Forum | Cable unit Egas, study and testing

I made a study and made some test to be able to reprogram the control unit of the gas in my car.
of course cable is not for sale to the public but are authorized installers ...
the system is "sly injection" of gas and @ of Rovigo, I found someone on ebay selling a kit to program the majority of units on the market (BRC, landi, lovato ..), but obviously not mine
-
http://cgi.ebay.it/AUTOGAS-LPG 45ec3f0b3b ... ..
-
the control units of the gas using the same standard machines until 2005/2006ciso9141
http://www.onboarddiagnostics ... ... 03.htm # ISO
which has 2 wires:
K-way line on which the signals pass tx and rx
The line to wake up the unit (optional)
where the state 0 and 1 is Vbatteria and gnd, it has 10.4kHz
if you look on google of ISO9141 will be out of everything.

Egas software uses only the serial port, then the interface cable will convert a signal from RS232 (9 pin) to K / L line
Tx and Rx data to go in the K line and the line L will have to be driven by some fucking remaining pins among others.
a classical scheme used for this application is the one on this link:
http://hem.passagen.se/tomasf/obd2/
is rudimentary and completely non-standard rs232 but it should work:
tx that drives the K-line x-ray line and receives the same signal that drives the serial port RTS line L.
the one on ebay was probably at least a circuit that logically should be done so '.
normally have 4-pin connectors of the ECU gnd Vbat KL, even for the machines is the same thing
on my pole for the former I have 3 gnd Vbat and K, and with a simple cable 9E bought in China can read the unit properly.

and here begins the cock: my 6-wire connector:
Red Vbat
Black gnd
K / L yellow brown (but do not understand is which of the 2 K)
Can-Can + gray green (you can 'understand X, which the tension at rest and 2.4V, exactly what the protocol can)

I've checked the oscilloscope signals coming out of the serial when I start the software Egas and find a signal on Tx 10.4kHz and a single 80ms INPULSES of DTR. supponco so that work on K / L and that DTR is used to drive the line L
for the remaining 2 wires can, probably in the cable seriale_pc / unit is not used
I made a cable that is logical at least to levels similar to what I linked before, but done with an RS232 transceiver and a language plus a discrete MOS to drive the line L.
put the scheme in the next post
---
I added some jumpers and a series of NAND to have the ability to reconfigure and conbinazione signals, with these jumpers can combine the signals from tx and dtr, also receive the L line, but unfortunately does not work.
I tried all possible conbinazione signal and reverse the wires yellow and brown, but there's no way to get a dick.
????????????????????????
someone for this cock dicavo hands?