Has anyone ever held one of those pico power devices in his hands? I am looking for a substitute to an MSP430F135 which is running an RTC with <2uA. An Atmega324P would be perfect, but is it available?
I do not believe in promised delivery dates from Atmel and its distributors any more.
Does anybody have practical experience with the P-chips or other devices running an RTC at less than 4uA?
Jörg.
Has anyone ever held one of those pico power devices in his hands?
I am looking for a substitute to an MSP430F135 which is running an RTC with <2uA.
Hmmm--I've gotten Mega48 below 1uA with many wakeup sources enabled, but that was with the (coarse) watchdog timer for timekeeping. Are there specs for any AVR that can keep a clock going at 2uA? Will the '164 be less than the '48?
Anyway, back to
An Atmega324P would be perfect, but is it available?
I do not believe in promised delivery dates from Atmel and its distributors any more.
The first thing you do is to sign up for the samples on Atmel's Web site. Yes, "enforcement" is spotty but it gets the official request into the distribution chain.
Now, in my area the rep is active. That may not be true in your area. After I put in my sample request, I contacted the rep and repeated what I needed. That does several things: I can find out from the rep (or the rep finds out from Atmel, or puts me in direct contact with the district Atmel person) on whether the part is "real" or not--as we put it, can SOMEONE hold one in their hands? The rep also knows, or will as the district Atmel contact, if there are customers in the area that have received a lot of samples and coupld perhaps spare a few. I've also had useful contact with these customers and worked together on applying a new model.
An active distributor will also do the legwork with the rep that I described above.
Be easy on the distributors. Remember that (in general):
--they can only sell (delivber) what they can get
--they'd be happy to sell it to you if they could get it
--the dates that they give to you for delivery are what Atmel tells them
DigiKey seems to expect delivery of that family in early February.
I received some '164s about a week ago, about a week after my request and contact with the rep. They were the exact package and model that I requested. They were not marked ES, and the lot was produced a while ago--week 23 IIRC.
We had a Mega16L app that was having problems at low Vcc levels. I asked for '164/'324 samples, and went ahead and got some (old style) Mega644V to try. The '644V solved the problem and was put into a small batch. The '164 was applied ansd seems to solve the problem as well, but not enough time on it before the holidays.
Given the size of the migration document, the port from Mega16 to Mega644 was relatively painless--a couple of hours.
Summary: I'd think that you can get small quantities to try in your app if you shake all of the trees.
[edit] The prelim datasheet that I have here does indicate less than 1uA in power-save with RTC running, but the detail under PRR indicates about 10uA for PRTIM2. So maybe you will be successful. It is a bit confusing. Report back on your results. (Try a '644V fro reference if you get a chance.)
Lee
Lee, thanks for the info. Problem is, I have to decide within the next days if I stay with the MSP or change to AVR for this project (~20k/year, production of the new version to start end of february).
The 48/88/168 in power save with 32kHz osc running should draw ~6-7uA accoring to datasheet, the P-chips should do that with <1uA. 6uA is quite a lot, using (the same) CR2032 lithium cell will give less than 4 years of backup time (compared to >10 years for the MSP).
Hmmm..., the MSP offers a 12-bit ADC, price is ok at ~$3.50, but 16k flash is a bit short of what I would like to have and I want to get rid of that controller family in my products. If only Atmels scheduled delivery times could be trusted.
Jörg.
If only Atmels scheduled delivery times could be trusted.
Lee
