Re: mega128 at 18.432 MHz

Jason Kyle <[email protected]> Thu, 08 Jan 2004 12:27:37 +1300
Newsgroups gmane.comp.hardware.avr.general
Message-ID <[email protected]>
At 17:34 07/01/2004 -0500, Dave VanHorn wrote:
>At 11:10 AM 1/8/2004 +1300, Jason Kyle wrote:
>>At 16:42 07/01/2004 -0500, Dave VanHorn wrote:
>>
>>
>>>>Would be quite useful to know if 18.432MHz is safely achievable.
>>>
>>>If it were, the part would be rated for it.
>>
>>Sure, no harm in exploring the possibility though.  It's not like there 
>>isn't any other CPU or micro in the world that hasn't been overclocked.
>
>Other than it being a colossal waste of time, no.

In your opinion, of course.  Like I said, no harm.



>>>>Of note is the datasheet USART baud rate table showing clock 
>>>>frequencies upto 20MHz.
>>>
>>>That is a holdover, from years ago, when atmel intended to introduce a 
>>>20 MHz version.
>>
>>And prior to that a 24MHz version.
>
>I didn't see that in anything since '95.
>
>>>What is it that you think you can't do at F<=16MHz?
>>
>>What is the engine capacity of your car?  You *could* drive to work with 
>>an 800cc 3 cylinder normally aspirated engine under the hood, but I bet 6 
>>jelly beans you don't!
>
>Clocking a micro faster than you need to is wasting power, and creating EMI.

Yeap.  So what's your engine capacity?



>>IMO you can't know what others are building and what their speed 
>>requirements are :)
>
>Nope. I know what the chip ratings are, and I know that going beyond the 
>speed rating is inviting problems.
>I have enough problems.
>I know that every time I've seen something that "had to be overclocked", I 
>was able to recode it to fit comfortably within the ratings.
>In one case, 4x under the ratings!

It's great you can achieve such good speed and code efficieny 
optimisations, there are however applications that just can't be squashed 
so readily (don't ask for examples as I too don't have time to 
waste).  Feel free to continue operating below the limits.  Some of the 
newer AVR's have clock source switching so you can really slow things right 
down when required which is great for power and EMI reduction.


>>Sure most micro designs are usually clocked faster than they need to be 
>>but what's wrong with some headroom?
>
>Buy fast chips, write faster code, clock only as fast as you need to.
>http://www.dvanhorn.org/Micros/All/Overclocking.php

Read it the first time you posted the link.

As a designer I often find there are requirements for extra features to be 
added to firmware and the only responsible thing I can do is allow 
headroom, my boss would be very unimpressed if I was to tell him I needed 
to re-write a whole bunch of code in order to add a feature because there 
wasn't enough instruction cycles available.

Luckily we all have different approaches and design methodologies otherwise 
we'd all be designing identical products!

Jason Kyle