MS3 Pro v1.5 Released Firmware

General support questions and announcements for MS3. See also MS3 manuals.

Moderators: jsmcortina, muythaibxr

Post Reply
R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

1) The Engine State function needs to be explained how the strategy works... more specifically how the algorithm is designed to detect idle state and overrun fuel cut state. The update to v1.5 wrecked the car's ability to enter the idle state, enable the IACV, and control the engine speed... it just stalls now. The state machine gives several parameters to calibrated and tips on how to configure them... but no mention of how they change the state machine's operation. With the removal of the idle control speed lockout... the idle control can become active before the engine is actually at or near the idle target speed. Overrun Fuel cut doesn't appear to have a hysteresis as braking while in gear will cause the car to toggle in and out of engine fuel cut off... causing violent bucking.

2) Why can't we log the individual PID contributors to the Wgdc?

3) Why can't we log the individual PID contributors to the IACV %?

4) Why does it appear that idle ignition timing, idle VE, Idle state, & Idle closed loop control all have different enable/disable conditions?

5) In the Megalog viewer software - why dont we have one status for engine state that toggles between which state it is in, rather than have an idle state that shows when it's on, along with the other states?
jsmcortina
Site Admin
Posts: 39587
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK
Contact:

Re: MS3 Pro v1.5 Released Firmware

Post by jsmcortina »

R/TErnie wrote:1) The Engine State function needs to be explained how the strategy works... more specifically how the algorithm is designed to detect idle state and overrun fuel cut state. The update to v1.5 wrecked the car's ability to enter the idle state, enable the IACV, and control the engine speed... it just stalls now. The state machine gives several parameters to calibrated and tips on how to configure them... but no mention of how they change the state machine's operation. With the removal of the idle control speed lockout... the idle control can become active before the engine is actually at or near the idle target speed. Overrun Fuel cut doesn't appear to have a hysteresis as braking while in gear will cause the car to toggle in and out of engine fuel cut off... causing violent bucking.
Improvements to behaviour have been made in pre-1.5.1 beta 1.
2) Why can't we log the individual PID contributors to the Wgdc?
3) Why can't we log the individual PID contributors to the IACV %?
Added in the latest beta. See http://www.msextra.com/forums/viewtopic ... 01#p503901
4) Why does it appear that idle ignition timing, idle VE, Idle state, & Idle closed loop control all have different enable/disable conditions?
Engine states reduces the number of settings, there were more previously.
5) In the Megalog viewer software - why dont we have one status for engine state that toggles between which state it is in, rather than have an idle state that shows when it's on, along with the other states?
I don't believe there is a way for the logging software to display it any other way. I wish there was.

James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
whittlebeast
Super MS/Extra'er
Posts: 2221
Joined: Tue May 04, 2004 8:20 pm
Location: St Louis
Contact:

Re: MS3 Pro v1.5 Released Firmware

Post by whittlebeast »

Please post a log and matching MSQ
R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

Re: MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

Thank you James!

I'll update and give it another run.

Is there a known bug with using the advanced PID settings for closed loop boost control?

The WGDC% doesn't appear to include Closed Loop corrections, not sure if that's because the cl control isn't working... or if it's just the output of the Wgdc pre-control map.
I suspect this issue may be corrected with the megalog viewer update.
R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

Re: MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

So...

The Boost Control - Advanced PID.

I ended up with
Sensitivity - 470
P -75
I -7
D - 200.

I still have some overshoots - needs more D term.

it would appear that the i term is exceptionally sensitive, the D - term is exceptionally NOT sensitive, and the P term is somewhat insensitive.

The slider bar for "sensitivity" does not react linearly either. The difference from 350-375-400-425-450-475 is pretty minimal.. however if I move that slider to 500, it's an exceptionally large reaction all of the sudden... dare I say logarithmic? I'm going to go play with higher settings than 470 this afternoon. I'll report back on my findings.

If you set the sensitivity bar at 250 (middle) and Set your P Term and I term to 100, and D term disabled. (without the ability to log you wouldnt know this...) but the I term is contributing +75% or more Wgdc and freaking out causing huge oscillations. Meanwhile the P term is contributing less than .100% wgdc contribution.

To me this shows that something is awry, and also you have to give the people the ability to log the PID outputs in standard mode. Having to turn off the car, download the log, and then import it every time makes the real time ECU take longer than it takes me to flash OEM ECU's (2-3 minutes per burn) aka HUGE waste of time. Also when you go to download the logs, the log file size is incorrect. It will use the previous log file size.

Do you have a change list or a post where I can read about the code changes to v1.5 beta 1 to beta 3?

I think the engine state control needs more parameters... for example why does the engine predict I'm entering the idle state when I make a WOT shift at 7k rpm? Because the rate of engine speed acceleration transitions through the idle state region? Why is there not an engine speed lockout for idle state? Same goes for fuel cut off... why not use the vacuum based switch for fuel cut off with a hysteresis? I dont care if I have to change 150 characteristics as long as it functions properly in the end. Maybe I'm just bitching. I'll go spend some time on it today and report back. Thank you guys for reading/listening.
R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

Re: MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

Re: MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

1) Boost Control PID
2) Boost Control - boost by gear
3) Engine State - Idle control - Overrun Fuel cut off

1) Boost Control
Ok I played with the Boost control sensitivity slider some more today to get more D term.
By going from 470 to 487 - the response from the controller was nearly doubled. This confirms my previous assumption that response is somewhat logarithmic.

My new settings are...
Sensitivity - 487
P - 35
I - 3
D - 195

*****

2) Boost By gear logic doesn't seems to meet either of my needs, and so i suspect it doesn't help anyone else either. It changes the the target boost over x% tps? per gear? So quickly, if we look at our target based on throttle position map, I would need a different throttle position enable for each gear if I wanted to run different amounts of boost in each gear... and maintain a linear boost to throttle position request.

IMO - This logic needs to be changed... to Target KPA change per gear. The amount of boost target per gear that is changed from the normal boost target table.
1 - (-50kpa)
2 - (-20kpa)
3 - (0kpa)
4 - (0kpa)
5 - (0kpa)

The Target Boost table (say its at 200kpa @ 100%TPS)
1 - 150
2 - 180
3 - 200
4 - 200
5 - 200

This works because our boost bias table has different boost target breakpoints and will reduce the target boost control duty cycle accordingly.

2.1 - part duex
The other part i want is a Boost Duty Cycle Bias adjustment per gear. For example... I need 3-4% more wgdc is 1st gear than I do 4th. So this would be a gear based compensation to my feedforward boost control table (Boost Bias in your terms)
Example
1 - (4%wgdc)
2 - (3%wgdc)
3 - (2%wgdc)
4 - (0%wgdc)
5 - (-2%wgdc)

This allows my feedforward to be more accurate and allows my Closed Loop boost control to work even better... in every gear.

3) Engine states - Idle - and overrun.
So,
I had some issues with the engine states being tuned properly... as I suspect some of you are also having. I wanted to share my thoughts, what I did, and why i did it. Might not be right, but worked pretty well for me.

My current settings
Fast acc = 1500
Slow acc = 600
Fast dec = -2000
Slow dec = -600
TPS closed = 1%
WOT = 95%
Overrun MAP threshold = 33%

So... what I found was that I could get the engine to enter the idle state and not miss any when I increased both my slow dec/acc bounds from 200 to 600. If these slow bounds were set too close to zero, the state machine would flip me in and out of idle state (wreaking havoc on the idle pwm controller, idle ignition controller, and Idle VE)

The idle VE map needed a little work - I messed around with this by using the AC to change load, alternator (headlights) to change load, and then simply by changing my target idle rpm to make this map "good" at nominal. I didn't disable my EGO correction, but I had it as a gauge to see what it was doing. I got my EGO correction to be within 2-3% percent and then moved on.

Idle ignition controller - I really like this feature. it allows you to pick your nominal ignition timing, and then pick the amount of timing you want to add/subtract from that nominal to stabilize the engine speed during idle state. I had orginally set this to 5 degrees and allowed up to 10degrees +/- for the controller to swing the timing. The issue I was having, when the engine state machine was flipping in and out of idle state, this was turning on and off (1 sec delay) and would cause... by itself and unstable idle speed (IACV was not moving) I lowered my basic ignition timing in that region sub 1500 rpm low load, and raised my nominal idle ignition to 10 degrees and not 5 degrees. That helped a little....

Next up was to setup the delays when entering the idle control. I had my ignition timing idle delay set to zero so when the state machine (uncontrollably) changed to idle and back out, the idle ignition timing would grab the engine speed and yank it down. This worked as a work around, but the issue was I needed my state machine to work properly. By widening the slow dec/acc thresholds I was able to detect idle state and stay in it.... and then I was able to put a 1 sec delay on the idle ignition timing control (which prevented false detection during shifts etc)

During decel I would have issues with the idle state enabling when it should have been in overrun. I thought this was related to the slow acc/dec thresholds... it wasn't. One of the guys here wrote in his notes for the release to set it as close to the lowest idle manifold pressure as you can. So I turned off all the electrical loads, the AC, etc and just watched the lowest MAP i could make idling and set it 1-2kpa below that. Now overrun works like a charm... this is just pure stupidity on my part to be honest. i forgot where the overrun map threshold was in the SW the other day. Prior to having it at 33kpa, I had it at 22kpa... so yeah go figure.

I still dont quite understand the fast acceleration/deceleration rates and what it has to do with the state machine, but for now it's working decently well. I think it helps identify slow accel from cruise, and fast accel from slow accel. WOT is obviously totally based on throttle position.

A great visual for this is to open the Engine State Settings window and drive around. Clutch in, coast, come to stops, drive around town or a parking lot and watch the states. When you find a state that isn't reporting properly, make some adjustments....
E4ODnut
Master MS/Extra'er
Posts: 406
Joined: Sun May 02, 2004 9:13 pm
Location: Gibsons, BC, Canada

Re: MS3 Pro v1.5 Released Firmware

Post by E4ODnut »

R/TEarnie,
Very well done! It's a pity that we don't get some more guidance from the developers in this area.

As far as overrun fuel cut is concerned, am I correct in assuming that your lowest idle manifold pressure is ~35KPA, and this would be with the engine fully warmed up and no parasitic loads? Also, I am curious as to what engine you are working with.
Robert
'53 VW Beetle Street Baja dual port 1641 MSnS-Extra
R/TErnie
MS/Extra Newbie
Posts: 8
Joined: Fri Jun 02, 2017 3:25 pm

Re: MS3 Pro v1.5 Released Firmware

Post by R/TErnie »

yeah 34-35kpa. Full warmed up, fans off, AC off, lights off, radio off.

It's a 1.8L Mazda Miata engine. Has a FlyingMiata turbo kit, MS3Pro, and some other tid bits. I've tuned quite a few supercharged and turbocharged Miatas in the area... and all of them have been on megasquirts.
E4ODnut
Master MS/Extra'er
Posts: 406
Joined: Sun May 02, 2004 9:13 pm
Location: Gibsons, BC, Canada

Re: MS3 Pro v1.5 Released Firmware

Post by E4ODnut »

Thanks. I'll experiment with my settings and see if I can get some over run fuel cut hysteresis
Robert
'53 VW Beetle Street Baja dual port 1641 MSnS-Extra
Post Reply