[McHUG] Butterfly update.

[Peter Morton]

Hi Rich-

My son, James, says TWI is much faster than bit-bang if the processor has 
the hardware for TWI.  With bit-bang the software must write each state of 
the clock and data to the SCL and SDA i/o pins.  With TWI, the software 
simply writes data to a register and the hardware does all the work of 
toggling the SCL and SDA pins.

For the SI570 output, you should use the AC connection when driving the 
SoftRock.  This will DC isolate the SI570 outputs from the injection point 
on the SoftRock board.

I haven't finished building the cables to interface the SI570 board with the 
power supply, parallel port, and scope.  I plan to do some work on it 
tonight.

I found software on the Yahoo SoftRock group site for controlling the SI570 
with the PC parallel port.  I didn't understand how it worked with the 
GTL2002 level translator on the SI570 board until I researched I2C and 
determined that the I2C outputs stay high until data is sent.  The Gref and 
Dref ports of the GTL2002 derive their input reference voltage from the SDA 
and SCL inputs that are high most of the time.

The software also supports WinRad (and PowerSDR) by placing a supplied DLL 
file in the same directory as WinRad.  I tried this but got an error message 
about a missing MSVCR71D.DLL file.  I searched, found and downloaded the 
missing file on the internet and placed it in the WinRad directory. Now 
WinRad starts without the error and allows the SI570 board to be selected as 
an input. (??)  I'll experiment with this when I finish building the 
interface cables assuming the SI570 works. The MSVCR71D.DLL file is a 
Microsoft C support file, probably a run-time library.

Looks to me like you are really mastering this technology in record time.

-Pete, W3GVX



----- Original Message ----- 
From: "Rich Mitchell" <geobra at att.net>
To: "MicroController Ham User Group - Physical Computing for Ham Radio" 
<mchug at mailman.qth.net>; "MicroController Ham User Group - Physical 
Computing for Ham Radio" <mchug at mailman.qth.net>
Sent: Sunday, March 16, 2008 12:52 PM
Subject: Re: [McHUG] Butterfly update.


>> If you send me a copy of your source code & makefile I'll give it a try 
>> and see
>> if I can figure out what's going on.
>
> Steve,
>
> Thanks for looking at the code.  A few Si570 questions since you and Pete 
> have built your Brainard boards.
>
> The schematic shows either an AC or a DC connection.  For the LVDS should 
> we use the AC connection?
>
> In the W0CNN code he has two I2C flavors:  TWI and bit-bang.  From 
> scanning the data sheets it appears that the ATmega168 has a full fledged 
> TWI but the ATmega169 is only partially implemented.  Should we use the 
> TWI only for th 168 and the bit-bang approach for 169?  Would it hurt to 
> use the bit-bang for both?  By the way I didn't know what TWI meant until 
> I wikied it, so I may have misunderstood the data sheets.
>
> Thanks and 73
> Rich, N3III
>
> --
> McHUG - Physical Computing ;)
> MicroController Ham User Group
>
> ______________________________________________________________
> McHUG mailing list
> Home: http://mailman.qth.net/mailman/listinfo/mchug
> Help: http://mailman.qth.net/mmmain.htm
> Post: mailto:McHUG at mailman.qth.net