elbekko Posted November 9, 2014 Share Posted November 9, 2014 We're working on megasquirting Escape's GEMS P38a. The car side of the wiring we have pretty much figured out (we think), but there's one issue we thought we worked out, but not so sure now... We read Zim's thread, in which he clocked the flywheel 90° in order to use the stock GEMS crank sensor. Now, this is a nice solution, but not really an option for us as it's a manual and thus a very different flywheel that can't just be redrilled. So a plan hatched of just moving the coil wires around to get the 90° rotation. For a firing sequence of 1 8 4 3 6 5 7 2 A standard install would have the coils as: A - 1 & 6 B - 8 & 5 C - 4 & 7 D - 3 & 2 And with our scheme it would become: A - 8 & 5 B - 4 & 7 C - 3 & 2 D - 1 & 6 Logically, this should net us the same 90° rotation of the missing tooth as redrilling the flywheel, right? It shouldn't matter for the injectors either, as it's banked fire. Now for the issue... it won't start. Plops vaguely in the exhaust, but makes no attempt at properly firing. So we started rethinking this, and came to the conclusion it's impossible. Which can't be, because it worked for Zim, so what's going on here? A quick rundown: * Standard GEMS missing tooth is 2 teeth in front of TDC. * EDIS wants 5 teeth behind. * Clocking the flywheel nets us 7 teeth behind. This means that without any correction, it's at 30° advance. So we need to set -20° trim angle. But according to the megamanual EDIS can't go below 10° advance, so there's no way to make this correction. Or can it, and am I reading the megamanual wrong? We're getting an RPM reading in megatune, so the VR sensor should be connected up right. Any input would be welcome, it massive screws with the mind, we've drawn so many diagrams of it all we can't make heads or tails of it anymore. Quote Link to comment Share on other sites More sharing options...
Bowie69 Posted November 9, 2014 Share Posted November 9, 2014 Do you think you maybe went 90 the wrong way, and now ~180 out? Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 9, 2014 Author Share Posted November 9, 2014 Maybe... Borrowing Zim's image and modifying it a bit: Since we want to get this result: That would mean changing B into A, C into B, D into C, A into D. Correct? And just to fully try and confirm, the same firing sequence transposed on Zim's clocked image: Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 9, 2014 Author Share Posted November 9, 2014 Just noticed I drew my lines through the wrong line on Zim's pictures, but the main concept still stands. Quote Link to comment Share on other sites More sharing options...
Bingy Posted November 9, 2014 Share Posted November 9, 2014 We're working on megasquirting Escape's GEMS P38a. The car side of the wiring we have pretty much figured out (we think), but there's one issue we thought we worked out, but not so sure now... We read Zim's thread, in which he clocked the flywheel 90° in order to use the stock GEMS crank sensor. Now, this is a nice solution, but not really an option for us as it's a manual and thus a very different flywheel that can't just be redrilled. So a plan hatched of just moving the coil wires around to get the 90° rotation. For a firing sequence of 1 8 4 3 6 5 7 2 A standard install would have the coils as: A - 1 & 6 B - 8 & 5 C - 4 & 7 D - 3 & 2 And with our scheme it would become: A - 8 & 5 B - 4 & 7 C - 3 & 2 D - 1 & 6 Logically, this should net us the same 90° rotation of the missing tooth as redrilling the flywheel, right? It shouldn't matter for the injectors either, as it's banked fire. Now for the issue... it won't start. Plops vaguely in the exhaust, but makes no attempt at properly firing. So we started rethinking this, and came to the conclusion it's impossible. Which can't be, because it worked for Zim, so what's going on here? A quick rundown: * Standard GEMS missing tooth is 2 teeth in front of TDC. * EDIS wants 5 teeth behind. * Clocking the flywheel nets us 7 teeth behind. This means that without any correction, it's at 30° advance. So we need to set -20° trim angle. But according to the megamanual EDIS can't go below 10° advance, so there's no way to make this correction. Or can it, and am I reading the megamanual wrong? We're getting an RPM reading in megatune, so the VR sensor should be connected up right. Any input would be welcome, it massive screws with the mind, we've drawn so many diagrams of it all we can't make heads or tails of it anymore. Just a suggestion as am normally wrong, but, do you not need to factor in the extra 10 degrees for the missing tooth on the trigger wheel? Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 9, 2014 Author Share Posted November 9, 2014 Hmm, maybe. Fiddled with it (a lot) more today. First found out the coils were wired wrong, for some reason C & D were labelled as "coil 1" in the loom... Still didn't work. Went over the plug wires again, they were in the wrong wrong order too So with that fixed, it sounded better. Still didn't fire properly, but some smoke from the exhaust (instead of the intake...). Sounded like it was either still a little out, or just flooded (which is entirely possible). So we gave up for the day. On the even less bright side, the new cam is probably also ruined with all the cranking. Some more things learned during the day's research though: 1) MS documentation is horrible. As in, truly damn horrible. We had to find out from a recent post of Fridge on here that the trim angle only does something when the fixed angle is set to -10. Argh! 2) EDIS should be able to go to 10° ATDC, which is in the ballpark of what's needed. 3) We're still not sure how Zim's engine ever ran properly More to follow... Quote Link to comment Share on other sites More sharing options...
TheRecklessEngineer Posted November 9, 2014 Share Posted November 9, 2014 Don't forget that it's two rotations of the flywheel for every firing cycle. Therefore, by shifting the plug leads around as you have, you've shifted the firing angle 180 degrees relative to the flywheel. (I've had several pints of home-brew, so the above might not be correct.) Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 9, 2014 Author Share Posted November 9, 2014 How's that? It's wasted spark and banked injection. Cylinder 8 is 90° from cylinder 1 on the crank. So letting TDC be at cylinder 8 instead of 1 should move it 90°, not 180°. Unless there's something I'm missing? Standard firing sequence is: 1 8 4 3 6 5 7 2 A B C D A B C D So the two rotations of the flywheel are already taken into account. Quote Link to comment Share on other sites More sharing options...
TheRecklessEngineer Posted November 10, 2014 Share Posted November 10, 2014 Yes, you are right. Obviously my home-brew did the job. The only other thought that occurs to me (now sober) is have you gone the right way? Have you tried: A - 3 & 2 B - 1 & 6 C - 8 & 5 D - 4 & 7 Quote Link to comment Share on other sites More sharing options...
Cheesy Posted November 10, 2014 Share Posted November 10, 2014 If you cant get this to work converting the ECU to 4 logic level outputs is very easy, 3 transistors 7 or 8 resistors and you are done. There are many cars that have ignitors that can be used, in fact I imagine that over that side of the world the Bosch ones are pretty common (and robust by the sound of it). You can use either a 4 channel one or two 2 channel units (this is what I did with Subaru ignitors). They should drive your existing coils fine, I think the coils I used are very similar, just from a Hyundai instead of a Range Rover. Quote Link to comment Share on other sites More sharing options...
Belgian_Dude Posted November 11, 2014 Share Posted November 11, 2014 It runs as a parallel project this week, we built and swapped in some 4,6L engine in my Classic (Thor 4.0 block & pistons, "TVR" 4,6 crank & conrods, Thor ancilliaries, GEMS flywheel, OE GM VR sensor stuck somewhere in the vincinity of it) Since swapping coil leads is easier than redrilling the flywheel, and it's nice to have both team cars running similar setups, we did the same on mine and it fired up right away this proves the MS manuals are, to say the least, unclear in the trim dept, and the setup is viable. Also proves engines run better when sparked at about TDC of the compression stroke rather than random however, the theory and coil order as explaine by Elbekko checks out + Zim's 20 deg trim in MS works as well we should get the P38a running soon, must be something small now, probably mapping / after start enrichment / ... grtz, Q. Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 11, 2014 Author Share Posted November 11, 2014 Indeed. Theory proven, and proven well. Let's hope we can get the P38 running soon, shouldn't be *that* hard now we know it can work. Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 21, 2014 Author Share Posted November 21, 2014 Right, did some more work on this. The Classic runs perfectly, apart from a PAS issue, which we still haven't fully figured out. So we tried to load that MSQ into the P38's ECU... and it didn't work. Took out the plugs, they were properly flooded, cleaned them up and refitted, re-triple-checked the spark plug wires... and it still didn't work. We're properly stumped now. The ECU and wiring loom are from Nige's 90 as far as we know, this was included with the Eales heads Escape bought from Nige a few years ago. It was all fairly straightforward apart from coils A & B being labelled as "coil pack 2" I've uploaded some data logs and the latest MSQ, we're really unsure what to make of it. MS seems to think it's running a few times, but instantly dies again. Some insight and help would be very appreciated. Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 24, 2014 Author Share Posted November 24, 2014 No-one with any ideas? Quote Link to comment Share on other sites More sharing options...
robertspark Posted November 24, 2014 Share Posted November 24, 2014 Haven't read the above and digested too much but have you put a timing light on it to see where it's actually firing on number 1 and in relation to tdc? Not worked on a 4.6 but presume there are timing marks somewhere Sounds like a spark issue if the plugs are fuel soaked. If it was getting spark during the power stroke it should try to start and at least be churning over Quote Link to comment Share on other sites More sharing options...
Bowie69 Posted November 24, 2014 Share Posted November 24, 2014 Put the plugs in the oven for 20 mins. Quote Link to comment Share on other sites More sharing options...
FridgeFreezer Posted November 24, 2014 Share Posted November 24, 2014 If the plugs are soaked it won't fire, had that before. Also, are they soaked because of a fault such as injectors being fired too much by a confused ECU, or fuel pressure issue, etc.? Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 24, 2014 Author Share Posted November 24, 2014 Plugs were soaked, took them all out, cleaned them with a rag, burned them with a lighter, cleaned them with a rag again. The filthiest plug sparked fine outside a cylinder, although I do realise that may be different inside the cylinder. Fuel system is stock GEMS, I think the plugs fouling is just because it won't start. Tried both Nige's standard map, and the map Quentin uses for his 4.6 (with the exact same setup except Bosch injectors, but at GEMS pressure). No timing marks on a 4.6, could try adding some if the plugs are coming out again. But should be fine, as the same setup works fine on the other engine, and we've gone over the plug wires many times now. Thanks for the help though, was beginning to think we were invisible here Have you guys looked at the datalogs? Quote Link to comment Share on other sites More sharing options...
Random_badger Posted November 24, 2014 Share Posted November 24, 2014 OK, i'm fairly new to this but have recently been looking at timing issues with my own, I've a couple of questions as I'm not totally familiar with your set-up: Is the VR sensor reading off the fly-wheel teeth? Have you tried (from the initial post) +20 degrees in the trim? From your description and looking at mine, if i went to 7 teeth that would put my missing tooth 20 degrees BTDC, hence the +20 How many teeth on your wheel ( would allow how many degrees per tooth and a refinement of the trim angle). This may sound simplistic, but sometimes the simple mistakes can muck it up for weeks. Good luck with it... Quote Link to comment Share on other sites More sharing options...
Escape Posted November 25, 2014 Share Posted November 25, 2014 Just to clarify, I'm using the mechanical parts of the GEMS fuel system (pump, regulator, rail and injectors) but driven by MS. And the engine was running, though with a floppy timing chain, when we fitted the MS. @badger, appreciate your input, but we know the setup works because the 4.6 in Quentin's RRC uses exactly the same: 36-1 standard GEMS timing wheel with missing tooth 70° BTDC and trim +20. And we've checked it over and over again to find any differences that could explain why one engine wont start... Quote Link to comment Share on other sites More sharing options...
zim Posted November 25, 2014 Share Posted November 25, 2014 Now i'm confused. (since the PM's lol) I thought the running engine was using a clocked wheel, didn't realise you'd got it running by swapping leads - if thats the case, then we know it works, you need to check & double check the non runner Have you got (or access to) an oscilloscope to check if the VR is working ? Or are you seeing RPM when cranking ? As said above, draw on some timing marks (the 4.6's don't have like the earlier models, but can be drawn with white pen) and see if / when it's sparking. G Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 25, 2014 Author Share Posted November 25, 2014 The running engine is also using swapped leads, indeed. Everything has been checked, double checked, triple checked, ECU swapped, ... It sounds like it's almost about to fire. The logs look like it's about to fire. But it doesn't. VR seems to be working fine, cranking RPM is between 200 and 300 as shown in MS. We've used a timing light (and a plug outside the engine) and verified it's actually sparking. But no clue when exactly, as we don't have any timing marks (yet). The wiring has been tied into the stock GEMS wiring, the ignition switched live etc are all stock, so should work fine. Same for fuel pump (and we're obviously getting fuel). Quote Link to comment Share on other sites More sharing options...
Bowie69 Posted November 25, 2014 Share Posted November 25, 2014 Plugs in the oven, I bet they are still fouled -sometimes you just can't bring them back. Can you swap parts over to the running engine and confirm your potentially bad parts are OK? i.e. swap ECU, EDIS, VR, fuel pump, loom even? Could it be that the fuel pump relay is only energising once, then going off when cranking? i.e. your ignition live is in the wrong place. Quote Link to comment Share on other sites More sharing options...
elbekko Posted November 25, 2014 Author Share Posted November 25, 2014 Plugs in the oven is something we can try, or put in the old ones. Swapping parts over should be possible up to a point. - ECU has already been swapped without result - EDIS can be swapped, but seems to work (gives spark, gives RPM) - VR can't be swapped easily, on the RRC we used a "normal" VR sensor on a GEMS trigger wheel because it's a Bosch block and they're not interchangeable. But again, there's a solid RPM reading in MS, so I doubt that's it. - Swapping the fuel pump is a no-go, and it did work fine a month or two ago. - Loom very maybe, would be quite a bit of work. Fuel pump relay is the stock P38 one, just triggered by MS instead of GEMS now. Quote Link to comment Share on other sites More sharing options...
robertspark Posted November 25, 2014 Share Posted November 25, 2014 Suggest timing mark to be drawn Plug no 1 out screwdriver gently placed in plug hole (not stuffed in and forced upward) and turn the engine by hand Put a mark on the screwdriver that you can align with something and a mark on the crankshaft Turn the engine clockwise past tdc until your screwdriver drops into the plug hole to the alignment mark. Place a second mark on the pully (that aligns with something) and tdc. Will be the midpoint of these two marks Yes there are issues with this method as I'm not sure if the pistons are dished or have valve recesses in them. But this should get you somewhere close. You can also do the method a few times to see if you get the same mark It sounds like your close to getting the right timing marks if it's on the cusp of starting I have no idea how you adjust the vr sensor to advance or retart the ignition by a few degrees Don't forget the more you turn it over the more you will flood it and need to dry off your plugs again Hope this helps Rob Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.