Are there any plans to add this to the official build? I saw Andrey has had this in his custom build for a long time now. I will be using ID injectors, which come with the deadtime specs.
http://www.youtube.com/watch?v=SbPqbvMfC6k
I second this to be done as soon as possible.
In the vemstune, there is a section for "firmware parameters", but the "dead time' setting is not highlighted/able to be changed. We should add this to the wishlist for new firmware.
looks like a great idea - would make the setup a little easier to understand i think ;D
Please post these videos in the video section, okay its in Russian, but we have visitors from .ru and its handy to have all the videos that we can find that are VEMS related in the one place.
The video is useless as far as being a tutorial to anyone who doesn't have his custom build :( It will become useful if the inj deadtime table is ever implemented. He does have an awesome video for using autotune to set your VE table from scratch but it's Russian, they'd probably see it from him first. I was thinking to download it and reupload it subbed in English.
That translation would be nice ;)
Hello
I did also look at that video and I think its must already be implemented.
I did following
Open the config file in notepad
Look for row
CompileOptions: 0000005F
Change it to
CompileOptions: 00000060
And then I have this option in injector settings.
I don’t know if this is the right thing to do no warranties for the result.
Micke K
How did you figure that out?! I wonder if it actually does what it seems. Any developers care to comment?
You need to have a firmware compiled with this feature to make it work. The CompileOptions only reveals what flags were used with the current firmware.
It might make it into the current branch. But honestly I have very little problems with the methods used today, the downside is that it might be harder to understand.
So if my injectors are supplied with deadtimes specified for each voltage, how can I use that with the current method?
It would be nice if there was just a dropdown switch to use either method if you feel the current way is more usable.
All this is also highly dependant on the flyback control solution used.
If you use unknown/new injectors with Autronic and they are not in the tuning software database, you have to send them in for testing to get their characteristics tested (they do this for free). They have no use for any information on the injectors, they test and create their own data for their electronics.
I usually experiment and start out with 0 ms rampup and open time, and increase rampup if I see a problem with the shape of the VE table and the general behaviour. It's not highly scientific, but spending countless hours trying to find the optimal setting that works outside of 13.5 Volt operating voltage is just wasted IMO. The smaller the injectors get the less time I spend thinking about it.
Thanks for the reply!
Does flyback affect injector opening? I mean the deadtime is primarily for the opening characteristics right?
Well these are 1000cc injectors and they come deadtime-matched, with the values defined from the manufacturer, so I would like to use them straight up without experimenting. Do you think it would be possible for the method to be configurable like I mentioned?
Current settings is absolutely user unfriendly. 99% of users use crappy settings.
And: F1 gives wrong/incomplete description about values.
And: even in wiki we can't find definitive description.
And: current method do not allow use non-linear voltage compensation. - way how to make tune stable at usual voltages, and at same time little protect engine when voltage drops catastrophically.
Mattias, has deadtime table method some downsides?
Gints
No, I don't think it has a downside.
What I meant with deadtime + flyback, opening and closing times are still a complex game. Changing the scheme will still require some brain-activity from the user.
Trouble is, we've seen some horrors comming off dynos by "professional" tuners, case in point that BMW with the 120% warm-up enrichment at operating temperature, and some others that I will not mention.
Giving people more features than is strictly needed can damage the reputation of the system even further - I can prove that peole claim that its never "their fault" and blame the "magic" ECU.
I think a standard set of defaults would work well, with an option to unlock the Expert features inside the ECU - things like injector characteristics and wideband controller PID settings should be set with good defaults, then hidden out of the way.
But then by the same token, we should just hide all configurations, and have an installation wizard instead of vemstune, like Microsoft. It asks you for the year, make and model of your car, loads the defaults and that's it. That way no one will ever configure their fuel injector output as the fuel pump and burn down their house, and damage the system's reputation as being a fire hazard.
I'm in software and I work for a major corporate wireless lan player. I cringe whenever I hear that we should make something easy to use. It's a half truth. A complicated product doesn't have to be easy to use. There are complicated things in life - if someone doesn't get it then that's just too bad. Now, making something more complicated than necessary is a different matter. It just needs to be as easy to use as possible - which might still be difficult or for some, even incomprehensible for some others. I'm no EFI expert, not by a long shot, but it seems the deadtime table is a pretty standard thing. We could just add a line in the help menu that says "If you don't know what this does, don't touch it".
I think it's more important to have a good explanation for the offered configuration (as Gints mentioned, many of the current F1 popups are either sparse, wrong or non-existent). But for this, I would be happy if there was any way to configure it, even if that means manually editing the config, as long as the firmware would know what to do.
For what it's worth, I'm impressed as all hell with how mature VEMSTune is as a product, for one as young as it is with such limited development resources (which isn't meant as a knock against the developers - they're just not a full-time army of managers, architects and developers, the way a corporate outfit would be).
Rob, in deadtime table we have 5 straightforward values. Now we have 4 absolutely non-understandable values for any tuner!
we still have possibility to give some suggested values. like:
8V 2000
10V 1100
12V 850
14V 550
16V 350us
In addition current suggested values are wrong.
4th
Bump - any word if this will be implemented/integrated?
For all interested, I just heard back that this will indeed be implemented, with a selectable strategy (so the old way will still be available). So cool!
Quote from: BigD on March 30, 2011, 03:28:41 PM
For all interested, I just heard back that this will indeed be implemented, with a selectable strategy (so the old way will still be available). So cool!
;D