Feature request: 16x32 fuel/ign table... flame suit, on!

Testing and development of Megasquirt 3

Moderators: jsmcortina, muythaibxr

Post Reply
krisr
Master MS/Extra'er
Posts: 799
Joined: Wed Aug 17, 2005 1:17 am
Location: Sydney, Australia

Feature request: 16x32 fuel/ign table... flame suit, on!

Post by krisr »

Hey guys, I know i'm going to stir up some controversy here and go right against the grain, but here goes. I think we need more fuel cells but only in the direction of RPM - maybe boosted guys want more MAP cells, but in my specific use case, 16 load cells works great for an NA application. I have found that I have a very extraordinary car and am just running out of control of the injector in the rev range as I cannot be as granular as I can be. The car is a 400 Pontiac, kinda tight 9.5" 5000 converter, 3.89 gears with a 28" tyre, street car, but does see the strip. Basically, 16x16 is not enough I feel to give proper control of idle, cruise, the middle and up top, so at the moment i'm kinda compromising with cell spacing on the RPM axis as i'm using more RPM points down low so the street drivability is fantastic, however after 2400RPM, i'm using 600rpm increments to the next cell up to 6600rpm.

I know it's kind of a controversial topic as the cells load/rpm points are user configurable and most people dont want to tune more cells than they have to, however I just don't feel we have enough rpm points for N/A cars that run open loop.

What is the consensus of others in MS-land? Would a user-configurable amount of cells be a better approach? i.e. the user could say "I want 10 load points and 25 rpm points", however the max points could be capped at the afore mentioned 16x32.

Lets discuss?

Kris
Sydney, Australia
1971 Holden Monaro HQ
MS3X Sequentially fuelled 400 Pontiac
Suprazz
Master MS/Extra'er
Posts: 499
Joined: Mon Jan 22, 2007 7:58 pm
Contact:

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by Suprazz »

You can use 2 or 4 tables to expand your ignition table to 16x32 or 32x32 if you want.
Best looking and most advanced CAN-bus gauge
Toyota Supra 7MGE, 7M-GTE and 1JZ-GTE Plug and play ECUs: http://www.perfecttuning.net
Serial to Bluetooth or Serial to WiFi
DIYAutotune Canadian, EFI Source and ECUMaster reseller!
Paul_VR6
Super MS/Extra'er
Posts: 1978
Joined: Wed Apr 15, 2009 11:31 am
Location: Oxford, PA
Contact:

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by Paul_VR6 »

Can you give more details of the engine itself? Compression, throttle type, cam, dyno graph, etc. Also msq and a log or graph of the afr error you are trying to fix would help as well.

I have tuned more than a few engines at or over 100hp/l na and never ran into a lack of cells. Most engines I could tune with less than 16 and many big v8s could probably use half that!

That being said, like the last poster mentioned, you can use table switching to extend your table. Say table 1 goes to 3500rpm, the next table would start there and go to redline.
-Paul
1992 Corrado SLC 3.6 VR6 11.38@120 - MS3 Pro Ultimate - Microsquirt I/O - Can EGT - Racepak IQ3s
kptuned.com - Megasquirting the World! Megasquirt Sales, Service, Tuning and More!
Image
whittlebeast
Super MS/Extra'er
Posts: 2221
Joined: Tue May 04, 2004 8:20 pm
Location: St Louis
Contact:

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by whittlebeast »

This comes up on motors with wide operating ranges. My Vtec Honda idles around 900 and pulls to 9000. Turbo versions of the same motor adds can easily go from a MAP of 15 on a down shift to a MAP of 200-300 KPA. That is a huge operating range to fit in a 16x16 table. Motec starts at 16x16 with the ability to insert rows and/or columns up to a total of 24x24 possible as I recall.

Andy
Reverant
Super MS/Extra'er
Posts: 1233
Joined: Sat Apr 15, 2006 12:39 am
Location: Athens, Greece

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by Reverant »

A 22x22 table would cover 99.999% of all engines and would fit nicely in the 1K RAM pages.
The man behind MS Labs
2005 Audi A3 2.0L TFSI DSG AWD - Extreme MS3
2002 Mazda Miata 1.8 6sp - Enhanced MS3 1.4.0, sequential injection, sequential ignition, big turbo, lots of boost
krisr
Master MS/Extra'er
Posts: 799
Joined: Wed Aug 17, 2005 1:17 am
Location: Sydney, Australia

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by krisr »

Howdy, below is the MSQ. I'll put a bit more context around the motor.

* 400 Pontiac, ported aluminium heads, Victor EFI intake, 4bbl 1000cfm TB
* 242/248 @ 0.050 cam, 110LSA, installed at 109ICL
* Turbo 400, full manualised, 5000 9.5" converter
* 3.89 trutrac 9" rear, 28" tyre
* Is setup for E85 w/flex fuel, have not had the change to tune the Flex fuel/spark tables yet

Basically, with the converter, manualised box, gears and the rev range of the engine, I can easily use the entire map with the exception of the highest rpm, most vacuum corner obviously. It will cruise at 35mph at around 2000rpm, 60mph at around 3000rpm, the converter will does feel a bit like a CVT as it will flare a bit if you give it a stab, but when on the dyno or at the track it will flash straight up to 5000 and out to 6500. This car being very street, you'll see i've put more focus into the lower rev range and fuel control is super close to target (not using EGO CORR) with MAT correction but yeah, the spacing between RPM cells from 2400 onward means I need to compromise a bit with interpolation.

I had thought about table switching, however i'm not sure how I can make rpm switching work when i'm looking to go Flex - but then will run into the same problem i'm guessing and will need to switch flex tables on rpm too so all 4 VE tables would be used in theory. I need to have a read of the manuals to see if this is even possible....

The dyno graphs i've got atm wont show much as I think i'm having valve spring issues or hyd roller lifter pump-up, it's a work in progress still :x
2017-09-06_14.33.34 - final 2.msq
Sydney, Australia
1971 Holden Monaro HQ
MS3X Sequentially fuelled 400 Pontiac
tpsretard2
Master MS/Extra'er
Posts: 662
Joined: Thu Feb 14, 2008 4:59 am

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by tpsretard2 »

Reverant wrote:A 22x22 table would cover 99.999% of all engines and would fit nicely in the 1K RAM pages.

totally agree on this, both road and race.
nathaninwa
Super MS/Extra'er
Posts: 2637
Joined: Tue Sep 14, 2004 4:17 pm
Location: Aberdeen, Wa

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by nathaninwa »

I do table switching at 3000rpm. Basically if you following the rpm curve of the engine upnand to the right I have a 32x32 table which comes in handy for lots of vacuum area and boost every 20kpa so the tune can get dialed in pretty good, it also allows break points in rpm to be really close around peak tq. Not that all of it is needed, but I like playing with it
Volvo 940, 2jzge, MS3Pro, daily
240Z, 2JZ, MS3Pro boost control
krisr
Master MS/Extra'er
Posts: 799
Joined: Wed Aug 17, 2005 1:17 am
Location: Sydney, Australia

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by krisr »

@nathaninwa do you use flex by chance? Mine being N/A I don't need load points. Just need to scale it horizontally.

Is there any reason why the table assignment is fixed and not flexible? I.e. if I configure switching on rpm I can only use VE1 and VE3. This seems to (even though there is no ini error) mean I cant use VE3 for flex even though it's enabled.

It'd be good to use ve1 and ve2 for pump gas switched and ve3 and ve4 for e85 switched.

Sent from my SM-G930F using Tapatalk
Sydney, Australia
1971 Holden Monaro HQ
MS3X Sequentially fuelled 400 Pontiac
krisr
Master MS/Extra'er
Posts: 799
Joined: Wed Aug 17, 2005 1:17 am
Location: Sydney, Australia

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by krisr »

Ohwell, i've decided to bite the bullet and just used RPM switching on tables VE1 --> VE3 and will not worry about E85 for now. If a table size change isn't on the cards then maybe(hopefully) more flexibility as to how flex blending is considered when I finish getting the engine sorted out on pump gas.
Sydney, Australia
1971 Holden Monaro HQ
MS3X Sequentially fuelled 400 Pontiac
AbatelliCristian
Super MS/Extra'er
Posts: 855
Joined: Sun Oct 10, 2010 6:33 am

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by AbatelliCristian »

Hi,
I see your msq, but I think that using breakpoints every 200rpm is useless. From your VE I do not see large nonlinear jumps to justify many breakpoints.


Regards
Cristian
Matt Cramer
Super MS/Extra'er
Posts: 17499
Joined: Thu Apr 16, 2009 8:08 pm

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by Matt Cramer »

AbatelliCristian wrote:Hi,
I see your msq, but I think that using breakpoints every 200rpm is useless. From your VE I do not see large nonlinear jumps to justify many breakpoints.


Regards
Cristian
Agreed, that VE table doesn't show much in the way of irregular spots. A good tip off that you don't have enough resolution would be if you took one of your data logs and did a scatter plot of AFR Error as a function of RPM and load. If it is near zero at the points covered by the grid in your VE table but has larger numbers in the gaps between the corners of the grid, then there's a case for a larger table - or it may be possible to solve it with different breakpoint spacing.

Which brings up another point... while 22 x 22 tables should work for most engines, will that be enough for people who think they need 32 x 32 tables simply because they're convinced they need 32 x 32 tables?
Matt Cramer -1966 Dodge Dart slant six running on MS3X
krisr
Master MS/Extra'er
Posts: 799
Joined: Wed Aug 17, 2005 1:17 am
Location: Sydney, Australia

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by krisr »

I decided that Flex is too good of a feature to forfeit so scrapped the table switch idea and just left the bins for now. When I get the new valve springs and lifters i'll try get more logs and maybe re-bin it again.

On the point of 22x22 or 32x32, I think setting and forgetting bin placement has it's advantages in that on a fresh engine, most people wont know where to put them so just take a guess and most likely 9 times out of 10 will just evenly space them out but TS I admit does a great job to re-bin and re-calc the values but i'm still curious as to why the number of breakpoints are even fixed in either axis at all and why can't it be user defined with a ceiling limit on both axis' to either 22 or 36 (or wherever works for the hardware architecture) especially if MS3's have quite a bit of ram that isn't utilised. In my mind, right or wrong, I think of increased resolution in things like fuel and ignition tables no differently to the reason why 36 tooth crank wheels are preferred over using a single pickup per cylinder (distributor) vs relying on between values - precision used in both input and output, can't be a bad thing?

Anyhow, i'm starting to think this is now a bit of a moot thread as it's swaying in the "not broken so it doesn't need to be fixed" column, but at least / hopefully it can be discussed (...hopefully).
Sydney, Australia
1971 Holden Monaro HQ
MS3X Sequentially fuelled 400 Pontiac
Paul_VR6
Super MS/Extra'er
Posts: 1978
Joined: Wed Apr 15, 2009 11:31 am
Location: Oxford, PA
Contact:

Re: Feature request: 16x32 fuel/ign table... flame suit, on!

Post by Paul_VR6 »

Its just that after tuning a lot, and I mean a lot, of engines I can't think of one time I thought to myself "I wish there were more boxes to fill in!" As that is about how I look at it now. With experience you can make a pretty good guess on breakpoints in the ve with cam and throttle choice (as those are the main drivers).

I don't disagree that more cells may seem easier at first, but having to do it that way on other systems can get quite infuriating.
-Paul
1992 Corrado SLC 3.6 VR6 11.38@120 - MS3 Pro Ultimate - Microsquirt I/O - Can EGT - Racepak IQ3s
kptuned.com - Megasquirting the World! Megasquirt Sales, Service, Tuning and More!
Image
Post Reply