XMEGA

Go To Last Post
440 posts / 0 new

Pages

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

Quote:

My supplier at EBV have placed the order for a few thousand units, starting with 5 samples. Hopefully I will get the confirm tomorrow.

LOL.
Quote:

But after reading this topic, I wonder if it's not that easy after all..

If the A1 models were filling in nicely at the distis, then I would expect the announced A3 and A4 models to follow along in due time.

But as the silicon for the first Xmega model has been very slow to reach "available now" status, when exactly would you expect to get a production supply of an A4 when even Atmel's marketing seminar says will not be available till next year?!? (And what makes you think that the '16A4 will be the first available model in the A4 series?)

So worrying about ISP programming that particular model will be the least of your worries. As the architecture is essentially the same I'd expect no significant ISP or code generation differences so I don't think compiler support or ISP support will be a problem--once you can get your grubby hands on some chips.

Re "can't use 100 pin" -- that seems to me to be the only reasonable alternative if you want to start dev. now. Perhaps a few 64-pin samples by the end of the year.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Wed. Sep 24, 2008 - 04:20 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Probably am I too optimistic..

;)

But can't you lay down a prototype design with pads for the 100 pin device just for development then later respin the board or even start out with a dual site (44 / 100) design?

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

Or a true bench prototype with STK600 and wires and cables to peripheral subsystems.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

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

Quote:
But can't you lay down a prototype design with pads for the 100 pin device just for development then later respin the board or even start out with a dual site (44 / 100) design?
Worth consider.
Quote:
Or a true bench prototype with STK600 and wires and cables to peripheral subsystems.
I don't think our beta testers likes the idea of having the STK600, wires and cables out in the field.. :)

All of our earlier designs have used well known avr devices. So I obviously did the mistake thinking that devices listed at avr.com/Products was availible.

We will have to rethink this project a bit. And saying a word to that EBV Field Sales Engineer woman...

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

If you have not already, look at the Nu Horizons Webinar slides, and/or listen to the presentation.

The US distis have listed the announced models. Although I don't expect stock to be there, at least we can get an idea of the pricing structure.

Hmmm--

Quote:
start out with a dual site (44 / 100) design?

Might be tight for chip-inside-a-chip as there will only be about 2mm between the inside of the pins on an A1 to the outside of the pins on an A4. But maybe everything will align nicely 1-to-1.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

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

I know it'd be a routing nightmare but I was actually thinking about two completely distinct IC locations - obviously board space is probably not the issue for a bench prototype that it is for the production design

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

@Diff
Perhaps you can use this workaround:

"Solder down adapters from XMEGA A1 to XMEGA A3 and A4 device footprint are now available from Winslow Adaptics.

With the adapters it is possible to take the 100-pin ATxmega128A1 device and solder it to a board with a 64-pin XMEGA A3 or 44-pin XMEGA A4 footprint. This is intended for customers that will use a future XMEGA A3 or A4 device in their production, but want to start development with the currently available ATxmega128A1. Some will design their final board now, and a 100-pin device will not fit so they need an adapter.

For one complete adapter they will need one 100-pin interface module where the ATxmega128A1-AU (TQFP) can be soldered on to, and one selectable solder-down module with a 64-pin or 44-pin TQFP or QFN/MLF

footprint.

Ordering and lead-time details:

W12061SDF - 100 Pin TQFP Interface Module:

http://www.winslow.co.uk/product...

W12059M - 44 Pin TQFP Solder Down Module:

http://www.winslow.co.uk/product...

W12062M - 44 Pin MLF Solder Down Module:

http://www.winslow.co.uk/product...

W12063M - 64 Pin MLF Solder Down Module:

http://www.winslow.co.uk/product...

W12060M - 64 Pin TQFP Solder Down Module:

http://www.winslow.co.uk/product...
"

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

Any one know how to migrate Atmega128L to Atxmega28i or where to find the instruction to do this.

Thanks
Hing

goujam wrote:
Any one know much about the AVR XMEGA chips, I have just had someone arrange to come and see me next week to show me them. It would be nice to have some background so i know what he is talking about.

thanks

James

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

This is the only way I see to figure out your program code migration are the documents below, your own research & time spent.

ATmega128 data sheet:
http://www.atmel.com/dyn/resourc...

Xmega data sheets:
http://www.atmel.com/dyn/product...

Xmega application notes:
http://www.atmel.com/dyn/product...

The application note AVR1000: Getting Started Writing C-code for XMEGA, explains the new naming conventions and module organization used in the Xmega.

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

[code]

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

An order I placed 6 months ago for some Atxmega64A1's just shipped from mouser today. Looks like they're in stock!

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

Laenfinehack wrote:
An order I placed 6 months ago for some Atxmega64A1's just shipped from mouser today. Looks like they're in stock!
They have listed a few hundred as in stock. If this is all they have they won't last long.

And Avnet has a few hundred 128A1 in stock at the moment. They mark them as "top sellers". No wounder they are top sellers, they are the only xm*gas Avnet has to sell. All other types are out of stock.

So *tmel's game of infinitesimal supply of this product of the (last) year continues. Maybe someone wants to run a betting pool? Where will the next xm*gas show up, what xm*ga type and in which quantity?

Stealing Proteus doesn't make you an engineer.

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

As mentioned Mouser have a stock of ATXmega64A1's (315 at present I seem to remember they ordered 380). Interestingly they have not re-ordered and the lead time is quoted as 38 weeks.

Digikey had some A1's a liitle while ago but they did not last long.

Nu Horizons have some ATXMega128A1's

This leaves a number of questions:
1) Are they the old Rev G with its bug list.
2) What of the export licence problem. Digikey still restrict orders to North American customers. Nu Horizons only list stock in North America and they are marked export controlled.
3) When can we expect to see some A4's

Trevor

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

Quote:
hopefully getting delivery dates for 16A4,32A4,64A4 or 128A4 samples. Probably am I too optimistic..
Welcome aboard you've joined the XMega club. Now all you need to do is wait a 'few' months and you just might get a 128A1, ONE of them, and it probably won't work. There's no rush for this project is there ?

<º))))><

I am only one lab accident away from becoming a super villain.

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

Quote:
I have just had someone arrange to come and see me next week to show me them. It would be nice to have some background so i know what he is talking about.
The new XMegas can run up to 1GHz, they have 4Gigs of RAM, a built in hard drive and they can talk to satellites. They are all in production, available for next day delivery and they come as standard encrusted in diamonds and precious gems.. trust the atmel sales team, trust them, why on earth would they lie to us ?

<º))))><

I am only one lab accident away from becoming a super villain.

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

Just got some chips today.

Where do I find the chip rev?

Atmel
ATXMEGA64A1
AU 0913

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

The last letter at the end of the second line on the underside of the chip.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

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

Okay H.

It's a good thing I didn't solder it to a board yet.

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

Quote:
Where do I find the chip rev?

Quote:
last letter at the end of the second line on the underside of the chip

I think on the Xmega's Atmel ought to print the errata (rev) level in BOLD on the top of the chip, so you can remember to look at the errata on the data sheet to figure out why your proto will not work right! :roll:

what happens if they run out of alpha characters before they get errata free silicon? 8)

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

The Alpha equivalent of the Dewey Decimal System, of course...

AA, AB, AC, ... ZZ, ZZA, ZZB, ZZC...

:)

JC

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

Just spotted the reference to ATXmega D series processors in the Studio release notes.

No info on the Atmel website. Any ideas ?

Could this be a version without the bits that caused the mysterious "Export Licence" problems?

I am not even going to ask when to expect availablility. A reasonable person might expect the "A" series before the "D" series is announced. :?

Trevor

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

In August 2008 Atmel offered us the 16D4 at an extraordinarily attractive price (undercutting any existing 16K part and even some 8K) but sadly (perhaps happily?) it was no good to us as the single sided board we planned to use (for cost purposes) was to be reflow soldered and we needed a DIP package. Maybe we got off lightly not waiting for the vapouware?

(I couldn't mention it previously until Studio recently shipped with a mention of the 16D4)

When asked for an 16D4 datasheet at the time they just gave us a copy of the generic "A datasheet" and the suggestion that, because of the cross family compatibility, we should just develop with 128A1 and then switch when the 16D4 became available.

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

Of the 14 xmega types now officially listed on Atmel's web site (all are A type), Atmel's inventory check only lists some kind of stock for two of them.

So, there are twelve vapor devices in the queue before the D devices.

And still Atmel doesn't have the balls to honestly communicate what's going on.

Stealing Proteus doesn't make you an engineer.

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

clawson wrote:

When asked for an 16D4 datasheet at the time they just gave us a copy of the generic "A datasheet" and the suggestion that, because of the cross family compatibility, we should just develop with 128A1 and then switch when the 16D4 became available.


Strange things happen when a product goes through the vapourware phase.

I have spotted small quantities of three different A1 chips with distributors. No idea what revision levels they are.

No sign of A4 I suspect the only warning we will get is when the data sheet gets information in the errata section.

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

Well also note that Atmel have a quite a different relationship with their "big customers" (Chinese Microwave Oven and washing machine manufacturers) than they do with small time outfits and sales through distribution. It could well be that the quickest way to get a 16D4 might actually be to go out and buy a cheap microwave oven and disassemble it! (also look under the hood of your car - there could be one lurking there). In fact it's likely that they prioritise the order in which they fab various models according to who's willing to place the 1m/5m/10m sized orders. If they fab 10m chips for some order it's possible a few 10K's get out to distribution too.

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

clawson wrote:
Well also note that Atmel have a quite a different relationship with their "big customers" (Chinese Microwave Oven and washing machine manufacturers) than they do with small time outfits and sales through distribution.
Atmel's SEC filing 10-Q indicate they make around 47% of their net revenue through distributors, and have (needed to?) change contracts with some European distributors so Atmel takes more risk. Distributor business is certainly no small fish for them.

If the twelve MIA A types and that 16D4 already exist in volume and the distributors don't have access to them, well ... I hope Atmel hasn't decided to piss of the outfits contributing 47% to their net revenue.

At least they know the risks. In the long litany of potential risks they list relations with big customers and with distributors.

Stealing Proteus doesn't make you an engineer.

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

Here in the UK Atmel keeps chopping and changing their distributors, which doesn't inspire much confidence. I had a very good relationship with one of their previous distributors, but I have had problems with the current batch, apart from MSC, which is a new one I've never heard of before. Perhaps I ought to try them.

[Edit]
I see that MSC is in fact a German company, so is probably as bad as the others.

Leon

Leon Heller G1HSM

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

Recently I sent a total of 3 sample requests to Atmel with respect to the ATxmega64A4.

Not a single person has sent me a courtesy email to acknowledge my request.

Why is it that I can order free samples from National/Maxim/TI and I will receive them within 3 days, yet Atmel drag their feet just trying to contact my local distributor who will take their time getting back to me.

They produce such great products but they seem to be spending more money on marketing than producing bug free silicon, creating good customer service and putting samples into the hands of developers like myself.

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

magellan,

I suspect Ostrich Syndrome. There aren't any 64A4's yet so they figure that if they bury their heads and pretend you never asked they won't need to explain their vapourware strategy to you.

If you want to work with Xmegas then start with 128A1's (which are available) then just down-size if/when the A4's finally arrive

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

I am inclined to agree with you.

I use the AVRs and the AT91 ARM9 series in several products. Atmel seem to be spinning their wheels a little when it comes to releasing bug free silicon.

I was hoping that the A4 series might have a much smaller (or empty!) errata sheet. AFAIK, all the AVRs have zero errata and that's the way it should be for a simple CPU. I can understand errata for a large and complex ARM9 chip though.

I could start my project with the A1 and then scale down. Size is fairly crucial, so I will need the smaller chip eventually. Trouble is that commercial deadlines can be fairly tight ... and we've already had to skip several Atmel CPUs because they just didn't appear in time. i.e. AT91SAM9260 rev B silicon ...

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

So, anyone have any more recent XMega news?
I REALLY like the looks of the Xmega16, though if it isn't going to be available any time soon, I may end up going with a Mega162.

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

I've given up with xmega and I have since switched to AVR32. Go 32bit! Not much difference in price or size, but a whole lot more speed and peripherals.

The only downside is AVR32Studio. It is based on Eclipse which I find to be the worst IDE that I've ever used. AVRstudio just "works", AVR32studio just confuses ... once I get over it I'm sure I'll like it. I just hate it at the moment.

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

Quote:
I may end up going with a Mega162
Unless you have some board that needs that pin out use the Mega164p as they are newer and have same pinout as the M324p, 644p, 1284p.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

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

Atmel just released, within the past few days, their Cortex-M3. I currently use a Cortex-M3 (STM32) and really like it. You could also look at that.

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

Does somebody have an idea, when the erreta of the xmegas will be fixed? Perhaps never? Because the idea is nice, to have 12 bit ADC, and DAC on the chip, but it has more limitations than use. So it I want a reliable soulution I have to use an external devie anyway...

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

No real updates on them. Lots of people seem to be having issues with buggy silicon these days - even TI's MSP430F5438 has some big "this feature basically is broken, don't use it" erratas.

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

I just requested a small quantity of the ATxmega64A4-MU and was told that this part is obsolete and I should rather design with the -MH part. But upon ordering the -MH, I was told it's being redesigned and will only be available in Q1/2011. I checked on Mouser and sure enough, it's marked "obsolete". WT*?

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

I voiced my suspicion that a major redesign was under way for the A4 chips a couple of months ago when supply difficulties first emerged.

Clearly Atmel were too ambitous in thier specification for the first generation Xmega's. Hardly surprising the result was:
a) buggy silicon
b) extremely buggy documentation.

Having said that my first project has delivered a product that has extremely impressive performance and if anything exceeds my expectations. Just a pity about the delays caused by the poor documentation.

Atmel now need to fix all the bugs and go for a relaunch as soon as possible. It would be a shame if delays damaged the reputation of the Xmega any further.

Meantime they should provide a full bug list for the silicon and state whether individual bugs will be fixed in the silicon or in the documentation. :wink:

OK I know I am being unrealistic but I can hope :roll:

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

Having so many bugs, I see as a good strategy to take it out from the market for a while in order to fix all the bugs, or at least most of them instead to keep it on the market and fill the market with lot of revisions, application notes with workarounds.

And yes, too ambitious for an 8 bit microcontroller.

George.

Pages