MHz as a function of VCC?

Go To Last Post
13 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So if its possible to run an xmega at 32MHz at 3.3V, why cant they make a mega avr with the same process and run it at 32MHz and 3.3V? What would you call it? a Ymega?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am a little confused at what you are questioning Bob. From what you illustrate above, I interpret the same thing that the Xmega will run at 32MHz at 3.3v. Or it can be said that the xmega can run at 32MHz and while it's powered by 3.3v.

IIRC Atmel datasheets always spec operating frequency AT(@) specified voltage.

Am I missing something? :?

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

To move a design to a new fabrication process is nqormally not a trivial matter unless the design is synthesizable. The xmega was the redesign for a newer process. With prcoessors like the ARM it is a bit like arduino for ics- you choose what you want from a library of bits and pieces and tradeoff, speed,power and size. You then press the compile button and out pops an ic design. You then go make a million of them.

Fpgas, ram and flash normally follow new process changes quickly as they are basically one cell copied millions of times.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So nobody thinks a 32MHz 3.3V AVR is a good idea?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sounds like a good idea to me. Would it hurt sales of xmega devices? Then how about a 50Mhz xmega, that might fix things for the 32Mhz mega.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Its a great idea. Even better at 64MHz. Even better would be 32MHz at 5V to 1.8V Problem is that the small geometry transistors needed for fast operation have low breakdown voltages.

Will it happen? Probably only in our fantasy.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
So nobody thinks a 32MHz 3.3V AVR is a good idea?

That is what a xmega is!
A redesigned AVR.
The RAM-buss is redesigned to deal with the higher speed.
And you can only get the higher speed with a PLL which only the xmega's have (from a x-tal), and with a intern PLL you can better fiddle with the phases of the clock, to match different delay's.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree that the speed performance comes down to the geometries and fabrication processes.

In fact, the A devices have got different speed ratings. I presume that the 'old' chips with V and L variants all come from the same wafer as the non-L chips. The old processes were less consistent. L and non-L were determined at the wafer testing stage.

What seems foolish to my mind is the way that Atmel publish the ratings. Personally, I would simply state 20MHz for 3.3V, derate for < 3.3V and not give any information about > 3.3V

So when they went to 'A' fabrication, they should have designed for 20MHz @ 3.3V
Distributor inventories are simpler with a single device.

In a stroke, customers would give up 5V.

AFIK, this is what M*crochip have been doing for the last 10 years. i.e. design for 3.3V operation with 5V tolerance.

Pure speculation. I am only a hobbyist.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That they still make "real" 5V chips is one of the main reasons that I still use AVR's.
And that is an other reason why they can't run as fast (the glass layer in the gate needs to be thicker to stand 5V).

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

That is what a xmega is!

+1

But what's the goal here? If you just want more CPU horsepower and 3.3V why wouldn't you just use Cortex? You can have 50-170MHz, 32bit operation, 2/3 stage pipelining and even DSP MAC and SIMD operations if you go for M4. In total I guess you can get performance that is 20-30 times that of a 20MHz mega.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That only applies to the pin drivers. The core can run on whatever voltage the designer chooses. Many 3.3V chips run with a 1.8V core.

I presume that M*crochip sort out their designs ok. AFIK, they are completely 5V tolerant.
I am always wary of ARM chips that claim to have 5V tolerant GPIO but are less revealing about 'alternate functions' of their pins.

It annoys me that a ATmega324PA will not run at 20MHz and 3.3V. i.e. it is a later design fabricated on a modern geometry.

Of course, 99% of chips will run fine at 20MHz and 3.3V. As a hobbyist I probably never meet the 1% that may be iffy at some extreme temperature. A commercial product can never take the risk.

I also presume that the ARM chip has no problem with the size of core or its complexity. Most of the silicon real estate is in the memory arrays and the pin-drivers. The AVR is in contrast a fairly low-tech chip.

I agree with you. AVRs are nice and simple to use.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
It annoys me that a ATmega324PA will not run at 20MHz and 3.3V. i.e. it is a later design fabricated on a modern geometry.
yes that is why it has a max speed of 20 MHz, and since it is a 5V process it is at that voltages(+-).
If you want 3(.3)V AVR's go for the xmega's.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Does anyone have applications for AVRs that need > 20MHz?

20M @ 0.5-1MIPS is a lot of activity.

regards
Greg