DIY_EFI Digest Wednesday, January 12 2000 Volume 05 : Number 017 In this issue: Knock Sensing 68HC11 Disassembler... Re: GMECM and Crank Sensors RE: DIY_EFI Digest V5 #16 Re: DIY_EFI Digest V5 #16:Knock Sensor Re: DIY_EFI Digest V5 #16:Knock Sensor Re: DIY_EFI Digest V5 #16:Knock Sensor Re: Knock sensor on dyno See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. ---------------------------------------------------------------------- Date: Tue, 11 Jan 2000 13:43:00 -0800 From: "Buchholz, Steven" Subject: Knock Sensing Some time ago I found a reasonable reference on knock detection on TI's website ... I checked and found it is still there ... here's an abstract: ENGINE KNOCK DETECTION USING SPECTRAL ANALYSIS WITH TMS320C25 OR TMS320C30 DSPS - An efficient method of detecting combustion engine knock is using spectral analysis. The detection process algorithm adapts to a no-knock reference at varying speeds and loads by using multiple frequencies. This document presents an problem overview, current technology, and two implementation examples are given to aid in the development of system specific hardware and software. The first system is based on the 32-bit floating-point TMS320C30 DSP with software written in C. The second is production oriented using a TMS320C25 16-bit fixed-point DSP and assembler software. The appendix contains the code for each implementation. http://www.ti.com/sc/docs/psheets/abstract/apps/spra039.htm I have been contemplating the addition of a knock sensor to an engine which did not originally have one so that I can start tweaking the performance. While this reference does give a good background into the whys and hows from a DSP standpoint, it doesn't go into strategies for mounting sensor(s) for best results. Are there any references that discuss this sort of issue? Does anyone know any companies who actually make the sensors? I'd love to see specs for typical sensors (frequency response and such). The TI reference states that there are two types of "indirect" knock sensors, ones that are tuned to specific frequencies and broadband units. From what I've seen I'm thinking that most if not all of the knock sensors I've seen have been tuned units to minimize the amount of subsequent processing required. It seems to me that if one were contemplating adding a knock sensor to an engine, the broadband approach might be the preferred solution. Steve Buchholz San Jose, CA (USA) ------------------------------ Date: Tue, 11 Jan 2000 23:47:39 +0000 From: Donald Whisnant Subject: 68HC11 Disassembler... For people who have wanted to download my HC11 disassembler during this past month-and-a-half or so and found that my website was down, I just wanted to let you know that it is back up and running and appologize for any inconvenience... I moved to switch jobs and am now on a different internet service, though from the user perspective, the site appears the same... The current address for the front-end of the site is: http://dewhisna.home.netcom.com The HC11 disassembler link can be found on the download page: http://dewhisna.home.netcom.com/download.html My old ISP still has my site active, though I don't know how long they will keep it active... There you'll only find a link to this new location... And unfortunately, all of the search engines still only have the old site listed... I will be putting up a mirror for this new front-end of the site once I get the details worked out with my new ISP... And I eventually plan to register a name for the site... But in the meantime, use the above address and enjoy... As for the next version (1.2), it is basically complete except for the manual re-writes -- I've been too busy with the job switch and relocation to get it completed yet -- but it's coming (I promise)... Donald Whisnant dewhisna@xxx.com ------------------------------ Date: Tue, 11 Jan 2000 19:02:23 -0500 From: "Bruce Plecan" Subject: Re: GMECM and Crank Sensors First one that jumps to mind is the late vortec. You really need to find a parts house that has some catalogs with photos for part ID'ing. They do exist from Delco, and Standard Electronics. Grumpy | For yet another project, is there a nice reliable, relatively easy to mount crank sensor that interfaces easily to the GM ECM's you favor? | In this case a dist is not an option. | I have provisions to build a mount, make/adapt a crank wheel of some type or precisely install magnets in a pulley or even a flywheel. | Does the GM DIS have 5 or 6k rpm limits, something I read somewhere indicated that, but they may have been talking about HEI coils? Nope, check out the GN Type list info.. Almost stockers can run to 6.5K. GM includes a "rev limit" to corral warranty costs, ie weak valve springs, limited injector size. Too often that gets credit as a weak ignition. All this stuff I am playing with needs to make it | to 6.5K, the race app at least 8500. | | Dopey | | >about the DIS, I'll scan a diagram showing the notches/offset/which notch | >| >fires which coil pack etc.. | >| >Grumpy | > | >------------------------------ | ------------------------------ Date: Tue, 11 Jan 2000 17:02:00 -0800 From: "Stowe, Ted-SEA" Subject: RE: DIY_EFI Digest V5 #16 howdy and thanks for the tip to go look at the fuel injection parts MSD has. (I was talking about mounting injectors. as some of you probably know, msd has some flexible fuel hose hardware, "injector manifold pocket, epoxy-in" pn 2145. and a range of hose hardware, also injectors. by any chance does anyone have 4 of these hanging around ? I really only need 2 right now, I'd hate to have to buy 8 (set). I suspect that I'll buy injectors from msd also, I have to figure out the flow. I was going to look up the flow rate of the bosch injectors I got from my donor 2.3 volvo and double that. does anyone know if bosch injectors are low impedence or not ? they were batch fired by the volvo jettronic lh computer, I'm trying to figure out if I buy msd injectors, if my computer will drive them directly or not. thanks. Ted Stowe - -----Original Message----- From: DIY_EFI-Digest-Owner@xxx.edu [mailto:DIY_EFI-Digest-Owner@xxx.edu] Sent: Tuesday, January 11, 2000 12:00 PM To: DIY_EFI-Digest@xxx.edu Subject: DIY_EFI Digest V5 #16 DIY_EFI Digest Tuesday, January 11 2000 Volume 05 : Number 016 In this issue: Re: DIY_EFI Digest V5 #15:Knock Sensors Re: DIY_EFI Digest V5 #15:Knock Sensors Re: TWM & GMECM Re: GMECM and Crank Sensors Re: Knock sensor on Dyno See the end of the digest for information on subscribing to the DIY_EFI or DIY_EFI-Digest mailing lists. - ---------------------------------------------------------------------- Date: Tue, 11 Jan 2000 04:52:37 -0800 (PST) From: Carter Shore Subject: Re: DIY_EFI Digest V5 #15:Knock Sensors Ouch! Geez Grumpy, gimme a break. I was replying to the original post by brucep@xxx.net, not to your reply, (got my Bruce's mixed up). You've probably forgotten more than I'll ever learn on this subject, and I appreciate the fact that you are willing to share it with us. But bear with me a second. I'm facing a similar problem to the dyno-building Bruce (he is testing tractor motors). The problem is that no one has applied knock sensors to my engine (at least not the manufacturer). Maybe there's a KS that will work, maybe not. I've seen a spec sheet for Ford KS that specifies the resonant frequencies for some, but not for any other manufacturers. Since I have worked professionally with noise and signal analysis, I looked to the tools that I used every day for years. Spectrum analysis is one of them. It's just a measurement tool, not abstract theory. When I was in the biz, the instruments cost thousands of dollars. Now it's free for the downloading. I believe that a lot of list members would find a simple cheap system for detecting knock very accurately on any motor a useful tool. Maybe you too. I'll share whatever I find with the list, if there's any interest. Even if turns out to be BS, at least we will have found out what does not work, and move on. Thanks, Carter __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com - ------------------------------ Date: Tue, 11 Jan 2000 04:52:37 -0800 (PST) From: Carter Shore Subject: Re: DIY_EFI Digest V5 #15:Knock Sensors Ouch! Geez Grumpy, gimme a break. I was replying to the original post by brucep@xxx.net, not to your reply, (got my Bruce's mixed up). You've probably forgotten more than I'll ever learn on this subject, and I appreciate the fact that you are willing to share it with us. But bear with me a second. I'm facing a similar problem to the dyno-building Bruce (he is testing tractor motors). The problem is that no one has applied knock sensors to my engine (at least not the manufacturer). Maybe there's a KS that will work, maybe not. I've seen a spec sheet for Ford KS that specifies the resonant frequencies for some, but not for any other manufacturers. Since I have worked professionally with noise and signal analysis, I looked to the tools that I used every day for years. Spectrum analysis is one of them. It's just a measurement tool, not abstract theory. When I was in the biz, the instruments cost thousands of dollars. Now it's free for the downloading. I believe that a lot of list members would find a simple cheap system for detecting knock very accurately on any motor a useful tool. Maybe you too. I'll share whatever I find with the list, if there's any interest. Even if turns out to be BS, at least we will have found out what does not work, and move on. Thanks, Carter __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com - ------------------------------ Date: Tue, 11 Jan 2000 08:32:54 -0500 From: "jtyler" Subject: Re: TWM & GMECM Well it certainly pays to ask first. Thanks a bunch. I could always go to a crank wheel later after all the otherstuff is sorted out, but for street use I certainly don't need to. I just wanted to eliminate the dist if possible. So back to the Crane. What electrical input is the ecm looking for from the trigger wheel? Do I build a circuit to condition the Crane box output to the GM ecm, or do you need to make a new driver for the optical sensor itself and have it trigger the ecm directly? Has someone been down this path...sounds like? If you can send me the wheel drawing that would be great. If a fax is less effort let me know. BTW there is absolutely no rush on this. I am one of those "Blessed" with too many projects. Dopey >------------------------------ > >Date: Mon, 10 Jan 2000 21:54:10 -0500 >From: "Bruce Plecan" >Subject: Re: TWM & GMECM > >Geez, not even in the same league. Protractor and X-Acto knife vs AL, and >measuring, and figuring out where the notches have to be and slotted for >enough adjustment. If I owned a machine shop and had the engine on a stand, >I'd still go the dissy route >Grumpy > > >| I like the "cut you own wheel" idea with the Crane, very clever. >| However, if I am going to go DIS, would it be that much harder to >| make a wheel for the crank pulley and use some kind of pickup down there? >Is there one you especially like? Thanks in advance. >| Jim - ------------------------------ Date: Tue, 11 Jan 2000 09:24:50 -0500 From: "jtyler" Subject: Re: GMECM and Crank Sensors For yet another project, is there a nice reliable, relatively easy to mount crank sensor that interfaces easily to the GM ECM's you favor? In this case a dist is not an option. I have provisions to build a mount, make/adapt a crank wheel of some type or precisely install magnets in a pulley or even a flywheel. Does the GM DIS have 5 or 6k rpm limits, something I read somewhere indicated that, but they may have been talking about HEI coils? All this stuff I am playing with needs to make it to 6.5K, the race app at least 8500. Dopey >about the DIS, I'll scan a diagram showing the notches/offset/which notch >| >fires which coil pack etc.. >| >Grumpy > >------------------------------ - ------------------------------ Date: Tue, 11 Jan 2000 12:11:58 -0600 From: steve ravet Subject: Re: Knock sensor on Dyno > Date: Mon, 10 Jan 2000 15:00:32 -0800 (PST) > From: Carter Shore > Subject: Re: DIY_EFI Digest V5 #14:Knock sensor on Dyno > > Bruce, > I believe that the sensors are built to be sensitive > to specific frequencies, to minimize false trigger, > etc. In other words, they are specific to a given > engine. > > I have been working with freeware audio spectrum > analyzer program that uses the sound card on a PC. By > connecting a wide band microphone to the motor, and > then simulating knock (resembles pounding on the block > with a hammer or broomstick), the characteristic > frequencies of a given motor can be determined. The > resonant frequencies show up as spikes. If you can > safely induce actual knock in the running motor, even > better results. > > Using that info, the PC software can be set up to look > for knock signals for that specific motor. > > All freeware, just need to bolt on a microphone > (guitar mic?) Save the results on a floppy. > > YMMV, I'm just playing with it right now. Carter -- what is this software called and where is it available from? This is very interesting. - - --steve - - -- Steve Ravet steve.ravet@xxx.com Advanced Risc Machines, Inc. www.arm.com - ------------------------------ End of DIY_EFI Digest V5 #16 **************************** 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". NOTICE: This communication may contain privileged or other confidential information. If you are not the intended recipient, or believe that you have received this communication in error, please do not print, copy, retransmit, disseminate, or otherwise use the information. Also, please indicate to the sender that you have received this email in error, and delete the copy you received. Thank you. ------------------------------ Date: Tue, 11 Jan 2000 17:04:24 -0800 (PST) From: Carter Shore Subject: Re: DIY_EFI Digest V5 #16:Knock Sensor Some requests for the software I'm playing with. It's a program called 'spectragraph', available for download at: www.monumental.com/rshorne/gram.html There may be other better ones out there, but this one seemed to fit my needs. If this works, maybe we could keep a library of 'knock signatures' by engine. There's an app note for a chip that does FFT for KS, but maybe overkill; if you can find the fundamental resonant knock freq with better than 6 dB S/N ratio, why look any further? On the other hand, multiple mics and holographic audio imaging could allow you to pinpoint not just knock to the individual cylinder, but maybe detect impending bearing failures etc. before they go catastrophic. What about an 'active' electronic crankshaft damper that attenuates 'on the fly'. Oops, sorry for the geek out. Anyway. I wonder what kinds of things might influence the knock signature? Maybe forged vs cast crank? Certainly aluminum vs cast iron block/heads. Maybe the intake manifold or headers. Aluminum flywheel? Different crank dampers? Would the KS change if you used filler in the water jackets? What about just bore/stroke changes? Operating temperature? Manual vs auto trans? Lot's of questions, no answers yet. Might be it just doesn't matter. We'll see. Could be fun. Carter __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com ------------------------------ Date: Tue, 11 Jan 2000 17:04:24 -0800 (PST) From: Carter Shore Subject: Re: DIY_EFI Digest V5 #16:Knock Sensor Some requests for the software I'm playing with. It's a program called 'spectragraph', available for download at: www.monumental.com/rshorne/gram.html There may be other better ones out there, but this one seemed to fit my needs. If this works, maybe we could keep a library of 'knock signatures' by engine. There's an app note for a chip that does FFT for KS, but maybe overkill; if you can find the fundamental resonant knock freq with better than 6 dB S/N ratio, why look any further? On the other hand, multiple mics and holographic audio imaging could allow you to pinpoint not just knock to the individual cylinder, but maybe detect impending bearing failures etc. before they go catastrophic. What about an 'active' electronic crankshaft damper that attenuates 'on the fly'. Oops, sorry for the geek out. Anyway. I wonder what kinds of things might influence the knock signature? Maybe forged vs cast crank? Certainly aluminum vs cast iron block/heads. Maybe the intake manifold or headers. Aluminum flywheel? Different crank dampers? Would the KS change if you used filler in the water jackets? What about just bore/stroke changes? Operating temperature? Manual vs auto trans? Lot's of questions, no answers yet. Might be it just doesn't matter. We'll see. Could be fun. Carter __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com ------------------------------ Date: Tue, 11 Jan 2000 20:28:55 -0500 From: "Bruce Plecan" Subject: Re: DIY_EFI Digest V5 #16:Knock Sensor | If this works, maybe we could keep a library of 'knock | signatures' by engine. Sure seems like putting alot of effort into a dead ended subject, thou The basic flaw with all this is that all the stuff you mention below effects an acoustic sensor. An acoustic sensor, is a half hearted attempt at knock detection, since so many items can false trigger it. Why not work on a Ion knock detection strategy. Something that looks for any knock on any engine, rather then some feable side effects of detonation. | | There's an app note for a chip that does FFT for KS, | but maybe overkill; if you can find the fundamental | resonant knock freq with better than 6 dB S/N ratio, | why look any further? | | On the other hand, multiple mics and holographic audio | imaging could allow you to pinpoint not just knock to | the individual cylinder, but maybe detect impending | bearing failures etc. before they go catastrophic. | What about an 'active' electronic crankshaft damper | that attenuates 'on the fly'. Listen for enough noises, adn the thing will always be triggering. | | Oops, sorry for the geek out. Anyway. | | I wonder what kinds of things might influence the | knock signature? Maybe forged vs cast crank? Certainly | aluminum vs cast iron block/heads. Maybe the intake | manifold or headers. Aluminum flywheel? Different | crank dampers? Would the KS change if you used filler | in the water jackets? What about just bore/stroke | changes? Operating temperature? Manual vs auto trans? And most any mechanical lifter cam drives it into fits....... Oh well. Grumpy | | Lot's of questions, no answers yet. Might be it just | doesn't matter. We'll see. Could be fun. | | Carter | | __________________________________________________ | Do You Yahoo!? | Talk to your friends online with Yahoo! Messenger. | http://im.yahoo.com | ------------------------------ Date: Tue, 11 Jan 2000 19:37:04 -0800 From: Bill Sundahl Subject: Re: Knock sensor on dyno Take a look at http://home.netcom.com/~bsundahl/knock/sound/KnockSounds.htm This shows some of the work that I have done on recording and analyzing the signal off of an Eagle Talon knock sensor. The sensor does not seem to be "tuned" that I can tell, it's just a wide band microphone. I am assuming that the GM sensors are very similar. The Talon uses a fancy filter circuit inside the ECU to detect knock, but I have never been able to figure out exactly what it looks for in the signal. - -Bill ------------------------------ End of DIY_EFI Digest V5 #17 **************************** 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".