MS3 w/ MS3X Install - L6 turbo, sequential injection & boost control

[Sam280Z]

Putting this here because its the first thing to read:

http://www.msextra.com/doc/index.html

 

 

I just got my MS3 with MS3X on version 3.57 board from diyautotune. I'm going to use this thread to keep track of my install and important information & links.

 

 

 

 

 

This is the engine it is going to control.

 

 

 

 

It is currently using a MS1 w/ HiRes code

Engine components (that matter):

63lb/hr Siemens Deka injectors,

TWM ITBs,

Holset turbo,

Crane HI6 CID ignition (similar to MSD 6AL),

Rebello worked P90 head w/ solid cam,

stock balanced rods/crank,

JE forged flat top pistons with valve reliefs (~8.5:1 compression),

280ZXT optically triggered distributor

 

These are BIG injectors and even running the HiRes MS code, I can only change fuel by ~10% at idle. 0.035ms is the smallest change supported by HiRes and my PW at idle is about 1.3xx ms. A 1ms injector opening time means that is 10% of my "injecting" pulse width (PW).

 

I'm getting terrible (<16mpg) mileage and will be looking to the MS3 to help with that. other than that, it is running very well.

 

More to come...

Edited September 20, 2011 by Sam280Z

[Sam280Z]

This weekend I swapped the MS1 out with the MS3 and after a lot of reading, finally figured it out enough to get it running.

 

First thing was moving what I could of the "tune" over. Limited instructions are here: http://www.msextra.c...s3/upgrade.html

 

I went table by table in TunerStudio and grabbed screen captures for information I could not export and import. I then attempted to put similar settings in MS3. This is tricky as many settings that use millisecond units in MS1 use % in MS3.

I swagged it based on my recollections of PW values. It is apparent that much of my tune will have to be re-done in MS3.

 

My next real challenge was that MS Extra used the FIdle output to trigger the ignition. Since MS3 can have it's own ignition driver, this connection to the Crane Hi6 unit was not going to work.

 

Remember, I'm using the optical 280ZXT distributor already set up for MS1 Extra per the instructions here: http://forums.hybrid...asquirt-wiring/

 

It took a while, but I found the instructions to trigger a CDI box here (It's almost at the bottom of the page in section "CDI spark box e.g. MSD") :

http://www.msextra.com/doc/ms3/ignition.html#ignout

 

As far as I can tell, the diyautotune MS3 w/ MS3X does not come with the ignition driver installed on the 3.57 board. I guess they expect you to use the MS3X tacho output.

 

One wire into the MS3X connector (pin 26) run to the tach input on the Crane Hi6 and it started right up.

 

I didn't run it long because the trigger information is handled different in MS3 that MS1. More to research here:

http://www.msextra.c...istributor.html

USE THESE SETTINGS AT YOUR OWN RISK

I DO NOT KNOW WHAT I AM DOING

THIS CAR IS NOT TUNED YET

Settings (all others grayed out):

Spark Mode: Trigger Return

Trigger Angle: 68.00 (this is unique to my car - It depends on where the dist is locked down.)

Use Cam Signal: Off **

Skip Pulses: 3 **

Ignition Input Capture: Rising Edge **

Spark Output: Going High (Inverted)

Number of coils: Single Coil

Spark Hardware in use: Tacho

Cam Input Used: JS10 **

Fixed Advance: Use Table (this is set to a specific value when timing the engine)

Cranking dwell: 6.0 **

Cranking Advance: 20.0

Dwell Type: Fixed Duty **

Dwell Duty: 25% ** read this value somewhere - need to find

Sprk Hardware Latency: 0 **

middle LED indicator: Off **

Overdwell Protection: Off **

Spark Trim: Off **

ALL SETTINGS REQUIRE A POWER CYCLE TO SET IN THE CONTROLLER.

** Do not fully understand this yet.

Edited September 20, 2011 by Sam280Z

[Sam280Z]

So that's where I am so far. Got it started. Went back to MS1 Extra for now because this is my daily driver.

Next step is to get the MS3 and MS3X wiring installed - and cleaner than I did with the MS1 install.

Good news is I used terminal plugs on all sensors, so I should be able to switch back and forth pretty quickly.

Major difference in this install and last (aside from many many more wires) is the relay board. On my MS1 I wired in all my own relays.

 

 

One thing that is needed is the trigger. The 280ZXT optical trigger is fine for batch injection. For sequential, the MS3 needs more timing information.

It appears that there are a couple of options shown here:

 

 

 

On the right is a CAD drawing of option 1. I have copied the design of a 24 hole + 1 cam pos hole trigger wheel from the diyautotune site. Theirs is the wrong diameter for the 280ZXT distributor. With the help of a fellow HybridZ member, I am going to see about getting one fabbed up.

 

.005" tolerance is between .01 and .02 deg accuracy.

Trigger wheel is .15mm or .006in thick.

Attached .txt doc is a dwg file. remove the .txt extension for use in CAD programs.

 

CAS_Trigger_Wheel.dwg.txt

 

On the left is option 2. The 300ZX optical trigger wheel I just found this tonight and don't know if it will work - its all in alpha stages of development.

 

Links:

Megasquirt-3 MS3 Ignition - Nissan 360 tooth CAS and Optispark: http://www.msextra.c.../nissancas.html

Megasquirt-3 MS3 GM Optispark: http://www.msextra.c...k.html#settings

MS3 Tach Input from Nissan Cas No Signal.: http://www.msextra.c...p?f=131&t=41816

Nissan RB25 CAS, how can I help?: http://www.msextra.c...p?f=125&t=40767

Optispark GM DiyAutotune: http://www.diyautotune.com/tech_articles/how_to_megasquirt_your_chevy_lt1_lt4_l99.htm

 

 

That's all for tonight...

Edited September 21, 2011 by Sam280Z

[Sam280Z]

Haven't done anything else yet (except reading). Just adding these links on wiring:

 

MS3 and MS3X: http://www.msextra.c...3/hardware.html

Relay Board: http://www.megamanua...assemble.htm#ew

 

280ZX Tach wiring: http://forums.hybrid...asquirt-wiring/

 

http://jbperf.com/#

Edited October 6, 2011 by Sam280Z

[Shemyazaz]

I'm researching this right now as well. What have you come up with for the low resolution signal? I'm tossing around several ideas, but so far making a new single slot wheel for the CAS is top of my list. Although, I have also considered using a crank trigger for the low and the 360* CAS for the high....but I'm a bit worried about the variance in the CAS.

 

Thoughts?

[Shemyazaz]

What are the dimensions for the l28 wheel?

 

Any other CAS/dizzy units fit in the L motors?

 

Matt at DIY says we might consider soldering 5 of the slots closed, leaving one for the cylinder ident, and using a crank trigger.

Edited October 20, 2011 by Shemyazaz

[aarc240]

On my MS3 install I've used a cut down Datsun four cylinder distributor body.

A steel plate mounted to the top supports a standard Ford CAS which detects a single steel post wheel attached to the original Datsun lower distributor shaft.

A VR interface from Jean Bellenger translates that signal for the MS3.

Unfortunately I didn't think to take photos while doing that one!

 

btw, for CAS I'm using a Ford sensor with a custom 36-1 wheel 180mm in diameter mounted to the back of the harmonic damper.

Edited October 28, 2011 by aarc240

[Sam280Z]

Got the MS3 installed and configured to run using the 280ZXT trigger wheel and the mainboard injector drivers. Next step is to install a different trigger wheel with both crank and cam position information so I can try out sequential injection.

 

Sheymyazaz, check out the links at the bottom of my post #3 in this thread. I hope to do all the triggering inside the dizzy since it has two optical triggers. Right now leaning towards using the one in the right hand side of the photo in post 3.

[Sam280Z]

First a little more info on the engine:

Cam Specs

Intake Cam Lift: .318 in

Exhaust Cam Lift: .315 in

Overlap 40deg

In opens 23deg BTC

Ex closes 17deg ATC

In closes 63deg ABC

Ex opens 51deg BBC

 

In: 266deg @ .015 in lift

Ex: 254deg @ .015 lift

110deg intake centerline

 

Managed to do some pretty cool things even without a triggerwheel that allows sequential .

 

Set the idle timing at 10deg BTDC while under no load at idle with a switch point to bump it to 20deg when the AC compressor loads the engine and lowers the MAP. This keeps the idle speed nearly constant at around 850RPM when the AC kicks on.

 

Set a more reasonable deadtime based on this: http://www.msextra.c...njdeadtime.html

I happen to be running the same DEKA injectors so I lucked out. My tune is much more stable with changing air temps now.

 

Corrected a wiring error. I had wired the 12V for the optical trigger to the Fuel Pump relay. The car would start fine if you started cranking as soon as you turned it on. If you waited it would not start. This is because MS powers the FP for 2 seconds and then turns it off until the car starts. Bumping the cranking speed up to 300 RPM from 100 RPM helped with starting a lot as well.

Edited November 5, 2011 by Sam280Z

[Sam280Z]

ITB w/ Boost

http://www.msextra.c...p?f=131&t=40879

 

http://www.msextra.com/forums/viewtopic.php?f=131&t=42501

 

Dual Table ignition:

http://www.msextra.com/forums/viewtopic.php?f=131&t=42373

Edited December 1, 2011 by Sam280Z

[mildsquare]

Flippin sweet man! What intake manifold are you using? I have a set of cartech preped mikuni's and the hks type 1 surge tank. I have my heart set on the blowthrough but after going for a ride in my buddies l28et 260 efi just keeps making more sense lol.

[Sam280Z]

Thanks.

It is a TWM manifold. From what I understand it is a Cannon copy.

 

[mildsquare]

I was thinking of doing that setup with my mikuni manifold and either the twm or o.e.r. itb set. Ive been doing so much research its makin me nuts. Im glad i came across this.

[Sam280Z]

Got the DIY trigger wheel. Install Instructions

Took just a few minutes to get it installed. Took a little while longer to get it started since I didn't put the center dist wire back.

It is critical to turn off and back on MS3 between changing trigger settings. When it did start, it was advanced 50 deg. After a couple of tries I settled in on a tooth angle of 355.4 - not too far from the initial setting of 345.

 

Idle timing is rock solid now. It used to bounce a lot. Setting the trigger angle was much easier and could be done at low engine speed.

Engine smoothed out noticeably at low speed, especially under load.

 

You can compare before and after graphs here:

Old:

New:

Edited January 12, 2012 by Sam280Z

[Matt Cramer]

Sam - thanks for confirming that angle works on the L28 as well! You're the first one to report back on the trigger angle used for this motor; I had based the article on accounts from KA owners up until then.

[Sam280Z]

While I was modding the 3.57 and MS3X boards according to the DIY instructions, I went ahead and switched out the 2.5 bar MAP sensor with the MAP Daddy 4 bar sensor that was in my MS1. This will allow real time baro correction.

When I calibrated the sensor I noted that the old settings did not match the 2.5 bar settings. I'm not sure if I had ever calibrated the sensor after I got the MS3. Settings for all three calibrations (4 bar, 2.5 bar, and custom - as delivered) are shown in the attached image.

 

The "custom" settings were not too far off of the 2.5 bar settings and resulted in the MAP readings of about 5 kpa more than what I'm seeing now.

Edited January 13, 2012 by Sam280Z

[Sam280Z]

Got it going full sequential today.

First impression: MUCH smaller pulsewidths needed for the same AFR. I should see a significant improvement in MPG.

Edited to Add: Had wrong (too long) deadtime settings for MS3X drivers.

 

 

http://www.msextra.com/doc/ms3/sequentialfuel.html

Edited January 17, 2012 by Sam280Z

[Madovic]

Are you still running the stock injectors or did you switch to some high impedance injectors? I didn't think the MS3 could drive 6 low impedance injectors independently (sequential rather than batch fire).

 

I'm just starting a conversion on my '75 (going from stock ecu to MS3 + MS3X), so I'm very interested in your progress and setup.

 

Tim

[Sam280Z]

Tim, I am using Siemens Deka 63lb/hr high impedance injectors. This motor is pretty far from stock.

 

I spoke too soon on the small pulse widths. It's not in the manual, but the PW1 calculation does not directly tie to the PW for each cylinder. My PW's are smaller, but not as drastically as I thought (10%-20% instead of 50%-60%).

Spoke too soon again. The difference I was seeing was because I neglected to put the correct deadtime value against the MS3X injector drivers. I had only updated the mainboard bank drivers. The MS3X drivers were using a value of 1ms instead of .495ms which is correct. Now PW1 and the seq PW's are nearly spot on.

Edited January 16, 2012 by Sam280Z

[Sam280Z]

I've been dealing with an intermittent miss for a while now. It happened mostly when hot and was mostly visible on the AFR gauge. It would indicate an increasingly widening range of AFRs with massive swings to lean. It was barely discernible in how the engine ran.

While poking around, I discovered that wiggling some of the spark plug wires (Magnecore) would make it better or worse and wiggling the #5 wire could kill the engine.

I figured I needed new wires and consulted the Crane HI-6 manual which called for low resistance wires (which the Magnecores were not). The recommended Crane Fire Wires were back-ordered until March. I called Crane tech support and was told that any high quality wound wires would work.

Bought a set of universal MSD wires and installed them on Monday. This completely cured the intermittent miss but now there was a seemingly random sync loss especially on first start up which became progressively worse over the week (from one or two to over 40 sync losses in a 20 minute drive).

Turns out, when changing the cap, rotor and plug wires, I must have inadvertently knocked the ground wire to the CAS loose. I'm surprised the car ran at all without the CAS ungrounded, but it did.

 

Now to replace the rattling Dynomax VT muffler. Put it on a year ago and thought the rattling was from hitting the differential. Fixed this and it turns out the early VT mufflers had problems. It is covered under warranty. I just have to swap it out.

 

Spoke too soon again. Today while driving was getting regular loss of sync to the point it was annoying. I'm pretty sure it is the MSD wires. A scan of my datalogs shows no loss of sync before they were installed on Monday. After messing with the wiring and double checking everything, I enabled interrupt masking and period rejection in the noise filter dialog (.1ms and 15%). Have not has a sync loss since then.

Edited January 29, 2012 by Sam280Z