Are there more I/O pins than I/O blocks?

[kryten_droid]

I have an XC2S300E-pq08 which the data sheet says has 146 user I/O max.

However, the PAR says:

Number of bonded IOBs: 144 out of 142 101%

There are still quite a few I/O pins left on my dev board.

I'm puzzled that it seems I can use any I/O pins but
less than all of them at the same time.

Is this the case or am I missing something fundamental?

Thanks in advance,

K.


Command Line : map -p XC2S300E-pq208-6 -cm area -k 4 -c 100 -tx off
Target Device : x2s300e
Target Package : pq208
Target Speed : -6
Mapper Version : spartan2e -- $Revision: 1.58 $
Mapped Date : Fri Oct 31 01:02:35 2003

 

[Andras Tantos]

"kryten_droid" <kryten_droid@ntlworld.com> wrote in message
newsHxob.1172$yH4.711@newsfep1-gui.server.ntli.net...

I have an XC2S300E-pq08 which the data sheet says has 146 user I/O max.

However, the PAR says:

Number of bonded IOBs: 144 out of 142 101%

There are still quite a few I/O pins left on my dev board.

I'm puzzled that it seems I can use any I/O pins but
less than all of them at the same time.

Is this the case or am I missing something fundamental?

Thanks in advance,

There are four dedicated clock pins.

Regrads,
Andras Tantos

 

[Simon Peacock]

I think you will find that you have over committed the pinout of the device.
For some reason this isn't always classified as a fatal error

Simon


"kryten_droid" <kryten_droid@ntlworld.com> wrote in message
newsHxob.1172$yH4.711@newsfep1-gui.server.ntli.net...

I have an XC2S300E-pq08 which the data sheet says has 146 user I/O max.

However, the PAR says:

Number of bonded IOBs: 144 out of 142 101%

There are still quite a few I/O pins left on my dev board.

I'm puzzled that it seems I can use any I/O pins but
less than all of them at the same time.

Is this the case or am I missing something fundamental?

Thanks in advance,

K.


Command Line : map -p XC2S300E-pq208-6 -cm area -k 4 -c 100 -tx off
Target Device : x2s300e
Target Package : pq208
Target Speed : -6
Mapper Version : spartan2e -- $Revision: 1.58 $
Mapped Date : Fri Oct 31 01:02:35 2003

 

[kryten_droid]

"Simon Peacock" <nowhere@to.be.found> wrote in message
news:3fa45ff8$1@news.actrix.gen.nz...

I think you will find that you have over committed the pinout of the
device.
For some reason this isn't always classified as a fatal error

Checked out the chip and there are only 146 I/O pins.

Although my user config file only said that only 123 were used,
the design still had more I/O blocks than the device could bond to pins.

So even though the design could theoretically fit, if the UCF was
considered, the place-and router would reject the design.

D'oh!

So I removed a load of I/O that was just redundant test signals
and the software has compiled and fitted the design.

 

[Simon Peacock]

I assume its still correct.. but there have often been FPGA's which are pin
bound.. that is there are more I/O's internally than there are pins.. that's
why the same device fits into a TQ144 and a TQ208 with no free pins.

Simon


"kryten_droid" <kryten_droid@ntlworld.com> wrote in message
news:hRipb.1394$wY2.963@newsfep1-gui.server.ntli.net...

"Simon Peacock" <nowhere@to.be.found> wrote in message
news:3fa45ff8$1@news.actrix.gen.nz...
I think you will find that you have over committed the pinout of the
device.
For some reason this isn't always classified as a fatal error

Checked out the chip and there are only 146 I/O pins.

Although my user config file only said that only 123 were used,
the design still had more I/O blocks than the device could bond to pins.

So even though the design could theoretically fit, if the UCF was
considered, the place-and router would reject the design.

D'oh!

So I removed a load of I/O that was just redundant test signals
and the software has compiled and fitted the design.