The new beta versions sure look much better!
Printable View
The new beta versions sure look much better!
thanks just dont get windows 10 !!!
seriously i have windows 10 in a VM now, i'll be testing with it. so far i dont know why everyone is having so much trouble with it. works fine for me
i'd really be interested to see if the gear selector and TCC lockup thing is working in the new beta.
the gear selector in beta 3-4 didn't actually update itself (just an oversight), so if you did something else (changed timing or whatever) it probably would have kicked into the selected gear.
i wish i had an auto to test with.
edit: i just found my REAL mistake with the gear selector. i'll fix it shortly.
Steve just a thought, I wouldn't spend any time optimizing that interface for tablets, etc. Your gifting and interest is in the LT1 interface. The trick is to find someone who's interested in working GUI development and let them optimize that side of the code. Any local colleges teaching courses on that stuff who have students who'd be interested in doing this as a project?
Sent you an email directly about Beta7.
There's no need for anyone to get their panties in a knot about Windows 10 tablets. A W10 tablet is basically a PC that has a touch screen instead of or as well as a mouse and keyboard. It's the exact same OS.
QT's info was rather sparse when doing a quick look besides it being libraries for some type of use. What you're saying is that QT is similar to using Visual Studio, but it's currently behind on supporting Windows 10. They'll probably officially support W10 before too long.
What resolution are you developing for? 16x10 ratio WXGA or 16x9 ratio 1366x768 are typical lower end laptop and tablet resolutions.
well i got the problems now too. every since beta 4 eehack wont show, it acts like wants then nothing. windows 10 user here preview build 10586. Who all is in the windows insider preview program?
back to beta 3 for the time being
oh i could careless about the touchscreen stuff too, Do what you do best Steveo Hack Away!
i'm not getting my panties in a knot. but interfaces designed for a mouse may not work well with a touchscreen.. and this is designed for a mouse. for example good luck using my sliders with a touchscreen. might work alright but i doubt it.Quote:
There's no need for anyone to get their panties in a knot about Windows 10 tablets. A W10 tablet is basically a PC that has a touch screen instead of or as well as a mouse and keyboard. It's the exact same OS.
http://doc.qt.io/qt-5/classes.html
i chose it not just because i hate writing gui code, but also because i'd never written with QT before and i figured now would be a good chance to try it out. it's worked well enough
my ui is 800x600 fixed, just so it'll fit on whatever crappy netbook without me having to worry about dynamically re-sizing anything.
I've been using both HPTuners and Tunerpro on the tablet with the touchscreen. It works well enough. 800x600 is an OK resolution.
are you sure you're running the right executable and library set? remove any 'compatibility' settings?
here's a clean copy, beta 8, i fixed the installer up a bit and made a few minor changes, http://fbodytech.com/download/223/
just to be sure i started fresh, uninstalled everything, installed it on windows 10, and it fired right up.
remove c:\program files\eehack, then use the installer, and make sure you're running only this current version and it's associated libraries....
edit: make sure you dont have any antivirus getting in the way too, maybe temporarily disable it or whatever?
ok so if i start task manager and then try to run eehack i can see it start in the process tab then quits. it shows 1.5 mb memory use beta 7
Attachment 10059
give me a min and ill try 8
:rockon:Success!! with beta 8 :rockon:Attachment 10060
awesome!
so at the other end of the spectrum from our fancy new windows 10 machines, i just tried the new beta on my netbook - a relatively old and slow intel ATOM with xp sp3), and although it worked, things did not go well.
it seems qt5.5's timers are different than qt5.4 somehow (or maybe the serial interface is faster or slower?) first thing i noticed is that i seem to have lost a few records per second of throughput, which speaks to some kind of timing issue.
then flashing, it failed during the erase stage, which is a horrifying thing, but luckily my flash routine kicked into recovery mode, increased serial delays, re-executed the erase program, and managed to salvage the situation. still doesn't explain why it happened in the first place.
i'd be interested to see if people with faster computers have this problem with the new beta (which uses newer QT libraries)
i might have to roll back to QT5.4 but considering the windows 10 issues, i'd rather stick with the current branch..
I ran the 4.0 Beta 8 installer, then asked it to run eehack.exe. It says
"Unable to execute file: C:\Program Files(x86)\EEHack\eehack.exe
CreateProcess failed; code 740.
The requested operation requires elevation."
UPDATE seconds later - I turned off "compatibility mode" and "run as administrator" and then clicked on eehack.exe. It started right up. I'm going to try to connect to a car tonight and try the interface.
UPDATE 2: GOOD NEWS: Key on / engine off (test car is loud and it was late):
- EEhack 4.0 Beta 8 starts instantly
- Connection works well.
- Fast mode gets me to about 13.7 records/sec.
- I tried all the auto and TCC control buttons and I could hear all the solenoids clicking appropriately in the pan. I would say this functionality works, though I was unable to test what happens when I return to "auto" mode.
Nice work Steveo!
it's only happening because it's trying to launch eehack (because you left the 'launch eehack' button checked in the installer)Quote:
CreateProcess failed; code 740.
The requested operation requires elevation."
... and windows 10 is too braindamaged to maintain backwards compatibility and sane default security settings. god knows why it's not ok for one process to launch another process with normal windows system calls.
it does work in previous versions of windows. i can remove the option for release and users can just click the icon themselves after.
edit: i did google a workaround though, i'll try it out.
workaround seems to work, and i'll leave the box unchecked by default for good measure.
thanks so much, that's such a relief it's working and you're getting good data rates! at least it's just my old winxp laptop having serial issues.
one thing i'd like if you don't mind, even without a driving test, if you'd see if the gear selection DISPLAY matches the gear selection BUTTONS. im only like 50% sure i got the solenoid combinations correct as they may be stored in a different order than the display bits. im totally in the dark on this automatic trans stuff.
please also play around and see if you can break it in windows 10
Will do. The gear indicators seemed to work correctly, at least I didn't notice anything out of whack, and I would have noticed that stuff pretty quickly given how familiar I am with the 4L60E.
BTW, sent you TWO additional emails on the injectors with links.
BTW my blue screens of death in my earlier tests with 4.0 beta 3 were likely due to my serial adapter, a tripplite USA19HS. Later attempts resulted in immediate BSOD so I repaired the driver installation and no more problems.
Changing gears works, manual engagement and disengagement of torque converter does not.
Wideband looks much better with smoothing, though I appear to have a slight lag, it appears that it changes just after the gauge. Probably not enough to matter though.
Temp readings appear much more sane. I was seeing a max temp of 145F on my commute to work before the update, now it's showing 170F, which seems more likely with a 160F t-stat. Haven't compared temps to datamaster. I haven't even used datamaster since I started logging with eehack, though I guess I should have done a comparison when the temps seemed off.
On a side note, when testing the gear selection I accidentally left it in fourth at a stop light, 4th gear start was very sluggish. LOL
I was shocked that it didn't buck or stall.
i'll look into it. glad the gear selector works, though.Quote:
Changing gears works, manual engagement and disengagement of torque converter does not.
yeah we had to give up a little bit of response time. how the 'smoothing' really works is averaging the last 8 records or so together. if i made it a weighted average it'd be a bit less 'smooth' and a bit more 'responsive' but i think i'll leave it how it is.Quote:
Wideband looks much better with smoothing, though I appear to have a slight lag, it appears that it changes just after the gauge. Probably not enough to matter though.
yeah that was a big blunder on my part, mis-reading disassembly. shoulda just trusted the datastream file.Quote:
Temp readings appear much more sane.
That's interesting that the TCC didn't work for you Brian - I could clearly hear a solenoid clicking and unclicking when I tested that last night. What's your application? On the '95s, there's TWO solenoids that work on TCC lock, one controlling on/off and the other controlling the apply rate; I'm wondering if this mode works both of them?
oh there's actually two solenoids?Quote:
I'm wondering if this mode works both of them?
i see my mistake now.
i finally got my sockets, so im going to heat gun this t-side chip off at work today and get my test ecm working again.
im getting sick of saying new beta
but here
NEW BETA
http://fbodytech.com/download/223/
might fix TCC lockup, please try my new cylinder drop test, also the installer is better, should launch on windows 10, also installs visual c++ libraries, no dependencies whatsoever afaik.
please uninstall all old versions before installing this one (or delete c:\program files\eehack), just so i can make sure it works as-is without any stale libraries in place
I have a 95 trans in a 94 car. The trans has two solenoids, one's a pwm lockup solenoid i believe. that valve is staked open and the solenoid is not hooked up, since there's no wire from my pcm
I need to do the same, have the sockets and air solder gun, but no time to get it done. it's a little scary to do it to a working pcm though, may wait til I brick it.
I've flashed about 10 times using v4.00 with no issues! Must faster using the bin compare action. Loving this thing and I know there are many people here that appreciate the hard work. I just wish I had some wide bands so I wouldn't be tuning blind so to speak.
thanks, this has totally spiraled out of control from a simple mode4 controller to an entire lt1 aldl suite.. im glad someone other than me finds it usefulQuote:
Loving this thing and I know there are many people here that appreciate the hard work. I just wish I had some wide bands so I wouldn't be tuning blind so to speak.
i tuned mine in speed density across the board the board first with narrowbands, then i switched to open loop and polished it up by feel. when i finally got widebands, i found i was only off by about 2-3% in some areas (which is even within the margin of error for a wideband sometimes)
maf narrowband tuning is stupid easy, since you're tuning two dimensionally. you just have to nail your trims per airflow, and your load (MAP) axis will pretty much line itself up, at least close enough to get you going. only downside is you can't push WOT AFR close to the limit.
you can tune very well with narrowbands if closed loop is working properly, especially with my analyzer, simply disable power enrichment, enrichen high end airflow by like 5% for safety, and just start logging. the analyzer will spit out a new maf table for you. that'll get you close.
problem is, aluminum headed LT1s want to run around 13.1-13.3 AFR for max power, but lets say you're off by 5% in your high map airflow, you're ending up with a 14.0:1 or worse, which starts doing nasty things at really high load and RPM.
so i'd target like 12.5:1 in power enrichment and hope for the best, you might leave a bit of horsepower on the table. with a wideband you can more safely push it into the low 13s.
my power enrichment analyzer is rigged for narrowbands too, it wont tell you AFR, but if you exceed 14.5:1 in power enrichment it'll let you'll know...
i would also like to take the time and thank everyone that's donated me anywhere between 2 bucks and 50 bucks and/or random parts to the fbodytech/eehack cause in the last couple months,
fbodytech's 'paypal beer n parts fund' assets now top out just over a hundred bucks sitting in my paypal account, a spare ecm, a nice dual wideband velcro'd to my dash, and all sorts of random sensors and junk in my closet.
just goes to show, the lt1 community definitely takes care of itself in the end if someone puts in the time
i'd like to pay it forward and throw a few bucks towards supporting this forum's hosting bills, since eehack/eex wouldn't be possible without it, does anyone know who pays the bills for this forum nowadays?
manual torque converter engagement works. TCC indicator is backwards when selected manually, but correct when on auto.
that shoulda been fixed in the latest beta..? maybe i reversed my already accidently reversed reversal or whatever. programming is hard.Quote:
TCC indicator is backwards when selected manually, but correct when on auto.
see you're taking advantage of the beer money...
lmao j/k
latest beta 4 tested this evening
If anyone hasn't donated something to Steveo, now's the time. There's an ebay auction for some repair parts his car could really use.. ending in the next 18 hours. If you can throw $50, 20 or even $10 his way, do it soon.
Definitely deserving!
can you confirm that this is still the case with latest beta?Quote:
TCC indicator is backwards when selected manually, but correct when on auto.
so if you click: On the converter is locked, but the indicator is OFF?
and if you click OFF converter the is unlocked but the indicator is ON?
or is it something else
im flying blind with this one
That's exactly what I saw yesterday.
I will try to confirm asap
appreciate it buddy, thanks. im hoping 94 and 95 don't react differently, can someone with a '95 trans/ecm confirm too?