The logging of injector pulsewidth (PW) seems to be in milliseconds in MT. This is fine, but it's only got one decimal place of precision, which leads to huge quantisation errors for something I'm working on.
I know that internally, VEMS works in microseconds so is there an easy way to get more accurate datalogs?
If not, I'll willingly listen to the difficult way!
May be try replace this entry in vems3.ini
entry = pulseWidth, "PW", float, "%.1f"
with this:
entry = pulseWidth, "PW", float, "%.3f"
Please report is it works!
Gints
So that should give me 3 decimal places if everything works - assuming it's the same convention as C...
Certainly will report back!
Quote from: dnb on November 29, 2007, 01:14:30 PM
So that should give me 3 decimal places if everything works - assuming it's the same convention as C...
Certainly will report back!
No it do not work. I just try it. This gives values like 3.600 and 3.700 in logs :(
I am interested too. I use MLV and little hacked vems3.ini. VE analyser works for me. Not so good as for Megasquirts, but I can use this feature with some precaution.
The number sent down the RS232 port from VEMS for pulse width is only an 8 bit number (0x00 - 0xff or 0 - 255).
MegaTune divides this number by 10 for display giving a min 00.0ms to max of 25.5ms.
Increasing the display resolution will not make any difference, this is why it threw 3.700 at the screen.
Hilly