MicroSquirt, BMW's, and Turbocharging

[Arktasian]

Crossover.pdf[/attachment]Actually, found this pretty quick on a web search - "Crossover"

Crossover.pdf

(31.79 KiB) Downloaded 471 times

Again, no expert by any means on most of this, but I have played around some taking gauge measurements on my boxer project. Seemed to log about equal numbers right up to roughly 15psid boost gauge reading.

[Charlie @ Evergreen]

You should find that with a diesel ideally a 1:1 boost vs back pressure ratio is desired. But most are not even close. If you want better boost response you will have more in the 2.5 :1 ratio or more. Gas applications the rule of thumb is closer to 3:1 so in other words if you have 20 lbs of boost you would have 60 lbs of back pressure. You can get a healthy 1:1 ratio using a much larger turbine section vs compressor but your performance would suffer. Most of the reason why the manufacturers are matching turbo's with turbines smaller than compressors now days. Its also how I build hot rodded billet turbos. Rule of thumb is keep the turbine /compressor within 25% of flow of each other.

[Charlie @ Evergreen]

If your going to the salt flats..then you can run wide open this is where you will find much larger turbine sections that only need to run WOT. Less heat and back pressure increasing the HP in upper RPM range.

[georgew4714]

Im lovin this thread

[Arktasian]

Im lovin this thread

George, I'm inclined to say the same thing about this newly discovered (to me) site

[Arktasian]

Variable geometry turbines seem all the rage on Diesel which I have more dealings with in my daily work, much as I would rather was smaller stuff. It does seem to bridge the transition between camps on the "crossover" phenomenon but wonder if they might have a place on our smaller platforms and limited real estate?

[Arktasian]

Here's an interesting link - axial flow turbine development
http://turbo.honeywell.com/whats-new-in ... ost-turbo/
Appears to be somewhat stale.

[Arktasian]

Megasquirt sounds kind of Mickey Mouse - don't let that fool you into thinking its lower affordable price comes at the expense of results - below a link to some Salt Flat activity with good movies and feedback on what was happening out there even back as far as 2012.
(it is a long running project and some of the content doesn't come up any more)
http://www.msextra.com/forums/viewtopic ... 04&t=46426

[Charlie @ Evergreen]

No Mickey Mouse..any system is only as good as the time put into setting it up and tuning. One of our customers used it on his Reno Sport Racer with twin turbos/intercooling/water meth injection etc... Think it was 2013 he won his class almost averaging 400mph.

[dazbo]

ark,
just an update on the ignition control on the 750turbo ecu/dfi system as im not sure you got your answer.
there is none.
that is to say ,the dfi is pretty much a standalone add on ,much like the original ms and controls only fuel.hence it has an aditional wiring loom for the injection system just like the early ms (but earlier)
the ignition system is wasted spark ,2 hall(i think) sensors( EDIT: sorry for any confusion, the sensors are actually VR (variable reluctance)) and one crankwheel with a single (large) tooth,which drives an ignitor unit with a preprogrammed advance curve,much like the standard gpz750 but with a different advance curve, i think 30 degrees total advance as opposed to 35 degrees for the standard gpz unit.
the dfi takes its rpm reference from the coil negatives as per the original ms before all the extra(pun intended) development of ignition control.
i succesfully transplanted a complete 750turbo injection system onto a 1986 gsrx1100 turbo ,which worked well until the crank snapped

[Sandman]

Would this mean that we would be able to take one of the hall sensors and use that as an input to the MS system for the crank position signal and then we'd be able to tweak the timing within the MS software from there? and with this we would be able to get rid of the igniter unit?

[Arktasian]

I hadn't noted the above posts and comments, as have been somewhat off line. (well, cleaning things up at work so that our family can do a weeks vacation without significant bother of office phone calls )
I was planning on following through on a bit of research as to the most advisable ignition triggering components and settings on the MS platforms when controlling a Kawasaki - thing is you can do stuff in many ways and some work " sorta okay", while others allow significantly superior results. The stock tone wheel does show in the manual image as a single tooth, and the dual sensors then work with the simplified DFI module (as described by dazbo) to provide wasted spark ignition with limited advance control.
Although it may prove possible to connect up the stock components directly to the MS ecu and get her running, typically it is better to utilize multiple teeth on the tone wheel for a cleaner rpm signal and sync.
I suspect I will be able to do that math and study to determine that a small tone wheel with perhaps 24 teeth total, one or two missing would work.
So, I'll just say that "Yes", with the addition of a cheap little replacement tone wheel on the crank, the stock sensors would allow MS full ignition and fuel control, and to do away with the original satellite modules and ignition bits etc.

[Arktasian]

Lucky me, on vacation with the family and have lots of time to sneak onto the forums here.
Below I am going to paste clip an excerpt from the DIYA site having to do with adding a controller "piggyback" fashion onto an existing system. For reasons of "maximum potential" I recommend going right to full fuel and ignition control if someone wishes to MS their Kawasaki - because there is so much to be gained in the timing department - but I also recognize there is a wish here to know "What is possible on my Kawi, to take a measure of control without getting to far into tuning and electronics etc." I note many posts where anxious owners wish to perform some upgrades but don't want to melt their pistons. I realize there are ways listed here to go about doing that pretty nicely, but not in a full control manner. There is nothing quite like taking full ownership of the program, knowing exactly your AFR states and deciding when and where things are going to happen. Not to offer offense here, but it could be compared to transitioning from black'n white to color.
Below the exerpt:
Installing MegaSquirt in Parallel With the Stock ECU



While there are many cars where MegaSquirt can completely replace the stock ECU, on some cars you’ll need to leave the stock ECU in place. There’s several reasons why you might want to do this. On some installations, you might want to have the MegaSquirt control just the fuel and would rather not deal with having to tune the ignition. Or you may have an ECU that controls the automatic transmission, has the voltage regulator for the alternator inside it, controls a drive by wire throttle, or has other features that aren’t easy to control with MegaSquirt. On a car like that, the solution is what’s called a parallel installation.

The whole idea behind a parallel installation is letting the MegaSquirt take control of what you’d like it to control (the fuel injectors and ignition system most commonly) without interfering with other systems that the factory ECU may control that may or may not really be engine management related. Sometimes this approach is called a piggyback, but that can cause a bit of confusion with systems like the Apex’i SAFC, which modifies signals to or from the stock computer. The MegaSquirt does not change any signals; instead, it will take over functions of the stock ECU outright.

Installing the MegaSquirt in parallel amounts to disconnecting the stock ECU from what you need to tune, and connecting these things to the MegaSquirt instead. You can do this with injectors, coils, IAC valves, and most anything the MegaSquirt can control. Often the toughest issue is what to do about sensors. These guidelines will help you deal with the sensors the MegaSquirt needs.


Sharing Sensors (or not)

MAP sensor: The MegaSquirt EMS has it’s own dedicated MAP sensor built in, so you will just need to add an extra vacuum line for it.

Throttle position sensor: Check the TPS signal running to the stock ECU with a volt meter. If it never goes above 5 volts, and if its voltage increases as the throttle opens, you can use this sensor with MegaSquirt. Just splice the Megasquirt TPS signal wire in to the signal wire to the stock ECU. You should not try connecting the Megasquirt’s VREF wire to the TPS, as tying two VREF signals together can cause problems.

Temperature sensors: These are a bit problematic. On some ECUs, like the early Miatas, it is possible to get the MegaSquirt and the stock ECU to both use the same sensor. To do this, you will need to remove the R4 and R7 bias resistors from the MegaSquirt and determine the bias resistors the stock ECU uses, then calibrate the MegaSquirt appropriately. This link provides more detailed information about how to share temperature sensors between the MegaSquirt and stock ECU. On some ECUs, like most GM computers, you cannot share temperature sensors with the stock ECU at all. IF your unsure of what to do here, your safest option is to provide the MegaSquirt with its own CLT and IAT sensors, separate from the stock CLT/IAT sensors that the factory ECU will continue to use.

Crankshaft position sensor / camshaft position sensor: These are sensors you can usually share with the stock ECU, but there are a few issues that can cause the MegasSquirt to not play nice with them. If you have a VR sensor and you connect the sensor to the MegaSquirt backwards, you could possibly ground the sensor and wipe out its signal. Solution– don’t wire it backwards! With Hall effect or optical sensors, sometimes the combination of pull up resistors in the MegaSquirt and stock ECU can interfere with each other. If this happens, one or both of the ECUs will not get an RPM signal. Often, you can solve this by removing the pull up resistor from the MegaSquirt.

Oxygen sensor: It’s perfectly OK to share narrow band oxygen sensors with the stock ECU. However, there’s a good chance you will want to add a wideband O2 sensor at some point, in order to give the MegaSquirt more tuning information. These often don’t work with stock ECUs, so your best bet is to weld in a second bung for the wideband sensor. Another case where it pays to have separate sensors for the MegaSquirt and the stock ECU.

So what have you really done here?

Basically, your goal in the above is to make sure that both the stock ECU and the MegaSquirt EMS have the INPUTS they need from all the necessary sensors so they can both independently make all of the calculations needed to control your engine. Both systems ‘think’ they are in control of the engine, as they are processing all of the incoming information from these sensor inputs, and they are sending the corresponding OUTPUT signals in the way of fuel injector pulsewidths and ignition pulses. So which one gets control of your engine now? Well… that’s completely up to you now: which set of outputs did you actually wire up to your fuel injectors and ignition system? That’s the system that’s controlling your engine; the other system just continues thinking it’s controlling the show when the MegaSquirt ECU is really pulling all the strings.

What if you get a Check Engine Light?

In many cases, the stock computer will continue to be dumb enough to never notice you cut it’s injector and ignition outputs. It doesn’t like its sensor inputs cut, but it’s outputs it’s not so picky about. In some cases, the stock computer might be smart enough to figure out that you’ve done something tricky here and it might trigger a check engine light. In some cases it may not matter as it often wouldn’t affect performance and you should only be implementing a system like this on a off road/race use vehicle anyways, where the CEL was of no other concern. But there are situations where this could affect performance, so you’d need to resolve the situation. When a stock ECU activates a CEL, it often enters a ‘limp-home mode’. If the stock ECU had control of your fuel and ignition, this would generally trigger conservative fuel and ignition maps to be used which would reduce performance. That’s very likely not the issue here though, as you’ve got your MegaSquirt EMS in parallel controlling the fuel and spark. But, that stock ECU may still have control of a variable intake system, or variable valve timing system, or throttle by wire system, and these could therefore also be affected by the stock ECU entering a limp-home mode restricting performance. So this CEL can still be a problem for you. So you basically need to determine the cause to the CEL and ‘trick’ the stock ECU into thinking everything is OK. The right way to do this is to read the codes the stock ECU is storing to see what it’s complaining about and then resolving the issue(s). Is it complaining that the injectors are disconnected? You can connect dead injectors to it’s outputs to trick it– or better yet, high wattage (50w) resistors of a similar resistance to your stock injectors.

[Arktasian]

So, there are several ways to go on this. It is apparent that the stock Hall sensors will jive with the MS to do "Fuel Only" control - direct plug'n play. Of course you need a wide band O2 sensor to make heads and tails of AFR's, but with that, and a VE table for fueling "auto-generated", and then some VEAL or self tuning while you ride you could get quite a bit for very little technical input and cost.
I'll paste the general blurb for that from the MS manual below.

Naturally, I recommend changing out the single toothed stock tone wheel with one having 12 or 24 teeth, I probably won't do anything on that unless I come across a suitable engine to test.

[Arktasian]

One of the squirters over on MS had a comment on engines of similar design to the Kawasaki- a Suzuki 750, that I am not at all sure applies but will mention for the fun of it.
I assume due to the Kawi having all its injectors wired together, and firing together batch style that there is not individual or "location sensitive" behaviour by the stock ecu attempting to balance out individual cylinder needs? In other words, at any given point on the stock fuel map there would be one injection duration or MS of fuel that operates each of the 4 injectors identically?
That's my take and confirms a rather easy approach towards alternate control etc.
Thoughts and comments?