Show or hide header
Welcome Guest, posting in this forum require registration.

Torque » Torque OBD ECU Scanner » Torque Discussion / Ideas » USB connection issue

Pages: [1]
Author Topic: USB connection issue
Posts: 5
Post USB connection issue
on: November 24, 2016 (GMT)


I have some issues while using a usb connection to Torque PRO.

1) First problem is very similar to this topic:
Despite of that the data sent to Torque are constant (I am pretty sure for this, cause I can sniff the data transfer via Hyper Terminal and a small adapter), gauges occasionally jump to top or bottom or any unpredictable value.

2) Second problem is that the baud rate that I choose from settings menu doesn’t correspond the real one. For example, I set up 38400, but the real speed is 19200.
Another example is 9600. The real speed with this setting is around 400 (!) baud. Yes I have an oscilloscope.

3) And the last one: data polling freezes about 1 time a second. And I am sure this is on the tablet side, because I see an adapter’s response. And after this response polling freezes. Maybe I should use some delay after receiving a PID request before the answer, because by now I send an answer almost immediately.

My config:
Head unit NAVITREK NT006 Android 4.4.2
FTDI FT232RL (original chip) adapter
self-made additional sensors polling device
Torque PRO 1.8.154
and a simple sniffer


Posts: 6073
Post Re: USB connection issue
on: November 24, 2016 (GMT)


Can you tell me if the xtal providing the clock to the ELM327 is spot on what it should be? (or if it’s RC driven, if it’s within the right frequency?)

The most common cause of comms ‘hiccups’ with clone ELMs is that the timing for them is off (or nutty capacitance on the tx/rx from the FTDI to the ELM chip causing issues) – these’ll need to be ruled out first as they’re the most common issue

That aside, if you’re able to see the comms, another telltale sign is when the app sends a command to the chip, and then nothing is sent back (resulting in a big pause). You should see this if you’re sniffing the data. Again, what happens is that the chip gets garbled data, and then is still waiting for a \r or \n to terminate the command, which never arrives.

Aaaand! another telltale sign of corrupt comms is the dials jumping as you say! – basically this time the ELM is sent something like an 0105, and instead it thinks it has another valid pid (eg, 0106) and proceeds to send back the wrong response

I’ll look into the bauds, though it’s using a common ftdi driver so the chip should be configured for the right baud rates (and the automatic option should always find the correct baud in any instance). It’s been a while since I last looked at the control registers of an ftdi – I’ll try to look at this next week to see if anything is wrong at this level!

Posts: 5
Post Re: USB connection issue
on: January 27, 2017 (GMT)


Just got back to my project. I tried another android device with my board and FT232 adapter. Now it’s Galaxy S5.
So, I captured the real speed depending on settings:
Setup: Real speed:
2400 9600
4800 57600
9600 460800
19200 2400
38400 19200
57600 115200
115200 300 (!!!)
460800 38400

This is the speed measurement directly on the tx pin of FT232 adapter.

Is Torque waiting for \n\r or it is no matter? Is there any difference if I send \n or \r only, or \n\r or \r\n?
Maybe there should be a small delay between request and answer?
Because as I can see now, there is a request from Torque and an immediate reply from ELM327, and a long silence interval after that…

Thank you.

Posts: 5
Post Re: USB connection issue
on: May 11, 2017 (GMT)


Pages: [1]
WP-Forum by: Fredrik Fahlstad, Version: 2.4
Page loaded in: 0.09 seconds.

  Follow me on twitter