Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

FPGA Recommendation

10 views
Skip to first unread message

Abby Brown

unread,
Jan 29, 2010, 10:08:13 AM1/29/10
to
Hi,

Could you give some FPGA recommendations? Requirements ...

About 2000 cells
Free design software with quick learning curve, data paths
Peripheral voltages compatible with a standard logic family
and RAMs
Easy programming (ICSP) or self loading from EEPROM
Minimal "gotchas" and quirks

All the families I've looked at have power consumption and
prices well within my needs.

This is for a PDP-1 CPU.

Thanks,
Gary

Peter Bennett

unread,
Jan 29, 2010, 1:00:02 PM1/29/10
to

I've used Altera parts - Max II or Cyclone II, depending on the
resources required.

Altera's free Quartus Web Edition development software is quite easy
to use, and allows you to use any combination of schematic diagrams or
HDL to define your design. (you can even use 7400-like symbols on a
schematic.)


--
Peter Bennett, VE7CEI
peterbb4 (at) interchange.ubc.ca
GPS and NMEA info: http://vancouver-webpages.com/peter
Vancouver Power Squadron: http://vancouver.powersquadron.ca

Nico Coesel

unread,
Jan 29, 2010, 1:53:00 PM1/29/10
to
"Abby Brown" <abby...@charter.net> wrote:

>Hi,
>
>Could you give some FPGA recommendations? Requirements ...
>
> About 2000 cells
> Free design software with quick learning curve, data paths
> Peripheral voltages compatible with a standard logic family
>and RAMs
> Easy programming (ICSP) or self loading from EEPROM
> Minimal "gotchas" and quirks

IMHO all FPGAs have gotchas and quirks epecially if you push pin
placement and operating frequencies to the max. I'd choose the older
Spartan3 series from Xilinx. Pretty fast and every pin can used as an
in/out.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

krw

unread,
Jan 30, 2010, 5:56:49 PM1/30/10
to
On Fri, 29 Jan 2010 18:53:00 GMT, ni...@puntnl.niks (Nico Coesel)
wrote:

>"Abby Brown" <abby...@charter.net> wrote:
>
>>Hi,
>>
>>Could you give some FPGA recommendations? Requirements ...
>>
>> About 2000 cells
>> Free design software with quick learning curve, data paths
>> Peripheral voltages compatible with a standard logic family
>>and RAMs
>> Easy programming (ICSP) or self loading from EEPROM
>> Minimal "gotchas" and quirks
>
>IMHO all FPGAs have gotchas and quirks epecially if you push pin
>placement and operating frequencies to the max. I'd choose the older
>Spartan3 series from Xilinx. Pretty fast and every pin can used as an
>in/out.

So far, I've found the Altera MAX-II CPLDs to be pretty good. They're
a little different than a true FPGA but the user really doesn't notice
the difference so much. One of the advantages of the MAX-II series is
that the config flash is built in. Max-IIs only need a 3.3V supply,
too (internal LDO). Cheap too.

Atsunori Tamagawa

unread,
Jan 31, 2010, 5:12:03 AM1/31/10
to
krw wrote:
>
> So far, I've found the Altera MAX-II CPLDs to be pretty good. They're
> a little different than a true FPGA but the user really doesn't notice
> the difference so much. One of the advantages of the MAX-II series is
> that the config flash is built in. Max-IIs only need a 3.3V supply,
> too (internal LDO). Cheap too.

Yup. I like the Max2 CPLD's too. I especially like their
embedded memory, which has plenty of space to store your data.
When you need to program RF PLL synthesizer chips and such using
the CPLD sequencer, you want to store the programming words in the
memory instead of declaring them as hard-coded constants in your
Verilog/VHDL code. That saves a lot of main logic gates.

One thing, though, I am concerned about the MAX2 CPLD products
is the limited programming cycles. They only guarantee 100 times
of erase/programming cycles.
I think I have done more than that or very close to that number
to one of my devices. So far I don't see any problems.

I'm simply curious about why they had to specify this 100 times
limitation for this device family. Usually, devices like Nor flashes,
NAND flashes, CPLD's, FPGA's with internal config memories, and
CPU with internal flash memories allow 10,000 times or so of
erasure/programming cycles.

Altera must have employed especially hottie electron girls on
thier MAX2 silicon so they blow quicker.

Atsunori

Nico Coesel

unread,
Jan 31, 2010, 6:28:50 AM1/31/10
to
Atsunori Tamagawa <tama...@att.net> wrote:

There usually is a relation between the number of programming cycles
and the data retention time. You can program a flash many times, but
the retention time will get shorter (no longer guaranteed).

0 new messages