MS2/Extra 2.0 beta 14 (20071020)

M52 POWER! · Post by **M52 POWER!** » Tue Oct 23, 2007 10:07 pm

I should have some time tomorrow to test this out again. Sorry If I make no sense I am still a newbie to tuning and overlook simple things like this sometimes. To me it just seemed my VE wouldn't be off that much, I've got it down pretty good in beta 8 oh well....

Is there any real reason to go to beta 14 from 8 if I like how the car is running? Any other serious bugs resolved since then?

-Chris

Jon k · Post by **Jon k** » Wed Oct 24, 2007 5:12 am

Chris -

One thing worth mentioning is that somewhere along the road VE calculations or something were altered because users going from a 2.5 bar to 4.0 bar map sensor were having issues with the fuel table - then, a later release came out that allowed better migration from one sensor to the other - I am not sure which version this entails, but it's something to remember. I'd say move to the latest beta's not because your car doesn't run but so we can test the code

Post by **muythaibxr** » Wed Oct 24, 2007 7:11 am

Jon k wrote:Chris -

One thing worth mentioning is that somewhere along the road VE calculations or something were altered because users going from a 2.5 bar to 4.0 bar map sensor were having issues with the fuel table - then, a later release came out that allowed better migration from one sensor to the other - I am not sure which version this entails, but it's something to remember. I'd say move to the latest beta's not because your car doesn't run but so we can test the code

I think the thing you're talking about here was MS1...

The old code was squirting 2x more than it should, and it was doing so unevenly... The engine may run better now that the code is working correctly.

Ken

Jon k · Post by **Jon k** » Wed Oct 24, 2007 10:02 am

muythaibxr wrote:
I think the thing you're talking about here was MS1...

The old code was squirting 2x more than it should, and it was doing so unevenly... The engine may run better now that the code is working correctly.

Ken

Crap that might be - I am trying my hardest not to confuse the MS1/MS2 cars I work on

M52 POWER! · Post by **M52 POWER!** » Wed Oct 24, 2007 3:58 pm

SCORE! Beta 14 works!!!

It was my stupid idiot fault the VE's were about half of what they should be so now I'm retuning again! If I had known there was a bug in beta8 that I tuned for maybe I would have caught it myself.

hassmaschine · Post by **hassmaschine** » Wed Oct 24, 2007 6:37 pm

yeah, you just got unlucky. The rest of us had our VE tables tuned before beta 8 & 9, so when the bug popped up we were able to just skip them until it was fixed without a significant retune.

Jamie · Post by **Jamie** » Wed Oct 24, 2007 8:32 pm

Mines up and running, neon420a mode, wasted running spark and fuel. runs just as good if not better. nothing broke from beta 13 either. Does the 4wire IAC need more testing? if it does ill stop dragging my feet and hook it up to help out.

Jamie

UnaClocker · Post by **UnaClocker** » Wed Oct 24, 2007 9:04 pm

Yeah, it needs more testing, but I think Ken (or was it James, I get the two confused.. heh) has a stepper and is doing some bench testing, so maybe the next beta will be a better version to do the testing with.

Post by **muythaibxr** » Thu Oct 25, 2007 7:33 am

Yeah, I'm going to work on that a bit more tonight. Basically the stepper I have is very unreliable unless it's "always on."

I think this is because there needs to be a short delay between turning on the stepper controller and actually trying to make the motor move, so I'm going to try adding that.

Once I get it working the way I think it should, I'll add the ability to control it to my PID algorithm.

Ken

masterx81 · Post by **masterx81** » Fri Oct 26, 2007 9:42 am

muythaibxr wrote:Yeah, I'm going to work on that a bit more tonight. Basically the stepper I have is very unreliable unless it's "always on."

I think this is because there needs to be a short delay between turning on the stepper controller and actually trying to make the motor move, so I'm going to try adding that.

Once I get it working the way I think it should, I'll add the ability to control it to my PID algorithm.

Ken

If i can help, on my code (with a pic) i've not done anything strange, and the stepper move well with both 'always on' and 'move only'.
I've tryied it hundred of times, forth and back, and it not loose one step...
No delay moving the stepper, as it's already slow itself :-p
I've found that with the right step time it move well in every condition...

Post by **muythaibxr** » Fri Oct 26, 2007 11:00 am

The delay I'm talking about isn't going to be noticable really... like 5 ms or so (user configurable) between when the code turns on the stepper controller, and the pins to control the motor are toggled.

masterx81 · Post by **masterx81** » Fri Oct 26, 2007 4:16 pm

muythaibxr wrote:The delay I'm talking about isn't going to be noticable really... like 5 ms or so (user configurable) between when the code turns on the stepper controller, and the pins to control the motor are toggled.

Ok, but i've not needed any pause doing my experiments...
With the pic i've bit bang-ed the outputs, this is the piece of code that i use:

if (Phase1 == 0 && Phase2 == 0) {
Phase1 = 1;
Phase2 = 0;
} else if (Phase1 == 0 && Phase2 == 1) {
Phase1 = 0;
Phase2 = 0;
} else if (Phase1 == 1 && Phase2 == 0) {
Phase1 = 1;
Phase2 = 1;
} else if (Phase1 == 1 && Phase2 == 1) {
Phase1 = 0;
Phase2 = 1;
}
//enable the stepper chip
IACDis = 0;

As you can see, i first set the right step (directly configuring the pin outputs), then enabling the IACDis pin, with no troubles in the moving of the stepper... This routine is valid with MoveOnly, AlwaysOn and 15Minutes IAC... If there is no more moviment to do, the iacdis is disabled...
I not think that you need some time between the stepper enable and the relative moviment... Probably the problem is elsewhere...

Whoo, i've tryied to take a look at the ms2 extra code... The little c that i know not help me to understand that!

Post by **muythaibxr** » Fri Oct 26, 2007 5:48 pm

I think it probably depends on the stepper.

Ken

UnaClocker · Post by **UnaClocker** » Fri Oct 26, 2007 7:01 pm

Seems to me, there's a setting in there (I don't have MT handy at the moment) that sets the time per step.. It's very important to have it right for your particular stepper. I know I read that the Chrysler ones take a longer step than Chevy ones.. Maybe that's the case with yours as well?
I say mine doesn't work right, you tell me it's my settings or wiring. Yours doesn't work, you start re-writing code.. hehe.. You sure you set it right?

Just givin ya a hard time.. ;)

masterx81 · Post by **masterx81** » Sat Oct 27, 2007 12:44 am

muythaibxr wrote:I think it probably depends on the stepper.

Ken

Have you already tryied with 15min iac?
As some steppers need to warm up before can go in 'move only'.
I've triyed my code with 3 types of stepper, and on everyone of them the code was working... One of them was also an unipolar wired leaving the center tap unconnected.
Probably if in 'always on' it's working, the step time it's right, but also with move only, while the stepper is moving (like when trying to get the rest position), the iacdis is always 'disabled' (enabling the current at the coil).
while thestepper is moving, i leave the iacdis disabled, and only when i'm sure that there are not more steps to do i enable it... so while is moving, it's acting like an 'always on'. you are doing the same, or every step completed you enable the iacdis then re-disable it for the next step every time?

Post by **muythaibxr** » Sat Oct 27, 2007 8:02 am

UnaClocker wrote:Seems to me, there's a setting in there (I don't have MT handy at the moment) that sets the time per step.. It's very important to have it right for your particular stepper. I know I read that the Chrysler ones take a longer step than Chevy ones.. Maybe that's the case with yours as well?
I say mine doesn't work right, you tell me it's my settings or wiring. Yours doesn't work, you start re-writing code.. hehe.. You sure you set it right? Just givin ya a hard time.. ;)

I tried values between 2.5 and 30 ms. None work reliably unless in the always on mode.

UnaClocker · Post by **UnaClocker** » Sat Oct 27, 2007 8:05 am

When I was using the AIC, seems I always used the 15 minute mode.. But I can see how the "always on" or "moving only" would be necessary with closed loop mode.

Luke · Post by **Luke** » Sat Oct 27, 2007 8:56 am

I'm unable to use this firmware, I get the following error:

ERROR: Stack pointer out of range, e=0xE3

I've installed the release version okay, are there further hardware mods that need to be made?

<edit>

scratch that, I answered "yes" to the are you upgrading from regular MS2 and it went through okay.

Post by **muythaibxr** » Sat Oct 27, 2007 10:15 am

I think my motor is one of those that requires jumpering around the current-limit resistors on the ms2.

Instead I'll just put a heatsink on the stepper driver, and develop the closed-loop code in that mode. I'll probably start working on it today (I'm away from home, but have my laptop) and hopefully test it tonight when I get home.

Ken

masterx81 · Post by **masterx81** » Sat Oct 27, 2007 10:58 am

Forgot to say, all the steppers that i've triyed are running with step times of 6ms! Lower than this i can't get a reliable movement...