bryant" Date: Tue, 3 Sep 1996 08:53:10 +1200 Subject: Re: Ancient History ... What I want to know, is what happens when you have radical cam timing? This nice puddle of fuel on the inlet valve ... comments? >>> Comments? OK! :-) Get some forced induction and say "later" to all the big cam blues, and make more power to boot! True story: Kent Rudbeck of Muncie, Ind (one of many) runs 10.3x at near 130mph in his 231" V6 Turbo Regal. Cam specs at 0.050"? Duration- 200deg; lift- about 0.430". My Turbo Regal went 12.59 at 108.9 a few weeks ago with a ridiculously slow launch (taking it easy on it); mods from stock are minimal at best. Engine internals are pure stock; these times are typical for these cars. Well, sorry about the lumpy cam diatribe... We are getting some interesting EFI data during these 1/4 mile runs. We all appreciate the control potential that EFI gives us over the various engine parameters. But in reality there are lots of unexpected things going on. For example from the runs I made a few weeks ago I found that 1) my MAF sensor was maxed at 255 (8 bit) the entire run. 2) my injector duty cycle was 100% almost the entire run, even with injectors that are 25% larger than stock (36lb/hr vs 28lb/hr). So this means that the only things controlling the A/F during the run were airflow (boost) and base fuel pressure setting. Basically I could have wired the injectors "ON" to a twelve volt source and had exactly the same results! These static MAF and injector duty cycles at WOT are not uncommon; it's amazing how fast you can go, even when the EFI is out of control! As for spraying the back of the intake valve- that's typical as well. Interesting that many of the fuel injectors used these days make little attempt to atomize- what they fire is a stream, not a mist. You can make loads of power either way! Todd Todd_King@xxx.com ------------------------------ From: cloud@xxx.edu (tom cloud) Date: Tue, 3 Sep 1996 14:44:25 -0500 Subject: Re: Ancient History >and gets 12mpg (city) using a Quadrajet 750cfm; I want to improve that to >15-17mpg. >Arnaldo E. >aec@xxx.net > I think 12 mpg from a 455 is fantastic. Can't see how you get this good -- FI or carbed. What's your rear end ratio? tom ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:01:00 -0700 (PDT) Subject: Cams and FI > >My '73 914 just happens to have a high lift cam with a lot of overlap. >It won't idle decent below about 2000 rpm. To pass a smog check, I set >the lifters out to .030", and then it idles fine. Clatters like hell. >I've never had backfires with it. > >An additional problem is the low vacuum, about 12", below 2000 with >that engine. The Djet responds to low vacuum by squirting more fuel, >which makes it richer until it dies. I have found a fix for the ECU to >alter the fuel curve, but haven't tried it in the car yet. > >RD > Don't know if it's relevent, but we've run more radical cams than that regularly and get good idle at 1200-1300 rpm with a pissing in the wind system.(k) The accepted mythology is this can't be done, but being as we do it I have a tough time believing it. Our theory sez that as long as the pulsation isn't large enough to pressurize clear out to the barn door(cone) it don't matter what kind of cam you have. One of the guys has a 94 corvette, and he commented that one of the rabbits was faster than his vette.(125 horse nitrous on a rabbit is a REAL ride!:-) Don't know if 914 ever used pissing in wind with barn door systems. ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:01:03 -0700 (PDT) Subject: Manifold/throttle bodys-ED:-) Hey Ed, if you're the manifold guru for Ford, how come the manufacturers haven't gone to throttle bodies for each cylinder like on some of the vettes? I know the vettes are two intake per cyl. but the advantages are many with one valve also, and it can't cost very much when they're already using port inj. Heck it might even be cheaper, being as the current massive aluminum castings could be eliminated in place of plastic plumbing. ??????????????? ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:00:50 -0700 (PDT) Subject: Bosch d disease > >From: rickydik@xxx.com (RD Rick) >Date: Mon, 2 Sep 1996 12:06:45 -0700 >Subject: Re: DIY_EFI Digest V1 #256 >I may be able to help you, Tom. I have a Djet tester, 17 years of >driving Djet 914's, lots of spares, psuedo-schematics, oscilloscope, >etc., and a distributor machine to do dynamic things. >What is the nature of your problem? > >RD What I've got is a Cadillac style V8-6-4. The way this happened is that the brain on my mercedes 4.5 started dropping a circuit or two of injectors, and now is to the point that it's so disgusting that I tore the rest of it apart for long put off tweaking. I came to this grandiose conclusion because I've checked everything else repeatedly and it checks good. My questions are a little ignorant as I don't mess with porsche's and air cooled vw's(they're to close to the ground):-> is are the d-jetronic systems they use equipped with four injector circuits? If so, do you think any of your schematics would be usefull in surgery? Dist. machine, ocilloscope... I'm jealous! ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:00:47 -0700 (PDT) Subject: Re: How tall is tall? > >From: Todd Knighton >Date: Mon, 02 Sep 1996 11:01:45 +0000 >Subject: Re: Knock sensors > >Tom, > My first question is, how tall are you really?!? ha, ha. Well, I'm getting old and shrinking, I'm down to 6'8" now. Being as you ask about me, The picture on the front of a $5 bill has been blown up and placed next to mine in a local newspaper, and I had long time friends insisting I am related to Lincoln. We use the knock sensors with the monitor from J&S to tune the engine >till there is no more knock showing, not to leave it run on it. A >properly calibrated engine should never see knock anyway! >Todd Knighton >Protomotive Engineering Yeah, from what I've heard about that J&S stuff it sounds pretty neat. Kind of afraid to ask price, but I did make note of the address, and someday when in pain will inquire. ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:00:52 -0700 (PDT) Subject: Re: d disease >The Djet and Ljet EFI in the aircooled opposed VW and Porsche 914 >engines sprays directly at the intake valve from three inches away. > >With the Djet, the fuel is squirted in at the beginning of the intake >stroke two of the cyls, and at the beginning of the power stroke on the >other two cyls, so its operation is compromised. > >People on the 914fans lists have been debating the injector timing >question, so your post is being forwarded there. Hope you don't mind. > >RD If it's any help Mercedes applications I believe inject just before or slightly after intake valve opening. Here I'm really going to screw myself, but here goes. It would seem to me that one could use a mercedes brain on a four cylinder and set the points up so that the sequence you want is accomplished. All I ask for the idea is enough of the mercedes brains to keep me supplied before you guys grab them all up.:-) Just for those who aren't thoroughly neaseated yet, maybe I should mention the fact that we haven't checked into DIGIFANT yet! :-> ------------------------------ From: talltom Date: Tue, 3 Sep 1996 13:00:55 -0700 (PDT) Subject: Mercedes mechanical > >How about the real mechanical FI systems? Like off a '69 Mercedes coupe. >No I'm not talking about K-Jet, I talking something that looks like a >diesel fuel pump/distributor. I've only seen it in a workshop manual. > >Anybody with any experience with these, care to comment? Personally I'd never attempt to use one without duress. If anybody is so inclined however there are a few laying around here that could be had reasonable. I have a turbo bmw 6, with l-jetronic that's much safer $$$ wise. ------------------------------ From: cloud@xxx.edu (tom cloud) Date: Tue, 3 Sep 1996 15:10:59 -0500 Subject: Re: FORTH >I've seen a few postings recently about FORTH, I've done a fair bit >of programming in FORTH, so as my first contribution here goes :- > >The microcontroller we used was an Hitachi H8, What is an H8?? I harken back to a Fairchild F8 (shudder). Seems it was either a 4 or 8-bit micro. ------------------------------ From: "Terry McLane 312.630.0533" Date: Tue, 3 Sep 1996 15:02:05 CDT Subject: Bosch to Hitachi MAF >> Andrew Rabbitt wrote: >If 220CFM is 0V, then how is the ECU going to determine anything more >than that (ie: negative volts). The way it looks to me is your ECU is >limiting your airflow as much as your AFM. The voltage figures I quoted were only for purposes of example, but it should only be necessary to measure the air flow up to the engine's peak volumetric efficiency. This should cooincide with peak torque (3200 RPM @xxx. I assume there is no need for air measurement beyond this point as VE falls off at higher RPMS. In otherwords, the engine may use more air per second, but not more air per cylinder per injection event. Therefore a richer mixture would not be necessary. >>Like all other MAFs, the Hitachi unit (I presume it's a bypass unit >>similar to that fitted to the previous model Taurus, and many other >>Fords), is non-linear and therefore a 2-point calibration will not be >>adequate. You're right, I am interested in the bypass unit, due to its low restriction (low pressure drop). The air meter on my car uses a logarithmic potentiomter which is hooked to a "swinging door" suspended in the air stream (resistance is 200-400 ohms). Needless to say, it's very restrictive. Having read a couple of articles on the Hitachi unit used by Ford, I was under the assumption that it also used a 5 volt reference and a logarithmic output. It also doesn't require any other circuitry for "burn off" like other hot-wire MAFS. I figured that the Hitachi would allow higher airflow potential so if I could determine voltage output equivalent to peak torque and idle in my engine for both units, they could be swapped if a suitable mapping implementation could be found. As a bonus, the air meter also uses a thermistor to measure the air temperature, the Hitachi adjusts for temp/pressure automatically, - I thought I could hook up a trim potentiometer to substitute, giving me a straightforward rich/lean adjustment mechanism. Always looking for more adjustability. (I imagine the ECU is fairly sensitive to air temperature in this application). >>Also, at large throttle openings, the signal jumps around a bit, and >>the only way I know of getting around this is to purposely alias the >>signal by sampling synchronously at a pre-determined crank position. >>(I'm assuming a digital controller) Is this at constant high flow, or sudden openings? At constant values, shouldn't the closed loop mode of the ECU adjust? >>Setting this up is no easy task (speaking from experience here) and >>if you get it wrong, you'll likely be out by more than 100% in some >>areas of the speed-load map. Bear in mind that I have only a passing knowledge of electronics, my original question was based on the observation that it should be possible to build a simple circuit to map the low voltage and high voltage of each unit at a given air flow, and then interpolate. Obviously if there are other electronic "gremlins" present in the Hitachi MAF, this may not be feasible. I'd like to talk to you direct- can you respond with your e-mail address? ...Terry tm8@xxx.com ------------------------------ From: Todd Knighton Date: Tue, 03 Sep 1996 14:03:15 +0000 Subject: Re: K_knock sensor Tom, For low compression or even mild compression that's okay. But, with high compression engines or turbocharged mild compression stuff your rattling will knock the rings right out of it. Sounds like you're doing this at partial loads, what does this method tell you about full throttle stuff. Is your vacuum advance, or computer chip for that matter, set up with the proper amount of retard so that your condition is correct at all other rpms and throttle positions, or only the one you happen to be at at that speed and throttle position. When I was 15, I used to set up my D-Jet 914 that way, and it seemed to work pretty good. Make a run, if it pings, retard it, if it doesn't advance it. You could get pretty good results. But then again that was a 2.0 liter 4 cylinder that made about 125 hp on a good day. Now I'm working with 3.6 to 3.8 liter 700 horsepower turbocharged engines that won't tolerate that kind of stuff. You'd grenade a high output engine if it pinged just once! Just ask Ford, when they were developing their 2.0 liter 4 cylinder turbo engine for GTP, they had a ignition glitch that only hit at high rpm and load and made the engine "rattle a little". Actually, they destroyed about 20 engines before they found the RF interference that was getting into the crank sensors and messing up the timing. But then again, I've heard a lot of Ford, Chrysler, and GM products running around the streets rattling like mad, and they seem to keep going. Low cylinder pressures I guess. The reason I told you about the J&S knock sensor is you can do exactly like your saying, but the knock sensor won't let it "rattle" it will just tell you how much it had to retart the timing to make it not "rattle". Then you can go retard the ignition by that amount. No damage done! Todd Knighton Protomotive Engineering tom cloud wrote: > > I don't do the race circuit, only want to go fast some times > and just drive the others. The way I tune my toy is to set the > advance until I get a little rattling on long hills on hot days > (we have some of both of those) and back it off a little. > > So, Todd, or anyone else who cares to chip in, what's wrong > widdat? > > tom cloud ------------------------------ From: "Edward C. Hernandez" Date: Tue, 03 Sep 1996 17:24:41 -0400 Subject: Re: Throttle Body Sizing Michael D. Porter commented on the throttle body sizing formula: > To further this thought, shouldn't the intake valves duration somehow >be included in this(calculation for throttle body sizing)? It is: that's why he includes volumetric efficiency(VE) in his formula. Problem is, without a dyno, you have to guess your V.E. >Who cares? Isn't the object to get as much velocity as possible? No, the object is minimize flow restrictions while optimising the opening rate. You are right: it is difficult to drive cars with big throttle bodies because at low throttle settings, a few degrees makes a huge difference in flow. OEMs get around this by designing cams into the throttle linkage to modify the opening rate at lower throttle settings. - -- Ed Hernandez Ford Motor Company ehernan3@xxx.com ------------------------------ From: rickydik@xxx.com (RD Rick) Date: Tue, 3 Sep 1996 14:54:56 -0700 Subject: Re: Bosch d disease talltom wrote: >>From: rickydik@xxx.com (RD Rick) >>Date: Mon, 2 Sep 1996 12:06:45 -0700 >>Subject: Re: DIY_EFI Digest V1 #256 > >>I may be able to help you, Tom. I have a Djet tester, 17 years of >>driving Djet 914's, lots of spares, psuedo-schematics, oscilloscope, >>etc., and a distributor machine to do dynamic things. >>What is the nature of your problem? > What I've got is a Cadillac style V8-6-4. The way this happened is that the >brain on my mercedes 4.5 started dropping a circuit or two of injectors, and>now >is to the point that it's so disgusting that I tore the rest of it apart for >long put off tweaking. > I came to this grandiose conclusion because I've checked everything else>repeatedly and it checks good. > My questions are a little ignorant as I don't mess with porsche's and air>cooled>vw's(they're to close to the ground):-> is are the d-jetronic systems they use>equipped with four injector circuits? If so, do you think any of your schematics>would be useful in surgery? Dist. machine, ocilloscope... I'm jealous! Yeah, the Djet tester I have was made by EFI Associates LA. It can test the 8 cyl Merc, but I don't know if my harness will fit. Is your ECU about 8" x 12" and have the 25 pin connector on the end? If so, my tester should work. What part number is the ECU? I don't have much in the way of schematics, but that vintage electronics is my cup of tea. I could check your dist, too, and dynamically test its inj. contacts. The Djet in 4 cyl engines has just two injector circuits. I'm 6' and fit the 914, but my 6' 4" son hates it. You would need a bubble canopy. :) RD ------------------------------ End of DIY_EFI Digest V1 #258 ***************************** 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".