[hpsdr] HPSDR Mercury Info

[Henry Vredegoor]

Hi Phil,all,

Thank you Phil for your response!

Sorry I did want to post to the reflector instead of your personal mail
address, instead I just hit "reply".....

Wouldn't it be a nice feature in this respect to have some degree of
"scalability" by e.g. full or partial population of the boards and/or
selection of part types? 

Maybe also an option for people who don't want to spend too much money first
time.

For example:
Single RX Channel  --> ONE FPGA populated, option to (have-) add(ed) a
second FPGA later
Dual RX Channel or more program resources --> TWO FPGA's populated

Or to go even further:
Possibility of population of the same "basic" board for different type$ of
FPGA's?
All of course being possible/doable/practical

Sorry, just venting idea's......
Not wanting to waste your time with my remarks......  :-)


Regards,


Henry.






> -----Original Message-----
> From: Philip Covington [mailto:p.covington at gmail.com] 
> Sent: zondag 18 februari 2007 20:57
> To: Henry Vredegoor
> Subject: Re: [hpsdr] HPSDR Mercury Info
> 
> 
> On 2/18/07, Henry Vredegoor <henry_vredegoor at hotmail.com> wrote:
> > > -----Original Message-----
> > >
> >
> > < snip >
> >
> > > The Digital Down-conversion (DDC) function will be done 
> in the FPGA.
> > > Without going into the details of CIC filters and their 
> compensation
> > > filters, there is enough room for most of the filtering and
> > > down-sampling functions in the FPGA except for the final step.  We
> > > need a compensation filter of about 300-512 taps for both 
> the I and Q
> > > data paths.  Filters of this size require a larger and 
> more expensive> > FPGA to have enough Logical Elements (LEs) 
> or Multipliers (MACs)  to
> > > implement them.
> > >
> > > Instead I have found a company that makes a serial FIR 
> engine in a 3mm
> > > x 3mm QFN package:
> > >
> > < snip >
> >
> > Pity for you not going into detail about this Phil, ;-) , hence my
> > (maybe-)stupid remark as a newbie to DSP:
> >
> > Would using TWO instead of ONE FPGA's solve the problem of 
> the number LE &
> > MAC elements? Or still not enough?
> > I don't know how this would work out technically or would 
> add up price-wise
> > (4 X QF1D512 against 1 extra FPGA), but one of the plusses 
> would be more
> > flexibility / programming resources?
> >
> > Sorry if this is a too obvious newbie remark!
> >
> > Regards,
> >
> > Henry.
> 
> Hi Henry,
> 
> We have actually thought about going with the EP2C20 Cyclone II
> device.  In HPSDR Mercury we will also have the EP2C8 device on OZY
> that can be partially utilized.  The problem is to get the 100 dB
> alias rejection we want in Mercury for HF use, we need a relatively
> high order FIR filter.  These tend to take up a lot of space in the
> FPGA especially when they are operating on two streams (I and Q) of 24
> or 32 bit data.  To fit everything in the EP2C20 device is pushing it
> for even one rx channel.
> 
> The Quickfilter devices are about $7 in single quantity.  If we make
> provisions for placing 4 devices we could leave two unpopulated for
> cost sensitive applications.
> 
> The advantage of having a larger FPGA gives us versatility and the
> ability to fit the processing to different applications of the
> Mercury.   We could always look at a Stratix or Stratix II FPGA but
> then we are talking about devices in the min of $250-$350 range.
> 
> Now if I was building just one receiver for my own use I'd say screw
> it and put a $500 Stratix II device on-board and then do about
> anything I'd want as far as processing.
> 
> 73 Phil N8VB
> 


 1171835030.0