shauer's MS3X Beta Testing

[shauer]

My MS3X just arrived today. A note to the other beta testers, the 37 pin male connector needed to hook up to the MS3X is not included. I ordered one from digi-key a few days ago and it is also supposed to arrive today. My plan is to have my board installed and running in my car by the end of the weekend.

The features I will be using, and the order I will be activating them are:
- Cam position sensor, I have a hall effect sensor installed on my engine
- Sequential fuel
- Sequential spark, I have COP with Glen's 4-coil driver board
- 3-wire PWM control

My engine is a BMW M10 engine, Inline 4 cylinder, SOHC 2-valve per cylinder. The engine has been stroked from 2.0L to 2.2L and is running a 306 degree cam and 40mm ITBs. Other modifications include the COP setup, idle air control, custom cam position sensor.

Ok, now for some basic wiring questions. I have not used sequential on MS2extra and the documentation is not complete so I want to make sure I understand this correctly.

Injector Questions:
My firing order is 1-3-4-2

It looks like I can wire up my injectors as follows:
- INJ A -> Cylinder #1
- INJ B -> Cylinder #2
- INJ C -> Cylinder #3
- INJ D -> Cylinder #4

Then it looks like I set up the firing order in the Engine and Sequential Settings as follows:
A -> 1
B -> 3
C -> 4
D -> 2

Is this correct?

What effect do the traditional fuel setting have in sequential? I'm thinking about the following:
Squirts Per Engine Cycle?
Injector Staging?

Do these 2 parameters need to be set to any specific values for sequential?

Ignition Questions:
It looks like the ignition order is hard-coded to run the ignition outputs in order, so I need to wire my coils as follows:
- SPK A -> Cylinder #1
- SPK B -> Cylinder #3
- SPK C -> Cylinder #4
- SPK D -> Cylinder #2

Is this correct?

That's everything for now. I will post updates and questions as I have them.

[jsmcortina]

The injectors need to be wired up in the same order as the firing order - just like the ignition outputs.

The list of "firing order" is so the trim tables for fuel and spark can be aligned with the correct outputs.

James

[shauer]

Thanks for setting me straight on the injector wiring, how about the "squirts per engine cycle" and "injector staging" settings? Anything special about them with sequential?

I just upgraded from 0.22 to 0.24 and there is really something going on with the communications on 0.24. I have never had a single problem with my USB connection on any previous version but I cannot get it to work on 0.24. TS keeps telling me it received invalid data and goes offline. 0.24 works fine so far with my USB to RS-232 adapter but will not work with the plain USB cable. All of these problems are with the engine not running. It took 4 tries to download my msq with the USB cable. It would give me the diff report, I would select the TS version, confirm, and then TS would immediately go offline.

I just pulled my MS3 from my car to install the MS3X board and discovered that my MS3 board was missing the 10 pin header (J1), James warned me this might be the case. I have some spare header pins and am in the process of soldering them onto the MS3 board. I will post pics shortly of everything before I power it up.

[shauer]

Ok, I finally got the missing header taken care of and took several pictures of the MS3X installed. I'll wait until I hear back before powering this up. I have double checked everything and it all looks good. I did notice that it is a little bit easy to mess up the installation of the wider ribbon cable because of the tension in the cable from folding it. You do need to pay attention to make sure you don't get off by a pin on the header connector.

Picture of my MS3 missing J1:


After soldering in the missing header:




The MS3X temporarily installed without the upper case:




Close-ups of the cables:

[muythaibxr]

Thanks for setting me straight on the injector wiring, how about the "squirts per engine cycle" and "injector staging" settings? Anything special about them with sequential?

I just upgraded from 0.22 to 0.24 and there is really something going on with the communications on 0.24. I have never had a single problem with my USB connection on any previous version but I cannot get it to work on 0.24. TS keeps telling me it received invalid data and goes offline. 0.24 works fine so far with my USB to RS-232 adapter but will not work with the plain USB cable. All of these problems are with the engine not running. It took 4 tries to download my msq with the USB cable. It would give me the diff report, I would select the TS version, confirm, and then TS would immediately go offline.

I just pulled my MS3 from my car to install the MS3X board and discovered that my MS3 board was missing the 10 pin header (J1), James warned me this might be the case. I have some spare header pins and am in the process of soldering them onto the MS3 board. I will post pics shortly of everything before I power it up.

James made a lot of changes to the code to catch serial errors (reads while trying to write, writes while trying to read, noise issues, etc...)

If it won't work it's because those sorts of issues are occurring.

Ken

[shauer]

James made a lot of changes to the code to catch serial errors (reads while trying to write, writes while trying to read, noise issues, etc...)

If it won't work it's because those sorts of issues are occurring.

Ken

I understand that part. The interesting thing from my end is that I was only attempting to connect with TS prior to starting the car. I was not doing anything at all except attempting to connect. No cranking, no running engine, no VEAL, no datalogging, no nothing and it refused to connect with the USB setup that had been running without any apparent issues for the past 3 months including multiple VEAL runs etc. I then reverted back to the RS-232 and that worked. I guess I have noise issues with the USB connection that are being detected but had no apparent effect on functionality or reliability in previous versions.

Anyway, Ken you asked me to post pictures of my MS3X install before powering it up, does everything look ok to you on the ribbon cable setup?

Thanks

[muythaibxr]

The ribbon cable connections look fine to me.

Ken

[shauer]

So far today I have managed to test my previous MS3 functionality with the MS3X card installed and hooked up to the MS3. No MS3X functions were being used yet but my engine ran and I managed to keep all the smoke inside of the various boards.

I then worked on getting the new 37 pin connector wired up for cam position, injectors, coils, and idle control. I have the 37 pin connector completed but still need to run the new ground wire to the engine and hook up to my engine's coil harness. Both injectors and coils are wired up in firing order to the MS3X.

I'm shooting for having the wiring complete and starting tests on my engine tomorrow. I don't have a stim so all my work is on the engine so I'm going slow and careful here.

[shauer]

Ok, I have my cam trigger wired up and I am attempting to see if it is working correctly.

What are the correct ignition setup settings for a 36-1 crank wheel and a single tooth cam trigger???

I have the MS3X set up for hall effect sensor, my sensor has an internal pull up so I do not have the jumper installed on the MS3X.

Here are the settings I have changed from my previous setup:
"use cam if available" set to "on"
"cam input" set to "ms3x"

trigger wheel arrangement is still set to "single wheel with missing tooth" do I need to change this?

What is the best way to prove the cam trigger is working? I am syncing up on cranking and seeing a reasonable cranking RPM, is that sufficient?

[muythaibxr]

You'll need to use dual wheel with missing tooth.

It shouldn't run in that configuration or even sync unless the cam signal is working.

With a hall sensor you might have to adjust the zero-crossing pot to get it to see the signal though.

Ken

[shauer]

You'll need to use dual wheel with missing tooth.

It shouldn't run in that configuration or even sync unless the cam signal is working.

With a hall sensor you might have to adjust the zero-crossing pot to get it to see the signal though.

Ken

Thanks, I changed to dual wheel with missing tooth.

The engine cranks and syncs with no lost sync during cranking.

I then disconnected the MS3X connector and MS will not sync at all. I guess this means my cam sensor is working. Next step is to get my coils connected and see if I can fire this thing up.

[jsmcortina]

Next step is to get my coils connected and see if I can fire this thing up.

Do check the MS3 manuals for settings and use the test mode.

James

[shauer]

Next step is to get my coils connected and see if I can fire this thing up.

Do check the MS3 manuals for settings and use the test mode.

James

Here is a question (yes another one)...

If I have been using the LED outputs in wasted spark configuration to drive my coils previously, the same "going high inverted" setting should still apply for MS3X, correct? I have not read any mention of the MS3X inverting any signals relative to the main board options, just double checking.

I will be sure to use the test mode on the coils. I don't have an injector test setup where I can do this easily with the injectors.

[shauer]

I was able to spend more time on this today than I planned.

After double checking everything, engine fired up first try!

Thanks James and Ken for answering most of my dumb questions, it helped a lot.

I have cam, fuel, and spark all running. Fuel timing is whatever the default was set to. I need to play with the injection timing tomorrow. My cam specs are:
duration - 306 degrees
intake - 45 degrees BTDC to 81 degrees ABDC
exhaust - 81 degrees BBDC to 45 degrees ATDC

It looks like I want to start with my injection timing set to finish the squirt at -45 degrees. Right???

I also still need to set up the 3-wire PWM idle valve. I am using the idle and boost medium current output pins on MS3X. It looks like I just set the output port to "idle" and the 3 wire mode to "boost".

With any luck I'll get out for some test drives tomorrow.

[jsmcortina]

Good work.

On my V8 I found that injection timing had a noticeable impact on engine idle quality.

I've tested 3 wire mode with a valve on the bench, should work as you say. With the Bosch valve I have, somewhere in the 50-100Hz range was about right. Too low and it is noisy, too high and it doesn't move.

James

[prof315]

I was able to spend more time on this today than I planned.

After double checking everything, engine fired up first try!

Thanks James and Ken for answering most of my dumb questions, it helped a lot.

I have cam, fuel, and spark all running. Fuel timing is whatever the default was set to. I need to play with the injection timing tomorrow. My cam specs are:
duration - 306 degrees
intake - 45 degrees BTDC to 81 degrees ABDC
exhaust - 81 degrees BBDC to 45 degrees ATDC

It looks like I want to start with my injection timing set to finish the squirt at -45 degrees. Right???

I also still need to set up the 3-wire PWM idle valve. I am using the idle and boost medium current output pins on MS3X. It looks like I just set the output port to "idle" and the 3 wire mode to "boost".

With any luck I'll get out for some test drives tomorrow.

Well done!

[muythaibxr]

As far as ignition outputs, if you were running on the negative lead of the LEDs before, then you'd have been using "normal" outputs and you'd have to switch to "inverted." If you were using IGBT's and directly driving the coils, then you were probably using "inverted" and you'll want to stay that way.

Good work getting it running!

Ken

[shauer]

Status update and 2 questions I could use some help on.

I upgraded to 0.25

3-wire idle air control works just as I described above. I'm using approximately 80hz. I had been using a fixed load resistor to ground on one side of the valve and driving it PWM on the other side. As a result, all of my PWM settings are now dramatically different and the valve appears much more responsive than before. I have had to cut my closed loop gains in about half to get a relatively stable idle. The closed loop idle still needs some tuning work though.

Injector timing question:
Ok, I have been reading up on the injector timing and my previous comments appear to be wrong. I could use some help reviewing my numbers here...

Question #1: TDC of the intake stroke (the "start" of the intake stroke) is at -360 degrees, correct?

Question #2: If I am correct above, then if my intake opens 45 degrees BTDC then I want to start with -315 degrees for my injector times (end of pulse), correct?

The reason I'm asking is that I do not notice a very dramatic effect in idle quality from changing the injector timing. This might be because my calculations are off, my idle control is still introducing too much variability, or my engine is less sensitive to injection timing.

Cam signal question:
I noticed on the wheel logger that the falling edge of my cam signal was dangerously close to the missing tooth on the crank wheel. The default value for the cam signal trigger was falling edge, I set this to rising edge just to be safe. The rising edge of the cam is several teeth prior to the missing tooth on the crank. Does this sound reasonable? Do the rising/falling edges in the logger match the rising/falling edge detection in the ignition settings?


I am going to remove all the old connections I used to use for ignition control on the main board, put the controller back together and try a test drive.

[muythaibxr]

Status update and 2 questions I could use some help on.

I upgraded to 0.25

3-wire idle air control works just as I described above. I'm using approximately 80hz. I had been using a fixed load resistor to ground on one side of the valve and driving it PWM on the other side. As a result, all of my PWM settings are now dramatically different and the valve appears much more responsive than before. I have had to cut my closed loop gains in about half to get a relatively stable idle. The closed loop idle still needs some tuning work though.

I noticed similar with mine, only now I can go to a high enough frequency that it doesn't buzz anymore, and it also responds better. I was able to make the open and closed positions much closer together which helped with the sensativity. Some of the change is because the flyback diode is different, so you should notice that your duty will move more towards the middle of the 0-100 range, and it'll close at a higher number and be fully opened at a lower number.

Setting the open and closed numbers to match here will make the code less sensitive. When I updated instead of changing the PID numbers I just changed those.

Question #1: TDC of the intake stroke (the "start" of the intake stroke) is at -360 degrees, correct?

Yes, but so does 360 (not negative). I recommend using that so you can get a smooth transition from before to after that point more easily. It really depends a lot on the engine though what number actually works best here.

Question #2: If I am correct above, then if my intake opens 45 degrees BTDC then I want to start with -315 degrees for my injector times (end of pulse), correct?

Yeah, or EDIT 405. This is just theory, but you'll probably want to make it higher than that to account for transport delay, and then as load increases and rpm increases, make it higher still. I've been thinking of making a transport-delay curve, but I think that might just be overcomplicating things a bit.

The reason I'm asking is that I do not notice a very dramatic effect in idle quality from changing the injector timing. This might be because my calculations are off, my idle control is still introducing too much variability, or my engine is less sensitive to injection timing.

I have noticed the same thing with my ITB-equipped engine at idle. However the timing seems to make a big difference in low-rev response especially, and general response all around. I have not messed around with it enough to figure out the best place for timing though.

Cam signal question:
I noticed on the wheel logger that the falling edge of my cam signal was dangerously close to the missing tooth on the crank wheel. The default value for the cam signal trigger was falling edge, I set this to rising edge just to be safe. The rising edge of the cam is several teeth prior to the missing tooth on the crank. Does this sound reasonable? Do the rising/falling edges in the logger match the rising/falling edge detection in the ignition settings?

Sounds fine to me. The missing+extra code just looks for the missing tooth after the extra tooth to determine "tooth #1" so as long as it is sufficiently far away from the missing tooth, you should be ok.

I am going to remove all the old connections I used to use for ignition control on the main board, put the controller back together and try a test drive.

Cool, let us know how it goes then.

Ken

[shauer]

I made a fair amount of progress during two test drives today. I figured out (I think) the injection timing and see some noticeable improvements. Thanks Ken, your explanation helped a lot.

TDC of the intake stroke is at 360 degrees. My cam opens the intake 45 degrees before TDC of the intake so that is 315 degrees. I picked 300 degrees as a starting value and noticed an immediate improvement in idle quality. So far I have dropped my idle speed from 1025 RPM down to 950 RPM. This was not possible before and is a very low idle speed for the high duration cam I am running. I think I can drop the idle another 25 or 50 RPMs further. That would put me at the OEM idle spec with an additional 42 degrees of cam duration. I'll do some more fine tuning of the injection timing at idle once I get the idle speed adjusted.

You actually went the wrong way. You wanted 360+45 not -45. Higher numbers BTDC mean earlier. That said, use whatever runs best.

In general the entire operating region of the engine where intake air velocity is relatively low appears to show improvement in quality and some power increase, especially in areas where the engine would bog down before.

That's similar to what I noticed mainly in response, but that was without having actually tried that hard to figure out what works best. I didn't spend much time on it.

Ken