cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
ccthx1138
Observer
Observer
2,735 Views
Registered: ‎04-27-2017

Powering up first ever FPGA based PCB - Spartan3AN - ramp up current manually?

Jump to solution

Hi guys.  New here.  I'm about to fire up my first PCB design with FPGA - Spartan3AN.  It's finally lunch day. 

I thought i better ask those who know, or can surmise, if i should ramp the current up from my test bench supply (5Vdc) to the PCB, or set the current to limit just over expected current (375mA quiescent on 1.2V, + a few mA for 3.3v and peripheral ICs),  so about ~400mA and apply full power instantly?  Other considerations?  TIA.

 

Cheers,

 

0 Kudos
1 Solution

Accepted Solutions
austin
Scholar
Scholar
4,612 Views
Registered: ‎02-27-2008

c,

 

Set current limit as you suggest.  Power on must not be slower than 50 ms (10% to 90%), and not as fast as below a millisecond or so (all devices are tested to these limits).

 

If you get higher current, check for assembly shorts, solder splashes.

 

 

Austin Lesea
Principal Engineer
Xilinx San Jose

View solution in original post

4 Replies
austin
Scholar
Scholar
4,613 Views
Registered: ‎02-27-2008

c,

 

Set current limit as you suggest.  Power on must not be slower than 50 ms (10% to 90%), and not as fast as below a millisecond or so (all devices are tested to these limits).

 

If you get higher current, check for assembly shorts, solder splashes.

 

 

Austin Lesea
Principal Engineer
Xilinx San Jose

View solution in original post

ccthx1138
Observer
Observer
2,721 Views
Registered: ‎04-27-2017

Hi Austin.  That's great.  Thanks for your help.  So, current-limit, instant-on it is.

 

I'm not sure how fast my bench supply reaches voltage on pressing the power supply 'output' SW. 

But I understand the tight constraints on dV/dT are for successful FPGA start-up, and not a constraint that must be met to prevent damage to the IC, is that correct? 

0 Kudos
austin
Scholar
Scholar
2,718 Views
Registered: ‎02-27-2008

Correct,

 

Only way to damage is to exceed the absolute maximum limits in Table 1 of the data sheet.

 

Or zap it with ESD.

 

I suspect you know how to prevent that.

 

 

Austin Lesea
Principal Engineer
Xilinx San Jose
0 Kudos
ccthx1138
Observer
Observer
2,715 Views
Registered: ‎04-27-2017

Good to know.  Thanks, austin.

 

0 Kudos