pengaru

No plans for a hawt boy version at this time, there are plans however for fully customizable 'clusters' using an xml-based gauge cluster schema I have yet to implement... so you could tweak the cluster editing a file that vaguely resembles what you'd edit to work on a web site (html) but for gauges. This is a bit down the road however.





I have a new image up, this one works much better: (right click - save as, just like the first post)

http://pengaru.com/~swivel/gdemo2.img



I have fixed the bug in the serial communications code so moving throttle now affects the TPS gauge, other gauges should be functional also. The TPS is supposed to be calibrated but after my last tests with this version it looks like it zero's properly but the 100% is incorrect, I'll be fixing that when I get home. Some other bugs like memory leaks and rendering bugs have been fixed too, next on the rendering bugs list is hatch-overwriting (you'll notice some of the hatches look like they arent straight, it's because a small hatch and a large hatch are both being drawn at the same place... I just have to implement some logic in the hatch drawing so they don't try to interfere...) and font metrics, the letters dance up and down a bit in words.. the visual quality hasnt been worked on since the early days of the project, most of the time has been spent figuring out the e6k since I made the initial gauge code. Expect visual quality to improve _significantly_ before final release.



Latency is still relatively high, as this is still a debugging version, but it's not that bad really. If you have > 1 bar map sensor the vacuum gauge will be wrong, I will implement the map sensor type detection and gauge-tweaks according to map sensor type soon... so if you have a 2 bar or 3 bar map sensor it will show vacuum in one direction and boost in the other. for now it's just vacuum and will be incorrect without a 1 bar sensor.



The instructions in my first post still apply for this new image, you must rawrite it to floppy .. have e6k connected and on when the software boots up etc. One change as far as the instructions go is you no longer have to hit enter at LILO:, but you still have to hit enter at 'VFS: insert root filesystem disk and press enter' prompt...



Hopefully turborotor can try this out and maybe some others? I know there are more people on this forum who have e6k's, cmon guys blow the dust off those floppy disks!



Thanks guys, good luck with the image if you try it.

pengaru

updated version, last one for a little while, give it a go if you're bored



(right click - save as, then use rawrite to create floppy as explained in 1st post)

http://pengaru.com/downloads/gauges_demo/predemo.img



this has throttle calibration routines corrected, otherwise identical to previous one.



any feedback appreciated.

turborotor

I'll test it in an hour.

Rob x-7

I wanted to check it out, but im a retart and cant/dont want to

figure it out, lol



it sounds nice though?

pengaru

It's ok rob



I don't think you have an E6K anyways? It's not really useful without one, though I will be adding E6X support soon... already have one for testing just waiting on a connector for it.

Mazderati

That looks like some seriously neat stuff.



Just to clarify, this is a program that creates a GUI/display for users of the Haltech E6K; literally, electronic guages?



I can appreciate the coding involved as I delve into my first programming class. :]

pengaru

Yes, these are electronic gauges. They are simulating analog gauges in realtime using the haltech e6k as the data source.



This is just the beginning, there will be many cool features added, especially in the full (pay) version. My car will be running this for gauges exclusively on a lcd panel replacing the normal gauge cluster.... this of course requires more inputs than the Haltech E6K provides (like fuel level and oil pressure...) but thats no big deal.

Rob x-7

I just wanted to check out the set up, I know its been

planned for a while.

turborotor

Much better, here are the few things wrong that i notice.



The vacumm was off by at least 10 units. I was pulling about 16 on the gauge and it was registering 25 as you can see.



The coolant temp was off by 9 degrees. I was running cooler on the guage.



I can't verify the AIT cause of no gauge. But todays weather is around 55 and is saying 31.



Not sure about A/F cuase the exhaust was still cold.



The rpm seem to be on the spot but i can't verify around the range cause is too late to rev the car in the apt complex.



The throttle position seem to be right on the money.



Are you going to add a boost guage?



Here is a pic.

pengaru

looks good



The vac gauge was expected to be off since you don't have a 1 bar map sensors (boosted), this will be corrected in the next image I put out... so to answer your question, yes, there will be a boost gauge... the vacuum gauge already there will be changed to show vacuum and boost when you have > 1 bar map sensor, it will just detect it when you run it.



looks like the TPS was calibrated properly too.



As I stated in the original post coolant and iat are being simply scaled for now, the output of thermistors is not linear so it requires a look up table to get back from the voltage to a temperature. This is why haltech tells you to use their sensors, they have the look up tables in their dos software setup for just the haltech/gm sensors. I will make it possible to select the look up tables, for now the software simply maps the 0-255 range for the sensors output to -40 - 260 which results in erroneous temp #'s. I personally need this functionality badly as I'm using the mazda coolant sensor, it works just fine.





Thanks turborotor for wasting some time on this, I owe you one.