MS3 new documentation discussion

Sam280Z · Post by **Sam280Z** » Tue Feb 03, 2015 8:29 am

OK. It wasn’t clear to me that the %age staged was %age of fuel flow. See if this is correct:

The ratio of fuel flow from primary and secondary injectors is determined by the Percent Staged value and how Transition Fully To Secondaries is set.

1)      Transition Fully To Secondaries ON:
a.       0% staged: 0% of fuel flow comes from the secondaries; 100% of fuel flow comes from primaries.
b.      100% staged: 100% of fuel flow comes from secondaries; 0% of fuel flow comes from primaries.
c.       50% staged: 50% of fuel flow comes from secondaries; 50% of fuel flow comes from primaries.

2)      Transition Fully To Secondaries OFF:
a.       0% staged: 0% of fuel flow comes from the secondaries; 100% of fuel flow comes from primaries.
b.      100% staged: Pulsewidths / DutyCycles are equal. %age of fuel flow depends on relative injector sizes.
c.       50% staged: I don’t know how to describe this. See below.

wrong -> Percent Staged = PW2/(PW*(Inj1flow/(Inj1flow+Inj2flow)) + PW2*(Inj1flow/(Inj1flow+Inj2flow))*100 <-wrong

This is based on my review of logs and is just a best guess as to the relationship between Percent Staged and PW.

As an aside, would it be possible to add a minimum pulse width setting to keep both sets of injectors in the linear region during staging?

noisymime · Post by **noisymime** » Sat Feb 07, 2015 3:46 am

Apologies if I'm missing it somewhere, but is there any info about injector staging / squirts etc in the new documentation? Section 2.12.13 (MS3) says:

"The most critical settings are on the left. Complete the Control Algorithm through to Injector size settings"

but then doesn't talk about any of these 'critical' options in any detail.

I vaguely recall in the old documentation there was a section on permissible squirt/cylinder combinations, the relative merits of alternating vs simultaneous, changing the number of squirts to achieve reasonable req_fuel etc. I found it really useful when I was first setting up, and just went to check it again now and couldn't find anything. Is it no longer needed?

Post by **jsmcortina** » Sat Feb 07, 2015 4:22 am

noisymime wrote:I vaguely recall in the old documentation there was a section on permissible squirt/cylinder combinations, the relative merits of alternating vs simultaneous, changing the number of squirts to achieve reasonable req_fuel etc. I found it really useful when I was first setting up, and just went to check it again now and couldn't find anything. Is it no longer needed?

I don't think that was in the MS3 manual or I'd have carried it over. I'll see if there's some text elsewhere that could be used.

James

hardline · Post by **hardline** » Thu Feb 12, 2015 2:19 pm

Might I suggest rewording section "4.1.1 Existing EFI Vehicle" to include the fact that control of Ford's returnless FPDM-based system works and has been tested on a number of working installs.

kaeman · Post by **kaeman** » Thu Feb 12, 2015 9:03 pm

I love the new manuals, and I am using them to do my toothed wheel install on the crank, I don't know if you have caught this, but in the edis section of the 3.57/mx3 hardware manual section 6.7.4 and later when showing the picture of the sensor in relation to the missing tooth and the rotation of the crank, the illustrations show the sensors mounted after tdc according to the arrow showing the wheels direction of travel. Thanks for all the hard work.

Post by **jsmcortina** » Fri Feb 13, 2015 3:13 am

kaeman wrote:I love the new manuals, and I am using them to do my toothed wheel install on the crank, I don't know if you have caught this, but in the edis section of the 3.57/mx3 hardware manual section 6.7.4 and later when showing the picture of the sensor in relation to the missing tooth and the rotation of the crank, the illustrations show the sensors mounted after tdc according to the arrow showing the wheels direction of travel. Thanks for all the hard work.

Those diagrams show the sensor aligning with the missing tooth when the engine is BTDC.

James

kaeman · Post by **kaeman** » Sun Feb 15, 2015 10:57 am

Ok, I was just confusing having sensor advanced, I am just going to use the advanced crank position and center the missing tooth on the sensor. Thank YOU.

seb209726 · Post by **seb209726** » Mon Feb 16, 2015 6:40 am

One small typo, in the assembly guides, when talking about the CAN jumpers. SPR pads have been mixed up. (CANH is SPR1, not 4, CANL is SPR2, not 3)

Post by **jsmcortina** » Mon Feb 16, 2015 8:23 am

seb209726 wrote:One small typo, in the assembly guides, when talking about the CAN jumpers. SPR pads have been mixed up. (CANH is SPR1, not 4, CANL is SPR2, not 3)

The photo under the text shows that to be in error too. Thanks for pointing it out.

James

Post by **muythaibxr** » Mon Feb 16, 2015 6:52 pm

Sam280Z wrote:OK. It wasn’t clear to me that the %age staged was %age of fuel flow. See if this is correct:

The ratio of fuel flow from primary and secondary injectors is determined by the Percent Staged value and how Transition Fully To Secondaries is set.

1)      Transition Fully To Secondaries ON:
a.       0% staged: 0% of fuel flow comes from the secondaries; 100% of fuel flow comes from primaries.
b.      100% staged: 100% of fuel flow comes from secondaries; 0% of fuel flow comes from primaries.
c.       50% staged: 50% of fuel flow comes from secondaries; 50% of fuel flow comes from primaries.

2)      Transition Fully To Secondaries OFF:
a.       0% staged: 0% of fuel flow comes from the secondaries; 100% of fuel flow comes from primaries.
b.      100% staged: Pulsewidths / DutyCycles are equal. %age of fuel flow depends on relative injector sizes.
c.       50% staged: I don’t know how to describe this. See below.

Percent Staged = PW2/(PW*(Inj1flow/(Inj1flow+Inj2flow)) + PW2*(Inj1flow/(Inj1flow+Inj2flow))*100

This is based on my review of logs and is just a best guess as to the relationship between Percent Staged and PW.

As an aside, would it be possible to add a minimum pulse width setting to keep both sets of injectors in the linear region during staging?

I think your math for c) above might be a bit wonky. The actual code is here:

Code: Select all

staging_percent = ((unsigned long) ram4.staged_pri_size * 100) /
                           (ram4.staged_pri_size + ram4.staged_sec_size);

calculated_base_pw = (base_pw * (unsigned long) staging_percent) / 100;

staged_transition_percent = <table lookup from staging table>

calculated_pw1 = base_pw - (((long) staged_transition_percent *
                                 (base_pw - calculated_base_pw)) / 1000L);

calculated_pw2 = (calculated_base_pw * staged_transition_percent) / 1000L;

So staging_percent is the percentage to scale down to when fully staged, calculated_base_pw is the pulse width at time of calculation that will result from the original base pulse with multiplied with the staging_percent, calculated_pw1 is the scaled down amount based on the commanded staged_transition_percent that the user specifies in the table, and calculated_pw2 is the amount of fuel the secondaries should be injecting.

The idea there is that at 50%, the PW of the primaries is reduced halfway to the final calculated value, and the PW of the secondaries is increased halfway to the final value. It's very similar to what was done with the old transition styles, but it lets you map the percentages in the table rather than automatically going fully staged or fully unstaged. It also keeps the total *amount* of fuel in grams being delivered the same as what would be getting delivered if you were staying entirely on the primaries. The idea being that the VE % should remain the same even if you change the staging % in the table.

The error in the current calculations (meaning error in the total amount of fuel in grams) is about 2% due to staging_percent being accurate to ~1%. If I changed the code so that staging_percent were accurate 2 ~.1% then the error in the total number of grams when comparing primaries only to primaries+secondaries would be around .4%. So there is a potential enhancement that we could make there.

Post by **muythaibxr** » Mon Feb 16, 2015 8:23 pm

Sam280Z wrote: As an aside, would it be possible to add a minimum pulse width setting to keep both sets of injectors in the linear region during staging?

Right now you can effectively get this behavior by setting your transition bins close together, and setting a relatively high number (20-30%) as the staging % in the staging table, or you can actually configure the injector non-linearity feature to factor this out.

Doing what you describe would require me to calculate everything in the code to get to the final pulse-widths, then compare that and "waste" the calculations and not start staging if the secondaries aren't above a specific PW. I'd rather not have the code go through all that math only to throw it out and use the original pulse-width when there are features that allow you to properly factor in the non-linear region.

Ken

mschlang · Post by **mschlang** » Wed Feb 18, 2015 8:26 pm

I noticed tonight that one page 216 of the ms3x/v3 manual, it states connect js7 to canh, but the photo caption shows js8 as does the wiring in the picture.

Eagle7 · Post by **Eagle7** » Sun Feb 22, 2015 12:38 pm

There's a conflict in the documentation of the rotary trailing spark in FC mode between two manuals.

MS3X/V3.0 Hardware Guide
Mode -> FC
Spark A Leading (IGt-L)
Spark B Trailing Select (IGs-T)
Spark C Trailing Trigger (IGt-T)

Megasquirt-3 Setting Up
Wiring:
Spk A Leading coil (IGt-L)
Spk B Trailing coil (IGt-T)
Spk C Trailing coil select

Can you confirm which one is correct for the 1.3.4 firmware?

Sam280Z · Post by **Sam280Z** » Tue Feb 24, 2015 7:20 pm

Ken,
Thanks for the explanation on staging. Sorry it has taken me so long to get back.
I'm trying to tune the table to do exactly what you suggest in your second post. The problem is that the %staged isn't in the logs, so I'm trying to calculate it from the PWs so I can see where to make my cutoffs.

I think I now understand what it is doing, and how it is doing it. It took a bit of algebra.

The staged % can be calculated from the PWs and the relative injector sizes like this:
set p as Inj1's percentage of total available flow: p=Inj1flow/(Inj1flow+Inj2flow)
then percent staged = 100*PW2/(PW1*p - PW2*p +PW2)

So for 2c in my original post, it could read:
c. 50% staged: with equal sized injectors, 75% of the fuel is provided by the primary injectors and 25% by the secondaries. {75% is half way between 100%(0% staged) provided by primaries and 50% (fully, or 100% staged) provided by the primaries.} PWs for unequally sized primaries and secondaries is adjusted accordingly.

Let me know what you think.

Sam

Marek · Post by **Marek** » Fri Feb 27, 2015 1:38 am

Section 17.11 of the Tunerstudio MS Lite Reference 1.3 pdf states that
"on the fly" switching between sequential and other modes can be used to test the effect on engine operation. In a recent thread http://www.msextra.com/forums/viewtopic ... 31&t=57570 , Matt says that this effect will be to stall the engine and don't do it!. James didn't reply.

Some months ago, I did attempt to use this feature on MS3/3X pre1.3a7 running 12 cylinder sequential switching from sequential to semi-sequential on the fly in the test mode as outlined in 17.11 and the engine did stall. I can't be certain, but * I think * it also knocked out the flyback circuit on my peak and hold drivers so I'm not keen on "retesting" this until it has been verified as working on 12 cylinders. (It'd be nice of way of checking dead time settings.)

If not all options are available for all engine configurations, perhaps the manual ought to qualify this more precisely.

kind regards
Marek

hardline · Post by **hardline** » Wed Mar 04, 2015 7:14 pm

Was my previous suggestion found to be unacceptable? I saw no feedback on it and I see no change made in the manuals.

242ATL · Post by **242ATL** » Tue Apr 21, 2015 2:52 pm

What happened to the MS manuals? I can't even find information on the sync error types.

Matt Cramer · Post by **Matt Cramer** » Wed Apr 22, 2015 5:31 am

hardline wrote:Might I suggest rewording section "4.1.1 Existing EFI Vehicle" to include the fact that control of Ford's returnless FPDM-based system works and has been tested on a number of working installs.

Do you have any example settings?

hardline · Post by **hardline** » Thu Apr 23, 2015 10:44 am

Sure do, I'll post up the necessary info tonight, as I am currently at work.

hardline · Post by **hardline** » Thu Apr 23, 2015 7:30 pm

Settings require either an inversion circuit for fuel pump polarity, or a pull-up resistor installed to the fuel pump output.

fpress.PNG

You then need to setup a generic sensor as the fuel pressure input, in KPA.

fpress2.PNG

Megasquirt Support Forum (MSEXTRA)

MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion

Re: MS3 new documentation discussion