|
Message
From: "Martin.J Thompson" <Martin.J.Thompson@t...>
Date: Wed, 26 Nov 2003 15:10:41 +0000
Subject: Re: [oc] OT : FPGA price
>On Tue, Nov 25, 2003 at 10:54:27AM +0000, Martin.J Thompson wrote:
>> >-no 2A power up spike,
>>
>> None of that in V-II or Spartan-III
>
>the virtex's are specified to draw a big power-on surge actually.
Correct Virtex, Virtex-E, Spartan-II and SpartanIIe, and probably Spartan-XL.. Not Virtex-II
Also, All the Altera FPGAs as far as I know exhibit this behavoiur, but its less well documented.
>Thats if you bring up the supplies in the correct order.
>No idea what happens in the 'wrong' order.
>
There is no wrong order. The amount of current does depend on the ramp time though.
>>
>> Not a good number to use. I *want* a Tjmax - then I can simulate the thermals of my system, with whatever heatsinks and fans I might or might not use and ensure
>that it is not exceeded. Or use the timing tool to derate. Of couse, this needs a good package model. Ambient temps are no use unless you are doing an average
>design.
>
>Last package model I saw was a theta-JA resistance, so max ambient
>still matters.
>As does Tjmax and the (probably data specific) power dissiption.
>
Tjmax is what matters to whether your design works or not. If you have a single FF toggling at 1Hz you can probably run in an ambient close to Tjmax. In a big design all clocking fast, your maximum allowable ambient temperature is lower for the same Tjmax. That's why there no point any vendor quoting me a Tambient max. They know nothing of my design.
Cheers,
Martin
--
Martin Thompson CEng MIEE
TRW Conekt
Stratford Road, Solihull, B90 4GW. UK
Tel: +44 (0)121-627-3569 - martin.j.thompson@t...
|
 |
Browse
