NIStune 0.90103 now available

[Matt]

- Preliminary Zexel ECU consult changed (polled mode) with increased timeouts and improved initlaisation sequence [not tested yet]
- Wideband/consult connectivity affecting log recording now fixed. Also improved wideband connectivity feedback when device not connected or powered on
- Improved keyboard only handling using CTRL-TAB to highlight map select options. Ability to use F5,F6 to pull up a hidden consult or wideband window already opened. Added ability to select already highlighted map from menu and redisplay it from underneath other windows
- SPACE bar to mark cells in grid

[Matt]

- Wideband update fixes
- Romulator crash fixes
- AFR graph RPM scale fixed
- Type 1 (Z31/R31 6802) Rev1,2 board ROMulator files increased to Patch Rev 11. Use these now for those particular boards with your romulator

[Matt]

Now 0.90103 available

- Added injector CC for Z31
- Merged in serial code improvements from consult comms to romulator, techedge and zeitronix. Tested romulator and Techedge but not zeitronix yet. Massive improvements with connectivity issues
- Change com port detection from 1-255 to get around USB-Serial converters trying to use the next unprobed port (and not getting found)

[Matt]

- Fixed stuff up made in last few days which affects consult map resync. (To do with address files being NULL meant invalid but some ecus have maps located at address 0000 which was the same value). Anyway fixed

- Added N14 SR20 and S13 SR20DE to the maptrace/consult table list with updated address files

- K11 table fixes with 2D timing tables having knock flag support

[Matt]

- Maptracing fixes added. When table maptracing was improved, affected 3D grid maptracing

- Injector adjustment now also changes latency using following forumla

((NEWCC/OLDCC * ORIG_LATENCY) - ORIG_LATENCY) / 3.5 + ORIG_LATENCY

This results in a similar value to live edit. originally the formula was /2

[TSL]

Only problem being injector latency isn't directly related to flow rate.

Simply looking at the factory nissan latency values for different sized injectors confirms this.

Then consider aftermarket injectors which have different pintel designs etc.

For that matter even the k-value calculations are only ever approximate.

One of the things that caused so many problems with live-edit users is the assumption that the calculations it made are exact.

[prushton]

i agree matt, i'd like to ensure i know what has been changed, the formulas, as im sure you are aware, on the k11 for example are very different to other cars.

Also the feedback switches have confused me, its like rom-editor and just shows a value, when really is should be a set of switches?

[Matt]

the original calcs for K const and Injector latency were based around the writeup devon did for the 300ZX code. That seems to be compatible for only RB/VG engines then?
http://www.redz31.com/pages/fuel/injectors.html

maybe I dont make this too asuming then, stick with just a K constant change (which also adjusts TTP min/max and TP scales). I can make obvious the multiplier used during the calculatinos for everything adjusted and only suggest a latency???

Peter - are you talking about the O2 feedback flags? if you use Filtered mode for the map these flags are not shown as 'raw values', then you can click the 'Flags' box to see where the O2 flags are enabled by colour. Please let me know if that is sufficient. 'O' will turn these on/off

[prushton]

no matt, the value im talking about is the

#Feedback control
#FB_SWITCH

As the name suggests this should be a binary switch variable, i haven't really had much time to play with the variables, and have never really needed to, but by setting the number very low as suggested by a japanese racer, i get rid of feedback altogether.

i am not sure where exactly the belief that its some coolant ref comes from, but im pretty sure its incorrect for the newer ecus at least

certain people / software use the raw value, others make use of the binary switches, the code used for the feedback switches makes use of

bbc / bbs - 0x80(MSB) of feedback control
bbc / bbs - 0x40(Second MSB) of feedback control
bbc / bbs - 0x20(Third MSB) of feedback control

Please note i am not "of techie"

there is no cmp - feedback value used as far as i can tell, which is why i think this is correct

[Matt]

no probs. I need to make a flag constants window then for checking flag 'bit' values

i was having a look at the O2 switch code yesterday and noticed some comparison being done in the RB30 code. I

I'll raise another change request for this

[prushton]

excellent matt, thanks very much, the feedback area i feel is a little misunderstood.


apologies if im "picking faults"

[Matt]

thats fine. the more things that get nailed down earlier the better

[Matt]

Hmmm. I'm not sure about this FB_SWITCH stuff now

6303 code treats it like a coolant variable, whilst 7790 code treats it as a flag

Z32 Code
ldab X140C (COOLANT_TEMP)
bcs L8C05
cmpb XFF91 (FB_SWITCH)

S13 SR20 Code
bbs #0x80,0x8038,0x9A82 (FB_SWITCH)
bbs #0x01,0x8038,0xA0A2
bbs #0x02,0x8038,0xA12D
bbs #0x08,0x8038,0xA1F4
bbc #0x04,0x8038,0xAAE5
bbs #0x20,0x8038,0xD254
bbs #0x10,0x8038,0xDC44
bbs #0x40,0x8038,0xE04E

Might have to split this into two different address types

[prushton]

matt, thats what i was saying above

[Matt]

yeah most of what I read before was heresay, so I had to look into the Z32 code myself today to confirm it

these are two separate items. I guess altering them has the 'same' effect which is why they have been confused as the same thing

Just wonder now if the 7790 code also has a CMP value on temperature similar to the 6303 bit which just hasn't been found

Or if the 6303 code has flags like the 7790 code...