MS3 pre-1.6 alpha 1 firmware (superceded)
Moderators: jsmcortina, muythaibxr
-
- MS/Extra Newbie
- Posts: 33
- Joined: Tue Mar 30, 2021 11:44 am
Re: MS3 pre-1.6 alpha 1 firmware
Hey James, thanks for the release.
I have been working for quite a while with load prediction, charge sensing, and transient combustion analysis in the industry pressure prediction plays a huge role in refining transient fueling/transient air calibrations. Can you go into more detail as to how this new feature was implemented and what was your intention with it?
In the German-based software, for example, we have manifold filling constants, prediction angles, throttle lag factors, and other calibratables that allowed me to fine-tune the load prediction phasing during transient maneuvers. Italian-based software on the other hand uses a fair amount of filters, and integrators to bias the predicted pressure closer to the estimated pressure via throttle angle or to the actual measured pressure via the TMAP. Interested in seeing how this could be implemented in the Megasquirt environment regarding transient fueling.
If transient fueling is open to discussion, I would very much so like to give my contribution and suggestions.
I have been working for quite a while with load prediction, charge sensing, and transient combustion analysis in the industry pressure prediction plays a huge role in refining transient fueling/transient air calibrations. Can you go into more detail as to how this new feature was implemented and what was your intention with it?
In the German-based software, for example, we have manifold filling constants, prediction angles, throttle lag factors, and other calibratables that allowed me to fine-tune the load prediction phasing during transient maneuvers. Italian-based software on the other hand uses a fair amount of filters, and integrators to bias the predicted pressure closer to the estimated pressure via throttle angle or to the actual measured pressure via the TMAP. Interested in seeing how this could be implemented in the Megasquirt environment regarding transient fueling.
If transient fueling is open to discussion, I would very much so like to give my contribution and suggestions.
Rafael
Renault Sandero RS F4R 2.0 NA-VVT
MS3X+DBWX2+CAN EGT
Renault Sandero RS F4R 2.0 NA-VVT
MS3X+DBWX2+CAN EGT
-
- MS/Extra Newbie
- Posts: 18
- Joined: Wed Dec 23, 2020 10:54 am
- Location: France
- Contact:
Re: MS3 pre-1.6 alpha 1 firmware
Hi can i ask if one day we gonna have a full 6 speed sequential gearbox support with air shift? That also make throttle blip and downshift with the DBWX2 drive by wire unit? Actually its good for drag racing but for track cars that's a missing feature.
Regards
Regards
Re: MS3 pre-1.6 alpha 1 firmware
I'm also disappointed as i'm waiting for closed loop sequential gearbox shift cut support since many years and didn't even get an answer on my request at all.
I think I will have to change my ECU im the near future because this is very important for me.
I think I will have to change my ECU im the near future because this is very important for me.
#1 MS3 Pro Evo, 1.5.2
Caterham with Saab B234, GT3071R, Sadev
#2 MS3 Pro, 1.5.2 , Saab 900 Turbo, GT2860RS
#3 MS2, MapDaddy 4bar, EDIS
Caterham with Saab B234, GT3071R, Sadev
#2 MS3 Pro, 1.5.2 , Saab 900 Turbo, GT2860RS
#3 MS2, MapDaddy 4bar, EDIS
-
- Site Admin
- Posts: 8230
- Joined: Thu Oct 14, 2004 12:48 pm
Re: MS3 pre-1.6 alpha 1 firmware
There are currently no plans to add support for this.
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
-
- Site Admin
- Posts: 8230
- Joined: Thu Oct 14, 2004 12:48 pm
Re: MS3 pre-1.6 alpha 1 firmware
What we're doing is a lot more simple than this.... I added a table with throttle position on the y axis, rpm on the x axis, and predicted load on the z axis.Motterdude wrote: ↑Mon Nov 14, 2022 6:07 pm Hey James, thanks for the release.
I have been working for quite a while with load prediction, charge sensing, and transient combustion analysis in the industry pressure prediction plays a huge role in refining transient fueling/transient air calibrations. Can you go into more detail as to how this new feature was implemented and what was your intention with it?
In the German-based software, for example, we have manifold filling constants, prediction angles, throttle lag factors, and other calibratables that allowed me to fine-tune the load prediction phasing during transient maneuvers. Italian-based software on the other hand uses a fair amount of filters, and integrators to bias the predicted pressure closer to the estimated pressure via throttle angle or to the actual measured pressure via the TMAP. Interested in seeing how this could be implemented in the Megasquirt environment regarding transient fueling.
If transient fueling is open to discussion, I would very much so like to give my contribution and suggestions.
I've never needed this on any "normal" throttle car, but on my ITB engine, it made a huge difference in throttle response.
Transient fueling is very much open for discussion.
Ken
Megasquirt is not for use on pollution controlled vehicles. Any advice I give is for off road use only.
Re: MS3 pre-1.6 alpha 1 firmware
What's the definition of closed loop sequential gearbox shift cut? Is it:
1) See request for upshift from shift lever strain gauge
2) Note current gear
3) Start cut when voltage from strain gauge increases beyond threshold
4) Keep shift cut active until gear increases by 1
5) Stop cut once gear has increased by one or the maximum allowed time to shift has elapsed.
If that's it, it is easy and I see no reason why it can't be implemented?
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
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
Re: MS3 pre-1.6 alpha 1 firmware
Unless you have an active gear indicator (I have my system set up to infer active gear based on RPM vs. wheel speed), you'd want to ramp timing back in based on the strain gauge and elapsed time only. As the revs fall, they'll pass through a window that will assume I'm in the next gear, though it may not be engaged.
Re: MS3 pre-1.6 alpha 1 firmware
I don't think anyone has a sequential without a potentiometer for gear sensing. Adding an ignition ramp on spark resume is trivial.
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
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
Re: MS3 pre-1.6 alpha 1 firmware
Thank you for keeping this alive. This is very important for me and it should be very easy to implement as everything is already there.
This is my idea for CL shiftcut, which is analog to geartronics. And yes, any sequential gearbox has a potentiometer which tells the dash which gear is selected and a seperate spark cut sensor which tells the ecu when to cut spark/fuel.
Now, the spark is cutted when the spark cut sensor goes down ->0V. This is perfect. The only huge difference between OL and CL is when the spark comes back.
OL is a fixed time. CL when the potentiometer is within a certain range. It's not very complicated, as everithing is already inside the MS.
The only things that have to be changed on the existing code is:
1. Small mV-window table for each gear where the gear is fully engaged, so that spark/fuel returnes:
#0 = 0,54-0,58mV
#1 = 0,88-0,93 mV
#2 = ....
.
.
2. A "failsafe" field for minimum and maximum spark cut, activated by the spark cut sensor.
So, that spark will never return below for example 0,03ms and will always come back after 1.0s. So you can still run the car if the gear sensor is broken or disconnected.
This is my idea for CL shiftcut, which is analog to geartronics. And yes, any sequential gearbox has a potentiometer which tells the dash which gear is selected and a seperate spark cut sensor which tells the ecu when to cut spark/fuel.
Now, the spark is cutted when the spark cut sensor goes down ->0V. This is perfect. The only huge difference between OL and CL is when the spark comes back.
OL is a fixed time. CL when the potentiometer is within a certain range. It's not very complicated, as everithing is already inside the MS.
The only things that have to be changed on the existing code is:
1. Small mV-window table for each gear where the gear is fully engaged, so that spark/fuel returnes:
#0 = 0,54-0,58mV
#1 = 0,88-0,93 mV
#2 = ....
.
.
2. A "failsafe" field for minimum and maximum spark cut, activated by the spark cut sensor.
So, that spark will never return below for example 0,03ms and will always come back after 1.0s. So you can still run the car if the gear sensor is broken or disconnected.
#1 MS3 Pro Evo, 1.5.2
Caterham with Saab B234, GT3071R, Sadev
#2 MS3 Pro, 1.5.2 , Saab 900 Turbo, GT2860RS
#3 MS2, MapDaddy 4bar, EDIS
Caterham with Saab B234, GT3071R, Sadev
#2 MS3 Pro, 1.5.2 , Saab 900 Turbo, GT2860RS
#3 MS2, MapDaddy 4bar, EDIS
-
- MS/Extra Newbie
- Posts: 18
- Joined: Wed Dec 23, 2020 10:54 am
- Location: France
- Contact:
Re: MS3 pre-1.6 alpha 1 firmware
This is not a problem Megasquirt handle upshift decently already the problem is downshifting with drive by wire or throttle actuator for down shifting.Reverant wrote: ↑Wed May 24, 2023 12:54 pmWhat's the definition of closed loop sequential gearbox shift cut? Is it:
1) See request for upshift from shift lever strain gauge
2) Note current gear
3) Start cut when voltage from strain gauge increases beyond threshold
4) Keep shift cut active until gear increases by 1
5) Stop cut once gear has increased by one or the maximum allowed time to shift has elapsed.
If that's it, it is easy and I see no reason why it can't be implemented?
Have not only drag racing in the world ! In motorsports when you have corners you need to brake and shift down.
And that's what is not easy to implement.