MS2/Extra 2.0 beta 14 (20071020)

Post by **muythaibxr** » Sat Oct 27, 2007 11:10 am

I tried several times...

All the times I tried worked with always on.

In any case, no big deal, I can get it to work the way I want.

Ken

woh · Post by **woh** » Sat Oct 27, 2007 12:11 pm

in file megasquirt-ii.ini.ms2extra under [Output Channels]; variable CheckSum and adc6 have the same memory location (141). Is that a problem? It is for documentation purposes.

Post by **jsmcortina** » Sat Oct 27, 2007 12:40 pm

woh wrote:in file megasquirt-ii.ini.ms2extra under [Output Channels]; variable CheckSum and adc6 have the same memory location (141). Is that a problem? It is for documentation purposes.

It is certainly an error! I'll take a look.

James

Philip Lochner · Post by **Philip Lochner** » Mon Oct 29, 2007 6:15 am

Does MS2/Extra support Dual WBO and Cont Baro simultaneously yet?

Post by **jsmcortina** » Mon Oct 29, 2007 1:41 pm

Philip Lochner wrote:Does MS2/Extra support Dual WBO and Cont Baro simultaneously yet?

Yes, it has done for quite a while.

James

Philip Lochner · Post by **Philip Lochner** » Tue Oct 30, 2007 8:37 am

jsmcortina wrote:
Philip Lochner wrote:Does MS2/Extra support Dual WBO and Cont Baro simultaneously yet?
Yes, it has done for quite a while.

James

That is very good news, I have to try it! Please could you just tell me where I would find the instructions as to how to wire it all - I've looked but I am blond.

woh · Post by **woh** » Tue Oct 30, 2007 9:26 am

I've been updating the documentation for RS232 and ran across these. Not sure if these are bugs;

1. When I get a table (e.g.VE1) using the 'r' command no data is returned (0 bytes) when I first go to a new page. After the first time it always works even if the table changes (but not the page). Is that a bug? Or am I expected to add a delay after the first three bytes of the command every time I change the page? It works fine with a delay after the first three bytes. I'd like to see the firmware keep track of page changes and add a delay. That would make the command definition much cleaner.

2. command "c" returns a single byte. It appears to be the 'high' byte of the variable 'seconds'. Is that as it should be? I expected the two byte variable 'seconds' or maybe a single byte that counts in seconds.

3. Should there be a "C" (capitalized) command?

4. The "A'" command works just fine returning 152 bytes. The "a\x00\x06" command returns 113 bytes. Does MS2/Exta support this 'a' command as an alternative to the "A" command? Or should I delete this in the documentation as an alternative to the "A" command.

Post by **jsmcortina** » Tue Oct 30, 2007 3:47 pm

woh wrote:I've been updating the documentation for RS232 and ran across these. Not sure if these are bugs;

1. When I get a table (e.g.VE1) using the 'r' command no data is returned (0 bytes) when I first go to a new page. After the first time it always works even if the table changes (but not the page). Is that a bug? Or am I expected to add a delay after the first three bytes of the command every time I change the page? It works fine with a delay after the first three bytes. I'd like to see the firmware keep track of page changes and add a delay. That would make the command definition much cleaner.

Could you post an annotated comm.log of what you describe.

2. command "c" returns a single byte. It appears to be the 'high' byte of the variable 'seconds'. Is that as it should be? I expected the two byte variable 'seconds' or maybe a single byte that counts in seconds.

3. Should there be a "C" (capitalized) command?

This is not present in any MS2 code. "C" only exists on MS1. We followed suit.

4. The "A'" command works just fine returning 152 bytes. The "a\x00\x06" command returns 113 bytes. Does MS2/Exta support this 'a' command as an alternative to the "A" command? Or should I delete this in the documentation as an alternative to the "A" command.

The "a\x00\x06" command is preserved in a failed attempt to support MV2.

mops · Post by **mops** » Tue Oct 30, 2007 7:07 pm

Hi guys.

Havent posted in a while, been busy with other thing, didnt really had time to play with my car or ms for tha matter.

anyways, did this AE feature made it yet to the code ? the one that shots extra squirts during ae, or rather for the durations of AE it changes to more squirts per cycle and then returns to what it was set for normal running.

thanks,
Mops.

P.S. with shame I admit that i'm still running B&G 2.687t5 code in my car, for like last 6 months, but i definitely wast to try out ms2/x asap. I'll be rebuilding engine soon aswell and got all sorts of new goodies to put on

M52 POWER! · Post by **M52 POWER!** » Wed Oct 31, 2007 5:06 am

I was experimenting with 6 squirts per cycle on my straight 6 the other day. In order to get it to idle and even show an AFR I had to set my req_fuel as low ass possible and my VE at idle from 45 to 8!!! Otherwise it was way way too rich.

Is this a bug in the firmware or am I not setting something up correctly. I am using brand new bosch 24# injectors.

Philip Lochner · Post by **Philip Lochner** » Wed Oct 31, 2007 5:22 am

M52 POWER! wrote:I was experimenting with 6 squirts per cycle on my straight 6 the other day. In order to get it to idle and even show an AFR I had to set my req_fuel as low ass possible and my VE at idle from 45 to 8!!! Otherwise it was way way too rich.

Is this a bug in the firmware or am I not setting something up correctly. I am using brand new bosch 24# injectors.

Based on my understanding, what you found is to be expected. Consider that the same amount of fuel must now be injected using 6 squirts as what previously was done most likely with 2. This means that the pw calculation should yield much smaller results, achieved by setting your req_fuel low and VE low.

But why would you want 6 squirts per cycle anyway, other than fooling around?

6040solder · Post by **6040solder** » Wed Oct 31, 2007 5:44 am

it probably means that the injector opening time is set very wrongly. the short pulsewidths make the problem much worse than with smaller injectors.

when opening time is set correctly, you should be able to switch back and forth between squirts with little to no change.

Philip Lochner · Post by **Philip Lochner** » Wed Oct 31, 2007 7:04 am

Philip Lochner wrote:Based on my understanding, what you found is to be expected. Consider that the same amount of fuel must now be injected using 6 squirts as what previously was done most likely with 2. This means that the pw calculation should yield much smaller results, achieved by setting your req_fuel low and VE low.

But why would you want 6 squirts per cycle anyway, other than fooling around?

Sometimes I regret opening my mouth!

6040solder is right, if your opening time is close to actual, you should have little difference between eg 6 squirts and 2 squirts. However, I find it hard to understand how opening time alone could account for such drastic differences in your fuelling parameters.

6040solder · Post by **6040solder** » Wed Oct 31, 2007 7:28 am

say you need 1ms opening time, and you have it tuned at 0.8ms opening time, your overall pulsewidth is 2ms to get your chosen afr with the incorrect opening time, so your actual fuel flow is for 1ms but the ecu thinks its 1.2ms

you are on one squirt

you change to 2 squirts and it gives each of them 0.6ms of squirt and 0.8 of opening time. this is 1.4 total each. because its actually supposed to be 1ms opening, you end up with only 0.4ms of fuel flow per squirt, and 0.8 total.

now change to four squirts and you get 1.2/4 = 0.3ms per event on top of the 0.8ms opening time minus the 1.0 actual opening and you are left with only 0.4ms total fuel flow.

in effect your initial tune was wrong because the opening time was wrong, but it worked until you changed squirts and ended up with 66% of the fuel, and then 33% of the fuel. obviously if you have it too small the opposite will occur.

i hope that helps :-)

6040solder · Post by **6040solder** » Wed Oct 31, 2007 7:30 am

thats actually wrong, but you get the idea. (woops)

it should be 80% and 40% respectively.

Post by **muythaibxr** » Wed Oct 31, 2007 8:06 am

M52 POWER! wrote:I was experimenting with 6 squirts per cycle on my straight 6 the other day. In order to get it to idle and even show an AFR I had to set my req_fuel as low ass possible and my VE at idle from 45 to 8!!! Otherwise it was way way too rich.

Is this a bug in the firmware or am I not setting something up correctly. I am using brand new bosch 24# injectors.

Which firmware? Alternating or simultaneous?

I've switched between all of the various numbers of squirts with no problem so there might be something else going on.

Ken

Post by **muythaibxr** » Wed Oct 31, 2007 8:06 am

mops wrote: anyways, did this AE feature made it yet to the code ? the one that shots extra squirts during ae, or rather for the durations of AE it changes to more squirts per cycle and then returns to what it was set for normal running.

That feature has been in and tested for quite some time now. So yes it's in

Ken

Post by **muythaibxr** » Wed Oct 31, 2007 8:13 am

Philip Lochner wrote: But why would you want 6 squirts per cycle anyway, other than fooling around?

Better throttle response is the main reason to do this.

I run 4 squirts alternating or 2 squirts simultaneous (with EAE lag compensation going to 4 squirts during accel) because the throttle response was *much* better that way. This will hold us over at least until we can do sequential injection... which will give the best throttle response (and is the best reason for going to sequential).

IF your injectors are too big or too small then you won't be able to just go to the maximum number of squirts for your engine so I added the lag compensation feature, which automatically goes to the maximum number of squirts for your engine when tpsdot or mapdot go over a certain threshold, then goes back to the normal number of squirts otherwise.

I also have some other features in mind to help throttle response, such as if you're mid-squirt when an accel event occurs, extend that squirt to the new value so that cylinders on those injectors will get close to the right amount of fuel that much sooner. I also plan on doing some MAP prediction using manifold volume, engine volume and throttle position along with an Alpha-N table (tuned properly). This should also help throttle response on certain engines.

Ken

woh · Post by **woh** » Wed Oct 31, 2007 10:29 am

This is not present in any MS2 code. "C" only exists on MS1. We followed suit.

OK, understood.

Could you post an annotated comm.log of what you describe.

Yes, I will when I get back home. Likely on Friday.
The basic problem is that after the first three bytes are sent the firmware does not appear to implement the remaining command if a new page is required. Two fixes on my end are adding a delay in the middle of the command or repeatedly sending the complete command until data is received.

The "a\x00\x06" command is preserved in a failed attempt to support MV2.

Will the "a\x00\x06" command survive long term? If it survives, will it be the same as the "A" command long term or only return a limited number of bytes?

If it is a failed attempt to support MV2, How about deleting it so there is no confusion with MS2std, or just making it the same as "A" (ie the variables).

M52 POWER! · Post by **M52 POWER!** » Wed Oct 31, 2007 11:20 am

How do I figure out my injector opening time. I had it set at 1ms a while ago and Jon suggested I change it to 1.3 ms and I did. Had to retune it ofcourse and now I'm experimenting with multiple squirts.

Is there a resource for bosch injectors that would tell me?