RF12 driver initialization delays - who is running into this?

Jee Labs Talk » How can I fix ... ?

RF12 driver initialization delays - who is running into this?

(13 posts) (4 voices)

Tags:

No tags yet.

jcw
(In response the a recent post, which appears to be related: )
I suspect that there is a problem with the RF12 driver power-up initialization, specifically with this code in rf12initialize():
```
    while (digitalRead(RFMIRQ) == 0)
        rf12xfer(0x0000);
```
It's only a problem after true power-up, as far as I can tell, not after a reset (including one caused by the serial/USB connection. Which is not really surprising, since the reset only affects the ATmega, whereas the problem appears to be the RFM12B (it only gets a hardware reset on power up).
I'd like to chase this bug and nail it down, but last time I looked into this, it turned out to be very hard to reproduce. It may be related to timing, voltage levels, or perhaps moon phases?
Maybe we can nail this down by working together on this. What I'm thinking of is to check in a number of different variations of the code, so that you can get different versions and help compare what's happening on your end. The basic idea is to let you follow the changes on the time line over here: .
All the changes will be in RF12.cpp - so if you want to get the latest version, which is revision 6548 right now, click on the number, then on "RF12.cpp" (not the "diff" link), then on the "Download" link at the top, to get that particular version.
I think I have one node which tends to take a few seconds when powered up. I'll look into this tomorrow.
If you've got a JeeNode, JeeNode USB, or JeeLink which seems to take a long time before it works properly once plugged in and powered up (seconds, perhaps even minutes?), then please keep an eye out for changes to RF12.cpp on the above time line.
I'd rather not clog this forum with the debug discussion we may need to figure this all out, so I've opened up an issue in Redmine which can be used instead for that discussion - it's at . There is one small inconvenience involved, though: you'll have to register on Redmine (top right), and I'll have to accept the registration (manually), to be able to add comments on the issue tracker.
I'll try to come up with some ideas on how to debug and fix this problem over the coming days. Given the erratic nature, the more people are willing to help with testing and feedback, the better our chances are of fixing this problem once and for all.
Posted 23/01/2011 01:20:04
tankslappa

Hi JC,
I'm willing to help test. I have JeeNode V4 and V5's available, and a "JeeNode on a breadboard" aka JeeNode circuit, but wired together on a bread board.
Not sure what you mean about the Redmine issue for comments, that's a bugzilla kind of thing isn't it, I don't think we can comment on that even once logged in.

Posted 23/01/2011 14:38:20
agorts

I have a v4 JeeNode with similar behaviour (http://talk.jeelabs.net/topic/737)...
The difference is that it keeps waiting for initialization for longgg... I let it for 50 mins but it did not continue...
If you think that this is one of this case, then I am at your disposal in order to track it down...

Posted 23/01/2011 18:11:30

agorts

I used the rf12demo sketch... and I put some trace commands (serial.println) in the RF12.cpp

Here it is the log:

[RF12demo.6-RFDebug]
-) [rf12config] Beg...
   A i1 g212 @ 868 MHz 
-) [rf12config] Before rf12initialize...
  -) [rf12initialize] Beg...
  -) [rf12initialize] Before wait for 'FM12B is out of power-up reset'...

The code stops in the lines:

while (digitalRead(RFMIRQ) == 0)
        rf12xfer(0x0000);
but it does not exit!!!


				Posted 23/01/2011 20:08:47


	
		
			
								
					hacker007

				
			
			
				
I had the same problem when using RF12, but not RF12B.



				Posted 24/01/2011 11:04:56
			
		
	
	
		
			
								
					agorts

				
			
			
				
I refresh the solders and now it works...!



				Posted 24/01/2011 17:49:36
			
		
	
	
		
			
								
					jcw

				
			
			
				
Ok - looks like I was chasing a false alarm, eh, or something...
See .



				Posted 24/01/2011 23:09:13
			
		
	
	
		
			
								
					tankslappa

				
			
			
				
That's great news. I'm glad you read cross-threads and gave it a try, I hadn't thought an iffy connection would have these symptoms.
Just out of interest, are you using lead free solder?
@JC back to fun stuff then :-D



				Posted 25/01/2011 01:32:12
			
		
	
	
		
			
								
					agorts

				
			
			
				
What is lead free solder? Is this a type of soldering?



				Posted 26/01/2011 21:23:31
			
		
	
	
		
			
								
					tankslappa

				
			
			
				
It's a newer solder (brought in by legislation not because it's better). It doesn't contain lead, so it's greener (although I'm sure the mining operations are anything but).
Anyway, I always groan when I have to repair a circuit that has been put together with it, because my soldering iron isn't really hot enough to melt the damn stuff properly!
I was just wondering if any of you guys were using it, and if that was causing the bad solder joins. Basically just trying to find a cause/pattern.



				Posted 26/01/2011 23:19:18
			
		
	
	
		
			
								
					jcw

				
			
			
				
> What is lead free solder?
First hit on google: 



				Posted 27/01/2011 09:01:13
			
		
	
	
		
			
								
					agorts

				
			
			
				
NO I am using plain 60/40 solder...



				Posted 01/02/2011 19:06:35
			
		
	
	
		
			
								
					tankslappa

				
			
			
				
Hmmm, no excuse then :o)
I was just trying to understand why we had a sudden influx of bad solder joints.



				Posted 02/02/2011 07:47:52