View previous topic :: View next topic |
Author |
Message |
hayee
Joined: 05 Sep 2007 Posts: 252
|
Zigbee damaging problem |
Posted: Mon Jan 03, 2011 1:08 am |
|
|
Hi,
I am using Zigbee's from CCS. I brought many Zigbee's transceivers time to time from CCS and I am stuck in a great problem that my Zigbee's are damaging one by one.
I checked the hardware connections lot of time but the connections are right.
I have a system in which there is one coordinator and two sleepy sensors.
When I TURN ON the system, the whole system works well for certain time (time is in hours or in days) but suddenly one Zigbee stop responding (it may be a sensor Zigbee or coordinator Zigbee).
For checking that the damaged Zigbee is working or not I connect that Zigbee in another system but no response.
Can the software create a problem?
I have no idea what to do, can you guys tell me some suggestions how to solve the problem? |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9244 Location: Greensville,Ontario
|
|
Posted: Mon Jan 03, 2011 6:40 am |
|
|
Never, ever seen code destroy a computer.It's just 1s and 0s !
More likely a hardware fault.Overvoltage to PICs, high ripple content,bad or missing grounds,induced HF or ringing. Without specifics of the actual schematic,components used and layout it's hard to say.
Are the Zigbee's 3 volt and the PIC 5 volts?
What transducers are hooked up to the remote sensors?
Any RF in the area(cell phones,CBs,Hams,etc.?
Are the units just laying round open or in aluminum case with proper grounding?
Do they 'die' randomly or say only between 3 and 4 AM?( if you think that's silly to ask, I had 30 remotes 'actup' at that time.Turned out to be crosstalk in the main telephone trunk between two cities.(Bank was updating data during that time period).Changing spanpairs cured the problem.
Can you test the system with just batteries?
The fact they work for awhile then 'die' could point the way to the source of the problem.
Obviously contact CCS to see if others have a similar problem,check for FAQ,etc.
When you find out what it is, post the solution for others. |
|
|
hayee
Joined: 05 Sep 2007 Posts: 252
|
|
Posted: Mon Jan 03, 2011 7:01 am |
|
|
Thanks for the reply
> Are the Zigbee's 3 volt and the PIC 5 volts?
Zigbee's and PIC both are operating on 3.3v
>Any RF in the area(cell phones,CBs,Hams,etc.?
I am testing the system in a room, there are mobile phones in the room (but i personally think that this would be not the problem, if it is the problem then it would be dangerous when i implement the system in the field)
> Are the units just laying round open or in aluminum case with proper grounding?
units are just laying round open.
> Do they 'die' randomly or say only between 3 and 4 AM?( if you think that's silly to ask, I had 30 remotes 'actup' at that time.Turned out to be crosstalk in the main telephone trunk between two cities.(Bank was updating data during that time period).Changing spanpairs cured the problem.
They die randomly.
> Can you test the system with just batteries?
I can test the system with batteries but i will test it when i find the proper cause for damage otherwise i will lose another zigbee module.
Some more suggestion so that i look into the system for that. |
|
|
SherpaDoug
Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA
|
|
Posted: Mon Jan 03, 2011 8:01 am |
|
|
I would try to find out exactly what in the Zigbee module dies. You probably don't have a schematic, but maybe you can look up part numbers and find power buses and clock lines. Is the Zigbee still regulating power? Is the clock still running? Is it generating RF? These questions may tell you what has failed and maybe why it has failed. _________________ The search for better is endless. Instead simply find very good and get the job done. |
|
|
jecottrell
Joined: 16 Jan 2005 Posts: 559 Location: Tucson, AZ
|
|
Posted: Mon Jan 03, 2011 8:46 am |
|
|
I think I recall I had a similar problem with XBees. I did a lot of work with them in close proximity and never reset their default power levels to the lowest possible. Once I changed them to the lowest possible power setting for prototyping, I haven't had the problem again. |
|
|
hayee
Joined: 05 Sep 2007 Posts: 252
|
|
Posted: Tue Jan 04, 2011 1:01 am |
|
|
Thanks for all your suggestions
The IC used in the module is EM260 processor (from EMBER).
I checked the power at the EM260 it is 3.3v
I checked the waveform of crystal, it shows waveform then shows 0v then shows waveform then shows 0v and it continues.
Due to this reason all other process are not working. Now what to do? |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19544
|
|
Posted: Tue Jan 04, 2011 5:21 am |
|
|
So you are saying that on pin 38 of the ember unit, you are seeing an 'intermittent' clock, with a burst of clock, then it stopping?.
Have you checked the voltage on the 1.8v regulator?.
What is the interval between the bursts?.
If the interval is about 2 seconds, it suggests the chip is performing a watchdog reset.
The 1.8v goes 'off', if the chip goes into 'deep sleep' mode. Hence testing this would be informative.
Best Wishes |
|
|
hayee
Joined: 05 Sep 2007 Posts: 252
|
|
Posted: Wed Jan 05, 2011 10:54 pm |
|
|
Quote: | you are seeing an 'intermittent' clock, with a burst of clock, then it stopping?. |
Its not exactly that. Its like fluctuation but very fast.
Quote: | Have you checked the voltage on the 1.8v regulator?. |
I haven't find any 1.8 regulator on the board but some of the pins of em260 shows 1.8v
but the values of these voltages also varying between 1.74 to 1.8v. |
|
|
bkamen
Joined: 07 Jan 2004 Posts: 1615 Location: Central Illinois, USA
|
|
Posted: Wed Jan 05, 2011 11:07 pm |
|
|
temtronic wrote: | Never, ever seen code destroy a computer.It's just 1s and 0s !
|
Back in the 80's -- Supposedly there was code that could cause a C64 to fry some critical components.
Since I owned an Atari800, I was safe. ;)
-Ben _________________ Dazed and confused? I don't think so. Just "plain lost" will do. :D |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19544
|
|
Posted: Thu Jan 06, 2011 5:24 am |
|
|
The 1.8v regulator, is part of the em260. Used internally for just about everything!. Only time it should drop significantly, is if the unit is going to deep sleep, or doing a power reset. It should be coming out on pin12, and this should be connected back to pin14, and to pins 1, 4, & 7, where it feeds the VCO, RF generator, and the IF stage, also to 32, 36, 37, & 40, where it feeds the crystal oscillator, the internal synthesiser, and the main logic. If this fluctuates significantly, it'll prevent large parts of the circuit from working (it is allowed to be as low as 1.7v, or as high as 1.9, but it needs to be _stable_ when the unit is running).....
Comes back to the question about 'intervals' on the fluctuations. Again, about 2 seconds, could imply a watchdog timeout....
Best Wishes |
|
|
hayee
Joined: 05 Sep 2007 Posts: 252
|
|
Posted: Fri Jan 07, 2011 4:31 am |
|
|
Quote: | about 2 seconds, could imply a watchdog timeout |
No its not 2 seconds. Maybe its time is less than 1 sec.
Actually what is happening is that the waveform is switching between the ground and the signal. I cannot measure the time delay by scope even if I increase or decrease the time division on the scope. Also the amplitude of the wavefrom is very low. Its amplitute is 680mV.
Maybe I am unable to be clear to you about what I am trying to say.
I don't know how to explain exactly. |
|
|
anh_risn
Joined: 05 Jun 2011 Posts: 4
|
|
Posted: Wed Sep 07, 2011 3:20 am |
|
|
Hi hayee, have you solve the problem?
I also encountered same problem like you do.
The zigbee suddenly broken after working for a while.
Seems like the zigbee cannot pass the EmberInitialization() function.
It gets stuck in the while loop inside EM260Reset() function.
I also check the regulator on the EM260 chip, it shows ok 1.8V |
|
|
bkamen
Joined: 07 Jan 2004 Posts: 1615 Location: Central Illinois, USA
|
|
Posted: Mon Sep 12, 2011 9:50 pm |
|
|
temtronic wrote: | Never, ever seen code destroy a computer.It's just 1s and 0s !
|
If I remember right, there used a virus for Commodore64's that could put the system in a state to damage chips...
I didn't own one though, so no verification on that one. :P
-Ben _________________ Dazed and confused? I don't think so. Just "plain lost" will do. :D |
|
|
|