[hpsdr] [OT] Shipping container overload

[Christopher T. Day]

Eric,

I am slogging though the USB spec [I actually _like_ reading specs,
another moral failing on my part] and the FX2 spec, but they are a
trifle long. However, it seems that that is exactly the design point of
the FX2.

Interesting about the USRP experience. I wonder about the "*can't*
control the stuff" comment. Isn't I2C multi-master? Couldn't the FPGA
also fiddle the chips? Note that that is strictly a technical question.
Believe me, I _know_ multi-master complicates the situation enormously,
kills off the FPGA space saving and is generally probably not a good
idea in this case. Just wondering if it's technically possible, that's
all.

Clearly, I need to work though the use cases for chip control and
interaction some more.


	Chris - AE6VK


-----Original Message-----
From: Eric Blossom [mailto:eb at comsec.com] 
Sent: Wednesday, March 15, 2006 12:31 PM
To: Christopher T. Day
Cc: High Performance Software Defined Radio Discussion List
Subject: Re: [hpsdr] [OT] Shipping container overload

On Wed, Mar 15, 2006 at 12:03:04PM -0800, Christopher T. Day wrote:
> ... - yield a device that worked with generic USB
> Audio Device drivers.

Makes sense.  
I suggest checking the USB specs for the audio device class.  It's a
standard.

> And speaking of commands, since the FX2 has a direct I2C master on
chip,
> it could be used to send configuration commands to the ADC/DACs and
save
> some space in the FPGA.

You could do that.  

That's what we did on the USRP, but in retrospect, I think it was a
mistake.  The problem is that if you hang it all off the FX2 I2C, then
the FPGA *can't* control the stuff.  It all depends on what you think
you might want to do.  If it's possible that you might want to run a
control loop in the FPGA that needs access to the I2C bus, then hang
it off the FPGA.  [Actually on the USRP, we controlled the ADC/DACs
over the SPI bus, which we bit-banged from the FX2.  Same conclusion,
however.]

Eric


 1142455458.0