I run the same headers. your knock is crazy. you cant tune around it. id delete the knock sensor.
Printable View
I run the same headers. your knock is crazy. you cant tune around it. id delete the knock sensor.
ok here we go with another one: http://fbodytech.com/eehack-2/beta/
added:
- graphing cursor! double click to seek to closest (but oldest) record possible
- load old eehack log files as msg0 (with a warning)
- minor analyzer bug fixes and enhancements
please try to break this and find bugs, almost ready for release!
Well that's definitely a huge improvement that really helps out!
it's nice to see where you are clicking in the graphing window, and bonus, it can scroll during playback. I don't know if that was intended as it seems to take a slight performance hit and has a stuttering effect, though if you choose fast playback it scrolls smoothly. It may be my old laptop though.
But so far so good. No flashing with the latest yet, but scanning works good for me so far.
I havent changed flashing so it should be fine
the overhead of drawing that cursor is high, the entire graph is re plotted. luckily this only happens when the graphing window is open, so it won't slow down normal operations.
stuttering when scrolling during playback or whatever is due to the fact that 'playback' isn't threaded (data acquisition is, though). i do plan to do log playback in a separate thread in later versions, so this will improve in later versions
One thing that I have noticed is when using the control window and trying to dial in an idle afr by forcing the afr, it would show the desired afr in commanded afr, but the wideband would get progressively leaner until the engine would die or I'd disconnect. I didn't realize until viewing a log of this that when I tried forcing an afr it was going into closed loop. I'm missing a O2 sensor so closed loop is a no go for me.
yeah man forced afr is weird.
it forces closed loop (since that's how the forced AFR works, it screws with the trim) then it clears BLMs and locks them.
but maybe the disable blm update isnt working?
that means its reading reference voltage and it's just dropping trims and cutting fuel. what happens to your trims when this is going on?
i might ask for kur4o's help with disassembly and try to hack it a bit in one of my patches so it behaves a bit better
I am thinking of eside mode 4 afr hack with all blm corrections set to default 128 value.
That way tside blm correction will still work but wont be applied.
So you will see the closed loop correction when your on custom open loop.
Delta AFR adder when in open loop is also possible from eside mode4.
Customs tables for afr correction can be patched. I`ve been thinking for cruise lean patch also but needs to dig in the v6 code first to see how it`s done there.
Might be good idea to clear all blm when leaving eside AFR.
Did anyone tried the idle open loop patch if it`s working you may add it as a custom patch dialog on eehack.
The problem with tside afr mode4 is huge and very strange. Never could make it work as it should. First thing is to reset BLM because when BLM update is disabled the trims are still applied even in Open loop.
Second disable blm update and closed loop.
Last set AFR target.
It is too much pain to make it work and eside mode 4 is the way to go.
Another thing to implement is second spark table switchable by mode 4.
I am working on the direct memory write hack and a hack to copy selected table to Ram location and run it through there. Used with direct memory write hack you can finely tune single cells. You can think on some algorythm for auto spark or some auto tune stuff.
I suspected this along time as some freeze frame data, now I am sure.
There are five sets of freeze frame data
first is at 1886
than 189a, 18ae, 18c2, 18d6 they have same structure and data as the first one.
You may add it as mode 3 request along with adaptive pressure modifier cells
1830 2nd
1841 3rd
1852 4th
and some more modifiers unindentified
at 1864 2nd
1875 3rd
These are the freeze frame data description
RESERVED:1886 unk_1886: fcb 1 ; DATA XREF: RESERVED:off_F7D3o
RESERVED:1886 ; 1914 value
RESERVED:1887 fcb $35 ; 5
RESERVED:1888 fcb $8E ; Î ; afr 23f
RESERVED:1889 fcb $A ; airflow 2c5
RESERVED:188A fcb $5F ; _ ; airflow 2c5+1
RESERVED:188B fcb 0 ; 1d65 value
RESERVED:188C fdb $166 ; LBPW 2b3
RESERVED:188E fdb $168 ; RBPW 2bd
RESERVED:1890 fcb $80 ; À ; LCORR 162
RESERVED:1891 fcb $80 ; À ; RCORR 161
RESERVED:1892 fcb $80 ; À ; LBLM 164
RESERVED:1893 fcb $80 ; À ; RBLM 165
RESERVED:1894 fcb $F ; 2bf airflow baro corr
RESERVED:1895 fcb $51 ; Q ; ad map 235
RESERVED:1896 fcb $33 ; 3 ; RPM*25 15e
RESERVED:1897 fcb 0 ; CLOSED/OPEN loop 0/1
RESERVED:1898 fcb $76 ; v ; coolant 1a2
RESERVED:1899 fcb 9 ; mph 19d
I see that these freeze frame data is only available on corvette PCM.
The Oem tool communicate with the corvette PCM on class 2 serial data or obd-2 standard and available data is much different[There is low resolution signal in ms and high resolution signal data I need to identify those and where they are located].
I flashed a factory corvette bin on 8051 pcm and the oem tool refuse to communicate over the aldl line[even with f and b body settings on the tool]. That was really strange and I make quick comparison of the bins and there is one timer used in aldl routine set differently at 2020 tside.
Corvette PCM will be very handy, unfortunatelly I dont have one to play with.
I found a great source for valid vin numbers at one junkyard site and will rip as much calibration as possible to fill the bin collection.
how about we poll some corvette eehack users and get a donated one to you in exchange for some vette feature research
i could probably make it happen
i wonder if i shouldn't rename the AFR TARGET and rework the controls to what it really is: trim skew
posted another beta. this should be a release candidate, i'd really appreciate notice of any bugs,
i've 'disabled' afr target for now since i've never seen it work as intended.
Im running eehack right now just downloaded my stock bin. Time for fun!
Steevo, how do I edit the .bin file?
i answered in your other thread.
Thanks steveo I see it now