UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

Reply

MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

Accepted Solution Solved
Highlighted
Visitor
Posts: 4
Registered: ‎02-22-2014
Accepted Solution

MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

What is the 7 Series MMCM's behavior after its PWRDWN port is asserted and then deasserted?

 

I have looked at UG472 (Clocking Resources Guide, Clock Management Tile chapter 3), and XAPP888 and searched the Xilinx forums, Answer Records, and web without finding answers to my questions.

 

I suppose it would restart (after an interval to re-acquire lock) but using what parameters?  Will it restart using the parameters from the most recent bitstream configuration?  That means the MMCM-associated configuration memory bits are not also powered-down by the PWRDWN port.  Fair enough if so.

 

What about the case where after the initial bitstream configuration, the MMCM has been reprogrammed through its DRP port with parameters (register contents) different from the bitstream configuration.  After a low-high-low cycle on the MMCM's PWRDWN port, will the MMCM restart using the original bitstream configuration or the more recent DRP programmed values.  

 

Is it necessary to reset the MMCM using its RST port after cycling its PWRDWN port?  Sure, anyone could guess it might be advisable.  I would especially like to hear from you if you know it to be necessary or unnecessary and why. 


Accepted Solutions
Xilinx Employee
Posts: 555
Registered: ‎10-11-2007

Re: MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

The DRP functional logic itself does not behave differently for PWRDWN or RST. The "registers" programmed previously through the DRP (or any other once) are not affected either way because they are configuration cells and are only overwritten if you re-program the part or by another DRP operation. Typically, from an application perspective, PWRDWN and RST are identical. The difference is obviously that in the PWRDWN case the MMCM completely shuts down for an extended period of time even if asserted only briefly. Takes a while to bring back the regulators vs simply reset. In addition, since power is turned of, it takes longer to reacquire LOCK vs RST because the VCO starts  from scratch.

 

View solution in original post


All Replies
Xilinx Employee
Posts: 142
Registered: ‎06-30-2010

Re: MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

effectively the question is what is the difference between PWRDWN and RST. With RST, we know it goes into its initial state, therefore, losing its DRP changes.
Does the same happen with PWRDWN or will it retain its last state when released from PWRDWN.

Would you agree?

I'll see what I can find.
Visitor
Posts: 4
Registered: ‎02-22-2014

Re: MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

Thank you for looking into this.

 

I do not agree.  With RST, I do not think it (the DRP) goes into its initial state.  Evidence of this is the recommendation (XAPP888 v1.7 page 12 bottom) to keep RST asserted for the whole duration of DRP programming.  Therefore I think that although RST will reset some things, the DRP registers are not among those things.

 

Xilinx Employee
Posts: 555
Registered: ‎10-11-2007

Re: MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

The DRP functional logic itself does not behave differently for PWRDWN or RST. The "registers" programmed previously through the DRP (or any other once) are not affected either way because they are configuration cells and are only overwritten if you re-program the part or by another DRP operation. Typically, from an application perspective, PWRDWN and RST are identical. The difference is obviously that in the PWRDWN case the MMCM completely shuts down for an extended period of time even if asserted only briefly. Takes a while to bring back the regulators vs simply reset. In addition, since power is turned of, it takes longer to reacquire LOCK vs RST because the VCO starts  from scratch.

 

Visitor
Posts: 4
Registered: ‎02-22-2014

Re: MMCM Behavior After Its PWRDWN Port Is Asserted and Then Deasserted?

Thank you ralfk, that makes sense.