cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
nagender_patel
Observer
Observer
525 Views
Registered: ‎03-21-2017

Floating point ip core

Hello, 

          I have developed a folating  point based IIR Filter design using a floating point multiplier and floating point adder. I'm operating at 100Mhz and I have given latency as 1 for multiplier and latancy of 2 for adder. When I'm using one channel Filter design it is working perfectly, but as per my requirements I have to design 52 channels filter design, but some channels are not working it's displaying value as NaN......I have doubt on floating point ip core .......for adder with 100Mhz is latency 2 is sufficient......?

0 Kudos
3 Replies
drjohnsmith
Teacher
Teacher
517 Views
Registered: ‎07-09-2009

put numbers up to a bigger number, see if that helps.

If so, then start reading up on what the numbers mean,

 

<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>
0 Kudos
nagender_patel
Observer
Observer
501 Views
Registered: ‎03-21-2017

If I go for higher latency I can't meet my sampling time requirement.....with one Filter module it is working perfectly.....but when I instantiate 52 modules some of the filter modules are producing output as NaN vlaue 

0 Kudos
drjohnsmith
Teacher
Teacher
461 Views
Registered: ‎07-09-2009

If some of th eIP blocks work, some dont, It does not sound like its the IP that has the problem .

Are you feeding them all correctly ?

Try feeding the same info to all the blocks in simulation,

 

<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>
0 Kudos