DIYEFI.org Forum

Hi All FreeEMS Brothers (and Sisters as well) !

I have a Datsun 280Z (6-cyl) with a FreeEMS all wired and connected. FreeEMS has a recent (early December 2013) firmware in it.
Mitsubish early 90 CAS (cam angle sensor) unit, Toyota / Denso Igniter unit to drive 3 (Land Rover) coils.
Injectors, Plenum and fuel rail are all stock Datsun.
Fred was visiting me end of last year, and we wired all up together - Fred has reviewed and "approved" everything.

Now to problems ...
While FreeEMS, Toyota Igniter and Coils were proven to give good strong spark (at plugs) in FreeEMS test mode, the real EFI
mode only gives an occasional spark, and logs show sync loss.

If anybody has ideas on how to proceed with my FreeEMS sync losses (or has any other input related to this situation).
I'm happy to give more details on my setup. I'm attaching the EMStudio logs from the cranking trials (with sporadic spark).
BTW: What are the time units on OpelLogViewer X axis ?

Note: The car was running well and smooth with earlier Megasquirt (however with traditional distributor and single coild), so plugs, plug wires, cylinder seal, fuel pump and other surrounding hardware that did not change (with swap to Jaguar FreeEMS board) should be up to the task.
Last summer I even taught my early-teen kids to drive stick tranny car with this earlier setup !

With Best Regards,
Olli

Olli! Welcome! Super glad to see this thread here after a couple of days away on the high seas! :-) Thank you!

Answers for questions:

I didn't take a look in your zip flie, however from the two logs you gave me, the small one had nothing distinguishable in it, and the big one was of the benchtest decoder, not the 4/2 decoder that you need. So to solve this, I would encourage you to review the metadata window in EMStudio before attempting any cranking etc as it's possible the car is still on that BenchTest firmware, in which case it can not run without first replacing it.

If/when you get proper firmware back in, and still have sync issues, follow this thread written for via and yourself a week or so ago: viewtopic.php?f=54&t=2369

OLV X axis is data points, so with a smooth stream that's about 80 times per second or 1/80th of a second intervals. Within the data is an accurate clock or two, and you can look at those to find exact differences and gaps etc, if you need to. Typically you won't need that, though.

I know I sat and helped build your branch on your machines, however it's be useful for me (and others) if you had that branch that you're trying to use for the car pushed to github so we can match log files to firmware revisions and suggest configuration changes based on that. Maybe I made a mistake/gave bad advice, who knows :-)

Fred.

Hi Fred and thanks for coming to help again !
I just committed the (config) file we changed to local Git and pushed the whole codebase to GitHub (also checked the change is there).
That should tell the complete "state" of the firmware we used.

I recall after we tried the BenchTest mode we compiled a more actual version (fuel disabled ? or maybe we did a HW based fuel disable).
Thanks for info on OLV X-axis units, makes sense now.

Is there a way to track which binaries (and when) we sent (uploaded) to FreeEMS if that granularity is needed ?

Meanwhile I'll study the config variables you mentioned, related to sync loss.

Olli

Olli! :-) We may well have disabled fuel in some cunning way, I don't recall. Nevertheless, the log you sent to me was of benchtest firmware, and thus not useful or relevant.

If you use EMStudio to log your data, the firmware build information is embedded within, and if you commit your changes before building and uploading, then that will match something in git and give us traceability, which we need. And yeah, I do need it :-)

It's late right now, so I won't be able to take a look at the config branch we did together tonight, however I will leave this tab open and take a look tomorrow and post again.

Thanks for playing along and making life easy :-)

If you're comfortable with it, maybe I could build you a binary from those sources, and you could download that from the internet and load it to your device? Then you could take a fresh log of THAT firmware cranking and we could take a look at THAT log and go from there? Just a brief 15 seconds or so of cranking should be OK to illustrate the issue or lack thereof, I guess. Thoughts? If the settings are aggressive still in your branch, we could tone them down for your next attempt before you do it, too.

Fred.

Hi Fred !
I'd really appreciate you building a binary with "best possible" (sync loss) settings to get 280Z started up.
While still experimenting CAS reference location I can disable injectors with means other than sw.
Hope you can give me FreeEMS Git version (hash), which will (through diff) highlight the changes
(so I see what all changed, by how much and learn about settings surrounding the sync loss).
Settings in my current branch as as-is form where we left off.

Hope you're doing ok in your "Home sweet home" !

Olli

No worries, Olli! Happy to provide you with git difftool output to satisfy your educatory needs :-D Just have to get around to doing it. Sorry for overly large delay. I've been dealing with an undue amount of bullshit recently, however that's now on the up/down so I'm happy enough. Yes, home sweet home is sweet indeed. Fresh fish, the best possible meat, fresh air, clean sand, clean ocean, etc, satisfy my minimum requirements for life.

Firmware S19 with easier-starting changes and fueling re-enabled: http://tmp.fredcooke.com/FreeEMS-0.2.0- ... nd2.s19.gz

commit 4a0677cf432d9d784b847d8c6c3661cf2f5268ca
Author: Fred Cooke <fred.cooke@gmail.com>
Date: Mon Feb 10 13:05:18 2014 +1300

Re-enable fueling in software, allow Olli to disable it physically instead.

commit aed19ce86b5cd9448a3ec8d59c2cc67b711b2f7e
Author: Fred Cooke <fred.cooke@gmail.com>
Date: Mon Feb 10 12:53:17 2014 +1300

Increase time tolerance to allow a bit more variation while cranking and reduce the number of confirmations after a sync loss, too.

Source code config pushed as an extension of what you pushed: https://github.com/fredcooke/freeems-te ... llisconfig

Detailed steps:

• <disable fueling in hardware via unplugging injectors or removing relay for them>
• git remote freeems-temp https://github.com/fredcooke/freeems-temp.git
• git fetch freeems-temp
• git merge --ff-only freeems-temp/ollisconfig
• git difftool 0c024e681a4e3e6cd2e47d9e4833fe964dad809f..4a0677cf432d9d784b847d8c6c3661cf2f5268ca
• wget http://tmp.fredcooke.com/FreeEMS-0.2.0- ... nd2.s19.gz
• gunzip FreeEMS-0.2.0-SNAPSHOT-232-g4a0677c-OLLIDATSUN-MitsiAndMazda-CAS-4and2.s19.gz
• FreeEMS_Loader
• <add jumper to device and turn on key>
• <select file with correct name and load it>
• <remove jumper and cycle key, observe fuel pump coming on>
• <kill the loader app>
• <launch emstudio and verify connection (don't ask me how, I don't remember how you set it up)>
• <attempt cranking and timing light checking with a trusty assistant>
• <bzip2 the log and attach to a post here with description of your findings>

If you think I forgot something, ask! Otherwise those instructions should get you a firmware more likely to allow you to rough-set the timing.

Fred.

Congratulations.

Not yet, mate, it didn't run quite yet :-) Soon though, with a little luck :-)