Page 1 of 1

Issue on single dizzy odd fire engines (patch included)

Posted: Tue Feb 07, 2012 3:49 pm
by karlo
We discovered an issue with single distributor odd fire engines. Specifically, this is a 90 degree odd fire V6. The code doesn't shorten dwell correctly at high rpms, causing dwell for two separate ignition events to overlap. This obviously causes an entire bank to drop out, in addition to overdwelling the coil. I attached some images to illustrate the problem. The following are the ignition settings:
ign-settings.png
Scoping the ignition output at 4500 rpms shows the spark output getting pulled down correctly and working just fine (potentially too short spark duration here, though):
ign-working.png
Scoping at 5300 rpms shows the overdwell and complete lack of spark for the one bank:
ign-failing.png
The following is a fix for this issue. The code for odd fire distributor mode gets this right, but the code is valid for any odd fire engine with a single coil:

Code: Select all

diff -uri orig/ms2extra_3.2.1_beta/ms2extra/ms2_extra_main.c ms2extra_3.2.1_beta/ms2extra/ms2_extra_main.c
--- orig/ms2extra_3.2.1_beta/ms2extra/ms2_extra_main.c  2011-12-08 16:04:05.000000000 +0100
+++ ms2extra_3.2.1_beta/ms2extra/ms2_extra_main.c       2012-02-07 19:53:53.756230800 +0100
@@ -3276,7 +3276,7 @@
                         /* rotary or DLI force 1 */
                         max_dwl_time = dtpred_local;
                     } else if (flash4.ICIgnOption & 0x8) {
-                        if (spkmode == 2) {
+                        if (num_spk == 1) {
                          // in odd fire set max dwell for shorter period
                             if (dtpred_last_local < dtpred_local) {
                                 max_dwl_time = dtpred_last_local;
Mind you, running a single distributor on a V6 90 degree odd fire engine is basically the same as on a even fire V8. The dwell time at 6500 RPM is < 2.3ms, so I'm not sure it's a great idea. However, a friend ran into this issue, so I put together this fix for him. The code hasn't been tested on the car yet, only scoped the outputs.

- Karl

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Tue Feb 07, 2012 3:53 pm
by jsmcortina
Thanks for the patch.
What code is this against?

James

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Wed Feb 08, 2012 6:26 am
by karlo
It's against ms2extra 3.2.1.

Thanks for the attention!

- Karl

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Wed Feb 08, 2012 6:35 am
by jsmcortina
karlo wrote:It's against ms2extra 3.2.1.
OK, good.
I'll add it to the TODO list. Any change like this requires a lot of testing in case the are unexpected side-effects in the many other modes.

James

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Mon Apr 02, 2012 12:16 pm
by sunfield
Hello James,

I understand this issue will put on your todo list.
When do think this solution will be integrated in a release?
I am interessed in this fix, because I have an odd-fire engine.

Regards

Otto

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Mon Apr 09, 2012 5:16 pm
by jsmcortina
This issue had slipped off the radar. I'll be looking at it in the next few days.

James

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Tue Apr 10, 2012 5:31 am
by jsmcortina
Karlo's fix looks fine. I've tested it on the bench and don't see why it should cause a problem. It is added ready for 3.2.2 which should be soon.

James

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Mon Jun 25, 2012 9:43 am
by ScoobyCoder
James,

To confirm 3.2.2 or above has this issue fixed and can be used on our oddfire engine (which was losing power above around 4k, which I think was exactly this bug... :)

Mark

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Mon Jun 25, 2012 11:58 am
by jsmcortina
Is that a question or a confirmation that the fix works?

James

Re: Issue on single dizzy odd fire engines (patch included)

Posted: Mon Jun 25, 2012 1:31 pm
by ScoobyCoder
james,

Was a question, but you answered via PM. I will check tomorrow if it works and report results after we get booked on the rollers towards the end of the week.

Thank you,
Mark