cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
matth32
Observer
Observer
333 Views
Registered: ‎06-05-2018

GTH Rx termination and Swing

Hello,

I have one question regarding the receiver interface of GTX/GTH of Zynq Ultrascale + and Zynq 7000.

In UG476 and UG576, we can see a protection diode connected to GND and MGTAVTT for P and N inputs.

If I take for exemple the use mode 2 of UG476 (p 172), with a term voltage at AVTT, a max swing of 1200 mVDPP is allowed. So this correspond to 600 mVPP in each single ended signal (P or N) which correspond to +/- 300 mV around the common mode.

If we have a common mode at AVTT on the RX side, we will have AVTT+300mV on each input pins. But on the absolute maximum rating the value is AVTT.

Could you explain me or point me what I'm missing, please? Is this 300mV are allowed due to the voltage diode?

Thanks

0 Kudos
6 Replies
eschidl
Xilinx Employee
Xilinx Employee
290 Views
Registered: ‎10-19-2011

Hi @matth32 ,

you are right, if you put the max values from datasheet together this way you run into a violation of the absolute maximum rating. And you should follow the absolute maximum rating. It seems to mean that you cannot have any swing at the input.

But consider, the termination is done inside the device and the datasheet values are valid for the pins. The situation at the pins will be build depending on the drive from TX and the internal RX termination.

You have to find out, what the situation in your setup will be at the pins and if you will violate the absolute maximum ratings. In most cases you will not see the mentioned swing at the RX inputs.
If the termination to AVTT would lead to a violation you should consider a different termination value.

------------------------------------------------------------------------------
Don't forget to reply, give kudo and accept as solution
------------------------------------------------------------------------------
0 Kudos
matth32
Observer
Observer
250 Views
Registered: ‎06-05-2018

Hi eschild,

Thanks for your answer. I have another question.

I know that we have to don't apply more voltage than the absolute max rating but due to the esd diode, is there any transients voltages allowed if there are current and time limited?

I search the maximum current for this diode in all documentation without success. My question is when the component is powered but also unpowered.

Thanks

 

0 Kudos
joancab
Teacher
Teacher
246 Views
Registered: ‎05-11-2015

I think the family's DC and AC datasheet mention the maximum time protection diodes can conduct for different current levels.

0 Kudos
matth32
Observer
Observer
231 Views
Registered: ‎06-05-2018

The datasheet mention a current (I_in parameter) for these diodes but I'm not sure if it's also applicable for the GTH diodes.

0 Kudos
joancab
Teacher
Teacher
229 Views
Registered: ‎05-11-2015

My approach is to never design stressing (putting any voltage on them, even before conducting) the protection diodes. My understanding is that they should only see EMI spikes and noise, so I almost never look into their ratings. Besides that, when they say the abs max current of something is 20 mA, it doesn't mean that 19 mA is perfectly safe and 21 mA will blow all of them up. Devices have variations, dependence on temperature, etc. Another reason to take these limits with a pinch of salt and keep an extra margin if what you want is to be on the safe side.

0 Kudos
matth32
Observer
Observer
152 Views
Registered: ‎06-05-2018

I agree, I never design a board with the goal to overstress any protections.

I just need to know if the diodes on GTH are the same than diodes on all other pins in order to evaluate impact about some failure on the board.

0 Kudos