From owner-diy_efi Tue Nov 29 15:43:12 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA07026; Tue, 29 Nov 94 15:43:12 GMT Received: from aztec.co.za by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA07021; Tue, 29 Nov 94 10:43:03 -0500 Received: by aztec.co.za (Smail3.1.28.1 #5) id m0rCTjF-000KdhC; Tue, 29 Nov 94 16:40 EET Date: Tue, 29 Nov 1994 16:40:25 +0200 (SAT) From: Wouter de Waal Subject: gcc, 68K and ignition advance To: efi list Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI OK, I (finally) got gcc to work. I compiled it under DJGPP, which means that it runs on a PC. So now I can develop code on my PC, then stick it in my 68K system. After I went through all the pain of building gcc, I heard that pre-ported versions are available from ftp.pooh.com. So if anybody is interested, you can (definitely) get the stuff from me, but it might not work 100% yet, or you can (maybe) get the stuff off the net, but I assume that if it's posted, it'll work. The car that is the target of all this fun is currently on blocks, I'm doing some fibreglass work on it to increase airflow to the (rear mounted) engine. The thing sucks +- 950 liters/second @ 6000rpm, so it needs BIG air channels. The first thing that I would like to start playing with is ignition timing. I've been thinking about why ignition advance is required, and I can't seem to make my understanding tie up with what is commonly done. So would some of you that have thought about this more thoroughly please enlighten me? My thoughts are as follows: * The air/fuel mix takes a constant _time_ to burn. * The max. pressure point has to occur when the piston is on it's way down, and when the conrod has a decent leverage on the crank. * Assume idle = 1000rpm, and advance = 10 degrees @ idle. 1000rpm = 6000 degrees per second, so the spark occurs 1/600 second btdc. * Assume that we want the maximum pressure at 10 degrees atdc. (Please, this is the wildest guess I've made in a long time) * The spark occurs 1/300 second before the peak in pressure. * At say 2000 rpm = 12000 degrees/second, 1/300 second = 40 degrees, so the advance should be 30 degrees. * In the same way, at 6000 rpm the advance should be 110 degrees (!). * Now obviously, you can't fire the plug before the inlet valve is closed. But this should have happened at say 160 degrees BTDC. So this is not why most advances seems to be limited to about 30 or 40 degrees. The obvious question is why _I_ think an advance of 100 degrees is fine, whereas my VW manual only allows 30 degrees, which is "all in" by 3000rpm. Somebody please clarify! * And the next tricky point - vacuum advance. The vacuum is an indication of engine load. But why is it a good thing to advance the timing when the engine is loaded? By how much should the timing typically be advanced for a given load? (And I mean _typically_, I know that the Bosch systems amongst others use complicated, dyno-derived maps. However, I am more interested in deriving the kind of curve a standard distributor gives.) OK. If that's not enough, I have plenty more questions :-) Thanks Wouter From owner-diy_efi Tue Nov 29 17:18:33 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA09594; Tue, 29 Nov 94 17:18:33 GMT Received: from wotan.compaq.com by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA09584; Tue, 29 Nov 94 12:18:30 -0500 Received: from twisto.eng.hou.compaq.com by wotan.compaq.com with smtp (Smail3.1.28.1 #12) id m0rCVyL-000vJlC; Tue, 29 Nov 94 11:04 CST Received: from bangate.compaq.com by twisto.eng.hou.compaq.com with smtp (Smail3.1.28.1 #10) id m0rCVy3-000uHSC; Tue, 29 Nov 94 11:03 CST Message-Id: Received: by bangate.compaq.com with VINES ; Tue, 29 Nov 94 11:03:54 CST Date: Tue, 29 Nov 94 10:39:29 CST From: Steve=Ravet%Prj=Eng%PCPD=Hou@bangate.compaq.com Subject: re: gcc, 68K and ignition advance To: diy_efi Cc: Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI Wouter de Waal Wrote: | | | OK, I (finally) got gcc to work. I compiled it under DJGPP, which means | that it runs on a PC. So now I can develop code on my PC, then stick | it in my 68K system. After I went through all the pain of building | gcc, I heard that pre-ported versions are available from ftp.pooh.com. | So if anybody is interested, you can (definitely) get the stuff from me, | but it might not work 100% yet, or you can (maybe) get the stuff off the | net, but I assume that if it's posted, it'll work. Post this, and send it to fridman@cpsc.ucalgary.ca so he can put it on the WWW server. Am I glad someone finally did this! I've been meaning to do it under linux, from the original source (very easy), but my linux machine is down for the time being. I would have thought it would be more difficult under DOS. Anyway, I don't think cross-compilers are pre-built anywhere, and if they are, they probably don't exist for DOS, since the FSF code doesn't support DOS. You might want to send a message to the GCC groups and 68k groups, I know when I was looking for a cross compiler like this a year ago, several people wished for a GCC cross compiler under DOS. is there really a site called ftp.pooh.com? | The first thing that I would like to start playing with is ignition | timing. I've been thinking about why ignition advance is required, and I | can't seem to make my understanding tie up with what is commonly done. | So would some of you that have thought about this more thoroughly please | enlighten me? | | My thoughts are as follows: | | * The air/fuel mix takes a constant _time_ to burn. I thought this also, another poster here said that burn time decreases with rpm, but *ignition* time is constant. So varying the timing is to account for the mixture ignition time, not burn time. Assuming that ignition time is constant (and short) compared with burn time, that explains why not as much advance is required as it would seem below. | * And the next tricky point - vacuum advance. The vacuum is an indication | of engine load. But why is it a good thing to advance the timing when | the engine is loaded? By how much should the timing typically be advanced | for a given load? (And I mean _typically_, I know that the Bosch systems | amongst others use complicated, dyno-derived maps. However, I am more | interested in deriving the kind of curve a standard distributor gives.) I think the purpose of vacuum advance is to advance the timing when the engine is *not* under load. ie when the engine is idling, lots of vacuum, lots of advance. Hit the gas, and the engine would start knocking, so to keep it from knocking the vacuum advance retards some. When you get to cruising speed, vacuum returns, the ignition is advanced some more. I'm not positive about the vacuum advance arguments, so feel free to correct me if this is not a correct description.0 | | OK. If that's not enough, I have plenty more questions :-) Lets have them.... :-) Ok, I looked in my email archive and found this response, if anyone is interested. It's a personal email that I am reposting without permission, but I think it's OK.... The heywood book, i believe, says that ignition time is constant and burn time varies. ---begin include Hi Steve, Somehow, I doubt this will get thru, as your address looks "unfriendly" to my mailer (my guess)..... Heywood is an excellent Combustion reference. The title is "Internal Combustion Engine Fundamnetals" by John Heywood (MIT). Two other references: Very general, with an excellent chapter or two on combustion, and lots of other source of energy discussion. Light Thermodynamics: DaRosa, "Fundamentals of Energy Processes" Good automotive (general) reference. Has everything from properties of auto-glass, to gasoline specs (originally written in German, so many of the specs are DIN): "Bosch Automotive Handbook", R. Bosch. I believe all three are currently in print, and should be available (order or in stock) in any good tech book store. The mailing list sounds like a good idea. Unfortunately, when ever I join a mailing list, I get flooded, and I end up missing an important message from my boss or something, and I get teethmarks on my ass. But feel free to forward any question, and I can try to BS my way through it :^) -Eric PS: The books I mentioned are not cheap. Heywood runs in the $75 range, Bosch in the $45 range, and DaRosa, I'm not sure, but probably somewhere in between. --end include From owner-diy_efi Tue Nov 29 17:22:13 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA09625; Tue, 29 Nov 94 17:22:13 GMT Received: from centipede.cec.wustl.edu by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA09620; Tue, 29 Nov 94 12:22:10 -0500 Received: by centipede.cec.wustl.edu (5.0/ECL-A1.27) id AA21315; Tue, 29 Nov 1994 11:22:03 -0600 Date: Tue, 29 Nov 1994 11:22:01 -0600 (CST) From: Matthew Noel Ettus Subject: Re: gcc, 68K and ignition advance To: DIY_EFI In-Reply-To: Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 46 Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI what kind of engine can pull 950 liters/sec? From owner-diy_efi Tue Nov 29 17:32:00 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA09676; Tue, 29 Nov 94 17:32:00 GMT Received: from maxwell.ee.washington.edu by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA09671; Tue, 29 Nov 94 12:31:53 -0500 Received: by maxwell.ee.washington.edu (1.37.109.4/UW-NDC Revision: 2.26 ) id AA22497; Tue, 29 Nov 94 09:31:42 -0800 From: Mike Gruber Message-Id: <9411291731.AA22497@maxwell.ee.washington.edu> Subject: Re: gcc, 68K and ignition advance To: DIY_EFI Date: Tue, 29 Nov 94 9:31:42 PST In-Reply-To: ; from "Wouter de Waal" at Nov 29, 94 4:40 pm Mailer: Elm [revision: 70.85] Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI > My thoughts are as follows: > > * The air/fuel mix takes a constant _time_ to burn. This is incorrect. A dense mixture burns faster than a mixture at idle or cruise, which was inducted at low manifold pressure. This is why we have "vacuum" advance, to provide additional time for the mixture to burn when it is not dense. Under full power, with atmospheric pressure in the intake manifold, the mixture in the cylinders will be dense, and take less time to burn. The density of the mixture will also vary across the rpm band of the engine, even at a fixed throttle setting, due to the change in volumetric efficiency across the operating range. This is one reason that engines don't require as much extra advance at high rpm as you thought. They burn the mixture quicker at high rpm than low rpm, because they fill their cylinders better at 6000 rpm than at 1000 rpm, for the most part. > * In the same way, at 6000 rpm the advance should be 110 degrees (!). See above... > * And the next tricky point - vacuum advance. The vacuum is an indication > of engine load. But why is it a good thing to advance the timing when > the engine is loaded? By how much should the timing typically be advanced > for a given load? (And I mean _typically_, I know that the Bosch systems > amongst others use complicated, dyno-derived maps. However, I am more > interested in deriving the kind of curve a standard distributor gives.) Vacuum advance gives more advance when the engine is under *light* load, ie., high intake manifold vacuum. When the engine is loaded, the vacuum drops, and advance will *drop*. -- Mike Gruber '88 Supercharged MR2 (ASP) '72 Datsun 510 (In progress ... perpetually!) From owner-diy_efi Tue Nov 29 21:00:47 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA10375; Tue, 29 Nov 94 21:00:47 GMT Received: from uahcs2.cs.uah.edu by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA10367; Tue, 29 Nov 94 16:00:41 -0500 Received: (from kking@localhost) by uahcs2.cs.uah.edu (8.6.9/8.6.9) id PAA15559 for DIY_EFI@coulomb.eng.ohio-state.edu; Tue, 29 Nov 1994 15:00:39 -0600 Date: Tue, 29 Nov 1994 15:00:39 -0600 From: Ken King Message-Id: <199411292100.PAA15559@uahcs2.cs.uah.edu> To: DIY_EFI Subject: comments on ignition timing Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI >The thing sucks +- 950 liters/second @ 6000rpm, so it needs BIG air channels. my math: ((950 litres/sec)*(60 sec/min)) / (6000 rev/min) =9.5 litres/rev! however, 4 stroke engine takes 2 revolutions to complete, 19 litres or 1178 cubic inches! yowsa! even pre-supposing a huffer/turbo doubling the atmospheric pressure, that makes for a 580 cubic inches! mid-engined! if i can get out there, can i get a ride!?! :) i'm presumming one of us muffed the math... :) still, sounds like quite a car. a gt40? just curious... >The first thing that I would like to start playing with is ignition timing. sounds like a great place to start... consider how a normal distributor does timing & use it as a baseline to make additional refinements. perhaps using something like msd's boost timer (timing adjustment from drivers compartment) for twiddling with the curve... >My thoughts are as follows: >* The air/fuel mix takes a constant _time_ to burn. not true (by my sources... :) the denser the charge, the faster it will burn. why? not sure. as the rpm increaces, the number of crankshaft degrees spent burning will increace, so timing must be adjusted to get peak pressure at teh proper time (see below), timing must be advanced. also, as the load increaces, the density of the charge increaces, speeding the burn, which requires *less* timing (again, see further below). >* The max. pressure point has to occur when the piston is on it's > way down, and when the conrod has a decent leverage on the crank. correct. recent mag. article (hot rod?) stated 14-18 degrees is 'optimum'. >* Assume idle = 1000rpm, and advance = 10 degrees @ idle. 1000rpm = > 6000 degrees per second, so the spark occurs 1/600 second btdc. sounds reasonable. our math matches on this one. :) >* Assume that we want the maximum pressure at 10 degrees atdc. (Please, > this is the wildest guess I've made in a long time) closer than you think (at least by my sources). [ munch ] >* And the next tricky point - vacuum advance. The vacuum is an indication > of engine load. But why is it a good thing to advance the timing when > the engine is loaded? By how much should the timing typically be advanced > for a given load? (And I mean _typically_, I know that the Bosch systems > amongst others use complicated, dyno-derived maps. However, I am more > interested in deriving the kind of curve a standard distributor gives.) higher loads (low intake vacuum signal) helps to compress the charge, which causes the charge to burn much more rapidly, requiring less advance. without *less* advance, the location of the crank at peak pressure moves towards TDC, and could even occur *BEFORE* TDC, which is not a good thing tends to make for unhappy cranks! :( also the higher cyl. pressure can help a local hot spot (carbon deposit, flashing seam,...) to pre-ignite the charge (again, not a good thing :) so how does it work? in the vac advance canister, a vacuum diaphram works against a spring. when there is high vacuum (low load), advance is made because the vacuum diaphram pulls on a rod that rotates the contact plate (for points distributors (proper term?)) or pickup plate (electronic) oposite the direction of rotor rotation, causing the event to occur sooner. when high load conditions exist (lots of pressure on the loud pedal :), the vacuum in the manifold drops off, and the spring gains the upper hand over the vac diaphram & pushes the plate back where it 'belongs'. this is part of the reason why you are suposed to disconnect the vac advance hose when doing a tuneup (to allow for the setting the base timing). hopefully i haven't propagated too much mis-information. corrections are expected if i blew anything, flames cheerfully ignored. :) later, kc From owner-diy_efi Tue Nov 29 22:01:14 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA10518; Tue, 29 Nov 94 22:01:14 GMT Received: from geni9.arl.mil by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA10513; Tue, 29 Nov 94 17:01:10 -0500 Received: by lamp0.arl.army.mil (5.65/DEC-Ultrix/4.3) id AA02814; Tue, 29 Nov 1994 17:00:15 -0500 Message-Id: <9411292200.AA02814@lamp0.arl.army.mil> To: wrm@aztec.co.za Cc: diy_efi Subject: gcc, 68K and ignition advance Date: Tue, 29 Nov 94 17:00:14 -0500 From: faustini@lamp0.arl.army.mil X-Mts: smtp Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI Here is what little I know about ignition timing. 1) -most- of the work done by the burning gas is done in the first 15-20 degrees of cranksaft ATDC rotation.I think this is because the piston can "out run" the speed of the flame front. (does this make any sense? what is the typical velocity of the flame front?) 2) I read that it typicaly takes about 1 milisecond for the flame front to reach the top of the piston in the average combustion chamber. 3) The vaccum advance only advances the timing during part-throttle operatopn, which is why you get significant manifold vaccum. During WOT you typicaly see 1" or less vaccum. Therefore, I conclude that during part throttle the cylinders arent "filling" so you can get a little more efficency by starting the burn cycle early. I hope these wild and reclesly gathered facts help. ---- Lou From owner-diy_efi Wed Nov 30 04:55:47 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA12280; Wed, 30 Nov 94 04:55:47 GMT Received: from localhost.eng.ohio-state.edu by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA12275; Tue, 29 Nov 94 23:55:44 -0500 Message-Id: <9411300455.AA12275@coulomb.eng.ohio-state.edu> To: DIY_EFI Subject: EFI68k update and gcc for host=msdos target=m68k-coff available now! Date: Tue, 29 Nov 94 23:55:44 -0500 From: John S Gwynne Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI ----- ===================== (1) gcc for EFI68k: I've rebuilt my software environment for EFI68k using m68k-coff with the Cygnus gcc patches. This includes: gcc-2.6.2 w/Cygnus gcc patches for embedded controllers gdb-4.13 bin-utilities 2.5.2 (gas/ld/ar/as/nm/objdump...) I've also ported Cygnus's standard C library (newlib-1.6) but have not fully tested it yet. glibc-1.09 is out with embedded system support, but it still has a few bugs. Once it's fixed, it will be an easy port to EFI68k The next version of EFI68k is drawn up, but I have not had time to build/test it. The changes worked out nice as now I have everything need to support the CPU on one 4"x6" board. There will no longer be a companion board with more memory or CPU related functions. It now has everything to support every function in RTEM including clock/calendar/timers... I now cautiously expect the entire EFI68k DIY_EFI project will consist of two 4"x6" boards. ===================== (2) this was posted to one of the news groups: Subject: gcc for host=msdos target=m68k-coff available now! | I promised to write a message on this newsgroup when I had it available. | | There are now binaries (and the changed source files for anyone interested) | available via anon. ftp. The binaries are gcc 2.6.0 for djgpp (msdos) host | and 68000 family embedded target. The ftp address is: | | ftp.lysator.liu.se in directory /pub/msdos/gnu/gcc-dos-m68k | | Read the textfiles in the directory for further information | | Have fun! | /Oskar Enoksson John S Gwynne Gwynne.1@osu.edu _______________________________________________________________________________ T h e O h i o - S t a t e U n i v e r s i t y ElectroScience Laboratory, 1320 Kinnear Road, Columbus, Ohio 43212, USA Telephone: (614) 292-7981 * Fax: (614) 292-7292 ------------------------------------------------------------------------------- From owner-diy_efi Wed Nov 30 09:33:16 1994 Received: by coulomb.eng.ohio-state.edu (920330.SGI/920502.SGI) id AA12803; Wed, 30 Nov 94 09:33:16 GMT Received: from aztec.co.za by coulomb.eng.ohio-state.edu via SMTP (920330.SGI/920502.SGI) for /usr/local/mail/majordomo-1.92/wrapper resend -p bulk -M 10000 -l Diy_Efi -f Diy_Efi-Owner -h coulomb.eng.ohio-state.edu -s -r DIY_EFI diy_efi-outgoing id AA12798; Wed, 30 Nov 94 04:33:03 -0500 Received: by aztec.co.za (Smail3.1.28.1 #5) id m0rClOO-000KdfC; Wed, 30 Nov 94 11:32 EET Date: Wed, 30 Nov 1994 11:32:03 +0200 (SAT) From: Wouter de Waal Subject: Re: gcc, 68K and ignition advance To: DIY_EFI In-Reply-To: Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-diy_efi Precedence: bulk Reply-To: DIY_EFI On Tue, 29 Nov 1994, Matthew Noel Ettus wrote: > what kind of engine can pull 950 liters/sec? VW 1700. No big deal. All aircooled VWs suck about that much air, the 950 liters/sec at 6000rpm comes straight from the manual. You need a _lot_ of air to keep these babies cool at high ambient temperatures. Wouter