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!

cancel
Showing results for 
Search instead for 
Did you mean: 
Observer squaringcircle
Observer
222 Views
Registered: ‎12-19-2018

Automotive Artix 7 temperature impact

Hello,

 

I'm planning to use a XA7A100T in an application with an ambient temperature of 125°C.

The design will be simple and small - just a serial interface at 20MHz controlling some IOs for LEDs, etc.

In addition, I'm planning to use an internal PLL to generate 400MHz from an external 50MHz source.

Now considering the self-heating of the FPGA it is quite likely to exceed the specified 125°C junction temperature.

 

Does anybody have experience or can anybodycomment on what the impact is if the junction temperature would reach e.g. 130 or 135°C?

As mentioned, the design is not complex except the internal PLL -increased jitter and frequency accuracy would not be an issue, as long as the clock keeps running.

 

Many thanks!

0 Kudos
7 Replies
Highlighted
Scholar dpaul24
Scholar
213 Views
Registered: ‎08-07-2014

Re: Automotive Artix 7 temperature impact

@squaringcircle,

@work I am also using Artix7 based devices targeted for automotive use. The max temp specified is 125 deg C.

I perform device qualification tests using the devices inside heated ovens where the A7 junction temp reaches 112 deg C (with environment temp = 70 deg C). More than this I have never observed. But since the specified max is 125 deg C, so no problem. It is a huge design with gates always toggling.

Makes me curious in what environment do you plan to use your design!

I guess at any temp over the specified limit, correct/stable operation is not guranteed.

 

As mentioned, the design is not complex except the internal PLL -increased jitter and frequency accuracy would not be an issue, as long as the clock keeps running.

For a simple design, if I assume the device environment temp to be 100 deg, it may be the case that the A7 cannot itself generate enough heat to cross 125 deg. Again your question highly depends on the environment temp where you wish to operate,

 

--------------------------------------------------------------------------------------------------------
FPGA enthusiast!
All PMs will be ignored
--------------------------------------------------------------------------------------------------------
Xilinx Employee
Xilinx Employee
176 Views
Registered: ‎06-30-2010

Re: Automotive Artix 7 temperature impact

The absolute max Tj for devices is 125C, if your Ta is 125C and you simply power the device on then teh static alone will self heat the device over the absolute max, so this would be operating outside of the specifications of the device. 

 

Is this Ta 125C an excursion or fail-safe spec, or is the device always going to be operating at this temperature? 

I assume it will be nowhere near the passenger compartment?

 

Thx,

John.

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
Observer squaringcircle
Observer
158 Views
Registered: ‎12-19-2018

Re: Automotive Artix 7 temperature impact

Hello,

 

Thanks both of you for your response.

@dpaul24  I appreciate your experience and description with self-heating of the device - really useful input for us.

@jheslip Thanks for your questions, I have been lacking a more detailed application description.

 

FIrst of all, the device will be used in a pure lab-environment for testing. No safety curcial application.

Worst case would be falsified measurement data.

 

The device will be exposed to the 125°C ambient for about 48 hours, then another 48 hours with 25°C, then back to 125°C etc. The entire cycle will be around 10 days.

As a backup we've now added a peltier controller for the worst case.

 

However, I do appreciate any further inputs / experience / assumptions of how the internal PLL would behave if operated above 125°C.

 

Many thanks!

0 Kudos
Xilinx Employee
Xilinx Employee
129 Views
Registered: ‎06-30-2010

Re: Automotive Artix 7 temperature impact

explain the process of what you are doing in more detail?

When Ta is 125C will teh device be powered and operational? or will it just be soaking at 125C?

If it is powered or operational then the absolute max will be violated (due to static alone, even more with dynamic) and so potential damage could occur so this will be run out of specification.

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
Observer squaringcircle
Observer
117 Views
Registered: ‎12-19-2018

Re: Automotive Artix 7 temperature impact

Correct - TA=125°C, the device will be powered and operational.

According to Xilinx power estimator consumption wilöl be in the range of 1.2W and TJ will exceed the 125°C.

In case we risk damage of the device we will think about a cooling option.

This basically covers my question quitee well - I mainly wanted to understand the impact of operating the device above 125°C.

If it was "only" performance degrade or inaccuracy, it would have been ok. But we do not want to risk device damage.

 

0 Kudos
Xilinx Employee
Xilinx Employee
107 Views
Registered: ‎06-30-2010

Re: Automotive Artix 7 temperature impact

once you violate the Absolute max then there is the potential for damage, if the device is run of out spec we can't quantify the impact either.

 

in terms of a heatsink, if Ta is 125C then you have 0C margin I am not aware of any heatsink that will be able to cool the device and ensure that the Tj is below teh Absolute max. Also, don't forget the sysmon error needs to be factored in so that will impact teh Absolute max, so that is 125 C - SysMon error. I don't have the datasheet pen right now but if teh error is +_3C then that means the Tj max is 122C.

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
105 Views
Registered: ‎09-17-2018

Re: Automotive Artix 7 temperature impact

sc,

Well, you are not the first to exceed the recommended ratings, nor even the absolute maximum ratings.  20 years ago, Xilinx devices were used in oil drilling, where down-hole temperatures could reach well beyond 125C.  Tricky, but the drilling companies figured out how to keep them alive long enough to be useful.  Obviously, Xilinx is not going to guarantee anything here, but they are certainly willing to sell you product that you will likely wear out or destroy eventually from the abuse.

I suggest you do the best job you can with heat sinking and airflow to limit the temperature rise above 125C.  If you stay below 135C you should be OK for long enough to be useful.  Above 135C and the package/die may fail, above 150C are you will likely fail.  Be sure you over constrain your timing by 20% at least (if you need to meet a 100 MHz clock, constrain to 120%.

Good luck.

l.e.o.

0 Kudos