DIY_EFI Digest Tuesday, 3 December 1996 Volume 01 : Number 368 In this issue: Re: [Fwd: Electronic Dynos] RE: Air Flow Measurement DIY v available systems Re: book re: ALDL to serial converter Re: FORSALE: Engine Management System, Haltech E6 Re: needhams emp-20 device programmer Re: [Fwd: Electronic Dynos] re: Schematic of a GM ALDL -> PC converter Re: Air Flow Measurement Re: VW Digifant EFI Re: Subaru Fault codes Re: FORSALE: Engine Management System, Haltech E6 Re: crankshaft/flywheel encoder patterns Injector Connectors DIYEFI FAQ Re: Injector Connectors Re: DIYEFI FAQ See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- From: Chris Wilson Date: Mon, 02 Dec 1996 11:16:09 GMT Subject: Re: [Fwd: Electronic Dynos] On Sun, 01 Dec 1996 20:07:52 +1000 Doug Robson apparently said in mail.DIY_EFI: > Ok so who want to discuss writing a program to convert the EMF against > an engine dyno to Horspower and torque. is it easy or am i silly Most rolling road dynos use a strain gauge system,and AFAIK,so do most engine brakes.Strain gauges are pretty simple cheap and robust.Why do it the hard way? - -- Chris Wilson Sun, 01 Dec 1996 22:30 GMT Gatesgarth Racing Developments Gatley,Cheshire,U.K. Race Car Preparation and Development http://195.102.33.46/~gategart mailto:kasper@xxx.com ------------------------------ From: Anthony Tsakiris Date: Mon, 2 Dec 1996 08:46:29 -0500 Subject: RE: Air Flow Measurement >>The response to a step increase is faster than a step decrease in airflow. > Ok, I though about it alot, and finally talked to all of my ME/ESM buddies and none of us can think of why this happens. If you have two Pt elements, one to measure ambient, and you heat the other to some constant delta-t above ambient and use the energy lost to find mass flow, there is no physical phenomenon to say this should work asymetrically with a step change in the airflow. < The dissipated energy is dissipated via heat transfer. The heat transfer can be both convective and radiant. The convective heat transfer component will change markedly with changes in flow; the radiant heat transfer will not. This might lead to different responses for increasing and decreasing flows. Anthony Tsakiris ------------------------------ From: Carter Hendricks Date: Mon, 2 Dec 1996 05:48:11 -0800 (PST) Subject: DIY v available systems On Sun, 1 Dec 1996 Ashe@xxx.au wrote: > > >>I'm wanting to put fuel injection (probably a GM 2.5L TBI unit) on > >>my 4cyl. 22R toyota motor. > > > > Want the easy way out? I would just go with a complete setup from a > >22R-E motor from a Celica... There are enough of them around... > >... No need to worry about interfacing the signals from the GM unit, you > >can have the (actually rather reliable) toyota unit that will nit perfectly > >in place with no fabrication. > > Look guys, > > I'm in kinda the same predicament as the originator of this query. You > fellas are way too up in the clouds for me. Does anyone have any > experience with a regular (ie major brands like Haltech, Injec, Motec > etc etc) aftermarket programmable efi system and Fiat twincams? > I'm great with swinging a spanner Same answer. Pick a L-jet 2.0 Fiat and transfer the stand-alone harness and mechanicals. Just clicks together and works, alot better than any of the Fiat carburator intake manifolds. To think that to start from scratch is "easier" is, well, silly. [Then, if you want a hobby, you can substitute controls for the Bosch system]. --Carter ------------------------------ From: George Lerm Date: Mon, 02 Dec 1996 17:58:23 +0200 Subject: Re: book Hi all I have found another book at my local book store. It is called Automobile Electrical & Electronic Systems ISBN: 0 340 58604 4 Frank F Parker wrote: > > > From: Frank Parker Date: Fri, 15 Nov 1996 08:38:25 -0500 > > (EST) Subject: Re: Lambda Sensors.... > > Much more in an excellent book by Ronald Jung, available from SAE, called > > "Automotive Electronics". LArge chapter on O2 sensors and all > > other auto electronics. Can get ISBN # of there is interest of group. > > > > Frank- I'm interested in any info on this book you could provide. Thanks. > > > > Todd, > > Well I was sorta close: The book is " Aotomotive Electronics Handbook" by > Ronald Jurgen, a McGraw-Hill book, ISBN #: 0-07-033189-8. It's pretty > expensive at about $90 but worth it for the excellent background on a > wide range of topics. Jurgen is the editor. The O2 sensor chapter > was written by chief O2 sensor engineer at Bosch. > Also excellent chapters on flow and knock sensors. > Should be available at good university library. > > Frank Parker > > > - -- George Lerm E-Mail: glerm@xxx.com Visit http://mickey.iafrica.com/~glerm ------------------------------ From: SRavet@xxx.com Date: Wed, 27 Nov 96 13:53:08 CST Subject: re: ALDL to serial converter Scott Feaver Wrote: | | What chip would be required to convert the ALDL signal from a 1991 ECM to a | standard serial connector? Would it be a matter of using something like an | 8031 or 68H11, reading them in on a i/o pin and some software to convert it? | | Scott | | I just got back from Thanksgiving holiday, and I see that I have 100+ new messges. Maybe this email has already been answered, but I had already typed most of this up so I'll send it anyway. Depends on the data stream. GM used two. The "C3" ECM uses a 160 baud stream, the "P4" ECM uses an 8192 baud stream. The 8192 stream is TTL level serial, 0=0 and +5=1. All you need to hook this stream to a serial port is a TTL to RS232 driver, like the MAX332ACP that the EFI-332 project is using. You can get this part for free from the Maxim WWW page. Then program your serial port to 8228 using a divisor of 14, and you should be good to go. This paragraph is "theoretical" in that I've never tried it, but is what I've been told is possible. I'm not sure how to tell if you have the 8192 stream or the 160 stream. Additionally, you have to trigger the ECM into sending this data, it doesn't send it all the time. I don't know how do that. The 160 baud stream is different. It's not really serial, in that each data bit is preceeded by a falling edge. The line is driven to the inverted data value 1.5-2.3 msec after the falling edge. After this sample interval goes by, the line goes back high, and 6.25 msec later another falling edge indicates that another bit is coming. This won't connect to any serial port, but it wouldn't take much to have a little programmable logic convert these into bytes and feed them to a UART. Here's a diagram: ______ _____________ | | | | |___|________| | T0 T1 T2 T3 T4 T0=falling edge that indicates the start of a bit (not byte). T1=rising edge, if the bit is a zero. If the bit is a one, the line stays low T2=sample period T3=rising edge to prepare for next bit T4=start of next bit T0-T2 = 1.5 to 2.3 ms T3-T4 = .5 ms minimum. Data is sent as nine bit bytes, with the first bit being a start bit. The remaining 8 bits are sent MSB first. The very first byte is a sync character. It has a "1" start bit, and all "1" data bits, ie 9 ones in a row. The data bits then are sent, each with a start bit of zero. When all the data has been sent, the whole thing starts over with the sync character followed by the data bytes. I don't have any information regarding what the bytes actually mean. - --steve Steve Ravet sravet@xxx.com Baby you're a genius when it comes to cooking up some chili sauce... ------------------------------ From: Jim Davies Date: Mon, 2 Dec 1996 08:24:00 -0800 (PST) Subject: Re: FORSALE: Engine Management System, Haltech E6 On Sat, 30 Nov 1996, Robert E. Yorke wrote: > My only problem with the Haltech, is figuring out how to allow the use of > the Torque Converter Clutch (TCC )signal, which changes the pitch of the > converter to provide a little better mileage. - --a minor point to note is that a TCC's function is to connect the trans. input shaft to the engine crankshaft. This is accomplished via a hydraulically actuated piston that is splined to the input shaft and is forced against the convertors front housing when lockup is desired. There are other variations on this theme, but this is the easiest system to control by computer. Stator vane angle switching has been used (64-68 GM are good examples) but this was to reduce idling-in-gear creep at traffic lights, although it was widely modified to be used as a variable stall speed convertor. Jim Davies ------------------------------ From: Stephen Dubovsky Date: Mon, 2 Dec 1996 11:53:02 -0500 Subject: Re: needhams emp-20 device programmer At 04:40 PM 11/28/96 -0800, you wrote: >I just bought a needhams emp-20 device programmer... > >But in order to program a very basic 27256 EPROM, it requires me to get >a additional card... >... Should have come w/ 3 cards to do all 8 and 16 bit 27xxx,28xxx parts. ------------------------------ From: Stephen Dubovsky Date: Mon, 2 Dec 1996 11:53:00 -0500 Subject: Re: [Fwd: Electronic Dynos] At 12:12 PM 12/1/96 -0500, you wrote: >Doug Robson wrote: ... >> Ok so who want to discuss writing a program to convert the EMF against >> an engine dyno to Horspower and torque. is it easy or am i silly >> >> regards > > >Hi Doug, > >I'm not an electrical engineer, but I think this is what you're looking for: > > EI = T*2*Pi*n/60 or > T= (60*E*I)/(2*Pi*n) > > where: > E is voltage (in Volts) > I is current (in Amps) > T is torque output (in Newton - Meters) > n is speed of rotation (in RPMs) > Pi = 3.14159 > >An alternative method to measuring torque is using a load cell, which is comprised of >two strain gages. While the generator is supported by its bearings, the load cell is >the only memberane that keeps it from spinning. You can get a pretty accurate reading >with a load cell within 0.5%. This is the method most water brake dynos use. > >Best Regards, >Mazda Ebrahimi > I also agree that using a strain gauge is more accurate, but there are 'gotchas' when using a load motor like this. Mazda is right about the formulas. EI=P (in V,A,Watts respectively). The only problem is measuring the Volts. The load motor being used (as a generator) has some internal resistance. When loaded, the I causes a V drop on this internal resistance that you cant measure. It causes a power loss in the motor (ever notice how it gets hotter when loaded, and not just turning at close to no load?). There are a few ways to get around this. 1) Measure motor resistance and compute what the real back EMF would be. 2) Map back EMF vs rpm and use that instead of the measured E. Both of these have accuracy problems in that 1) the rotor R changes a lot w/ temp 2) the back EMF will change slightly under load due to non-linearities in the B vs H cure of the iron in the motor (scarry magnetics stuff here). If your motor is a field excited machine (adjusts the load) then all of this becomes more complicated. The back EMF is also a function of your applied field (not constant as w/ a permanent magnet machine). The first method would probably be the best way then... Or just go find a strain gauge from Omega;) SMD - does it sound like I know too much about electric machines??? - -- Stephen Dubovsky dubovsky@xxx.edu 95 Yamaha FZR600 83 Porsche 911SC 84 Jeep Cherokee ------------------------------ From: SRavet@xxx.com Date: Mon, 2 Dec 96 11:05:57 CST Subject: re: Schematic of a GM ALDL -> PC converter Darrell Norquay Wrote: | | At 10:16 AM 11/26/96 CST, you wrote: | | Markus Strobl Wrote: | | >| Does anyone have a schematic of a GM late model ALDL to | >| PC converter? By late model I mean the late 80s/90s | >| 8192 baud ALDL. | | Steve Ravet wrote: | | >The Diacom software uses the parallel port to watch the serial data pin of | >the ALDL. It implements a UART in software. It does have timing problems, | | Seems to me these guys are going about this all wrong. The 68HC11 has a mode | where it's serial port talks at 8192 baud. I'm pretty sure GM EMU's are | 68HC11 based, thus the oddball baud rate. In the words of the prophets, you | need a lawyer to talk to a lawyer. It would be a simple matter to setup a | small 'HC11 module to either log the data, display it, or both. You could | add a second UART to the 'HC11 and just use it as a buffered baud rate | converter to talk to a laptop. Could do that also. The thing is that there are two baud rates available, 8192 and 160. As I wrote in another post, the 160 data stream will never hook directly to any serial port, because every bit has it's own start bit. Using the parallel port makes sense for diacom, because then it's an "all software" solution that works for both data rates. - --steve | | regards | dn | dnorquay@xxx.com | | Steve Ravet sravet@xxx.com Baby you're a genius when it comes to cooking up some chili sauce... ------------------------------ From: kleenair@xxx.com Date: Mon, 02 Dec 1996 11:47:06 -0800 Subject: Re: Air Flow Measurement Anthony Tsakiris wrote: > > >>The response to a step increase is faster than a step decrease in airflow. > > > Ok, I though about it alot, and finally talked to all of my ME/ESM buddies > and none of us can think of why this happens. If you have two Pt elements, > one to measure ambient, and you heat the other to some constant delta-t > above ambient and use the energy lost to find mass flow, there is no > physical phenomenon to say this should work asymetrically with a step change > in the airflow. < > > The dissipated energy is dissipated via heat transfer. > The heat transfer can be both convective and radiant. > The convective heat transfer component will change > markedly with changes in flow; the radiant heat > transfer will not. This might lead to different > responses for increasing and decreasing flows. > > Anthony Tsakiris Also, the rate of the hot wire temp. rise is limited by heat generated by the maximum current minus heat transfer by convection and radiation as explained above. But rate of temp. fall is limited only by heat transfer. The two quantities are not necessarily equal. ------------------------------ From: Todd Knighton Date: Mon, 02 Dec 1996 09:10:03 -0800 Subject: Re: VW Digifant EFI markh wrote: > > Hello, > > My name is Mark and I am working on a engine managament system based on > the VW-Digifant, it's a EFI from a '92 8v GTi with a 68HC11 > microcontroller inside. > My problem is that I don't know how the look up tables are build up and > where the look-up tables are staying in the EPROM (256 K). > I am very interested in all info about the Digifant EFI, my idea is to > make the Digifant PC configurable via a RS232 connection. Mark, I see a company in European car all the time that does this stuff for the GTi's. They've got EPA approved turbosystems as well. I believe it's New Dimensions or something like that. They've obviously got the code figured out because they're even setting fault codes for boost related problems, etc. Sorry I don't have better info, but it's a start. Todd Knighton Protomotive Engineering ------------------------------ From: Doug Robson Date: Tue, 03 Dec 1996 07:14:09 +1000 Subject: Re: Subaru Fault codes John Firkins wrote: > > Thanks to Doug & Chris > > Unfortunately, it turns out that the fault codes were misleading. I've > traced the problem to a worn bushing in the torque converter, which allows > the mounting bolts to come into contact with the adaptor plate. This > contact was enough to be picked up by the knock sensors, fooling the cpu > into believing the engine was about to self destruct! > > Cheers > John Firkins Holy Sh.. Batman, thats got to be the most abstract problem resolution Ive evr seen. Obviously next time its starts to overheat we should check the sun roof relays first hey!! wel done , but how did you work that out??? - -- |===============================================================| | When I die, | | I want to go in my sleep, like my grandfather, | | not screaming like the passengers in his car. | |---------------------------------------------------------------| | Doug Robson mailto:doug@xxx.au | | mailto:Doug.Robson@xxx.com | | Sydney, Australia http://www.cia.com.au/doug | |---------------------------------------------------------------| | Club Car Racing Register of NSW | Thank God | | 1992/93 Under 2 litre State Champion | for | | http://www.cia.com.au/doug/ccrrnsw.html | Gravity | |===============================================================| ------------------------------ From: "Robert E. Yorke" Date: Mon, 2 Dec 96 14:30:18 PST Subject: Re: FORSALE: Engine Management System, Haltech E6 Thanks for the info, Jim!! At 08:24 AM 12/2/96 -0800, you wrote: > > >On Sat, 30 Nov 1996, Robert E. Yorke wrote: > >> My only problem with the Haltech, is figuring out how to allow the use of >> the Torque Converter Clutch (TCC )signal, which changes the pitch of the >> converter to provide a little better mileage. >--a minor point to note is that a TCC's function is to connect the trans. >input shaft to the engine crankshaft. This is accomplished via a >hydraulically actuated piston that is splined to the input shaft and is >forced against the convertors front housing when lockup is desired. There are >other variations on this theme, but this is the easiest system to control >by computer. Stator vane angle switching has been used (64-68 GM are good >examples) but this was to reduce idling-in-gear creep at traffic lights, >although it was widely modified to be used as a variable stall speed >convertor. > >Jim Davies > > > Riverside, CA 2 '88 GTs! ------------------------------ From: Mazda Ebrahimi Date: Mon, 02 Dec 1996 19:27:45 -0500 Subject: Re: crankshaft/flywheel encoder patterns RABBITT_Andrew@xxx.au wrote: > > can anyone out there tell me what encoder patterns people are using > (especially on production engines) and what sensors are used also? > > ie: > > Ford EEC-4's, 36-1 pattern, inductive crank sensor, hall effect cam > sensor. > > I'm trying to find out what's out there and how common-place 60 tooth > encoders are. (trying to justify a 60 tooth design). > > thanks (in advance) for your help > > Andrew Rabbitt > Orbital Engine Company Hi Andrew, Dodge uses 8 pulses on the crankshaft along with one pulse on the camshaft sensor on their 5.2L and 5.9L Magnum engines (I hope this is helpful?!) I'm currently building a fuel injection box. I'm basing my design on 180 teeth on the crank and one pulse on the cam (rising on TDC on #1 and falling on #6). My justification for this decision was that the crank tooth wheel will generate a pulse for every degree of rotation (rising or falling). I'm doing this to make my programming somewhat easier. In effect, the toothwheel will serve as my physical system clock. Instead of using PWM channels (I have eight available on the processor), I'm using counters driven directly by the tooth wheel. So when I figure out my injector duty cycle (let's say 50%), I will load the counter for every injector with a value of 360 (half of 720). There will be no further interruptions to the processor. When the count is over, the injector will turn off and continue its count to 720, when it will reload again. I am interested in your thoughts on this approach. Best Regards, Mazda ------------------------------ From: Jennifer Rose Date: Mon, 2 Dec 1996 18:35:41 -0800 (PST) Subject: Injector Connectors Hi All Need to find a good source for the connectors that plug into injectors on a GM TPI unit. Have been scrounging the local u-pull-it, nothing left. Got the rest of harness parts - will start building soon. Thanks Vance ------------------------------ From: Stuart Baly Date: Tue, 3 Dec 96 15:40:07 EST Subject: DIYEFI FAQ DIYEFIers, OK, this is something that has been playing on my mind for a while now - I believe it's time for a DIYEFI FAQ. A week or so ago I had a think about writing a FAQ, and this is the basic outline I came up with: ************************* 0. How does my engine work? 4-stroke Basic anatomy - crank conrod piston cyl valves Basic operation - 4 parts of cycle Valve operation Side valve OHV pushrod OHC DOHC 4v vs. 2v Ignition How it works (Kettering, CDI) Importance of timing, significance Mixture rich/lean effects on power, fuel cons, emissions Throttling Significance of throttle pos./ manifold vac. on power Cooling system Why, how Exhaust Why, how turbo 4-stroke Basic supercharging. Boost control - wastegate. 2-stroke basic operation advantages and disadvantages from 4-stroke rotary basic operation adv and disadv from 4-s diesel 2-stroke diesel other weird ones 1. What is EFI? Mech. systems, analog elec. systems, digital elec. systems How does the XXXXXX sensor work? Input sensors - TPS, MAP sensor crank, cam sensor mass air flow sensor intake temp, coolant temp, cyl head temp sensor exhaust gas temp sensor exhaust gas oxygen sensor Outputs - Fuel injectors, ignition module Boost control MY CAR IS A XXXXXX. WHAT TYPE OF MANAGEMENT HAS IT? List of cars, systems DESCRIPTION OF VARIOUS SYSTEMS K-Jetronic, L-Jetronic, Motronic, Monotronic, LH-Jetronic ECCS, EEC-4, etc. I'D LIKE TO MODIFY MY XXXXX SYSTEM. WHERE DO I START? Sources of info. IF LIKE TO DESIGN MY OWN SYSTEM. WHERE DO I START? Existing systems. Starting pts. e.g. 80552 boards, EFI332 proj, 68332 boards 68HC??? boards Hints from those who have done it before e.g. power supply probs, temperature extremes, filtering etc. Info on sensors - sources Circuit cookbook - ign. and inj. drivers, temp. sensor circuits Info on injector flow rates etc. - sources. LIST OF AFTERMARKET ENGINE CONTROLLERS web sources GLOSSARY crank fire, open loop, closed loop, wastegate, speed density n-alpha, alpha, BMEP, V.E., advance, variable reluctance, VR Hall effect, absolute pressure, pressure, EGO, TPS, IAT, MAP BAP, any others? LIST OF REFERENCES THANK YOUS **************************************** I'm sure there are many other subjects/questions to be answered, but these are the ones I came up with in the hour I spent on it. This is a fairly big task, even though many of these questions are probably answered in FAQs elsewhere, and there is already a good list of references, parts etc. in circulation. Here's the deal - I've decided against writing it, for the following reasons - 1. I haven't got time 2. I haven't got the experience. I've got a fairly good understanding of the concepts etc., but I'm yet to lay a hand on an EFI vehicle in anger, which kind of puts me at a disadvantage. 3. A curly one - which audience should it be written for? I can see conflict arising over the need to be easily understood by the average Joe, and the desire to be 100% technically correct. We've all seen seemingly simple statements stretch out into week-long threads discussing the minutiae of the argument down the the nth degree (not that I'm complaining) - - just imagine what would happen with a 100k document full of simple statements. So there it is - maybe someone will take the ball and run with it, in which case I'd be happy to do my little bit, or maybe this post will be ignored. Apologies to all for bombing the list with such a long post, but I really think it's time for a FAQ. Cheers, Stuart. ========================================================= Stuart Baly (S.Baly@xxx.au) '71 Datrat 1600, '89 Kawasaki GPz900R, '81 Yamaha RD350LC ========================================================= ------------------------------ From: dzorde@xxx.au Date: Tue, 03 Dec 96 14:20:47 Subject: Re: Injector Connectors AMP and Bosch will do the job. Dan dzorde@xxx.au ______________________________ Reply Separator _________________________________ Subject: Injector Connectors Author: diy_efi@xxx.edu at INTERNET Date: 12/3/96 1:31 PM Hi All Need to find a good source for the connectors that plug into injectors on a GM TPI unit. Have been scrounging the local u-pull-it, nothing left. Got the rest of harness parts - will start building soon. Thanks Vance ------------------------------ From: Rob Skala Date: Tue, 03 Dec 1996 15:37:35 +0800 Subject: Re: DIYEFI FAQ At 03:40 PM 3/12/96 -0500, you wrote: Information regarding a very good idea deleted in the interests of preserving sanity........ Hi All, This I think is a very good idea. I have been looking at either buying and aftermarket injection system or designing my own from scratch....not sure which route to take at the moment. Although my knowledge of EFI systems is a little limited at the moment (it grows with each new message from this list !), I do have a working knowledge of the rotary engine and would be happy to put together a little something on that aspect if this idea gets off the ground. If anyone out there decides to get this going, let me know if you want any information about the rotary engine. Regards, Rob Skala Robert D Skala Materials Research Group School of Physical Sciences Curtin University of Technology Perth, Western Australia Tel: +61 9 351 2331 Fax: +61 9 351 2377 Mobile: 0412 061 060 email: skala_rd@xxx.au ------------------------------ End of DIY_EFI Digest V1 #368 ***************************** 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".