DIY_EFI Digest Tuesday, 5 January 1999 Volume 04 : Number 010 In this issue: Re: Water Injection Thread Re: Injector Duty Monitor incoming RE: Postscript aldl protocol Re: aldl protocol Re: Injector Duty Monitor Re: Water Injection Thread Re: aldl protocol Re: aldl protocol Re: aldl protocol Re: Injector Duty Monitor RE: ECM Wiring Diagram Injector Duty Monitor Re: Injector Duty Monitor Re: Injector Duty Monitor Re: Injector Duty Monitor Re: Water Injection Thread Slight corrrection on VSS Re: Water Injection Thread Turbine Discharge (was Water Injection Thread) Re: Slight corrrection on VSS Re: Slight corrrection on VSS Re: Slight corrrection on VSS Wide Band Lambda Sensor, where can I locate? GM PCM Auto Trans control Re: Slight corrrection on VSS Re: Water Injection Thread /turbo turbin See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- From: "Gary Derian" Date: Tue, 5 Jan 1999 08:02:27 -0500 Subject: Re: Water Injection Thread Thanks, Greg, for the good discussion. I never thought about the Mach 1 stuff in the exhaust housing. At Mach 1, though, flow through a nozzle is only dependant on density (upstream pressure). Velocity cannot increase with pressure drop but flow can still increase as density increases. In other words, sucking harder will not produce more flow but pushing harder will, albeit at a reduced rate (kink in curve) since velocity cannot increase. Gary Derian > >DO NOT misunderstand me, these wanderings are theoretical conjecture at >best--if I shut up and build something, test it, and it proves out, then I >would (and will!) be more forceful about it! >> >Regards, Greg > ------------------------------ From: "Gregory A. Parmer" Date: Tue, 5 Jan 1999 07:24:31 -0600 (CST) Subject: Re: Injector Duty Monitor > >For a relative sort of deal, see > >http://sura1.jlab.org/~bowling/fieye.html > >It is a *very* simple idea. > > I've seen the diagram and still confused somewhat. Anyway, is it > possible to wire a LED to the ECU side of the injector wire and just > look at the amount of time the LED is lighted. Perhaps, a continuous on Even the color route is relative. Brightness could be done, but I suspect it'd be less than useful. The bi-color LED is only a coupla bucks (unless you have to pay S&H). - -greg ------------------------------ From: rauscher@xxx.com Date: Tue, 05 Jan 99 08:29:48 -0500 Subject: incoming Clarence wrote: >Subject: Re: IAC controller circuit > > Wonderful! But, how do I access the 'incoming' site? > Point the 'ole browser at this: ftp://efi332.eng.ohio-state.edu/incoming/ Lots of goodies in there... BobR. - -- ------------------------------ From: Don.F.Broadus@xxx.com Date: Tue, 5 Jan 1999 07:56:58 -0600 Subject: RE: Postscript Hello David; Sorry it took so long to get back on line, the move was a mess, If you could let me know how much the OTC 2000 with shipping will be I will cut a check for you as soon as I get your address off list Thanks Don -----Original Message----- From: David A. Cooley [SMTP:n5xmt@xxx.net] Sent: Wednesday, December 30, 1998 10:32 AM To: diy_efi@xxx.edu Cc: frederic.breitwieser@xxx.com Subject: Re: Postscript At 09:23 PM 1/7/90 -0500, you wrote: >Not to be annoying, > >However I'm trying to view John Gwyne's 68HC000 schematic, as well as >several of hte EFI332 schematics off the DIYEFI website. I installed >Ghostview, a popular and robust postscript viewing application, as well as >attempted to import into visio, MS word, Excel, Powerpoint, etc, with no >luck. The schematic is entirely unreadable. > >has anyone converted John's diagram to a lame format like GIF, JPG, or even >better, PDF? > Some of the stuff on the FTP site I think was mis-named... I renamed some of the unviewable .ps files to .pdf and acrobat read them perfect! =========================================================== David Cooley N5XMT Internet: N5XMT@xxx.net Packet: N5XMT@xxx. Member #7068 I am Pentium of Borg...division is futile...you will be approximated. =========================================================== ------------------------------ From: Pat Ford Date: Tue, 5 Jan 1999 09:10:56 -0500 Subject: aldl protocol Hi Folks; does anybody have the protocol used by gm on the aldl port. i have a 92 tracker and want to play around with some settings THANKS Pat Ford email: pford@xxx.com QNX Software Systems, Ltd. WWW: http://www.qnx.com (613) 591-0931 (voice) mail: 175 Terrence Matthews (613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 09:56:01 -0500 Subject: Re: aldl protocol - -----Original Message----- From: Pat Ford To: diy_efi@xxx.edu> Date: Tuesday, January 05, 1999 9:33 AM Subject: aldl protocol There is lots of info about the ALDL, in the archives, but, are you trying to access the calibrations thru the ALDL?. The only ones that I know that have that access, are the late Flash Prom series. If your's does allow that would you share how it's done. The ALDL can be used for diagnostics, and a scanner can be had for less than the time to invent the link, well again in the archives is alotta material. Bruce > Hi Folks; > does anybody have the protocol used by gm on the aldl port. >i have a 92 tracker and want to play around with some settings > >THANKS > >Pat Ford email: pford@xxx.com >QNX Software Systems, Ltd. WWW: http://www.qnx.com >(613) 591-0931 (voice) mail: 175 Terrence Matthews >(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 > ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 10:48:17 -0500 Subject: Re: Injector Duty Monitor - -----Original Message----- From: Gregory A. Parmer To: diy_efi@xxx.edu> Date: Tuesday, January 05, 1999 8:46 AM Subject: Re: Injector Duty Monitor On the ecm bench I use LED's and much over idle they are just pretty much on. IMHO, to do anything meaningful means actually measuring the pulses width. Bruce. > >> >For a relative sort of deal, see >> >http://sura1.jlab.org/~bowling/fieye.html >> >It is a *very* simple idea. >> >> I've seen the diagram and still confused somewhat. Anyway, is it >> possible to wire a LED to the ECU side of the injector wire and just >> look at the amount of time the LED is lighted. Perhaps, a continuous on > >Even the color route is relative. Brightness could be done, but >I suspect it'd be less than useful. The bi-color LED is only >a coupla bucks (unless you have to pay S&H). >-greg > ------------------------------ From: bearbvd@xxx.net (Greg Hermann) Date: Tue, 5 Jan 1999 08:45:43 -0700 Subject: Re: Water Injection Thread >Thanks, Greg, for the good discussion. I never thought about the Mach 1 >stuff in the exhaust housing. At Mach 1, though, flow through a nozzle is >only dependant on density (upstream pressure). Velocity cannot increase >with pressure drop but flow can still increase as density increases. In >other words, sucking harder will not produce more flow but pushing harder >will, albeit at a reduced rate (kink in curve) since velocity cannot >increase. Also dependent upon temperature, since Mach 1 increases with increasing temp, but you knew that--- And since increasing the back-pressure on an engine will increase the EGT (due to less downward T/P shock during blowdown), at least with a turbocharger turbine nozzle, the flow will increase a bit more linearly with increasing backpressure than your statement implies when dealing with an engine /turbocharger pairing. (I have a FAIRLY decent handle on what is happening qualitatively here, but PLEASE don't ask for NUMBERS!!!) The "pushing harder" (higher density) to get more flow through the nozzle part would mean more back-pressure on the recip part of the engine, which would mean more pumping losses, and hence less net output. (which, in turn, translates to lower efficiency), which brings us back full circle! Exactly why I SUSPECT that a well done pressure recovery accumulator on a good set of turbo headers might make for a net gain!! Thanks for the discussion, Gary. The back and forth definitely stretches the knowlege envelope quite a bit!! WHICH IS FUN!!! I think the REAL unexplored gain is in (more efficient) pressure recovery from the Mach .75 velocity in the gas spiral at the turbine outlet, though. The circumferential component of the flow in that gas is distinctly greater than the axial component. Some way of straightening the helical flow out, and turning all (more) of that velocity back into static pressure would truly be a FREE increase in the pressure ratio across the turbine, with all of the performance improvements that that implies!!! Plus, I do not think that a device for doing it would take up all that much space! Also, no moving parts involved and decently low temps, so ALMOST as reliable as taxes! > Greg >Gary Derian > > > >> >>DO NOT misunderstand me, these wanderings are theoretical conjecture at >>best--if I shut up and build something, test it, and it proves out, then I >>would (and will!) be more forceful about it! >>> >>Regards, Greg >> ------------------------------ From: Pat Ford Date: Tue, 5 Jan 1999 11:09:25 -0500 Subject: Re: aldl protocol On Tue, 5 Jan 1999, Bruce Plecan wrote: > Date: Tue, 5 Jan 1999 09:56:01 -0500 > From: Bruce Plecan > To: diy_efi@xxx.edu > Subject: Re: aldl protocol > > > -----Original Message----- > From: Pat Ford > To: diy_efi@xxx.edu> > Date: Tuesday, January 05, 1999 9:33 AM > Subject: aldl protocol > > There is lots of info about the ALDL, in the archives, but, are you trying > to access > the calibrations thru the ALDL?. The only ones that I know that have that I'm trying to down load the calibrations, planning on eventualy doing a custom controller > access, > are the late Flash Prom series. If your's does allow that would you share > how it's > done. when i pop the ecu open I'll report whats inside ( I'll try to draw a schematic) > The ALDL can be used for diagnostics, and a scanner can be had for less > than the time to invent the link, wheres the fun in that 8*) , actualy this is a learning project as well, I am a class a mechanic and now a programmer, and want to try programming this beast >well again in the archives is alotta is there a search engine, and what are any helpful shortcuts thanks > material. > Bruce > > > > Hi Folks; > > does anybody have the protocol used by gm on the aldl port. > >i have a 92 tracker and want to play around with some settings > > > >THANKS > > > >Pat Ford email: pford@xxx.com > >QNX Software Systems, Ltd. WWW: http://www.qnx.com > >(613) 591-0931 (voice) mail: 175 Terrence Matthews > >(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 > > > > Pat Ford email: pford@xxx.com QNX Software Systems, Ltd. WWW: http://www.qnx.com (613) 591-0931 (voice) mail: 175 Terrence Matthews (613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 11:36:19 -0500 Subject: Re: aldl protocol - -----Original Message----- From: Pat Ford To: diy_efi@xxx.edu> Date: Tuesday, January 05, 1999 11:22 AM Subject: Re: aldl protocol Go to http://efi332.eng.ohio-state.edu/diy_efi hit archives, and seard title or content for aldl gcar diagnostics programming 101 programming 808, then just read the whole archives a couple times, HOHOHAHA, welcome Bruce > >I'm trying to down load the calibrations, planning on eventualy doing a >custom controller > >when i pop the ecu open I'll report whats inside ( I'll try to draw a >schematic) > >> The ALDL can be used for diagnostics, and a scanner can be had for less >> than the time to invent the link, > >wheres the fun in that 8*) , actualy this is a learning project as well, >I am a class a mechanic and now a programmer, and want to try programming >this beast > >>well again in the archives is alotta > >is there a search engine, and what are any helpful shortcuts > >thanks > > >> material. >> Bruce >> >> >> > Hi Folks; >> > does anybody have the protocol used by gm on the aldl port. >> >i have a 92 tracker and want to play around with some settings >> > >> >THANKS >> > >> >Pat Ford email: pford@xxx.com >> >QNX Software Systems, Ltd. WWW: http://www.qnx.com >> >(613) 591-0931 (voice) mail: 175 Terrence Matthews >> >(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 >> > >> >> > >Pat Ford email: pford@xxx.com >QNX Software Systems, Ltd. WWW: http://www.qnx.com >(613) 591-0931 (voice) mail: 175 Terrence Matthews >(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 > ------------------------------ From: "Ronald T. Webb" Date: Sat, 04 Jan 1997 07:56:33 -0900 Subject: Re: aldl protocol I've got the engine/ecm/etc for a '92 tracker, that I will be using in a homebuilt aircraft project. The ECM is NOT GM. It is made by Mitsubishi, based on the MELPS 7700 microcontroller (6502 upgrade popular in Japan). The components are proprietary, and I cannot even identify the EPROM with certainty. Mitsubishi's web site is quite good, with a lot of data sheets for these chips, but the chip that I think is the EPROM does not compute...It may be a RIOT chip (ROM-I/O-Timer). It would be interesting to follow your project, and I may be able to help... Pat Ford wrote: > On Tue, 5 Jan 1999, Bruce Plecan wrote: > > > Date: Tue, 5 Jan 1999 09:56:01 -0500 > > From: Bruce Plecan > > To: diy_efi@xxx.edu > > Subject: Re: aldl protocol > > > > > > -----Original Message----- > > From: Pat Ford > > To: diy_efi@xxx.edu> > > Date: Tuesday, January 05, 1999 9:33 AM > > Subject: aldl protocol > > > > There is lots of info about the ALDL, in the archives, but, are you trying > > to access > > the calibrations thru the ALDL?. The only ones that I know that have that > > I'm trying to down load the calibrations, planning on eventualy doing a > custom controller > > > access, > > are the late Flash Prom series. If your's does allow that would you share > > how it's > > done. > > when i pop the ecu open I'll report whats inside ( I'll try to draw a > schematic) > > > The ALDL can be used for diagnostics, and a scanner can be had for less > > than the time to invent the link, > > wheres the fun in that 8*) , actualy this is a learning project as well, > I am a class a mechanic and now a programmer, and want to try programming > this beast > > >well again in the archives is alotta > > is there a search engine, and what are any helpful shortcuts > > thanks > > > material. > > Bruce > > > > > > > Hi Folks; > > > does anybody have the protocol used by gm on the aldl port. > > >i have a 92 tracker and want to play around with some settings > > > > > >THANKS > > > > > >Pat Ford email: pford@xxx.com > > >QNX Software Systems, Ltd. WWW: http://www.qnx.com > > >(613) 591-0931 (voice) mail: 175 Terrence Matthews > > >(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 > > > > > > > > > Pat Ford email: pford@xxx.com > QNX Software Systems, Ltd. WWW: http://www.qnx.com > (613) 591-0931 (voice) mail: 175 Terrence Matthews > (613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8 ------------------------------ From: steve ravet Date: Tue, 05 Jan 1999 11:02:44 -0600 Subject: Re: Injector Duty Monitor New pwmeter.zip file uploaded to the ftp site in incoming. This one has a .gif and .pdf version of the schematic in addition to the .ps version. Plus, Georg wrote me a note explaining a little more how the circuit works, which I've put into the pwmeter.txt file. I'll also include his note here: - ------------------------------------------------------- As for the description, it might be a long one, but i will try and keep it short. If you look at the schematic, you will see that the signal can be taken from across the injector or the ecu[switchable]. It is optically isolated from the vehicle to eliminate any problems. The signal is inverted by the mosfet and the original and inverted signal is fed to the interrupt pins of the micro. The reason for this is that the int' pins are neg. edge triggered, and because we want to measure pulses widths and period, this is the way it has to be configured. Now all the micro bit does is when it gets triggered when the injector is switched on , a previously configured 16 bit counter starts a count with 1us rate. When the injector is shut off, the counter stops. This is all done with interrupts, so that all the other tasks the micro has to do can carry on without spending time in timing loops, ie. handles display routines when not measuring. Now that the counter has a count eg. 12354 counts x 1us = 12.345ms This get saved and the counter reset to for the next injector pulse. The saved value gets displayed on the LCD display in milliseconds. Updating of the display is done as and when pulse arrive. If no pulses arrive the display will hold the last pulse width. If you look at the code you will see the bulk of the code is to handle the lcd wrt. to formatting the output of it. Now we could modify the code to measure the period between injector pulses, we could calculate the duty cycle from that in terms of on/off time and percentage. I nfact you could display all of this on a 2 line display.That is on-time[ms], off time[ms] and duty cycle[%]. If you have eperience with micro's, it should be easy. - ------------------------------------------------------------------ So, this is a circuit that measures the pulse width and displays it on an LCD screen. Some small changes would make it calculate duty cycle as well. The LCD is part number M1632. I found 49 of them in stock for $13 ea at http://www.eio.com/lcd.htm I've added all this info to a new pwmeter.zip and uploaded it to ftp://efi332.eng.ohio-state.edu/incoming/pwmeter.zip. You still need a device programmer to program the atmel microcontroller though... - --steve ------------------------------ From: Don.F.Broadus@xxx.com Date: Tue, 5 Jan 1999 11:04:38 -0600 Subject: RE: ECM Wiring Diagram Hello David; Just got back on line today. Off line could you please give me the cost for the OTC 2000 including shipping And I will cut you a check. Thanks Don -----Original Message----- From: David A. Cooley [SMTP:n5xmt@xxx.net] Sent: Tuesday, December 29, 1998 3:03 PM To: diy_efi@xxx.edu Subject: Re: ECM Wiring Diagram At 03:46 PM 12/29/98 -0500, you wrote: > >I'm confused by this. If the white wire doesn't go to the ECM, where does it >go and what is it's purpose? Can I still get it from inside the car somewhere >like I originally thought, or will I have to run that dreaded wire out to the >coil? It runs from the ignition module to the ignition coil. There is an ignition pulse feed to the ECM in the car, but it is a VERY low current signal and loading it with an accessory such as a remote starter could cause problems (IE Burn up the ECM)... If there was a factory option for a tach, the wiring harness should have a wire leading to the dash just waiting for a tach... they liked to use one harness for all the vehicles so as not to have different parts for the same make/model. A wiring diagram would help you a lot! If there was no in-dash tach option, then you will need to tap into the white wire that goes from the distributor module to the - side of the ignition coil. When you run it through the firewall of the car, make sure you use a grommet! Later, Dave =========================================================== David Cooley N5XMT Internet: N5XMT@xxx.net Packet: N5XMT@xxx. Member #7068 I am Pentium of Borg...division is futile...you will be approximated. =========================================================== ------------------------------ From: rauscher@xxx.com Date: Tue, 05 Jan 99 13:10:00 -0500 Subject: Injector Duty Monitor On this subject, what is it that we really want to measure? Duty cycle or the pulse width? The two designs presented (B. Bowling and C Hack), are nice. But getting access to a PIC programmer can be a problem. JDR carries a bunch of PIC's, and the OTP 16C56 can be had for $4.50US, not bad. A AT89C2051 w/flash is $7.00US, plus Christian's design reads out on an LCD screen, nice again. If there's enough interest, would someone with a PIC programmer burn up a bunch for a group purchase? The other thing that can be done, is to use some analog stuff, integrate the pulse, and read it out with a DVM. Have it able to measure from say, 0 msec up to a max of 20 msec? If that's what we need, I'll do the design. Anybody? BobR. - -- ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 14:04:02 -0500 Subject: Re: Injector Duty Monitor - -----Original Message----- From: rauscher@xxx.com> To: DIY_EFI@xxx.edu> Date: Tuesday, January 05, 1999 1:25 PM Subject: Injector Duty Monitor IF your doing anything serious ya need pulse width, with at least 20.x millisecond maybe higher (more that 20), and a resolution of at least .x millisecond. I'd perfer .xx, with possibly an adjustment for dampening the flickering of numbers. Reading with a dwell meter is OK, and 1000x better than nothing but that's about it. Bruce > > > >On this subject, what is it that we really want to >measure? Duty cycle or the pulse width? > >The two designs presented (B. Bowling and C Hack), >are nice. But getting access to a PIC programmer >can be a problem. JDR carries a bunch of PIC's, and >the OTP 16C56 can be had for $4.50US, not bad. A >AT89C2051 w/flash is $7.00US, plus Christian's >design reads out on an LCD screen, nice again. > > >If there's enough interest, would someone with a >PIC programmer burn up a bunch for a group purchase? > > >The other thing that can be done, is to use some >analog stuff, integrate the pulse, and read it out >with a DVM. Have it able to measure from say, 0 msec >up to a max of 20 msec? > >If that's what we need, I'll do the design. > >Anybody? > > >BobR. > >-- > > > > > ------------------------------ From: steve ravet Date: Tue, 05 Jan 1999 13:10:03 -0600 Subject: Re: Injector Duty Monitor rauscher@xxx.com wrote: > > On this subject, what is it that we really want to > measure? Duty cycle or the pulse width? Duty cycle, right? The time available to inject fuel decreases with increasing RPM. So (for example) a 1 ms pulse at idle, every 20 ms is a lot different than a 1 ms pulse at a higher RPM every 5 ms. Just picked those numbers out of the air. Don't think of it as PWM, where the frequency never changes. > > The two designs presented (B. Bowling and C Hack), > are nice. But getting access to a PIC programmer > can be a problem. JDR carries a bunch of PIC's, and > the OTP 16C56 can be had for $4.50US, not bad. A > AT89C2051 w/flash is $7.00US, plus Christian's > design reads out on an LCD screen, nice again. It's Georg's design, if you're talking about pwmeter.zip. Sounds like the whole thing could be built for $40 or so. I'd be willing to buy one if someone can program the atmel parts, and if someone can modify the program to display both on time and duty cycle. - --steve ------------------------------ From: Orin Eman Date: Tue, 5 Jan 1999 11:45:24 -0800 (PST) Subject: Re: Injector Duty Monitor > If there's enough interest, would someone with a > PIC programmer burn up a bunch for a group purchase? I can burn PICs if necessary. Orin. ------------------------------ From: Clarence Wood Date: Tue, 05 Jan 1999 15:24:09 -0600 Subject: Re: Water Injection Thread At 08:45 AM 1/5/99 -0700, you wrote: (snip) >I think the REAL unexplored gain is in (more efficient) pressure recovery >from the Mach .75 velocity in the gas spiral at the turbine outlet, though. >The circumferential component of the flow in that gas is distinctly greater >than the axial component. Some way of straightening the helical flow out, >and turning all (more) of that velocity back into static pressure would >truly be a FREE increase in the pressure ratio across the turbine, with all >of the performance improvements that that implies!!! Plus, I do not think >that a device for doing it would take up all that much space! Also, no >moving parts involved and decently low temps, so ALMOST as reliable as >taxes! >> >Greg I was just reading about the "Banjo-type" turbine discharge, in Hugh MacInnes's Turbochargers. On page 72 he states that Dennis Sevier, Chief Engineer of Bristol replaced an elbow, which cost power on acceleration and at top end, with the banjo-type fitting and found that it "worked at least as well as a straight exhaust". Does anybody know where a feller could get a banjo-type fitting? IZCC #3426 1982 280ZX Turbo GL 1966 El Camino 1982 Yamaha Maxim XJ-1101J Motorcycle 1975 Honda CB750 SS (black engine) 1986 Snapper Comet lawn mower Clarence Wood Software&Such... clarencewood@xxx.net Savannah, TN. ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 17:36:34 -0500 Subject: Slight corrrection on VSS Some days, weeks, months ago, some one mentioned that the VSS has nothing to due with engine fueling, well I wanted to make sure, but yes it does on the BCC AUJP. Closed throttle with a VSS signal decreasing, does do a 0 injector oulse width, according to diacom. Also. appears in some 87-89 165 Vette stuff. Those were not stock (vette) but I'd doubt a chip burner had the knowledge to invoke it. It also, did this open or closed loop. Cheers Bruce What's that Doc?, if I hit send it goes to the list?. OK, let's see delete or send, oops ------------------------------ From: bearbvd@xxx.net (Greg Hermann) Date: Tue, 5 Jan 1999 16:01:07 -0700 Subject: Re: Water Injection Thread >At 08:45 AM 1/5/99 -0700, you wrote: >(snip) >>I think the REAL unexplored gain is in (more efficient) pressure recovery >>from the Mach .75 velocity in the gas spiral at the turbine outlet, though. >>The circumferential component of the flow in that gas is distinctly greater >>than the axial component. Some way of straightening the helical flow out, >>and turning all (more) of that velocity back into static pressure would >>truly be a FREE increase in the pressure ratio across the turbine, with all >>of the performance improvements that that implies!!! Plus, I do not think >>that a device for doing it would take up all that much space! Also, no >>moving parts involved and decently low temps, so ALMOST as reliable as >>taxes! >>> >>Greg > > I was just reading about the "Banjo-type" turbine discharge, in Hugh >MacInnes's Turbochargers. On page 72 he states that Dennis Sevier, Chief >Engineer of Bristol replaced an elbow, which cost power on acceleration >and at top end, with the banjo-type fitting and found that it "worked at >least as well as a straight exhaust". > Does anybody know where a feller could get a banjo-type fitting? Nope, but I would be interested too!! I have been pondering on those banjos too, also thinking about straintening vanes inside a backwards megaphone with a cone inside, and some other ideas that would really cause a fabricator to cuss loudly!! All thoughts on this welcome!! To give you all some idea of what I am talking about, on a 350 HP engine, there is about 29 HP worth of energy just in the momentum (calculated above the momentum of the gas in an exhaust pipe flowing at 200 fps.) of the gas leaving the turbo turbine!!! (Not talking about heat energy here, just momentum!!!) If you could recover just recover 60% of that , that would be about a 5% improvement in output, essentially for FREE !!! Let's get some ideas flowing!! Regards, Greg ------------------------------ From: Aaron Willis Date: Tue, 05 Jan 1999 15:04:56 -0800 Subject: Turbine Discharge (was Water Injection Thread) Clarence Wood wrote: > > At 08:45 AM 1/5/99 -0700, you wrote: > (snip) > >I think the REAL unexplored gain is in (more efficient) pressure recovery > >from the Mach .75 velocity in the gas spiral at the turbine outlet, though. > >The circumferential component of the flow in that gas is distinctly greater > >than the axial component. Some way of straightening the helical flow out, > >and turning all (more) of that velocity back into static pressure would > >truly be a FREE increase in the pressure ratio across the turbine, with all > >of the performance improvements that that implies!!! Plus, I do not think > >that a device for doing it would take up all that much space! Also, no > >moving parts involved and decently low temps, so ALMOST as reliable as > >taxes! > >> > >Greg > > I was just reading about the "Banjo-type" turbine discharge, in Hugh MacInnes's Turbochargers. On page 72 he states that Dennis Sevier, Chief Engineer of Bristol replaced an elbow, which cost power on acceleration and at top end, with the banjo-type fitting and found that it "worked at least as well as a straight exhaust". > Does anybody know where a feller could get a banjo-type fitting? > IZCC #3426 > 1982 280ZX Turbo GL > 1966 El Camino > 1982 Yamaha Maxim XJ-1101J Motorcycle > 1975 Honda CB750 SS (black engine) > 1986 Snapper Comet lawn mower > Clarence Wood > Software&Such... > clarencewood@xxx.net > Savannah, TN. Damn, Clarence, you beat me to that one. This is something I'm very interested in too - any leads?? Aaron ICQ # 27386985 ------------------------------ From: steve ravet Date: Tue, 05 Jan 1999 17:39:18 -0600 Subject: Re: Slight corrrection on VSS Bruce Plecan wrote: > > Some days, weeks, months ago, some one mentioned that the VSS > has nothing to due with engine fueling, well I wanted to make sure, but yes > it does on the BCC AUJP. Closed throttle with a VSS > signal decreasing, does do a 0 injector oulse width, according to diacom. > Also. appears in some 87-89 165 Vette stuff. Those > were not stock (vette) but I'd doubt a chip burner had the knowledge to > invoke it. That's decel cutoff, isn't it? I used to own an '89 vette with the digital dash. One of the many things the dash would do was read out instantaneous mpg. I distinctly remember that if you let off the gas, say in second gear and around 30 mph (like slowing for a stop sign), about a half second after getting off the gas there would be a small lurch, and the mpg meter would peg at 99. When you slowed to the point that rpm was almost idle, like 1200 or so, there was a much more noticeable lurch and the mpg meter would drop to a reasonable value. The service manual also confirmed that the ECM would shut off fuel delivery under some deceleration conditions. Mine was standard, does the automatic bin do this? seems like an auto could stall without the drivetrain keeping the engine turning. - --steve > It also, did this open or closed loop. > Cheers > Bruce What's that Doc?, if I hit send it goes to the list?. OK, let's > see delete or send, oops ------------------------------ From: Roger Heflin Date: Tue, 5 Jan 1999 18:06:11 -0600 (CST) Subject: Re: Slight corrrection on VSS On Tue, 5 Jan 1999, steve ravet wrote: > > > > That's decel cutoff, isn't it? I used to own an '89 vette with the > digital dash. One of the many things the dash would do was read out > instantaneous mpg. I distinctly remember that if you let off the gas, > say in second gear and around 30 mph (like slowing for a stop sign), > about a half second after getting off the gas there would be a small > lurch, and the mpg meter would peg at 99. When you slowed to the point > that rpm was almost idle, like 1200 or so, there was a much more > noticeable lurch and the mpg meter would drop to a reasonable value. > The service manual also confirmed that the ECM would shut off fuel > delivery under some deceleration conditions. Mine was standard, does > the automatic bin do this? seems like an auto could stall without the > drivetrain keeping the engine turning. > > --steve I did notice there is some code in the TCC lockup routines that do things different based on coasting above or below a certain speed. Also torque convertors work just fine in reverse, so at higher speeds the torque convertor will still be turning the engine just not a 1:1 speed. At lower speeds the computer may be doing something to keep it from stalling. Mine has not dropped below idle speed while I am coasting so it is doing someting. Roger 93 Z28 A4 ------------------------------ From: "David A. Cooley" Date: Tue, 05 Jan 1999 19:16:31 -0500 Subject: Re: Slight corrrection on VSS At 05:39 PM 1/5/99 -0600, you wrote: >That's decel cutoff, isn't it? I used to own an '89 vette with the >digital dash. One of the many things the dash would do was read out >instantaneous mpg. I distinctly remember that if you let off the gas, >say in second gear and around 30 mph (like slowing for a stop sign), >about a half second after getting off the gas there would be a small >lurch, and the mpg meter would peg at 99. When you slowed to the point >that rpm was almost idle, like 1200 or so, there was a much more >noticeable lurch and the mpg meter would drop to a reasonable value. >The service manual also confirmed that the ECM would shut off fuel >delivery under some deceleration conditions. Mine was standard, does >the automatic bin do this? seems like an auto could stall without the >drivetrain keeping the engine turning. Yep.. The auto's do it as well. The MPH/RPM points are different to ensure the engine won't stall though. =========================================================== David Cooley N5XMT Internet: N5XMT@xxx.net Packet: N5XMT@xxx. Member #7068 I am Pentium of Borg...division is futile...you will be approximated. =========================================================== ------------------------------ From: "Ezra Hall" Date: Tue, 5 Jan 1999 19:27:42 -0500 Subject: Wide Band Lambda Sensor, where can I locate? This is my first post to this list. I am on digest mode. Nice to join this group of knowledgeable folks! I am an electrical engineer by profession, but I am working on an automotive project (in my spare time,haha ) and need a wide range Lambda sensor for it. The Bosche web site describes a Planer wide-band Lambda sensor, but all of my calls to their corporation have proved useless with regards to obtaining a P/N for a sensor that will work on my vehicle (the electrical connector is not important, I can solder). Any suggestions as to which vehicles use Lambda sensors with a wide range? I would like it to extend at least to a 12.6:1 Air/Fuel ratio. I know these beasts exist. Thanks for any assistance you can provide! Ezra Hall ehall@xxx.net ------------------------------ From: "Ross Myers" Date: Tue, 5 Jan 1999 15:41:52 +1100 Subject: GM PCM Auto Trans control Has anybody been playing around with the P6? PCM's that also control trans shifts, not just TCC lockup, but 1-2, 2-3 etc shifts?. If so, is the calibrations for shift firmness & shift points like a standard table?. It seems most cars have a Econ/Pwr setting for trans control, although, maybe I shouldn't find this out cause I'd have axle snapping / front wheel launch setting instead ;-) Ross Myers ------------------------------ From: "Bruce Plecan" Date: Tue, 5 Jan 1999 19:35:13 -0500 Subject: Re: Slight corrrection on VSS - -----Original Message----- From: steve ravet To: diy_efi@xxx.edu> Date: Tuesday, January 05, 1999 7:20 PM Subject: Re: Slight corrrection on VSS snip The AUJP is an auto calibration Bruce Mine was standard, does >the automatic bin do this? seems like an auto could stall without the >drivetrain keeping the engine turning. > >--steve > ------------------------------ From: "Espen Hilde" Date: Wed, 6 Jan 1999 01:30:35 +0100 Subject: Re: Water Injection Thread /turbo turbin > > >Thanks, Greg, for the good discussion. I never thought about the Mach 1 > >stuff in the exhaust housing. At Mach 1, though, flow through a nozzle is > >only >> I have allso never thought about this point my self, thanks Greg. I think that the design of manifolds must follow application. in a regular standard car with backpressure exiding boost pressure at high rpms, where you need fast spool up as possible I think your better of with a manifold with as little volume as possible,without destroing the flow to much. This means little heat loss from manifold ,(fast spool up)When the exhaust has passed the turbin it can not do any negative backpressure work anymore. There fore the small volume of manifold is a bennefit in the high backpressure aria.At least I think it will work over a greater aria than the timing of pulses in a tubular manifold.It would be interesting to discower how a turbo with one turbin for each cyl.mounted just outside the exhaus outlet of the engine head would work.As you say , the pulses will be strong and evently break the mach 1 . many small turbines calls for high leakage .maybe not a god Idee after all. In a turbo aplication with focus on maximum output I would go for headers. When the turbine is mached for max efficensy at high rpm.At what point do we want the pulses to help us?At the point just before the boost starts? Long not isolated headers will lose a lot of the precius energy we have when we need it as most, when boost is starting , this means less efficient aria, at high rpm this lost energy does not play a major role, maybe we want it to cool the exhaust as much as possible , because we have to much of it.(We have discussed earlier the possibility of insulating the inside of the exhaust tubes....was it possible?) What about using a twin entry turbo and connect the two enties with an extra wast gate walve at the point where backpressure exides boost pressure.This will increase the turbine envelope,and that is the main problem with turbo chargers .The efficient turbine work aria is to small . Maybe thats why chryslers turbine car was no sucsess. Garret has started massproducing turbines with wariable nossle, you will not need a wastegate any more... No more dumping of energy. Many do a modification to the turbines( to expand their ranges?) some cut/clip the wings on the outlet part of turbine wheel , some increase the outlet circumference diameter to a bigger dimension than the turbine wheel. Its a pitty that turbine efficency charts is difficult to get ,all I have seen is compressor maps , their range of efficency is much greater..... Backpressure will increase faster if you have a little turbine wheel with the greatest nossle/exhaust hausing (mach 1 at more places on the way trou the turbine?), then is better to have one of the smallest exhaust housings on a big turbine wheel, I think this gives you not the best peek efficiency but a wider aria. I have seen two stroke engines with tuned expansion cambers and turboes no the end, tremendus output! Drag snowsleds with efi and tuned expansion cambers and turbo.....WOW Maybe we must alter the engine to give a strong exhaust puls to take full advantage of this in a 4 stroke. With the hat safe on my head.(cone shape) espen Please run an english spell check before reading.....I dont have one.....it shows...... ------------------------------ End of DIY_EFI Digest V4 #10 **************************** 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".