CC2530: Device doesn't respond on time when it's supposed to

Jack Loop

Part Number: CC2530
Other Parts Discussed in Thread: TIDA-00600

Hi. Whenever I send an on/off command to my "ZStack 3.0 - SampleLight - ZigBee End Device," it often doesn't respond within 1-2 seconds. It usually takes about 3-10 seconds to respond. Below are images of my situation with a poll rate of 1 second. How can I fix this in the code so that the CC2530 will respond within 1-2 seconds? (Note: I tested this with the default code)

over 8 years ago

0 YiKai Chen over 8 years ago

Guru 735695 points

Do you use Ubiqua Protocol Analyzer to check what happens?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

I just checked and I can confirm that this issue is happening.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Can you attach your sniffer log?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Here it is.

Tests.zip

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

I see the signal strength of your device is not so good (It's -59dbm) and sometimes there is no MAC ack when device does polling. I think that causes your delay issue.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

How might I fix these 2 issues?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

How far is it between your coordinator and device?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

About 35ft.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

If you put your device next to coordinator, do you still see this issue?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Yes.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Can you put device next to coordinator to do sniff again and attach sniffer log?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Here it is.

Test.zip

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

The signal strength is still low. How far is it between your sniffer HW and coordinator/device?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

I placed them right next to each other.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

I think this is about your HW. If it's your custom HW, you should check the RF performance.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

I don't think it is a hardware problem because I have tested it with a "router" configuration and it responds immediately when I send an on/off command to it. I think it has something to do with the End Device configuration. What do you think?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Router keeps receiver always on so it's not the same. I judge this according to your sniffer log and my experiences.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Can you explain why it is a hardware problem if response time is fine in router mode, but not in end device mode?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

You can try the same thing on TI CC2530Dk and I believe you won't see the same issue. I have implemented lots of end device with short polling to receive data and none of them have this issue.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Can you show me how you see the signal strength of my device in Ubiqua?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Is this value the RSSI value of the End Device to the Coordinator, or is it the the RSSI value of the End Device to the protocol analyzer device?

0 Jack Loop over 8 years ago in reply to Jack Loop

Guru 17870 points

I noticed that when I power the CC2530 board through the CC Debugger, it has a RSSI of -30dBm, but when I power the CC2530 from a full CR123A battery, it has a RSSI of -60dBm. Why does the battery make the CC2530 RSSI lower? How could I fix this?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

As I had replied you, there must be something wrong on your HW. You should check it thoroughly.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

1) What I am saying is that the RSSI changes upon different power sources: a CC Debugger and a plain battery. Is the problem that I don't have a voltage regulator for the battery (which might be why the RSSI is lower) whereas the CC Debugger does have a regulator (which might be why the RSSI is higher)?

2) Would adding the LDO from www.ti.com/.../LP5907 (mentioned from the TIDA-00600 site) resolve my issue with the battery negatively affecting the RSSI?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

If you use power supply to give power, do you still see the same issue?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Yes. The power supply gives the sames results as the battery.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Then, I don’t think it’s regulator causing the issue.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

I currently don't have a LDO on my CC2530, but the CC Debugger does. With the CC Debugger as a power supply, the RSSI on the CC2530 seems to be fine. So why isn't this a regulator issue? What else might it be then?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

I suspect it’s related to grounding. If you try to provide better grounding, does it improve?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

Below is an image of my setups. How do I improve grounding here?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

I think you should provide a larger common ground to your CC2530 custom board.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

So are you saying that the because there is a large common ground on the PCB of the CC Debugger, that is why the CC2530 has better RSSI?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Yes

0 Jack Loop over 8 years ago in reply to Jack Loop

Guru 17870 points

So I add a large common ground, but it didn't make a difference. What else do you suggest?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

How do you add large common ground?

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

I had a PCB with a large ground and attached my CC2530 to that. In addition, I also soldered a wire from the GND of that PCB to a big piece of sheet metal to enlarge the common ground plane. This large common ground didn't seem to make a difference.

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Try to connect all jtag pin to VDD.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

That didn't seem to fix it either. What else should I try?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

Run out of idea.

0 Jack Loop over 8 years ago in reply to YiKai Chen

Guru 17870 points

There some strange that I found out. When I have the CC Debugger (powering the CC2530) plugged into another computer USB port, the RSSI significantly decreases as well. I think it might have to do with the fact that the CC Debugger and the protocol analyzer device sharing the same power from the same computer USB ports. What do you think? Also, is this RSSI value about the receiver of the End Device to the coordinator or is it to the protocol analyzer device?

0 YiKai Chen over 8 years ago in reply to Jack Loop

Guru 735695 points

RSSI is about Protocol Analyzer received signal strength from end device.

Zigbee & Thread

Zigbee & Thread forum

CC2530: Device doesn't respond on time when it's supposed to