cancel
Showing results for 
Search instead for 
Did you mean: 
4,273 Views
Registered: ‎06-29-2009

ICAP destroy FPGA ?

Jump to solution

Hi,

 

Is it posible to destroy the FPGA when using the ICAP ....  supose bad signals or data is asserted into ICAP ? 

If yes:  is there an acceptable solution for not destroying the FPGA while experimenting e.g. limit the current when reprogramming? (although I thought the FPGA needs enough current when starting...)

I would like to experiment on multiboot with the ICAP on the SP601 board (with a PCORE in xps).

 

Serge

Message Edited by serge.kubera@ehb.be on 02-23-2010 03:18 PM
Tags (4)
0 Kudos
1 Solution

Accepted Solutions
Highlighted
Scholar
Scholar
5,152 Views
Registered: ‎02-27-2008

Re: ICAP destroy FPGA ?

Jump to solution

Serge,


With ICAP, as you suggest, it is possible to write pure garbage into the memories.


Will this "destroy" the device?


It may cause massive contention, which wil cause the device to get very hot.

 

If you let it sit there, beyond the absolute maximum junction temperature limits, for too long, yes it will be destroyed (if the power supply is capable of supplying enough current).  This usually takes minutes to hours, so you are in no rush to save the device, just turn off the power, and let it cool down (and do not put THAT garbage into it again).


Just watch the current into the Vccint, and do not exceed the junction temperature recommended, and you will be just fine.

 

The massive contention is caused by millions of ~50uA (worst case) contention paths in the interconnect multiplexers, so you have not caused any electromigration failures, nor affected the component long term reliability, until you let it cook for too long at a temperature greater than the absolute maximum in the data sheet... and then after awhile you have made "toast" and the device will be long gone.

 

 

Austin Lesea
Principal Engineer
Xilinx San Jose

View solution in original post

0 Kudos
2 Replies
Highlighted
Scholar
Scholar
5,153 Views
Registered: ‎02-27-2008

Re: ICAP destroy FPGA ?

Jump to solution

Serge,


With ICAP, as you suggest, it is possible to write pure garbage into the memories.


Will this "destroy" the device?


It may cause massive contention, which wil cause the device to get very hot.

 

If you let it sit there, beyond the absolute maximum junction temperature limits, for too long, yes it will be destroyed (if the power supply is capable of supplying enough current).  This usually takes minutes to hours, so you are in no rush to save the device, just turn off the power, and let it cool down (and do not put THAT garbage into it again).


Just watch the current into the Vccint, and do not exceed the junction temperature recommended, and you will be just fine.

 

The massive contention is caused by millions of ~50uA (worst case) contention paths in the interconnect multiplexers, so you have not caused any electromigration failures, nor affected the component long term reliability, until you let it cook for too long at a temperature greater than the absolute maximum in the data sheet... and then after awhile you have made "toast" and the device will be long gone.

 

 

Austin Lesea
Principal Engineer
Xilinx San Jose

View solution in original post

0 Kudos
Highlighted
4,245 Views
Registered: ‎06-29-2009

Re: ICAP destroy FPGA ?

Jump to solution

That is good news.

 

Thanks for the explanation!

 

 

Serge

0 Kudos