Xilinx IMPACT Problem... detects 101 unknown devices

Hi all,
I have a custom designed hardware.... when i ran the xilinx iMPACT to
configure the device, it ended up detecting the FPGA and a 101 unknown
devices and then just stopped dead? has anyone ever had a similar
problem? Please do help.
Thanks

Subhasri krishnan wrote:
> Hi all,
> I have a custom designed hardware.... when i ran the xilinx iMPACT to
> configure the device, it ended up detecting the FPGA and a 101 unknown
> devices and then just stopped dead? has anyone ever had a similar
> problem? Please do help.

Sounds like your JTAG chain is screwed up.

-a

please tell me what should i do now to set it right? should i read
something or should i check for something? why does it detect so many
devices when i have only one FPGA in there?
...thanks

Subhasri krishnan wrote:
> please tell me what should i do now to set it right? should i read
> something or should i check for something? why does it detect so many
> devices when i have only one FPGA in there?
> ...thanks
> 

Scanning the JTAG chain for devices is done by reseting the TAP controllers
which loads the devices IDCODE into the TAP output register and then shifting
the register out and waiting until you get a set of 32 ones (zeros??)
indicating the end of the chain.

The most likely cause is that you have connected the TDO cable pin to
something other than the TDO pin of the last device in the chain that is
generating bad data back to the cable.

Ed

Subhasri krishnan:
> please tell me what should i do now to set it right? should i read
> something or should i check for something? why does it detect so many
> devices when i have only one FPGA in there?
> ...thanks
I've had problems with too many detected devices when the cable
frequency was too high. You can change it in the cable setup settings.
If you use a Parallel Cable IV and it is properly detected, iMPACT
usually sets it to a default of 5MHz. Try lowering the frequency and see
if it gets any better.

cu,
Sean

I have had some problems similar to that in that I would get a strange 
number
of devices rather than the six I have on the board.  The Xilinx people 
pointed
out the signal integrity issues I designed.  Mostly the TCK line was the 
culprit.
How is your TCK line routed? Is there a series resistance? Or perhaps you
put an end Thevenin termination at the end? How long are the lines?

> Hi all,
> I have a custom designed hardware.... when i ran the xilinx iMPACT to
> configure the device, it ended up detecting the FPGA and a 101 unknown
> devices and then just stopped dead? has anyone ever had a similar
> problem? Please do help.
> Thanks
> 

My ISE7.1 says parallel cable drivers installed, but refuses to recognize my old cable III and tries using anon-existent multilink. Seems to be reading Prom & Virtex Id's, but can't start them after program. How do I get IMPACT to accept a Cable III ?

Subhasri krishnan wrote:
> please tell me what should i do now to set it right? should i read
> something or should i check for something? why does it detect so many
> devices when i have only one FPGA in there?
> ....thanks
> 

I've seen this happen with a dodgy JTAG cable.

Jeremy

some parallel cable III have the wrong logic to use with low voltage
parts... mine wont work below 3.3V.. what is your Jtag?


Simon


"Klaus Bickertt" <kfb@mpe.mpg.de> wrote in message
news:ee90993.5@webx.sUN8CHnE...
> My ISE7.1 says parallel cable drivers installed, but refuses to recognize
my old cable III and tries using anon-existent multilink. Seems to be
reading Prom & Virtex Id's, but can't start them after program. How do I get
IMPACT to accept a Cable III ?

Thanks for the input everyone.The "101 device" problem has been fixed.
It turned out that the voltage on the download cable is too low. should
be 2.8v and above, the design was only 2.5v. Now it is 3.3v.

But there is the error of "impact 583:the idcode read from device
does not match idcode in bsdl file". We think that the TDO's voltage is
too low (3.3v), while the TDO requires TTL > 4v. We are going to try
this tomorrow. But can anyone here think of another reason?

Thanks