Right, where to start. This is going to get very long, there is so much going on in my head, as well as on the engine building.
I will start this thread by setting some back ground and then expanding onto the 1400 TBi engine thats currently in the car. I will start another thread for the 16 valve turbo engine that is the current build project.
Most people will know of the Mini, most will know that its unrefined, noisey, uncomfortable and slow. Three of the former are fact, the latter is not! Dont get me wrong there are fast cars out there, but when most people see a Mini they asume its slow.
I have a personal attachment to the Mini, don't know where it came from, maybe when I was a kid, I dont know, but i've always, when i had the chance, mucked about with Minis. However, i will point out this, I am not so sad that I have named any of my minis, or called it a He or a She. Its rather disturbing to think people do :D
Anyway. I've had four Minis and the current one was rebuilt last year in 14 weeks, that included a new engine and gearbox. This Mini is one of the later Single Point Injection veriety.
The original engine was Boggo standard and the injection system was running pants. I bought a Crypton ACT service tool, a complete new injection maniflod and set about fixing the thing. Since then I have bored out the throttle body and fitted a larger throttle disc, fitted a larger injector from a Rover 214, increased the engine capacity from 1275 to 1399cc,fitted an injection Kent MD274 cam, a Group A rally cylinder head, a Mini Cooper ECU, a K&N and a 2" exhaust system. I dropped the Final drive down to 3.76:1 and stripped out the interior partialy. Needless to sayits feckin quick, lol, up to 90mph where it tops out, lol.
The reason i'm here With a VEMS ECU is two fold. The current engine spec outside the capebilities of the standard ECU, and I will also need a programable ECU for the 16 valve turbo.
I did try and overcome the knock issues with the standard ECU by retarding the reluctor ring on the flywheel. Unfortunetly at that time i didn't know that the cold crank degrees were 5. Guess how much i retarded the reluctor ring, yes, 5 degrees. Starting on a cold day is rather difficult. It will turn over and start first time, but then almost imediately the reves die off, and with foot flat to the floor, the reves gradualy rise untill about 3000rpm where it then seems to recover and then all is as normal. Its fine once its running and will start with out problem once warm, or warmer than cold. It does drive a little boggy in the lower rpm but pulls strong from 3k right through the red line on the rev counter without complaint.
I think the standard ECU is not releasing the engines potential anyway, so what with the other engine build, i decided to look into the DIY programable ECU market. On a budget there only seemed to be one option, Megasquirt, that was until i was pointed in the direction of VEMS by a fellow Mini entusiast in Norway. Looked into it and the rest you can work out for yourselfs.
On this engine, the 1400 TBi, i'm looking for 100bhp, thats the target, though if i get into the 90s i'll be a happy bunny. I'm currently pulling appart two wiring looms, one from a Mini and the other from a Rover 820 Turbo. All the screened cables and connectors are terminated, making it easy. This custom loom needs to be a quick swap for a standard loom and will use the same main loom connector to facilitate this.
The ECU itself is being fitted to a Rover MEMS case to give it more of a standard stealth kind of look, even though the original location is being abandoned and being mounted inside the car either under the dash or on the bulkhead. Ther wont be any room in the engine bay with the 16v engine, and as it stands now, the inner wings are removed, where the original ECU resided, and the ECU is now just cable tied to the engine itself, which is far from ideal!
Any way, i think thats enough for now, but i will be back soon for an update.
Engine Specification
Austin Rover A Series Transverse engine http://www.austin-rover.co.uk/index.htm?engineaseriesf.htm (http://www.austin-rover.co.uk/index.htm?engineaseriesf.htm)
5 Port EX Works Group A cylinder Head http://www.wpm.co.uk/worksminis/ (http://www.wpm.co.uk/worksminis/)
8 Valves
Inlet valve Diameter: 35.4mm
Exhaust Valve Diameter: 30.1mm
Stroke: 81.33mm
Bore: 74mm
Compression Ratio: 11.2:1
Cam spec: Kent MD274
K & N 57i Induction Kit
44mm Throttle Body
611cc Throttle Body Injector
Slark Race Engineering Re Worked Inlet Manifold
Maniflow Injection LCB exhaust Manifold
PlayMini 2" Big Bore 304 Stainless single box Exhaust System
Gearbox Final Drive Ratio: 3.76:1
(http://www.turbominis.co.uk/_photos/members-motors/293/2664.jpg)
(http://www.turbominis.co.uk/_photos/members-motors/293/2666.jpg)
(http://www.turbominis.co.uk/_photos/members-motors/293/2675.jpg)
(http://www.turbominis.co.uk/_photos/members-motors/286/3049.jpg)
Retarding the reluctor ring in the flywheel
(http://www.turbominis.co.uk/_photos/members-motors/286/3070.jpg)
(http://www.turbominis.co.uk/_photos/members-motors/286/3072.jpg)
(http://freespace.virgin.net/colin.milne/VEMS/DSCF2332.JPG)
(http://freespace.virgin.net/colin.milne/VEMS/DSCF2333.JPG)
Nice project, thanks for putting it on the forum.
I've always had a soft spot of the Mini as it was the first car I modified.
It was a 1380 I build using Vizard's book
(http://www.vems.co.uk/MiscPics/RHMiniNorthWield.jpg)
This was taken at a sprint at North Wield about 18years ago.
What 16V head are you going for? The BMW one or one of those exotic CNC'ed ones?
Nice project! Im gonna keep a close eye on this thread as i have a 1275 midget that will get the VEMS treatment at some point :) Vizzards book is on my shelf too ;D
Vizzard wrote some good books. :) (There's even a Rover v8 one...)
Rob: You're making me feel young! ;)
Quote from: [email protected] on February 08, 2007, 08:42:12 AM
What 16V head are you going for? The BMW one or one of those exotic CNC'ed ones?
Its the BMW head, those KAD and JKD heads must be gold plated :o
I'll get that topic started soon. There is a growing interest in this conversion, so plenty of potential for some VEMS sales, if you can turn them from Megasquirt (you know my opinions on that Rob).
Jim Lyons of Sheepspeed Racing in South Wales got the first ever 16 valve turbo A series running at the British Mini Showdown last year, July 15 and 16th. After a marathon build and three full days over the weekend of the racing trying to get the engine to fuel/ spark right, other things hindered a run down the drag strip. However history had been made and witnessed by virtualy all those interested in Minis and engines that were at the strip, a 16v turbo A series Mini driving under its own power. Later development saw Jim at a Mapping session on Emeralds rollers. There were still some issues, namely the lack of control of boost, Dave Walker ended the session with 200+bhp and 178lbft torque at 6500rpm and 21psi boost with the power curve still ging in the upwards direction :o. All that from 1330cc :D This was reported by Dave Walker in his column in Practical Performance Car in the last few months. Other than that, Jim is keeping the final figures under the hat for this years British Mini Showdown. There is going to be a few of these engines knocking about soon. >:(
Quote from: Maff on February 08, 2007, 01:58:16 PM
Nice project! Im gonna keep a close eye on this thread as i have a 1275 midget that will get the VEMS treatment at some point :) Vizzards book is on my shelf too ;D
If you can bear butchering a block, the BMW head is the only way to go. It would certainly put you in standing on the Midget arena ;D recon on 130bhp+ normaly aspirated from 260 duration cams. right up to 170bhp on 300 duration cams. These heads flow more standard than a fully modified offset valve five port head considered the best possible.
Like i said, im going to start another topic soon
Right then ???
I am close to finishing building the VEMS. I am half way through the spaggeti that is the wiring loom. I now want to clear up some uncertanties. Now, to some, these questions may be, and have, a simple answer, but I am either asking because I am seeking confirmation or I just plain dont know or understand. Please bear with me, I do learn quick though, however my mind is a fuzz with three projects on the go :-\
I shall post individual topics regarding each area of questioning/ discussion and then link them on here. So off to start my questioning ;D
Dual chanel knock detection http://www.vems.co.uk/forum/index.php?board=5.0
Tacho options http://www.vems.co.uk/forum/index.php?topic=41.0
Crank sensor positioning http://www.vems.co.uk/forum/index.php?topic=42.0
Wide band Lambda sensor loction http://www.vems.co.uk/forum/index.php?topic=43.0
I Button http://www.vems.co.uk/forum/index.php?topic=44.0
Extra Input Chanels http://www.vems.co.uk/forum/index.php?topic=45.0
Output chanels http://www.vems.co.uk/forum/index.php?topic=46.0
I think thats it for now. My head is mashed so no doubt i have missed loads i want to ask. ;D
Right, because the engine is only going to be in the car for another two or three weeks, and ive had other issues, namely a bent valve, i cant be arsed pulling the flywheel again to alter the relucter ring in the flywheel ::) So, i fitted the trigger wheel i bought for the 16 valver and made up a bracket for the sensor. This bracket actualy doubles up as a lower engine steady as well for the 16 valver.
(http://freespace.virgin.net/colin.milne/VEMS/DSCF2443.JPG)
I have also modified the stepper motor. The motor is a 6 wire Unipolar motor and obviously the VEMS stepper driver is a Bipolar. This was simply a matter of seperating the two common center tap connections that the Rover MEMS used as a +ve. This was simply done by scratching the track away on the small connecting PCB on the motor itself. I also removed the wire.
(http://freespace.virgin.net/colin.milne/VEMS/DSCF2445.JPG)
Ive finished the wiring loom pending the arival of a wideband lambda and its connector and im also waiting on a two wire knock sensor turning up.
Heres the ground wiring. I actualy found this a bit of a bugger TBH, it was re hashed a few times until i felt happy that it was adequate. I followed Robs instructions, 0.75mm wires for the grounds as short as possible, all grounds connecting at the same common ground point, sensor grounds attaching to sensor ground wire at aproximately 20mm away from the common ground point, cable screens attaching to the common ground point at the body (Rob recommended the engine, but the starter is mounted to the engine as is the alternator and its a small engine, I chose to follow the original loom). The common ground point is then extended with 4mm wire to the ground point on the car. The original factory location.
(http://freespace.virgin.net/colin.milne/VEMS/DSCF2440.JPG)
LCD is now up and working, just need to find a place in the car to fit it, lol there is no dash, just the clocks. Config of the VEMS is underway. Woo its nearly there
It wont be long now ;D
ok
Update on this.
I have been fiddling around with this for a few nights now. Its going to be the last week with the 1400 in the car so trying to get it to a point where its driveble so i can use the config as a base start on Jackmans engine. On that note Ive persuaded him to go OEM system untill the engine is run in and he has returned from the International Mini Meet in Denmark. He still has a fair bit of the car to finish assembling and the time left to map the engine is going to be very short, going OEM will elimniate this time consuming event in what is a very tight schedule.
I have got the stepper motor up and working and tuned, works really well actualy. Steps to extend set to 190, steps at 77c 80 and -40c 180, follow TPS enabled and step sequence set to 216, as for the wire configuration, i'll have to get back to that, but its easy enough to work out any way. Very important to mod the motor as in the above post. This also works on the idle valves with the same mod to the motor.
Finaly got the warm crank PW to 1.5, afterstart and idle VE set up so that it idles as even and smooth as its ever going to be and starts without giving it gas first try. Just need to see tomorrow if the cold crank PW of 4.5 is about right. Ignition table is prety generic but from what i have seen of others maps of varying levels of tune and cc for the A series, its fairly standard across the board, so i think its a good base to start.
Warm up enrichments i have yet to tune.
All the outputs other than the emissions purge valve work correctly, but still think the misc outputs needs a temperature refference to make them truely universal and versatile.
I have set the acceleration enrichment amounts to 0 for the momment as the higher VE load sites are way off causing any sort of hard acceleration with acc enrichment to overfuel and almost kill the engine.
I have tuned the EGO, whether its correct and what i should do i dont know but before i did, the engine speed fluctuated with EGO correction, I altered the cycles before change, slowing it down a bit.
One thing i noticed was I cant get the idle MAP down below 30kpa, it hovers around that point, maybe the valves are bolloxed again, or is this about right??? VE at idle is 18% with Lambda 1.00 and 15 degrees advance and 900rpm, TPS is 2% and PW of 1.3 ( big injector) Im running a low impeadance single TB injector with a resistor.
All is going well as far as i can see, hoping to go for a short drive with VE learn to get the VE table closer than it is and then hopefully set up accelerator enrichments by feel.
Then its on to the Knock set up :-\ and then engine out and 16valver in and start all over again :(, well at least ill have a base to start from ;)
LOL i spent a few hours in the car just trying to get it started and idling well, playing with numbers, thats all its down to now, having to charge the battery twice, setting up the primary trigger, stepper motor and the outputs, it all takes its toll on the battery :D
On a note, i have been reading of people complaining that the over run fuel cut causes the engine to go verry lean, this is not so, there is no fuel there so no burn the engine becomes a vacuum pump, pumping air. The lambda sensor sees it as very lean as there is 20.9% oxygen present. This is not bad and the OEM ECU also pics up over run fuel cut as off the scale lean. On investigating this on the OEM ecu, i noticed just before fuel resume the lambda voltage of the narrow band rose to 206mv from the base line of 200mv during fuel cut, maybe a pre stage fuel resume or some thing?? possibly to prevent the engine undershooting the idle rpm, stepper follow TPS enabled helps prevent this as well.
Quote from: Sprocket on May 02, 2007, 12:32:21 AM
One thing i noticed was I cant get the idle MAP down below 30kpa, it hovers around that point, maybe the valves are bolloxed again, or is this about right??? VE at idle is 18% with Lambda 1.00 and 15 degrees advance and 900rpm, TPS is 2% and PW of 1.3 ( big injector) Im running a low impeadance single TB injector with a resistor.
I'd be suprised to see vaccum below 30kpa at idle, what sort of value were you expecting?
Rob
Quote from: [email protected] on May 02, 2007, 08:27:48 AM
Quote from: Sprocket on May 02, 2007, 12:32:21 AM
One thing i noticed was I cant get the idle MAP down below 30kpa, it hovers around that point, maybe the valves are bolloxed again, or is this about right??? VE at idle is 18% with Lambda 1.00 and 15 degrees advance and 900rpm, TPS is 2% and PW of 1.3 ( big injector) Im running a low impeadance single TB injector with a resistor.
I'd be suprised to see vaccum below 30kpa at idle, what sort of value were you expecting?
Rob
LOL
Well all the gauges i have seen and the OEM ECU have values in mmHg with idle above 500mmHg (or is that below :P) I haven't really bothered to refference it across to absolute scale. I understand the metric values well as these are what i use every day, i was just sumising that it would have been nearer the 25 to 20 kPa.
Coolio, all is well then ;D
http://www-users.med.cornell.edu/~spon/picu/calc/pressure.htm
Enjoy :P ;) ;D
Coolio
Only problem with that is that the conversions are not for absolute pressures, which is what we are dealing with. It can and does get confusing more so when converting from an atmospheric scale to an absolute scale
0mmHg = 100kpa absolute
Therefore 735.559/100*30=220.668mmHg.
If you are not carefull and look at that, you may be confused into thinking that it is wrong, but the numbers when using an atmospheric scale are inverted so its actualy 735.559-220.668=514.891mmHg
Basicaly the numbers on a atmospheric scale rise with vacuum , and fall with an absolute scale
I find the whole range of SI units upsetting.
Imperial is too random for my liking ;D
I mean, who would convert the "Hg to mmHg to metricise it ::) As far as im concerned its still an imperial scale :D and still random
The Germans knew what they were doing when they invented Metric :D How simple is that when everything is divisable by 10 :D
However, each to thier own. I was taught metric so there :P
Dear god, I'm not an imperialist.
Metric units everytime, but my point is... Which one to choose from?
Lol sorry dude, I didnt mean to offend ;) :P
As for which ones, it has to be Bar and KPa. The two are easily refferable across. 1 bar is 100Kpa, 50MilliBar is 5 kpa . I use nothing else at work even when talking about deep vacumm, Screw the Micron vacuum scale, thats just for scientists dealing with perfect vacuum and wanting to sound flash down the pub :D
You got to love matric for its simplicity ;D
Anyway, back on subject.
I went for a drive last night with VE learn enabled. Is this slow or is it me? I drove around for about an hour and managed to get the low load sites pretty good and drove well, but i thin because im using such a big injector that the calculated VE table especialy at the higher load sites are still way off the mark. if i tried to keep the speed going up a slight hill the lambda went off the scale rich. LOL I need to knock the VE entries back about 20% and have a co driver on the Megatune ;D
I am a happy man ;D its all down to numbers now, oh and sorting out the knock sensor. LOL I'll have to start again with the 16v but at least i should be closer than i was when i started out :)
Following on from the 16 valver thread as I have now put this 1400 back in the car and took it to the mapping session intended for the 16v :-\
What fun was had getting it all back up and running. Infact i managed a good 2 foot flame out the exhaust, LOL. What had caused that was i left the hi flow fuel pump in the tank that i had fitted for the 16v. I had also fitted a smaller injector. I thought id try it and see if it wasnt going to be a problem, how wrong!! swapped out the pump, fitted the bigger injector and uploaded the original config i new was good to go on this set up. LOL first crank and it fired.
Im still concerned that the injector is too big, and the idle pulse width shows this at 1.4ms and also the VE map after mapping doesnt really get above 30. It drives real well though. Im going to try the smaller injector which is 10% smaller so, i presume i can scale up the VE map by 10% and hopefuly the idle PW will come up with it. Im going to fit a standard size TB as the modified one now gives 2% TPS at idle and drops to 1% at worst. Im having some issues when i dip the clutch at a junction, the engine stalls, and theres now hesitation about it, it just stops. Starting it is also a little difficult so I know the crank and afterstart settings need looking at. I think the stalling is not being helped by the low PW and 1% TPS. I'll start another thread about that.
Another thing that im a little concerned with is the EGO conrrection, it seems to be running the Lambda a little lean at crusing speeds and not bring it into line. The lower end of the VE map I need to tweek but its not too bad, some areas wer left as time was getting on.
At the end of it all, its not going to det, and im getting 14 lbft more torque than i had on the OEM ECU, its down by 5bhp but im not concerned with that as this dyno can not be compared to a rolling road. everything is happening just that bit lower in the rpm as well which is good, i dont have to rev the balls off it to get it.
Results, Flywheel power 95 bhp, flywheel torque 101 lbft, bhp at the wheels 80 bhp, torque at the wheels 86 lbft (460 lbft measured in third). It drives much better than the OEM ECU and deffo feels a hell of a lot quicker, but i do now have a lightened crank and ultra light flywheel in there. This may account for the bigger torque numbers and lower down the rpm?? Off to the drag strip to actuly see if it is quicker :P
The ignition map is what was generated at the same time when using the VE calculator, it was there when i looked at the table, looked pretty good and the dyno operator said it was pretty good also, only needing a couple of extra degrees at the top end, nothing more was gained worth worying about by going further. The map does look pretty representative for the old A series :D
Lambda map again is pretty standard,
(http://freespace.virgin.net/colin.milne/DSCF2651.JPG)
(http://freespace.virgin.net/colin.milne/DSCF2650.JPG)
(http://freespace.virgin.net/colin.milne/Plot.JPG)
(http://freespace.virgin.net/colin.milne/IgnitionMap.JPG)
(http://freespace.virgin.net/colin.milne/LambdaMap.JPG)
(http://freespace.virgin.net/colin.milne/VEMap.JPG)
Hi,
101 lb ft is good going! 8) My 1380, 286, 45 DCOE is only 92!
Like you say there's no two rollers/dyno going to read the same anyway!
Cheers,
Gavin :)
Nice results.
I'm interested in the issue you're reporting regarding the lambda correction, does VEMS report that its lambda 1.0 or does it know that its at 1.x and is not putting more fuel in? Are you using the correct wideband settings ( http://www.vems.co.uk/forum/index.php?topic=55.0 ), IIRC you built your own board did you get the pump pw zero and nerst values 100%
Rob
Nice results there! That type of dyno is the one I am going to use when I am getting mine mapped too, did you feel it was better than a RR?
That all looks nice. Good smooth curves, if a bit strangled at the top end. Do you have any provision for altering the cam timing to see if you can free up some high end horses, or do you think its something else?
One point on your target lambda map. You've got the right kind of numbers for high RPM, but at low rpm the engine won't be making the full 100kpa at WOT. Therefore you need to plot the WOT pressure line across the map, and then enter the target 0.86 along that line, and blend back to your cruise sites from there. Just a small point, but will help you pick up a little bit more low end torque. And if you're detonation limited at lower speeds, the extra fuel will help and maybe allow you to advance the ignition a bit too, so even more torque. :)
Nice work so far :)
Quote from: [email protected] on July 09, 2007, 08:56:24 AM
Nice results.
I'm interested in the issue you're reporting regarding the lambda correction, does VEMS report that its lambda 1.0 or does it know that its at 1.x and is not putting more fuel in? Are you using the correct wideband settings ( http://www.vems.co.uk/forum/index.php?topic=55.0 ), IIRC you built your own board did you get the pump pw zero and nerst values 100%
Rob
The sensor is reading correct, what i meant was that the lambda reading is way off the lambda target and doesnt seem to pull it back into line.
However having said that, I checked the factory defaults for WBO2 and found the Heater PID Ri target was set to 50 and not 150, probibly a key stroke mistake or just my dyslexia.
I did build the board and the first thing I did was calibrate the WBO2, it was reading correct. I thought I posted the values used, but maybe not. The sensor is reading right as it was comparable with the WBO2 on the dyno.
If you look at the log files I posted in the other topic about the stalling you should be able to see that at cruise its around 1.19 with a target of 1.00, always seemed to be lean.
I posted on your stalling thread about the EGO correction values, might be worth giving them a try.
Well, recovering from a week on holiday in the Mini ;D
Covered 1200 miles with only a few minor issues, none really to do with the ECU.
The only real issue was that the engine decided to stop of its own accord, and not restart. This was in the second lane on the River Avon bridge crossing on the M5, on the up side, in busy traffic :D what fun. Managed to push the car and trailer ;D over to the hard shoulder and set about working out what the problem was. This was also the time i noticed the screen started to play up. anyway, cut a long story short, swapped out the injector fuse and we were on our way.
During the week I left the VE learn on and encountered virtualy every driving style, from motorway, round town crawl to single track roads only wide enough for the mini :D with lots of first and second gear corners, hills of varying degrees and length. The VE table is now pretty spot on, though the numbers look a little low and flat. I suspect this is as a result of using a low ohm injector with the resistor, the characteristics of the injector are some what different now. I intend to half the req fuel and double the VE values to give a little better resolution, before trying the PWM, for which im going to try and get hold of some accurate information.
Acceleration enrichments were an issue with AE happening where not required. The mini is quite a bumpy ride and what you would consider a steady foot, is actualy inducing AE with the suggested Acceleration Bins 1 5 20 40. Changing the first two bins to 3 and 7 as suggested by dnb seemed to sort this. The hard acc amounts still need tuning.
Idle is pretty steady though i had to set the idle speed to 950rpm due to the ultralight weight flywheel and high compression. I have been playing with the EGO numbers to get the lambda to be fairly constant only fluctating about 0.07 max above 1.00. It never seams to go below the 1.00 even with the target at 1.00, i assume this is to do with the wideband lambda sensor PID??
Im still encountering the over run situation but no to the point of stall. I have tuned the engine so that its reasonably OK with no load, but in the recent weather with the heater, the wipers and lights on, it is a problem. Idle drops to below 800rpm before the idle PID activates the stepper from its lowest position after over run. Stepper speed set to 5 wich is bloody quick. IT is a problem!!! >:( Its not how the OEM working and i feel it needs resolving by the powers that be :-*
Other than that system works a charm ;D
Over the 1200 miles 650 of those with an extra 300kg payload/trailer fuel consumption averaged 36mpg, remembering that this has a 3.76:1 final drive where 60mph is 4000rpm. I dont think that is bad considdering the extra weight and is only down 2mpg from the OEM ecu without the extra weight of the luggage and trailer :D
Oh, one thing a really do need to sort is the coolant temp bins, or at least double check. The engine temp when driving is 80 yet i know the cooling fan switch comes on at 97, the VEMS indicates 86. TBH i havent realy calibrated this, oopsie ::) this will no doubt screw up all my crank, afterstart and warm up enrichments, after getting them pretty spot on ::) :-X ::) ::)
Im gooing to also have a play with full throttle shift and launch control whel i get the clutch switch fitted ;D ;D
still ongoing ;D