[chirp_users] Icom IC-T90A: Rolling an OPC-478 cable with a CP-2102 USB board

[Robert Terzi]

On 11/29/2013 4:00 AM, Chris Suleske wrote:

> Now I'm moving on to my real radio, an Icom IC-T90A.  I found the
> following info:
>
> http://highfields-arc.co.uk/constructors/other/opc478.htm#usb

> There's a note in the instructions above about connecting the CTS
> and RTS pins together... which I can't do on the 2102 board I have.
> The pins are on the IC of course, but not connected to anything on
> the board.  The diode I used appears to be good and electrically,
> the cable seems sound.  Any ideas?

Connecting CTS and RTS together is done to ensure that hardware flow control
using the Request-To-Send, Clear-To-Send handshake, doesn't prevent any
data flow.  In some cases, the serial API can be told to ignore hardware
flow control, so it's possible two different pieces of software would
give different results as far as being stopped by flow control.  Check the
settings for the Icom software you are trying to see if it has any options
for disabling hardware flow control.

However, since your adapter PCB doesn't have traces for those lines, I would
hope that the CP2102 internally pulls those lines to keep them from floating
and interfering.  Given your other results, I'd put the CTS/RTS issue aside for now.


> But it has not produced fruit with the Icom CS-T90A software.
> Specifically, it errors "Cannot access COM port."  If I remove the
> cable / diode from the header and jump TX and RX pins together, the
> CS-T90A software responds differently, indicating "No answer from
> the transceiver."  So, I know the software and the CP-2102 are
> talking.

If I had to make a guess, the diode is connected incorrectly, or doesn't
meet the needs for that circuit.  (I also wonder if there shouldn't be
a resistor in series with the diode like in the RS-232 circuit.)

It looks like that radio is one of the type that has transmit and receive
data share a pin.  Therefore, every bit transmitted will be echoed back.
The software has to expect to read back every byte it sends the radio in order
to differentiate what is being echoed versus an actual response from the radio.

So I'm interpreting your results as:  When you've got TX and RX shorted the software
is seeing the data (probe) it sends to the radio back and but isn't seeing any
response from the radio.   However, when the diode is being used to connect TX and RX
together, the software isn't seeing it's data being echoed.

On 11/29/2013 8:41 AM, Jim Unroe wrote:

> What I found out by using CP-2102 modules to make programming cables
> for my UV-5R (and other radios) is that the module ports are not
> consistently labeled. Some labels indicate "what the port is,
> receiver or transmitter" and other labels indicate "what the port
> connects to on the device".

Jim might have the right answer here, which would also mean that your diode
wound up connected backwards so it wouldn't copy the computer's transmitted
data back to the RX pin.

A voltmeter is helpful here.  Generally, you should see the TX line pulled
to either the circuit voltage 5V / 3.3V (or ground if it is active high instead
of active low).  The receive line will generally be floating so you should easily
be able to tell them apart.   If you use a terminal program that can send a
Break signal (which is defined as 500 ms of signal) you should be able to see
the transition even with a voltmeter.

Hope this helps,
--Rob