ATMEGA809 EVSYS Errata

Log in or register to post comments
Go To Last Post
5 posts / 0 new
Author
Message
balisong42
balisong42's picture
Level: Hangaround
Joined: Fri. Mar 20, 2009
Posts: 401 View posts
Location: Seattle, WA, USA
#1
Posted by balisong42: Thu. Mar 24, 2022 - 01:52 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've come across an issue with the event system on ATMEGA809. It seems only channels 0 & 1 are function. Channels 2-7 are non-functional.

These two code snippet both work (I see the timer tick on EVOUT):

	EVSYS.CHANNEL0 = EVSYS_GENERATOR_TCA0_OVF_LUNF_gc;
	EVSYS.USEREVOUTF = EVSYS_CHANNEL0_bm;

	EVSYS.CHANNEL1 = EVSYS_GENERATOR_TCA0_OVF_LUNF_gc;
	EVSYS.USEREVOUTF = EVSYS_CHANNEL1_bm;

Higher channel numbers output nothing. Implication are ports C/D/E/F pins cannot be used as event generators since these are not covered with CH0/1.

 

Chip details:

ICE/Studio reports RevA

Chip markings indicate birthday week 25/2019

Errata sheet suggest RevA is equivalent to RevB of larger siblings

AFR part suffix is tape&reel but was deliver on tray (from Digikey) -- probably just supply-chain issue.

 

I have 808s and 4809s I can test with to investigate the breadth of the issue.

 

 

 

 

This topic has a solution.

Tags:

AVR Microcontrollers, AVR and XMEGA, solved
Last Edited: Thu. Mar 24, 2022 - 08:34 PM
Top
El Tangas
El Tangas's picture
Level: Raving Lunatic
Joined: Thu. Jul 14, 2016
Posts: 3715 View posts
This reply has been marked as the solution. 
Posted by El Tangas: Thu. Mar 24, 2022 - 06:08 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You are not doing it right.

 

The channel registers don't take masks like EVSYS_CHANNEL1_bm. The correct way is:

 

	EVSYS.CHANNEL0 = EVSYS_GENERATOR_TCA0_OVF_LUNF_gc;
	EVSYS.USEREVOUTF = EVSYS_CHANNEL_CHANNEL0_gc;

	EVSYS.CHANNEL1 = EVSYS_GENERATOR_TCA0_OVF_LUNF_gc;
	EVSYS.USEREVOUTF = EVSYS_CHANNEL_CHANNEL1_gc;
Top
balisong42
balisong42's picture
Level: Hangaround
Joined: Fri. Mar 20, 2009
Posts: 401 View posts
Location: Seattle, WA, USA
Posted by balisong42: Thu. Mar 24, 2022 - 08:49 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yup, that's it. Guess I shouldn't blindly accept IntelliSense suggestions.

 

 

IntelliSense would certainly be more useful if the io.h files had the peripheral structs-of-register-types unioned with structs-of-enum-types. <sigh>

Top
KIIV_cz
KIIV_cz's picture
Level: Hangaround
Joined: Sat. Dec 13, 2008
Posts: 272 View posts
Location: Czech Republic
Posted by KIIV_cz: Thu. Mar 24, 2022 - 09:26 PM (Reply to #3)
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Lucky me that I always confirm the values of those constants - either by going to the definition, or by pop up info (in microchip studio you have to set cursor in it and look for the value somewhere on top of the code)

Computers don't make errors - What they do they do on purpose.

Top
balisong42
balisong42's picture
Level: Hangaround
Joined: Fri. Mar 20, 2009
Posts: 401 View posts
Location: Seattle, WA, USA
Posted by balisong42: Thu. Mar 24, 2022 - 10:26 PM (Reply to #4)
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'll certainly be using the drop-down list of typedefs in the future. I use it all the time for functions, just ignored it in header files.

 

Goto Definition is kinda buggered for me though. I have five different processors used in the solution. It's hit-or-miss if the correct io header opens.

 

Multiple procs may contribute to my issues with code-completion as well. Across the processor lines the event system naming is pretty divergent.

Top