wrs

Nice, My D1 Racing Clutch & Lighted Flywheel is coming out and getting a replacement like this although it will use the 240mm disk and standard M3 pressure plate. I've had nothing but trouble with the D1 clutch not fully disengaging since it went in. I suspect there is some issue with the D1 pressure plate requiring additional travel which the S6S 420G slave + release arm cannot provide. However, they cannot offer any technical details or drawings showing the required travel. The lightened flywheel is also causing a lot of gearbox chatter at idle. That flywheel looks pretty nice.

I spent a little time today tidying up some of the text. Long strings tend to flicker a little if the whole string is re written - like the time 00:00:00. Now I'm writing each set of digits individually and only if they change so there is no more flicker. Also, played around with some backgrounds, font colour changes and changed the icons for the Halo & Time. Here's the new screens (but almost anything is possible). Main Screen: Time Setting Screen: Halo Control Screen:

Eh, my Vanos has started to rattle. This began a couple fo weeks ago and has been getting slowly worse. The seal and rattle kit arrived during the week from Beisan Systems so it's going in on Monday to get fixed. Hopefully the timing chains and guides are ok... I was going to do this during the install but thought I might get away with it. Lesson - old vanos, don't use it without replacing the seals and doing the rattle fix first!!!

I have it running in it's first cut form. The photos below are the unit out of the car on my test bench - so no comms with the ECU. In this state all the values in the receive buffer are zero and display as such with the offsets etc. At present all screens will likely need a bit of prettying up and I want to change some of the icons to get a better looking theme. I'm also open to suggestions for layout. RPM is in the top left corner in the bigger text with the Coolant Temp next to it - these are the two main items of interest for me. The unit is based on the Freetronics Ethermega, 4D Systems uLCD-32PTU, Freetronics RTC Module, Freetronics Light Sensor Module and a Custom Isolated RS232 Module using a ADM3251E chip. It samples from the ECU every 100ms and updates the LCD if values have changed. It has a full-colour Halo controller built in. It remembers the last state of the halos and the colour settings using the spare non-volatile memory in the RTC. The RTC was added in case logging is ever added to the design + to show the time + to give a small non-volatile storage space that has no write limitation. The halo's also dim with the ambient light level so are always the same relative brightness whether on at night or during the day. The brightness is also held constant with varying battery voltage levels. The LCD in the car also dims to match the ambient light level so it doesn't blind you at night. This requires a custom mod to the LCD as the uLCD-32PTU does not support dimming standard. It only requires removing one chip resistor from the LCD board and tacking on a new resistor to a spare pin on the connector. This pin then gets linked to a PWM output on the Arduino board to directly control the backlight LED. At present it's only monitoring for over-temperature and jumps to a over temp warning screen if the event occurs. I haven't got a screen shot of this in the pictures below. Ethermega Board Adapter Interface with the RTC module and RS232 module + off-board connectors Main Screen Time Adjustment Screen Halo Control Screen

Yep, analogRead is very slow and I'm pretty sure it's blocking (stops all other program execution and waits for the ADC). I haven't used the Arduino serial library that much but will check how it goes before committing to it. I usually use a timer to generate an interrupt to trigger an ADC conversion. On the ADC interrupt I store the value into an array and increment the mux. This way the ADC is sampling at a fixed rate in the background with very little input from the CPU and the values are stored into an array of integers. You can do the same for the serial port too - keep the overhead very low. From memory the Arduino analogRead takes some very long time like 100us. If it is actually 100us then several thousand processor cycles are wasted... Conrolling the ADC, Serial Port and SPI for the SD card using interrupts is very much faster than the standard libraries. I'm certainly not that good a programmer and mostly self taught - just been doing it for a very long time.

Thanks for the info. Re the LCD - looks like a good project. The main reason I like the 4D display is it's uses it's own uP to do all the work - which has its adavantages and disadvantages. It wouldn't be too difficult to create a bunch of different forms (say 6-8 for each layout options) and just reference their objects using an array. It would mean all the object parameters are fixed as they are coded into the LCD. What I like about your project is the ability to contral all aspects of the object (eg colour) to allow for personal preferences. Re logging - I haven't had any problems logging data to a SD card before. The basic Arduino @ 16MHz seems to be able to write around 8kb/s to the SD card without problems (not using the Arduino library). This is a lot of data... For all high-speed applications I don't use the Arduino libraries (have been using the Amtel ATmega range for many years before Arduino came along). While the Arduino libraries make life simple for beginners the code is usually very slow (1/10 of what is possible on many library functions). If I need high speed I either directly control a peripheral or create my own small task specific library. If speed isn't important I just use the Arduino library - they're pretty good for most cases. I'm nearly done the first cut for my application. After that I'll look at what it will take to mod for others to use. By next weekend I'll be able to put up some photos of the working system.

Some good progress has been made on the project. I've got the colour LCD formating done and talking to the uP on the Arduino board. I'm getting the data from the ECU ok and have some of it mapped to the LCD. I'm trying to figure out a way to make it configurable for anyone who wants to use it - even if they're not software savy. The Arduino side is easy as a bunch of remarked 'define' statements at the beginning of the software can enable the values you want to look at so any or all can be configured to suit a personal requirement simply by un-remarking the define statements for the options you want. The difficulty is with the LCD - it requires a uniquie setup for every possible combination of data. Maybe the answer is to just have a bunch of preset options (maybe 10) and the user can pick the closest to their needs if they can't make the modifications themselves. I've also included a colour controller into the design so anyone with colour Angel Eyes or other colour LED lamps can control the colour, brightness and on/off of the LED's. It will adjust the RGB in 100 steps for each colour and 100 steps for overall brightness so any colour is possible. This code came straight from my house automation project. I'm not considering data logging yet but it would be possible as the Arduino supports SD cards in FAT16 or FAT32 so not too much limitation on storage - certainly enough for months or even years of log data...

Yeah, I suppose I should put up some pics of the car. It's nothing flash and needs some work to tidy a few things up which is why I haven't put any up yet. I guess I'll have to wash it and take a few...

Thanks muchly for the offer but I have already committed to changing. There is supposed to be quite a big difference between the 220 & 420 - not so much the flywheel/clutch/pressure plate but the release arm and slave - apparently...

Thanks. Mmm, yes and no... It has plenty of grunt and does go really well (especially compared to how it was) + it's a real hand-full in the wet. But, the additional torque is seriously showing up all the old stretched suspension components. I think it won't just be an external tidy-up of the car required - new tight suspension components all round too + changing the 15's to 17's on the next tyre change. I'd like to go to the early style M3 double spoke rims. Also looking at suspension options to lower it just a little and get Bilsteins all round + matching springs (has Sachs now). It will be a while before I can actually do any of these upgrade though - will just have to dream about it for a while :-)

Time for an update. Over the last few weeks I've been tidying up the last few things that were a bit rushed during the initial install. I was also having some trouble with the water temperature in the head fluctuating which may have been related to where the expansion tank had been installed temporarily during the install. It's now been relocated to the front again on the corner of the radiator and the temperature issue appears to be resolved - possibly small air pockets getting trapped?? The other major is the clutch. It's never worked correctly since installed. It disengages most of the time when the pedal is right at the floor but sometime doesn't quite fully disengage. I've had to remove the second carpet and the rubber stop on the clutch pedal which gave the pedal an additional 5mm movement. It disengages a bit better now but is still crap. What's worse is when it's hot, changing gear becomes a roulette game - will it crunch into gear or not!! I purchased a NZAD D1 Racing Clutch with lightened 6.5kg Solid Flywheel. They have not been able to help at all and cannot supply any form of support (dimensions, drawings etc) and just keep stating they've never had a problem before. It may just be my specific configuration but lack of support means I can't really figure out what is wrong either. So, I'm dumping the D1 Racing Clutch and getting a custom one built at KBM. It will be early next year before I can get in to have it installed so I'll just have to suffer in my jocks in the meantime. The carbon granules in the evaporative emission control canister have been replaced + the canister has been relocated to make space under the inlet manifold - much better. Here's what the engine bay looks like now: The new mount for the expansion tank. This doesn't fit the Fenix all-ally radiator and an ally bracket has to be fabricated + high quality cable ties with a stainless tongue used to hold it on. I stll need to add the low-level switch back - is ordered and should arrive next week. Relocation of the evaporative emission canister. The tape is only to hold it together while the glue sets. About the only thing left in the engine bay to fix up is the crankcase ventilation pipe - M50 vs M52 has a few differences... I need to find a solution to tidy up this make-it-fit solution: Once this is done it will be time to focus on the outside of the car. It's a bit rough in places. I had it repainted about 5 years ago but the guys doing the painting turned out to be sharks. The resulting job initially looked ok but isn't lasting. It stone chips really easily and now looks worse after 5 years than it did on the original paint at 18 years... I'd like to do another paint job at same stage + fix up all the window rubbers and other old parts showing their age. This will have to wait for about a year though - no surpises, funds are bit low after the engine job...

At this stage I think I'll try a 325 slave to get slightly more travel on the pin. M3 Slave = 22.20mm 325 Slave = 20.64mm Area diff = 22.20²/20.64² = 1.1569 So there should be about 15.7% more extension on the 325i slave. If it moves about 20mm then that's an additional 3.13mm travel on the pin. This might be enough...

Today I purchased a USB camera on a flexi-lead to see if I could get in through the slave cylinder hole to see the realease arm pivot pin. I got close but just couldn't see it (saw the spring clip a couple of times) and then the camera got stuck!!! Managed to get it out after about a minute of careful wiggling but decided that would do - don't it to get stuck and break off leaving loose bits in the bell housing. So, I'm still none the wiser. I want to find out what the problem is withough taking the gearbox out. It could be the pivot pin but I'd rather not take the box out to find it's actually ok. Here's a picture of the geometry of the clutch - does anyone know if this is correct (or prepared to measure on their car)? This picture shows the slave cylinder unbolted and sitting at the point where it takes pressure to push the slave rod into the cylinder.

Do you mean the metal release arm the release bearing clips onto? If so, do you know which cars/engines used the different arm?

Thanks. I'm pretty sure it's correctly seated - can just see it using a flashlight and mirror looking through the slave cylinder hole. Is sitting in the notches. Have been considering using a 325 slave cylinder instead of the M3. The smaller bore-size of the 325 slave means the rod will travel further for the same amount of pedal depression. This won't help if I'm right on the travel limit of the M3 slave already and may lead to the pin blowing out on a 325 slave. I guess the real question is - is 29mm of engagement of the slave pin the correct amount or should there be more? If there's supposed to be 35mm + of engagement when fitting the slave cylinder then the likely culprit is the fork pin. The fork pin looked ok when everything was fitted to the car but the stage II clutch kit may apply more pressure to the pin than the original clutch did and being the original old and possibly brittle pin it may have collapsed. Looks like an inspection camera is required... The distance from the mounting flange on the bell-housing for the slave cylinder to the dimple in the release bearing fork is 73.5mm. I'm not sure if this distance is correct - there's bugger-all dimensional info available to help diagnose clutch problems on BMW's!!!

Recently I replaced my engine and gearbox. The car is a 92 E36 325i Coupe which had the old engine and Jatco gearbox removed. It was replaced with a M50B30 + S6S 420G Getrag six speed manual conversion. The kit I purchased included a dual mass flywheel and standard M3 clutch. I decided to by a NZAD BMW E36 M50 M52 S50 Solid flywheel & Clutch kit to get the lighter flywheel and all new parts. The flexi-line to the slave cylinder is braided. My problem is the clutch doesn't disengage until the clutch pedal is about 10mm from the floor. It also ocassionally doesn'tt disengage fully leading to a bit of a crunch into the next gear. It does help a bit if I quickly tap the gas to rev match. The first 25mm of my pedal travel is very light then it goes stiff for the rest of the travel (pressing by hand you can feel the pedal spring first, then the hydraulic pressure). Today I decided to try to find out what the problem is. I've bled the brakes several time, at least twice each time using the pump the pedal - release the bleed valve method, press the rod on the slave cylinder method and the reverse flow method. For the later 2 methods the slave cylinder was removed and held vertical with the bleed valve toward the top or bottom depending on the bleed metohod (top for press the rob method and bottom for reverse flow method). I got no bubbles and no change in the behaviour of the clutch. I've checked the pedal pivots and everything is reasonably tight with very little slop. So, since all the usual issues appear to be covered (have I missed any??) and started doing some measurements. The slave cylinder rod appears to have approximately 45mm of available travel. When I put the slave cylinder into the bell housing the rod engages into the release bearing fork with 29mm of the cylinder still sticking out. You then push on the cylinder to push the oil out and get the slave cylinder into position to put on the nuts. This then means I have a maximum travel on the slave cylinder rod of 29mm of the 45mm possible. Could this suggest the fork pivot pin is buggered? Does anyone here know what the correct travel should be on the slave cylinder rod. I would have thought it would be around 35-36mm leaving 9-10mm spare for self adjustment to clutch wear. If the pivot is ok then what's the solution - extend the rod? I may have to get a fibre-optic inspection camera to check the fork pivot pin? Any ideas?

Now that's a good idea - hadn't considered putting it on the steering wheel. I like it!!

Will probably just keep the Arduino base so it's simple for anyone to modify. Link publish all the data you need to make your own system in their help menus. When I approached them to ask if comms was possible to their ECU they pointed me to everything I needed. Anyone wanting a pre-developed LCD display can just use one of theirs. In my case I wanted to monitor and alarm on bad situations hence developing my own (plus I wanted a colour LCD). I susect there won't be an enormous amount of interest - I'm just making the offer in case anyone has been thinking they want their own custom display with data and/or monitoring not currently available directly from the Link products. I'm certainly not trying to make a commercial product to compete with theirs...

I'm currently developing a colour LCD monitor for the Link G4+ ECU. To keep it simple I'm using the serial port on the ECU (considered CAN but I don't have a developed C library for that). 4D Systems in Ausi make a series of fairly smart touchscreen LCDs in various sizes. I've used these a lot for home automation systems and various other small projects. One of the easiest small embedded development platforms is the Arduino kits. The current prototype uses a Freetronics Ethermega (same as an Arduino Mega but with Ethernet) and the 4D Systems uLCB-32PTU. I've set up the Link G4+ to talk at 115k baud and Long Datastream to the Ethermega. The Ethermega sends the data request then receives the data block and translates it to display data for the LCD - really simple. Included in the design is a real-time clock + you can add a SD card to trend all engine metrics to a file (haven't developed this yet but easy enough to add). The current design shows time, date, RPM, IAT, ECT, MAP, TPS and Battery Voltage values as these are my requirements. A red warning screen pops up for warning messages (eg ECT too hot). It's currently sampling at 250ms intervals (4Hz) but could easily sample at 100ms (10Hz) or even faster. This makes a much cheaper and completely configurable colour LCD version of the Link display for a fraction of the price. The full retail cost of the prototype is around $280 *buying online). If anyone is interested in the project and/or has ideas to add please let me know. I'm happy to send any forum member the design details and software (when complete). Depending on interest I may be able to make a production version of the board to sell to memebrs at cost. I could even build to order with a LCD and layout of your preference - again at cost. Alternatively, since I have a development platform here, if someone wants to buy all their own bits I can help them get it going. I can send them the files for the LCD and the Arduino .ino file...

Thanks. From what I can see as a difference between our setups is: M3 Cams No MAF Cold Air Intake M3 EVO Headers Link G4+ ECU with a Dyno Tune Not sure what exhaust you're running - I'm using a M3 Exhaust with CAT delete + AdrenalineR 2 in + 2 out resonator. To add M3 cams you'd have to add vanos or make a modified vanos to non-vanos adapter for the inlet cam. Everyhting else is pretty easy - although the M3 EVO Headers may not fit the E30... It's all possible - you can do it too!! (just need a nice fat wallet).

Yep, exactly. These are the M52 manifolds of my M50B30 conversion - were on the engine when I purchased it. I dumped these for M3 Euro headers. They probably won't sell - unless someone with a M50 wants to change from the cast manifold. Can't comment on the M52 vs M54 sorry. Also, I didn't try the M52 vs M3 Euro headers to see how much difference there was either.

Today I replaced the trans fluid. It had some standard ATF fluid from the shop in it (it was red) and it was notchy and not nice to change most of the time - you never knew when it was going to bite back. Today I changed whatever it was in there to Amsoil MTF (5W30) full synthetic. So far it's excellent. Cold shifting is somewhat improved and while some mild notchiness is still there sometimes it's improved significantly. It's also a lot easier to get into 1st, 5th and reverse. The gearbox is much quieter too - no more mesh type sound. It's only done 100km on the new oil so far so the jury is still out - will see how it goes with 2000+k's on it. The Getrag S6S 420G appears to be very oil sensitive... I have a coupe of bottles of RP here too and may give it a try if the Amsoil ends up being a dog. Anyone tried Motul NISMO 2189E 75W140 (used by the GTR guys in the Nissan version of the 420g) - it's pretty expensive but supposed to be very good?

Here's links to full-sized images. It appears replacing the flexi on M3 Evo headers isn't too difficult - you just need someone who can weld in tight spaces and who can do tidy welds. All welds in the new exhaust are tig welds - no mig!!! http://i64.tinypic.com/15cf1np.jpg http://i67.tinypic.com/2ij0ya9.jpg http://i64.tinypic.com/2qjzsj6.jpg http://i64.tinypic.com/1zdut6g.jpg

Yes, I will be fixing the excess bolt length - need to replace them with stainless bolts but didn't know what length I'd actually need. Just bought some cheap set screws for now until I can measure the length required...

Today my nice new exhaust was fitted. There were a few very minor as-built changes during the fit but overall everything went pretty close to plan. The stainless tube was welded using the 'purge' method where the tubes are purged of air and filled with low-pressure argon. It stops the weld bubbling into the internal part of the tube to keep the flow undistrubed. The only untidy area was rear of the AdrenalineR resonator to get the pipes through the hanger area and mating up to the original rear exhaust section. Since putting the new engine in the car only a short time ago I've been chasing a very small exhaust leak - hadn't been able to find it. On Wednesday this week the leak suddenly got much worse which allowed me to finally find it. It was the flexi-joint on the rear headers. Information on this site led me to Autobend to buy a replacement. This was welded in during the fitting of the new exhaust. The results are excellent. The car now idles with a very nice sound and is very quiet in the cruise. Under power it has a very deep growl which is very nice. Under decel it has a deep warble a bit like a race car. Overall I'm very happy with the change. The exhaust put in by the local shop resonated and vibrated badly with a lot of odd harmonic frequencies - probably due to all the restrictions and expansions from the crush-bends in the pipes. All these odd sounds are now gone and there's a very regular sound across the whole rev range. During the change we also move the resonator back about 250mm from the position the local exhaust shop fitted it. The car now feels responsive again at low RPM. Another good find during the flexi-joint replacement was a big blob of weld where the old flexi was welding in. Whoever fitted it had to be the worlds worst welder - they'd blown a very big hole in the pipe and just poured in more and more weld to fix it. The result was an 8mm deep blob of weld 20mm long protruding into the 51mm diameter exit tube on the headers. This was removed with a rotary file to restore the pipe to it full internal diameter. New Exhaust: New Flexi: