MyE28.com

you'd have to put a 60-2 wheel and sensor from an e34/e32 m30 engine to get megasquirt with spark working on an e28.

MS isn't fast enough to read and count the flywheel teeth.

I'm resurrecting this thread since I'm going to start a Megasquirt project on my 528e in the near future.

altus22 wrote:MS isn't fast enough to read and count the flywheel teeth.

Doesn't the hardware that Ely's running have a faster processor than the original MS, and wouldn't that be fast enough?

shifty wrote:I'm resurrecting this thread since I'm going to start a Megasquirt project on my 528e in the near future.

altus22 wrote:MS isn't fast enough to read and count the flywheel teeth.

Doesn't the hardware that Ely's running have a faster processor than the original MS, and wouldn't that be fast enough?

I don't think it's that MegaSquirt isn't "fast enough" but that installing the 60-2 wheel and sensor is the easier way to make it work without modifications.

Brad,

I'm just curious, what plans do you have for your 528e? Unless you have more in mind than throwing MS at it, I wouldn't bother. It'll take a fair amount of work to get MS to be as good as the factory stuff. You can probably get a little bit better performance out of it after a lot of tuning, but it seems like a lot of hassle unless you've got "plans."

JM2CW

Skeen wrote:Brad,

I'm just curious, what plans do you have for your 528e? Unless you have more in mind than throwing MS at it, I wouldn't bother. It'll take a fair amount of work to get MS to be as good as the factory stuff. You can probably get a little bit better performance out of it after a lot of tuning, but it seems like a lot of hassle unless you've got "plans."

JM2CW

Skeen,

Remember everyone talking to Duke about how he shouldn't have started learning to tune with his Tec3 and his turbo M535i? I'm taking those words of advice to heart and starting on my POS 528e that I've owned for almost 4 years. I could do it on the 535iS, but I still need to do a lot of work on it (5-speed conversion, etc) before it's ready for yet another project.

I'm doing this mostly to learn more about my car and about tuning Megasquirt, so I'm going to stick to as many stock components as possible. This will make the install more simple and will allow me to upgrade things as I go, if I want.

Long term, I'd like to do a 2.8 rebuild with i top end components and a TCD M20 kit, but, again, that's long term. For the next, say 3 years, it's going to be mostly stock on the 528e.

shifty wrote: Remember everyone talking to Duke about how he shouldn't have started learning to tune with his Tec3 and his turbo M535i? I'm taking those words of advice to heart and starting on my POS 528e that I've owned for almost 4 years. I could do it on the 535iS, but I still need to do a lot of work on it (5-speed conversion, etc) before it's ready for yet another project.

I'm doing this mostly to learn more about my car and about tuning Megasquirt, so I'm going to stick to as many stock components as possible. This will make the install more simple and will allow me to upgrade things as I go, if I want.

Long term, I'd like to do a 2.8 rebuild with i top end components and a TCD M20 kit, but, again, that's long term. For the next, say 3 years, it's going to be mostly stock on the 528e.

Good plan

M635CSi wrote:I don't think it's that MegaSquirt isn't "fast enough" but that installing the 60-2 wheel and sensor is the easier way to make it work without modifications.

I don't remember how many teeth are on the stock flywheel, but I know it's >100. So if you're spinning 6500rpm, MS has to read 650,000+ pulses/sec. Just to count, it has to do at least one operation per tooth that it sees. That is 650kHz right there. Now you have to add in all of the floating point operations it has to do for fuel calculation, let alone spark. It adds up really quickly. The MS1 processor is only 4 MHz. MS2 might be able to, but that assumes that your VR sensor circuitry and signal wires are fast enough. I really would like to see how Bosch did it. I'm guessing they used a flip flop or 2 bit counter. If someone does get MS to run while counting all those teeth, I will bow down.

I'm just using ms1 as at the time i started doing megasquirt ms2 hadn't been developed enough to do wheel decoding. My engine is basically an eta, eta crank, rods, pistons, block, and head, but with an aftermarket I cam, dual springs, port work, I intake, headers etc... I first tried a maf swap on the car with stock 325i electronics and that was great, but I felt there was still a little more power to get. So I went with the megasquirt. To get it to do spark it has to have someway to know where the engine is in it's rotation. So I installed the 325i toothed 60-2 main harmonic balancer, but the pulley size is different then an eta, so you have to also install a 325i water pump pulley to take up the belt slack(both cars use the same size belt). Then of course the 325i crank (VR) sensor and bracket, which there are already bolt holes for on the block.

On a stock eta what you could do is just unplug the speed or reference sensor and plug the 325i VR sensor into one of those plugs, then splice into the wires that run to the plug at the ecu. Or run all new wiring, but the 3 wires need to be shielded to isolate noise from interfering with the signal to the MS brain.

I somewhat agree with not doing MS to a stock engine, but it easier to setup to run a stock engine then modify the engine with boost etc... then to set it all up from the get go and have it all running correctly.

The only points MS is a real comprimise on would be things like idle control and closed loop operation. I'm already running my motronic on open loop, so thats no big deal, but I do appreciate what the ICV does for driveability and idle quality. On a stock engine, theres nothing to gain and some to lose wih MS. On a turbocharged m20, I can definitely see the advantage of being able to map timing and fuel however you like, for any size injector. I salivate at the thought of it...

rs4pro3 wrote:On a stock eta what you could do is just unplug the speed or reference sensor and plug the 325i VR sensor into one of those plugs, then splice into the wires that run to the plug at the ecu. Or run all new wiring, but the 3 wires need to be shielded to isolate noise from interfering with the signal to the MS brain.

Sounds easy enough. This would keep you from having to install too many extra things.

rs4pro3 wrote:I somewhat agree with not doing MS to a stock engine, but it easier to setup to run a stock engine then modify the engine with boost etc... then to set it all up from the get go and have it all running correctly.

And, it makes it easier to learn MS than to have to learn MS AND learn a turbo AND whatever else you've added at the same time.

turbodan wrote:The only points MS is a real compromise on would be things like idle control and closed loop operation.

It shouldn't be as much of a compromise if you're running a WB O2 sensor, right?


I think I'm still going to stick to my plan of installing this on my stock eta. I'm just going to find a cheap, used wiring harness to butcher up and install the correct connector for the MS box and use the stock BMW sensors like Ely is doing. I can go with the 325i 60-2 wheel, water pump pulley, and VR sensor; install a WB O2 sensor; and I should be good to go running on mostly stock pieces. Am I missing anything, here?

That would make the install the least complicated and the best platform on which I can learn. I'm going to try to do some dyno runs with the JC chip as a baseline, first. We'll see how long it takes me to meet the baseline figures. My guess is 18 months.

there are boards out there to regain the factory style Idle control.

Yeah, makes sense. If you have plans to turbo down the road it will certainly help to learn how to tune first.

You will definitely have a hard time getting idle and driveability up to factory levels, but there is room for improvement in WOT. You should certainly work on ICV operation or cold starts will be terrible (ask me how I know).

turbodan wrote:The only points MS is a real comprimise on would be things like idle control and closed loop operation.

I have both on my MS units supplied by Peter Florance (SEM).

T_C_D wrote:

turbodan wrote:The only points MS is a real comprimise on would be things like idle control and closed loop operation.

I have both on my MS units supplied by Peter Florance (SEM).

And knock sensing

shifty wrote:I'm resurrecting this thread since I'm going to start a Megasquirt project on my 528e in the near future.

altus22 wrote:MS isn't fast enough to read and count the flywheel teeth.

Doesn't the hardware that Ely's running have a faster processor than the original MS, and wouldn't that be fast enough?

Shifty, I cant stand u for winning that megasquirt unit! On the other hand it couldnt have gone to a more deserving individual!!!

Nebraska_e28 wrote:Shifty, I cant stand u for winning that megasquirt unit! On the other hand it couldnt have gone to a more deserving individual!!!

I'm stoked I won it, too. It's something that I was planning on doing a while down the road, but this just helps to push the project up by a couple of years.

Oh, and it COULD have gone to plenty of more deserving individuals! There are so many people out there that contributed to this year's 5er Fest in bigger ways than I did. I'm sure there are some people that bought more than 10 tickets for this box. I consider myself lucky to win it having invested just $15 in my three measly tickets.