MS2/Extra 2.0 beta 17 (20071227)

[muythaibxr]

After more than a month, here is the latest beta:

Beta 17 release notes:
Some of these are a work in progress
-more Microsquirt compatability work
-timing prediction (for greater timing stability during accel/decel)
-added Honda VTR-1000 mode
-added fuel only mode
-added CAN polling of second MS2 to read ADC channels
-added EGT in Megatune
-added battery voltage logging in Megatune
-closed-loop idle fixes: PID numbers should be much larger (requires retune), Added slow close on throttle press
-Make the MS2 not hang when trying to run in rotary mode. With this, rotary mode seems to work fine.
-Re-indented main.c and misc.c
(EDIT) - fixed up second trigger so that the 2nd wheel speed can be specified and the number of triggers is correctly calculated
(EDIT) - Fixed always-on vs on while moving for stepper IAC motors.
(EDIT) - Get 12x12 dual table and table switching working again

NOTE:

The prediction code is a new implementation, please test with care and give us feedback on whether it is working correctly. First test with prediction off. Set fixed timing and rev the engine. Under accel the timing will likely retard a little, under decel it will advance. Enable prediction and compare. The timing should be more stable.

(EDIT) If you have tried one of the alpha codes released after beta 16, please upgrade as there is likely to be a nasty timing bug in those code.

I've not done much with the bugs in the bug list, nor have we really gone through msoultan's list of ini fixes. We'll be doing those for the next beta, then hopefully releasing.

James and I also intend to start releasing quicker again.. work has been nuts for both of us for the last several months.

The beta is here:

http://www.msextra.com/ms2extra/files/b ... 071227.zip

Ken

[Tjabo]

Ken or James,

Is it always assumed that a new MSQ needs to be created for every new Beta, just import maps and start over with everything else? Or not?

Thanks,

Thad

[jsmcortina]

Not really no. Megatune is designed that it will cope with minor changes to data structure. However, DO read the audit.log that is created to warn you of any issue as correct those.

James

[Asteris]

Archive is corrupted. Fix it, fix it, fix it...

[muythaibxr]

Just re-uploaded... must've transferred it without being in binary mode first.

Give it another shot.

Ken

[6040solder]

:-)

Just re-uploaded... must've transferred it without being in binary mode first.

oldest trick in the book? i've been caught like that too :-) i always wondered why its not the default?

[muythaibxr]

It is in most Linux distros, FreeBSD, MacOS X, etc...

I got caught b/c I used cygwin's ftp client, which I don't normally use (due to using the rest of the above on a regular basis.

Ken

[6040solder]

Code: Select all

Remote system type is UNIX.
ftp> status
Connected to 91.155.xxx.xx.
No proxy connection.
Mode: stream; Type: ascii; Form: non-print; Structure: file
Verbose: on; Bell: off; Prompting: on; Globbing: on
Store unique: off; Receive unique: off
Case: off; CR stripping: on
Quote control characters: on
Ntrans: off
Nmap: off
Hash mark printing: off; Use of PORT cmds: on
Tick counter printing: off
ftp> bin
200 Type set to I
ftp> status
Connected to 91.155.xxx.xx.
No proxy connection.
Mode: stream; Type: binary; Form: non-print; Structure: file
Verbose: on; Bell: off; Prompting: on; Globbing: on
Store unique: off; Receive unique: off
Case: off; CR stripping: on
Quote control characters: on
Ntrans: off
Nmap: off
Hash mark printing: off; Use of PORT cmds: on
Tick counter printing: off

thats a linux client (debian) and a win xp server.

do you mean the server? i swear the slackware server at least is ascii default too, or is that a client side setting?

[jsmcortina]

do you mean the server? i swear the slackware server at least is ascii default too, or is that a client side setting?

It is a client side setting.

Anyway, never mind that, get testing the new beta ! (joke)

James

[muythaibxr]

thats a linux client (debian) and a win xp server.

I believe newer fedora cores, and ubuntu both use binary as the default. I'm sure FreeBSD and MacOS X do as well.

I meant client side in any case.

But as James said, grab the code and test it!

Ken

[UnaClocker]

Thanks guys! I'll get on testing this one out. Still on B15 myself.

[UnaClocker]

I've GOT to get a different USB-Serial adapter, mine takes about 45 minutes to do the flash.. Strange!

[Philip Lochner]

I've not done much with the bugs in the bug list

Should dual table work in Beta 17?

[jsmcortina]

Should dual table work in Beta 17?

Yes. Dual table and table switching should work again with 12x12.

James

[Gokart]

Ignition Table 3 is present for use with table switching but the 3D Ignition Map 3 is not available.

[Keithg]

Flashed this morning.

Started fine. Drove odd. I looked through the settings and the VE table was whacked out. Imported a new VE and it seemed fine. I'll have to look at the spark map this afternoon.

I think I screwed up something with 2nd trigger. It no longer works. I had been running W/S lately for the snappier starts, and checked out COP yesterday with 20071122 and it no longer worked. Flashed 20071227 and it also does not work. This is not due to the code, but something I did. I hope it is not a dead input pin on the MS2.

I noted one odd occurrence with 20071122 and it may be the same here as well, I do not know. I have not looked on the bench with this version, but with the original stim hooked up, MT shows no sync, but the fuel pump is running. I pulled the processor and verified that the transistor is working fine to turn it on and off (+5 on base resistor turns on the LED and closes the relay), but when the MS2 is on board, the stim is plugged in and powered, the FP led is always lit.

KeithG

[matt_fulghum]

For the new PID pwm idle values, what would the range be for a baseline on each setting?

I just upgraded to the latest beta code from an ancient one where the PID idle control was still pretty sketchy, so I went back to using the normal PWM warmup, but it sounds like it's much better now, so I'd like to move back to PID idle.

So far, I've had to keep the idle pretty high to keep the car from stalling or at least lugging when turning the wheel or pressing in the brakes fast, and I assume PID idle should help both of those issues?

[muythaibxr]

I think I screwed up something with 2nd trigger. It no longer works. I had been running W/S lately for the snappier starts, and checked out COP yesterday with 20071122 and it no longer worked. Flashed 20071227 and it also does not work. This is not due to the code, but something I did. I hope it is not a dead input pin on the MS2.

The settings related to second trigger have changed. You may just want to check them.

I noted one odd occurrence with 20071122 and it may be the same here as well, I do not know. I have not looked on the bench with this version, but with the original stim hooked up, MT shows no sync, but the fuel pump is running. I pulled the processor and verified that the transistor is working fine to turn it on and off (+5 on base resistor turns on the LED and closes the relay), but when the MS2 is on board, the stim is plugged in and powered, the FP led is always lit.

KeithG

This is b/c you still have a signal coming in on the VR input. As long as there's a signal of some sort, the fuel pump will turn on, even if there's no sync.

I also think I fixed one of the bugs you reported with the idle valve slow close feature. It shouldn't go below the last-known-good PID value when you do a quick blip now...

Ken

[Keithg]

I think I screwed up something with 2nd trigger. It no longer works. I had been running W/S lately for the snappier starts, and checked out COP yesterday with 20071122 and it no longer worked. Flashed 20071227 and it also does not work. This is not due to the code, but something I did. I hope it is not a dead input pin on the MS2.

The settings related to second trigger have changed. You may just want to check them.

Will do. I think I did this morning, but it was pretty quick. I'll check again. How were they changed? I get no RPM and no Sync with 'dual wheel plus missing' and '2nd trigger falling'

(edit) Dead Hall sensor verified. Have not retested with new sensor, yet.

This is b/c you still have a signal coming in on the VR input. As long as there's a signal of some sort, the fuel pump will turn on, even if there's no sync.

So this is a feature, not a bug, cool!

I also think I fixed one of the bugs you reported with the idle valve slow close feature. It shouldn't go below the last-known-good PID value when you do a quick blip now.

I thought I noticed that, thanks!

Keith

[Philip Lochner]

Should dual table work in Beta 17?

Yes. Dual table and table switching should work again with 12x12.

James

Unless I managed to screw something up it looks like its not working James. MAP on VE table 2 still sits at 0KPA. MT confirms that beta17 is running on the MS. Should I have to change anything else other than "single table" to "dual table"?

Here's some extra info which MAY be relevant:
1) The file download-MS2-firmware.bat for beta 17 would not work so I copied ms_extra.s19 into the beta16 directory and flashed it from that directory;
2) loaded the beta16 msq into the beta17 flashed MS;
3) changed the text in my existing beta16 ini file to reflect b17b so MT would not complain - I change too many parameters in the ini file that its a great effort to change to a new ini file.