DEANO3528 Posted August 22, 2005 Share Posted August 22, 2005 Started by hendrik Hi there, currently I´m running my 3.9 Efi with the standard hotwire system (14CU). In case I manage to rebuild the engine (once I have time and money to do so) I would be happy with a cheap, programmable, understandable, reliable and somewhat simple EFI System like the Megasquirt. What I´d like to know first is - can someone tell me in which way the fuel economy and emissions were influenced ? Have you had any problems in long term use (I might take the Rover to an expedition in a few years and I´m pretty much interested in the reliability of non-standard parts). If I read the MS - homepage it appears that the MS works without an MAF-meter, puzzles me as this was such a step forward on the RoverV8-fuel management. Many thanks in advance, Hendrik geoffbeaumont Fridgefreezer's your man for Megasquirt. Waiting (impatiently) for mine to turn up at the moment fridgefreezer You could use the MS to trigger the LPG switchover, not a major headache. As for the delay, it would be a simple code change from X milliseconds to X revolutions or X ignition events (sparks ) No, MS won't care about "fake" injectors. In reply to: How did you mount your IAT sensor on the plenum? Well mine (verson 1) was just sort of "mounted" in a fresh-air environment under the bonnet utilising advanced flexible fixing technology. Basically it was cable-tied to the washer pump 'cos that's where the old AFM wire appeared. New version (on the bowler I'm rewiring & MS'ing this weekend) the owner has made a piece to replace the 9th injector that is a plate with a tube going where the nozzle was into the plenum with the IAT epoxied in, then a solid 8mm bar protruding where the fuel barb of the injector was to blank off the fuel line that would otherwise be left dangling. It's really very smart, photos in due course. ...and one more question (have I had my twenty yet?) - did you replace your stepper motor with a fast idle solenoid valve or a simple auxiliary air regulator? Never had a stepper motor but I'm pretty sure MS can control those now, or MS-II / UltraMegaSquirt will. My extra air valve is blocked off although I may reinstate it yet. You know the little Squirt's right! geoffbeaumont Kind of thinking aloud here...is there any reason why this would be a bad idea: No Image For those that can't decode my incredibly clear and self-explanatory technical drawing what I'm thinking of is a sheet of steel or alloy mounted to the side of the plenum using the middle and rear bolts just below it (front one has a vacuum pipe coming from it, but could maybe be used as well?) and lugs onto the three plenum bolts at the top. The two coil backs would be mounted on that, with the plug leads facing the right hand side of the car. Main thing I'm worried about is whether the plenum could take the stress? By the way, what's the plastic globe on the left wing that the vacuum pipe goes to? -- Geoff Beaumont Jeremy Can't really comment on how sturdy the plenum is as I don't have one on my current engine - seem to remember they are fairly solid though. If it's any help I do know that on the Cadillac Northstar V8 the coil packs are mounted on one of the rocker covers (which is at the back of the engine bay because the engine is mounted transversely). As you say I think keeping the ignition leads as short as possible is a good idea. The vacuum reservoir (globe thingy) on the wing is connected to the heater I believe - used to operate some of the flaps inside the heater I think, though I've never investigated mine. Whilst I'm here, a question: Just been vaguely looking at 'injection' inlet manifolds with a view to my own future Megasquirtification, and wondered if there was any real difference between the manifolds used on 3.5s, 3.9s and 4.6s. Anyone know? Also, whilst on the subject, I've never understood the current fashion towards throttle bodies (including the 'injected four-barrel' that Fridge showed a picture of above). What is the supposed advantage of injecting into a throttle body rather than injecting directly into the inlet manifold? Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk Geoff Beaumont I came across a really nice table comparing all the slightly different methods of fuel injection and their advantages and disadvantage while I was searching for something else the other day. Blowed if I can find it now though Surprisingly, some of the simpler systems had advantages other than just low cost - I think one of the arrangements that injected well before the intake ports gave better power at high revs, for instance (but don't quote me on that...!). If I come across it again I'll post it here. -- Geoff Beaumont Geoff Beaumont In reply to: If it's any help I do know that on the Cadillac Northstar V8 the coil packs are mounted on one of the rocker covers I thought of that, and it'd work on some cars, but on mime there's too much clutter in the way - air con pipes and the like. -- Geoff Beaumont fridgefreezer The manifolds on 3.5 to (early) 4.6 are all the same, the 3.9's have "3.9" cast into the plenum and the 4.6's have "4.6" There are detail differences - early flapper types have injectors which have a short fuel hose tail on the end whereas later versions have the fuel rail to "clamp" the injectors down, these later type injectors are more universal (you can walk into a scrapyard and pick new injectors off of pretty much anything). BTW the injectors seemed to stay the same (in flow rate / electrical characteristics if not mounting) all the way though from 3.5 to 4.6. For comparison photos, clicky here: http://www.spagweb.com/v8mini/gallery/05/0...01_efi_compare/ You know the little Squirt's right! Jeremy Thanks - having looked at that website I now understand why you are looking for a Thor manifold - compared to the 'orrible old aluminium plenum they look almost modern, and apparently give some useful torque increases too (9% increase at 500rpm lower on a Disco 4.0 compared to a 3.9 I've just read). Must say I do also quite like the carbon plenums shown on the RPI site (http://www.rpiv8.com/carbs-plenums.htm), especially the one shown on the Marcos. I could knock something like that up myself quite easily (my old SIII Lightweight had carbon fibre bits believe it or not!). I notice these plenums incorporate throttle bodies as well though, so that would be more expense if you were going to copy that design - and presumably throttle bodies would require the MS to inject the fuel slightly earlier than with normal injection just above the port? The one advantage of the 'injected 4-barrel' idea is that I would be able to retain my existing lpg mixer, which is a special thingy with vacuum-operated flaps, designed specifically for 4-barrel carbs Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk geoffbeaumont I like the idea of a Thor manifold too, but then I'd really be stuck for somewhere to bolt the coil packs -- Geoff Beaumont fridgefreezer I don't know, I think this will look quite good under the bonnet of my 109": The throttle bodies just replace the Rover throttle plate, the injectors don't move as they're on the inlet manifold. There's no timing on when the injectors fire, for the same reason they haven't implemented sequential injection: It doesn't matter. Most of the time the injectors are firing at the back of a closed valve anyway, but this is not a bad thing. There have been many posts about it on the MSEFI.com forums, just search the FAQ or "sequential". You know the little Squirt's right! Jeremy Thanks - that makes things easier in both respects. Regarding the throttle body, I had previously understood the term to refer to a sort of 'injection' version of a carb - ie. both the injector(s) and butterfly valve in one unit - but if (can) just mean the butterfly valve by itself I guess that makes things less complicated in terms of plenum design. I will have to start collecting parts and having a play. I take it the Thor manifold is horribly expensive and difficult to find? Also, who do I speak to to find out about the injected four-barrel thingy? Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk fridgefreezer Speak to Bill Shurvinton, tell him I sent you (my real name is John Utteridge, but fridge is less weird ) his e-mail is bill (at) shurvinton.fsworld.co.uk or PM me for his phone number. The 4-barrel throttle bodies he's selling are a bargain, you are actually better off restricting them (he sells drop-in restrictors) for the Rover V8 as they flow 750cfm. By contrast the 4-barrel Weber carb flows 500cfm and the Holley 380cfm You *will* need a MegaSquirt to drive them, and you'll want to find a set of 4 injectors from a vehicle that puts out the same power as a Rover V8. Most 2.0+ saloons or hot-hatches of recent years should suffice. You know the little Squirt's right! hendrik Hi, I´m still not determined what to do to the ignition. I see the EDIS system uses the star wheel (pictured further above) - isn´t that prone to collecting mud and bending the teeth ? On a automatic gearbox´ed car and with a 4.6 crossbolted block (given one finds one or gathered enough cash) wouldn´t it make sense to use the OEM 4.6´s flexplate and original crank-pos-sensor ? Could the EDIS be made to work with that, or (as I´ve been told) could the Megasquirt´nSpark system be the solution ? What is the preferred setup within the RoverV8 enthusiasts ?? Jeremy Thanks for the contact info Fridge I would also be interested to know more about the stock sensors on later engines. My 4.6 has provision at the rear of the block for some sort of device that faces the flywheel. I think I was told it was for a knock sensor, but I suppose it might be for a crank position sensor - although presumably you would need a matching flywheel (or whatever the crank position sensor looks at) as well. Being supplied as a short block I had to fit the flywheel from the old engine to mine. (I am assuming it is the flywheel you need rather than the flexplate, because it is the flywheel that faces the sensor). Regarding the MS'n'Spark etc - Fridge will give you a better answer, but basically a complete system requires 3 things: Sensors (either aftermarket or OEM), EDIS (to actually send the sparks), and a ECU (MegaJolt, modified Megasquirt or whatever). The ECU reads the signals from the sensors, and tells the EDIS what to do accordingly (although if the ECU breaks the EDIS will continue to send sparks at a fixed 10' advance by itself). On a different flight of fancy, I'm still thinking about intake manifolds for an injection set up. For no good reason other than for the hell of it, I would like to do something that avoided having to use the standard rectangular plenum. I remember seeing an article about one of the new TVRs recently (not sure which model), on which the 'induction tubes' to each inlet port were wide-diameter silcone hose. Looked very cool and would be comparitively easy to do as a cheap alternative to the Thor manifold. No idea of the science behind how to determine what width / length to make the hoses though, and you would have to figure out how to plumb in the throttle body too. Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk Geoff Beaumont Seems there's not too many of us playing with it yet, at least among the Land Rover fraternity, but I everyone I know of is using or planning on using EDIS. That doesn't mean it's the best solution though - as you've observed, Megasquirt'n'Spark definitely has potential and should allow the use of the crank sensor on later RV8s. I think it would also be possible to go coil on plug (no HT leads ) with this. Whether it's possible for EDIS to work with the 4.6 crank position sensor would depend on what type of sensor it is - EDIS uses a VR type sensor which gives a waveform that has negative and positive parts, and the EDIS module uses the wave passing through 0V as its trigger. A lot of the other manufacturer systems seem to use Hall effect sensors, which generate a square positive only waveform and different voltages. EDIS also expects a 36-1 timing signal from the sensor, and a lot of the other systems use a smaller number of inputs per revolution. There has been some discussion on the Megasquirt forums about building circuits to modify the waveforms from other sensors to create suitable inputs for EDIS, but this definitely isn't something I'd want to get into unless I was already confident with the rest of the system. For me, EDIS made sense for the following reasons: § It's well proven and reputedly very water resistant § It should be pretty simple to set up, as the interface to Megasquirt is very simple (one input line and one output) § I had to fit a timing wheel of some kind anyway, as I have an older (3.9) RV8 -- Geoff Beaumont Geoff Beaumont I'm not 100% certain about this, but I think the Megasquirt'n'Spark firmware contains provision, possibly experimental, for firing coil packs directly, which would remove the need for the EDIS module. -- Geoff Beaumont hendrik Geoff, needless to say that I´m not skilled enough to alter the code of the MS software on my own or adapt something radically new. Therefore I shall stick to what is common and experienced and known working! Even though I notice there are a lot of clever and helpful people around (worldwide) that could help and do help. AFAISee the distributor is obsolete with the EDIS ignition. What are you doing with it ? Removing it seems to be the best way but what about the oil pump ?? How do coil-on-plug things look like ? Wouldn´t they always be where you wouldn´t normally put a coil - next to the hottest area (exhaust). Geoff Beaumont I'm just leaving the distributor in place with no leads attached - it shouldn't be a problem. Like you say, there needs to be something there to drive the oil pump. You can get distributor blanks which I guess are just a plate with a spindle on the back for mounting the drive gear, if you're fussy about it. In reply to: How do coil-on-plug things look like ? Wouldn´t they always be where you wouldn´t normally put a coil - next to the hottest area (exhaust). Don't know what they look like, or whether they suffer from heat problems, but I think they were used on some of the later RV8 powered Land Rovers, so I guess it's not too much of a problem. -- Geoff Beaumont Jeremy Just received my serpentine/distributorless front cover/oil pump/waterpump in the post today! Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk Geoff Beaumont Hmm, Fridge, your RR is older than mine ('93) isn't it? What about yours Jeremy? I've just been looking at the wiring diagrams (for a '95, which is there nearest I can get my hands on), and the following are all tied in to the Hotwire ECU: § Fan control module (for the small ones in front of the radiator, I guess) § Air con clutch on compressor § Instrument cluster (speedo? something other than just the EFI light, anyway) § Fuel pump relays § Evaporative emission canister purge valve § Heated front screen (seriously...) This is starting to look a little more tricky than I'd bargained on... What inputs does Megasquirt need if it's only controlling the ignition? Throttle position, intake pressure, EDIS PIP. Anything else? -- Geoff Beaumont Jeremy Mine's an '87 as far as the electrics are concerned; although the EFI has been disconnected for at least two years nothing else on the vehicle seems to have noticed (at least, everything else still works) - but I dare say that since mine was built only 2-3 years after EFI was first introduced it is a fair bit simpler than the later versions. If you're just running MS to control the ignition, surely you will be leaving the orginal ECU well alone? Apart from maybe sharing the TPS I don't think the two systems will be aware of each other's presence at all. Jeremy Geoff Beaumont Leaving the Lucas system in control of the fueling seems to be the way forward for the moment, but I was only intending that to be 'stage one' - the idea being that once I had ignition sorted out I'd hand fuelling over to Megasquirt as well and get rid of the Lucas system completely. I take it you have the older flapper system in your car? At least some of the issues on my list are solvable. The fans, for instance, can be controlled by Megasquirt, but there may be some modifications required - it's impossible to tell from the wiring diagrams how they're controlled. I need to spend some time reading up on Megasquirts capabilities to figure out what can and can't be done. -- Geoff Beaumont 02GF74 Just received my serpentine/distributorless front cover/oil pump/waterpump in the post today! Jeremy, I hope you have also received the crank to go with the above, I'm pretty sure they are different, could be wrong tho' I'm trying to think but nothing's happening.... Jeremy Mine would have been a flapper, yes, though I flogged all the bits (and the complete aircon system) on Ebay soon after taking them off the car. Going through your list I would surmise the following: Fan control module - not sure if MS knows the engine temperature, so these may have to be switched seperately either with a manual switch or an aftermarket Kenlowe-type thermostat somewhere Air con clutch on compressor - this will be to increase idle revs to counteract the load of the aircon being switched on - goes back to the earlier discussion about stepper motors etc Instrument cluster - I would guess that this is the rev counter as this 'runs' off the sparks produced by the coil. Whether the Lucas EFI has some connection to the coil/distributor to know when sparks are being produced I am not sure. Wiring the rev counter to the EDIS coil packs is fairly straightforward I believe Fuel pump relays - obviously. Presumably MS has a connection to switch the fuel supply on - if not just wire them to the ignition I guess, as you would on a carburettor vehicle Evaporative emission canister purge valve - I assume this is the charcoal canister thingy. Probably doesn't work any more anyway, so just disconnect the whole thing. Heated front screen - Another big load on the engine when turned on, so answer as air con That's my tuppence worth anyway Jeremy 1987 Classic with 4.6 Visit my website! www.magazineexchange.co.uk Jeremy All part of the fun of discovery! There are differences because of course the oil pump drives directly off the crank, and the camshaft no longer has the gearwheel on the end to drive the distributor. It LOOKS like the oil pump will engage onto the crankshaft using the same grooves that the front pulley uses - whether it does, or whether the diameter / length of the crank is different only time will tell. I am also not sure if the existing camshaft (which is now longer than it needs to be) will fit behind the new front cover or not. I was thinking just last night that maybe what I ought to do is get an old 3.5 block just to do a dry-run of fitting all the parts before I try it for real on the 4.6. There are various parts of the serpentine system which I don't yet have, and I might well have a play around with EFI plenums as well. Jeremy Geoff Beaumont Megasquirt does know engine temperature, and using the MegasquirtnSpark-extra firmware (which you need anyway for EDIS) it can control up to four outputs based on a selection of criteria (rev limits and temperature limits - not sure if there's anything else). That takes care of the fans, possibly with a few modifications depending what signal the fan control module expects (I'm hoping just a simple voltage signal). It might also do for the heated screen, as I'm guessing that's just a signal to only allow it to be switched on if the engine is running. This might go for the air con clutch, too. Megasquirt does have a standard output for switching the fuel pump, which I should have remembered Looks like the Lucas system switches it in two places, but I'm sure that can be sorted out. I thought the tach input came from the alternator, but I could be completely wrong there... That's definitely one to look into. I guess the Lucas system would have to know something about the engine speed? Aren't you supposed to renew those carbon canister thingies when they're life expired? No, it probably doesn't work any more, but I suspect the cats don't either, so I'm sure I can live without it -- Geoff Beaumont fisha In terms of the Engine ECU having links to the heated from screen and air-con compressor switch. Its probably something very similar to the P38A setup. If the heated screen wants to go on, or the aircon compressor, the systems ask the engine ECU if there is enough engine capacity first to run them. If there isn't, the air con or the heated screen wont go on. So i guess that would be similar to what you describe Geoff where a few outputs lines could be triggered to say that when in this range, you can have air con or whatever. I suppose you would want to disable them below a certain rev limit to prevent stalling and also if the throttle is to the floor. Cause if your foot is to the floor, your obviously wanting full power out the engine and the air con will just have to wait. Highway_Star IIRC it's a pin to ground affair from the air-con, heated screen etc, which raises the idle speed slightly. Wait for the ricochet.... Geoff Beaumont In which case it should be possible to bypass them providing Megasquirt can maintain a stable idle with the loads kicking in (it's supposed to be pretty good at it if you do the fuel map right). -- Geoff Beaumont ariane44 Hi! Screen and Airco on the classic just indicate they need power to the ECU. Same as the "Drive" or "Neutral" signal from the autobox. No asking if they can have it in the Classic. The rev signal for the ECU is taken from the coil. See http://www.rangie.com/articles_topic.php?i...cat=6&subCat=30 For the "vaporative emission canister purge valve" - do RRC built before '95MY have these at all? Cheers Carsten ;-) Range Rover Addict Geoff Beaumont In reply to: For the "vaporative emission canister purge valve" - do RRC built before '95MY have these at all? No idea - I've got workshop manuals up to '91 and from '95, but I've no way of telling whether they where introduced in between those years (other than taking a look at the actual car, which I haven't done yet ). I've been told you need it if you've got catalytic converters, though, which my car does. Looking at that rangie.com article, I've a feeling that it'll be easier to do the Megasquirt install in one go, rather than 'fix' inputs for the old Lucas system. In reply to: Screen and Airco on the classic just indicate they need power to the ECU. Same as the "Drive" or "Neutral" signal from the autobox. No asking if they can have it in the Classic. That's good - Megasquirt should be happy enough not knowing about them, then, same as on an older car. It's not like the Lucas system seems to pay any attention anyway Cheers, -- Geoff Beaumont fridgefreezer Geoff - I suspect the AirCon / Screen wires tell the ECU that they're being used and that it should therefore "up" the idle (IIRC you have an idle stepper motor type affair on your vehicle) by letting more air in through the idle circuit. Since you are leaving fuelling to the Lucas ECU the MS will just detect the drop in air pressure and adjust ignition accordingly, since there is no throttle change the correction should work out fine. One thing you do need to check is how the Lucas fuel system is triggered - if it's triggered from the coil (as the rev counter I think will be) you'll need to find an output that mimicks the now redundant coil. Or just leave the dizzy in place, replace the coil with a 10 Ohm+ resistor and let that trigger the ECU & rev counter. I don't know if EDIS has a "normal" TACH output? My Range Rover is, err, a 1978 Series 3 with an unknown flapper EFi loom chucked inelegantly in. Still, all that will be changed soon, I'm going to ditch the Lucas loom entirely and re-wire specifically for MegaSquirt. You know the little Squirt's right! Geoff Beaumont In reply to: My Range Rover is, err, a 1978 Series 3 Doh! I knew that really In reply to: I don't know if EDIS has a "normal" TACH output? Neither to I, but I'll check tonight. Hadn't thought of replacing the coil with a resistor...might do that rather than attempt the whole Megasquirt'n'EDIS installation in one go. The longest I can manage to take the car off the road is over a weekend, so I don't want to bite off more than I can chew... -- Geoff Beaumont jwriyadh Be cautious on the likelihood of a replacing the coil with a resistor and triggering a 14CUX with the resultant 0-12-0-12 volt pulses. When I first ran my "kitchen table" system I had difficulty in getting the injectors to fire. Eventually solved with electronic ignition system running a coil and sparking properly. jw 1978 2.25 Lightweight 1997 V8 Defender 1998 V8 Discovery fridgefreezer I know you are supposed to put a 10k resistor in series with the TACH feed to the ECU when you feed it straight from the coil, presumably the ECU likes a slightly less "robust" signal than 0/12/0/12 volts then? You know the little Squirt's right! jwriyadh More like the opposite. If you put a scope on the coil primary you will firstly notice the expected 0-12-0-12. Then look at the spikes generated when the current stops and the inductive bit of the coil does its stuff. Spikes peak at about 60-70 volts. The ECU obviously has to deal with this under normal conditions so may not respond properly to 0-12-0-12. I have seen a small transformer used to generate the same pulses for a rev counter that required the standard coil waveform. An alternative is to burrow into the ECU and remove some of the circuitry so that 0-12-0-12 works. Shall have to look into that. jw 1978 2.25 Lightweight 1997 V8 Defender 1998 V8 Discovery fridgefreezer ...or say b*ll*cks to it and fit a proper ECU You know the little Squirt's right! geoffbeaumont Megasquirt's designed to work off the coil pulse isn't it? And, IIRC (which I might not...), the EDIS PIP line is connected to the same input, so how does MS deal with this voltage variance? -- Geoff Beaumont fisha I would suspect ( dont know for sure ) that the megasquirt ECU actually is 5V logic based so any 12V signal may need dropped to 5V to be passed to the ECU. If thats the case, then it probably goes through a transceiver of some sort and if so, your average transceiver can handle realtively large input voltages whilst always putting out a standard low voltage. fridgefreezer MS on a "normal" dizzy & coil system does just use the coil -ve end, clamped to within sensible limits on the input by zener diodes. With EDIS I don't honestly know what it does, if EDIS has a "normal" TACH output or what - I've not researched the inner workings of it although there's a lot of info on the MS'n'EDIS pages. You know the little Squirt's right! Geoff Beaumont If the coil isn't actually connected to any plugs, will you still get reliable tach signal from the distributor? Looks like I'm going to have to do Megasquirt fuel first of all in one go anyway, since no one seems to be able to tell me what inputs MS needs connected for ignition only use. I guess there isn't anyone who's done it that way round. -- Geoff Beaumont fridgefreezer Geoff, Lots of people have done MS for ignition only, the ECU doesn't need many inputs whatever you're using it for: RPM, Coolant temp, Air temp, MAP Sensor (vacuum), throttle potentiometer. You know the little Squirt's right! llsteve On the subject of serpentine covers, there are many issues with these that make them awkward to retrofit to 3.9 and 3.5 engines with standard short nose cranks. 1) The camshaft will not fit, you must use the serpentine type cam and it's driving gear and ideally, fit the anti walk or thrust plate behind it. If using Cloyes gears instead, there can be an interference between the thrust plate bolts and the back of the gear. 2) The crank can be made to drive the pump by extending the keyway in the nose and fitting the appropriate key. If you are exceptionally skilled with a grinder you may be able to do it this way, but you'd be better off either getting it machined in or use the 1995/6 Discovery crankshaft with the extended keyway. 3) The serpentine pump on this cover runs in the opposite direction. Not and impossible problem but a bit tricky to solve in a straightforward fashion. Obviously the viscous fan screws on the other way as well. 4) The oil pump pickup pipe is attached to the timing cover rather than the block. You will need this and the parts that attach it to the No4 bearing cap. 5) You will need to find a reliable way to blank the cam sensor hole. An aluminium plug epoxied into place seems to do the job. 6) The pattern of the timing cover is different where it seals to the sump so you will need to use the matching sump pan or get the cover modified. 7) Oil cooler unions and pipes will need to be made to suit. Shall I go on... Geoff Beaumont The catch being that I was planning on using the existing sensors, so if I'm still using the Lucas system for fuel, it will be using the coolant, air and throttle inputs. I'm not sure how well either ECU would take to being wired in parallel across the sensor outputs (which I presume would be the way to do it?). If I do things the 'sensible' way and implement Megasquirt fueling first, I get rid of this issue. With the mostly drier weather now (and copious quantities of WD40) my ignition is behaving itself fairly well at the moment, so I shouldn't have any problems doing it that way now. Mind, it's running much worse on LPG than petrol (more so than I'd expect from a weak spark), so I suspect the LPG kit needs some tweaking too -- Geoff Beaumont jwriyadh The solution to generating Ignition pulses for the 14CUX is shown at the end of the megasquirt.info/ms2/ page. jw 1978 2.25 Lightweight 1997 V8 Defender 1998 V8 Discovery Geoff Beaumont I can't see it, but that's a frames based site so I might not be looking at the same page as you. Do you have a direct link? It might not be relevant anyway, as that site is for MegasquirtII, and I'm running (or will be) the original MS. -- Geoff Beaumont fridgefreezer Geoff, I think the MS-II EDIS info is pretty much the same as the "-extra" version for the MS-I (Or "Classic" as I'm now going to refer to it) since the MS-II has basically incorporated a lot of the add-on code from -extra and others. The link is here, there's a Tacho circuit at the bottom of the page that looks like it should work OK for the Lucarse ECU too: http://www.megasquirt.info/ms2/EDIS.htm You know the little Squirt's right! hendrik Things got frightful specific on the way ... How easy is it to get the correct fuel mapping done ? I know a mixture too lean will cause overheating and that is said to be one cause of the leaking liners .. I know someone could provide his code for download, but in case the engines are too different (c/ratio, displacement, camshaft, ..) it might be necessary to modify it in DIY-manner. fridgefreezer Hendrik - if you can play playstation games you can tune a car on a laptop Basically with the car running and the lambda sensor working you have someone drive the car and someone tune it (because driving AND using a laptop don't mix ). On the MegaTune tuning screen you have the fuel map and a few gauges telling you what the engine's doing. You also have the Lambda sensor voltage displayed. As you drive, a cursor will move around on the fuel map showing what point on the map the engine is operating at. For each point you watch the lambda voltage and aim to keep it just above 0.5v by raising or lowering the amount of fuel at that point. That will get you a very usable "rough" map that runs and drives as well as the Lucas version. For fine tuning you can use MSTweak3000 - you drive around with the laptop plugged in and record data logs from the ECU. MST3K then reads these logs and does exactly what you did to get the rough map, but much more accurately. The longer you spend logging, the more data it has to use and the more accurate it will be. With a "normal" lambda sensor MST3K can only tune your map to 14.7:1, but it's an easy step to take it's suggested fuel map and richen it by a few % to get 13:1 or 12:1 in the high load areas. In fact to change the fuel map value from 14.7:1 to 12.5:1 you add VAT (17.5%) to the value As with many things this is harder to explain than it is to actually do it! You know the little Squirt's right! Geoff Beaumont Cheers, Fridge - that'll be very handy. Even if I get rid of the Lucas system at the same time as the dizzy I'll still need that for the tacho. -- Geoff Beaumont fridgefreezer Hendrik - just to add to my last post: You will find yourself looking at the fuel map and wondering "but is it right, is it OK here, there, etc." - just cast your mind back to the days of carburettors (remember those ) when all you really had to go on on was your sense of smell and the seat of your pants. Just make sure you're not running lean at any point and let the Lambda sensor correct the mixture under cruising conditions and you can't really lose! Oh and read the MegaSquirt FAQ on "how to tune" before trying - you need the lambda sensor turned off (0% correction allowed) while you're manually tuning it. You know the little Squirt's right! hendrik Hi Fridge, thank you for the encourageing words !!! Very much appreciated. I´ll go ahead when time allowes. Just one thing that always puzzled me: In my very personal ECU (brain) things work like = more fuel =more power=rich mixture=high fuel consumption. I understand that the OEM fuel mapping is causing the engine to run lean on certain conditions (speed limit on highway, cruising speed; rpm´s emissions tests are done at). Professionals who reprogram the EPROMs of the Lucas ECU´s say that they correct this lean mixture (so there is no danger of block cracking any more) and give more power whilst giving a better fuel economy that will recoup for the price of the better chip within a certain mileage. How does that work ? fridgefreezer Well more power means your engine is working more efficiently, so you then have the choice to either: 1) Drive the same speed as before but more efficiently 2) Use the extra power to drive faster using the same amount of fuel You are not necessarily going to be putting in more fuel across the board - you are (hopefully) tuning the ECU right for your vehicle and driving style. Manufacturers don't tune every ECU to every individual engine, they buy 1000's of ECU's and fit them to 1000's of vehicles at a time. You know the little Squirt's right! chrisstevenson 60 - 70v spike on the -VE coil terminal? My Mallory installation has a ballast resistor fitted and the pulses measure 4.8V, seems to be enough to trigger the ECU Roger_Warren To whom it may concern As an interested spectator (who has a suspect Distributor and a misfire ) can I ask that this excellent thread ends up in the Tech Archive. As I have no fear of electronics it is only a matter of time before I go EDIS and MS. Roger W fridgefreezer Chris - how are you measuring the spike? Unless you've got an oscilloscope on it with a fairly fast timebase relative to the coil pulses you may well miss the initial spike. If you're using a multimeter - well, You know the little Squirt's right! DEANO3528 In reply to: To whom it may concern As an interested spectator (who has a suspect Distributor and a misfire ) can I ask that this excellent thread ends up in the Tech Archive. As I have no fear of electronics it is only a matter of time before I go EDIS and MS. Roger W Don't worry Roger it will, but not until it's run its course and is no longer attracting any posts. Rather a bottle in front of me, than a frontal lobotomy! SLROC Hybrid_from_Hell 2011 then ? Nigel If I say anything, and there's 2 ways of taking it, if one way annoys you then I meant the other chrisstevenson What do you mean by initial spike? Is there a particular part of the output from the -ve side of the coil that the ECU uses to start the injectors? Reason I ask is because I've had problems getting some multi-spark amps to work on Efi vehicles as they all seem to have a slightly lower voltage from the tach out terminal than what the Lucas efi system likes, maybe is massivly lower if a standard coil --ve terminal output pulses at 60V!?? fridgefreezer Chris - I've actually no idea about the Lucarse ECU triggering, but to illustrate what I'm talking about check this waveform of the -ve side of the coil as it fires: http://www.picotech.com/auto/waveforms/pri...y_ignition.html Note the duration of the spike (200V+), spark burn (40v) and then how long the points are actually open for (the 12v area on the right) in relation to the timebase. The spark is a tiny proportion of the dwell, and the spike is a fraction of THAT. You know the little Squirt's right! Highway_Star Is that your waveform or Picos? reason I'm asking is that I'm right now looking for either a portable LCD 'scope or a pc based solution. I've trialled one from CPC which was worse than useless, trying to persuade Pico to loan me one for appraisal. As we're talking about essenbtially a DC source being switched on oand off to a large coil of wire, could the 'spike' be back 'EMF' Wait for the ricochet.... fridgefreezer It's Pico's and yes, it would be back EMF. You know the little Squirt's right! Geoff Beaumont Have fun editing this leviathon, Deano... -- Geoff Beaumont Geoff Beaumont Fridge - think you said you're locating your EDIS unit inside the cabin. Do you know if there are any RF interferance issues with doing this (I'm thinking noise on the radio as much as causing problems for ECUs)? Asked on the MegasquirtnEDIS forum, but someone has ventured an opinion, so I guess they've all sited theirs in the engine bay as normal. I'm thinking about burying mine in the dash, on the grounds that keeping electrical things dry is a Good Thing, even if they are supposed to be waterproof... Hopefully spending a bit of time this weekend working out what sundries I don't have and ordering them, so I can get on with installing everything -- Geoff Beaumont minivin aircraft methods are to keep the comms, ac and DC grounds seperate to reduce such instances of interference on the common lines, however a car has them all linked to the chassis so you will always have some form of noise present. All you could do is to mount the case on isolation blocks and run a earth line direct to the battery, but then that line could also act as an antenna and induce noise simply, plug it in and see if it works, if you get noise then an isolation capacity might be needed to put the noise down to earth. "land rover owners do it on all fours" fridgefreezer Geoff - I think there's only so dry you need to keep the EDIS bits, they're all sealed anyway. If you pop the front cover off the EDIS module you'll find it's potted in with some clear gel stuff. Very robust and unlikely to be troubled by water. The coil packs are similarly well sealed, as long as the HT lead boots seal OK I can't see much benefit to plumbing them in indoors. In fact I'd wager it'd be more hassle than it's worth - the system is known to work underwater: http://www.yorkshireoffroadclub.net/vehicl.../megasquirt.php Recent testing has shown that the ignition is quite happy running submerged although a problem with the cold idle did cause the engine to stall - the water was so cold (we had to break the ice in parts) that with the engine immersed in near freezing water the engine temperature dropped rapidly triggering the cold start enrichment function in the ECU. I was too late to catch it by increasing the revs and the engine stalled. As the exhaust manifold was still under water at one side I couldn't risk restarting it so had to be pulled out of the water where it fired up first time. A second test, where I kept the revs up a bit, saw the water up over the bonnet with the vehicle stationary (no bow wave to artificially lower the water level) for a period of time and the engine never missed a beat. You know the little Squirt's right! Geoff Beaumont I know...I just like the idea of less clutter under the bonnet . Plus I'm being perfectionist (completely pointless - the car will never be right...) -- Geoff Beaumont Geoff Beaumont If I'm thinking this one through correctly (which I'm probably not ), that circuit works for EDIS4 but will need modifying to take an input from all four coils for EDIS8? Why would you bother to build this for EDIS4 when it has a clean tach output anyway, unlike the other versions of EDIS? Is there any reason why I couldn't use the IDM output from the EDIS, described as follows on www.dainst.com: In reply to: The IDM signal present on all EDIS modules is used as the factory tach output to the ECU. This signal is used to determine the welfare of the ignition system. If a coil primary opens or does not fire, the IDM signal is not emitted, and thus the coil is faulty. The IDM signal is triggered by the flyback voltage of the coil being fired. The EDIS module processes this into a 512us pulse on the IDM line during operation. When the engine is stopped, but the Key is turned on (KOEOff), the EDIS module emits a 64us pulse every 262.144ms to indicate proper operation. Because of this, any tachometer operation from this line must filter out the 64us pulses. What would be the side affects of using this output unfiltered (once MS is in, it would only, so far as I'm aware, be supplying the tachometer itself)? The only thing I can think of would be a small false reading on the tach when the engine wasn't running, which is hardly the end of the world. What about if I did temporarily feed this to the Lucas system? Would it care when the engine isn't running? -- Geoff Beaumont jwriyadh Don't know what MS would do with the IDM signal. IF the IDM signal is big enough for the Lucas System to react to it(doubtful @ 12 volt) then the ECU would produce injection pulses and supply fuel to a motionless engine. Not good???? jw 1978 2.25 Lightweight 1997 V8 Defender 1998 V8 Discovery Geoff Beaumont Hmm, possibly. MS doesn't need a tach signal in EDIS mode - it uses the PIP signal from the EDIS controller, so once MS is in place only the actual tachometer will be using the tach signal. A bit of experimenting today (pulling wires off the coil while the engine was running) suggests that even with the coil not in use, the dizzy should still produce the tach signal, so I don't need to worry about it for the time being. -- Geoff Beaumont Niall DEANO Please save this thread, even if it's just long enough for me to get enough time to read it all. Answered so many questions I had about LPG / ignition maps already without trawling through the miles of info on MS pages. Just one gripe, can you please downsize the picture of that manifold (gorgeous though it is) it's giving my scrolling finger a real pin. Thanks to all so far for the excellent information, if we can just work Civet de renard into this it will go on forever. RRR2864 One is never enough! But just one running would be nice. DEANO3528 Don't worry Niall, it's on my hit list, but I can't put it into the Tech Archive until it's run it's course. You can't add to it once it's been moved there I'm afraid and it would seem a shame to cut this one short and begin 'Son of Squirt', as a lot of the stuff might be repeated. Rather a bottle in front of me, than a frontal lobotomy! SLROC fridgefreezer So what you're saying, Deano, is that Geoff needs to pull his finger out? You know the little Squirt's right! Geoff Beaumont Yeah, all right...suffering from lack of time to work on it at the moment. How's your's coming on then Fridge? -- Geoff Beaumont fridgefreezer My MegaSquirt is working perfectly thanks - it's just the rest of the landy that's in bits! You know the little Squirt's right! Niall You will know this because you can see the tarty LED's flashing through the panel gaps Just kidding, excellent site write up, thanks. Now can you just fit EDIS, dual mapped for LPG and write it all up cause I'm not able to get home to play and I'm an idle git too. RRR2864 One is never enough! But just one running would be nice hendrik I would like to resume to one topic - altitude correction. I remember that MS does a calibration at start-up. But what happens if I desire to drive up any high mountain without a pause? fridgefreezer Try these: http://www.msefi.com/viewtopic.php?t=8346&...ight=barometric http://www.msefi.com/viewtopic.php?t=5707&...ight=barometric http://thegebharts.com/megabaro.html I'm not convinced you'd ever really need that much barometric correction for driving up a mountain (after all, most other cars seem to manage OK), the lambda sensor will stop it running lean under most conditions (provided you are using it and allow the MS a good % range of adjustment). The cheap version would be to use the Extra code and dual tables, one tuned for sea level and one for 10,000ft, when you hit 5,000ft flick a switch and start using the high altitude table. The high alt version would be exactly the same fuel map, just shifted up a few % as required. Not he most elegant solution but it'd work. According to Lance: "Megasquirt-II also has the capability for real-time baro correction, but it hasn't been released yet." - I think MS-II's are actually shipping now. You know the little Squirt's right! fridgefreezer I've found you can now do real-time barometric correction by fitting a second MAP sensor that the ECU can then compare it's readings with: http://megasquirt.sourceforge.net/extra/barocorrection.htm You know the little Squirt's right! Link to comment Share on other sites More sharing options...
Recommended Posts