Hi friends,
I'm using a NRF51 IC with this 4.0BLE Modul PTR5518 and i have big problems with Range. As a protocoll i use gazell. In air it is possible do get some correct signals up to 6-7 m but that it don't work properly anymore. If you have a human body in between it is a disaster, you only get good signals with a range <1m. Can i somehow approve this? Please help me if you can. I'm very desperate.
Thank you in advance. Tanja
Hello tanja
RF systems are very sensitive to their surroundings. Walls that are not in the direct line of sight can reflect the RF signal causing multipath induced attenuation, especially if there is any metal within the walls like pipes or reinforcement bars. Objects in line of sight would typically attenuate or scatter the signal. Other wireless systems operating in the same band can worsen your SNR, and objects close to the antenna will typically alter its characteristics. Testing the range of a system indoors and outdoors can yield very different results.
I would like you to test the gazelle code that comes with SDK 12.3.0 and see how well your system does. I do find it a bit strange that you only get around 7-9 meter range with 4dBm output power and 250kb/s datarate, in my opinion you should have more than that.
I did a quick test with two nRF51 DK's and a slightly modified version of the gzll_ack_payload example. I simply changed it so the host retransmits the button presses made on the device, that way I could see on the device DK that the transmit was successful both ways. I still had a connection around 20 meters away through two glass doors. This office has a lot of metal pipes, some walls of plaster and some of reinforced concrete.
If you have any of our devkits available I would also suggest trying your code on those. This would help eliminate any hardware issues.
EDIT, Additional answers
The example is a part of the SDK, you can download it here. The latest SDK version to support the nRF51 series is 12.3.0. The gazell example code can be found in the \examples\proprietary_rf\gzll\gzll_ack_payload subfolder.
I think we would need to see the design to properly evaluate it. If you would like for us at Nordic to evaluate your schematic design and/or layout you can either upload it to your original post here on DevZone, or if your design is confidential you can create a mypage case at nordicsemi.com.
When frequency hopping the timeslots per channel option affects the number of retransmits before it changes frequency. If many of your channels are filled with noise from other sources this can affect your range. The nRF51's radio has a sensitivity of -96 dBm (0.1% BER) at 250kbps, but if there is a lot of noise present at the receiver you will need a stronger signal, or shorter range, to achieve the same BER.
How much power do you measure? Do you have a picture of the measured spectrum?
If I completely cover the antenna I have around 10-13 meters range, also through a glass door.
If you still have poor range with the SDK examples there might be a problem with the module you are using.
Best regards
Jørn Frøysa
Hello tanja
RF systems are very sensitive to their surroundings. Walls that are not in the direct line of sight can reflect the RF signal causing multipath induced attenuation, especially if there is any metal within the walls like pipes or reinforcement bars. Objects in line of sight would typically attenuate or scatter the signal. Other wireless systems operating in the same band can worsen your SNR, and objects close to the antenna will typically alter its characteristics. Testing the range of a system indoors and outdoors can yield very different results.
I would like you to test the gazelle code that comes with SDK 12.3.0 and see how well your system does. I do find it a bit strange that you only get around 7-9 meter range with 4dBm output power and 250kb/s datarate, in my opinion you should have more than that.
I did a quick test with two nRF51 DK's and a slightly modified version of the gzll_ack_payload example. I simply changed it so the host retransmits the button presses made on the device, that way I could see on the device DK that the transmit was successful both ways. I still had a connection around 20 meters away through two glass doors. This office has a lot of metal pipes, some walls of plaster and some of reinforced concrete.
If you have any of our devkits available I would also suggest trying your code on those. This would help eliminate any hardware issues.
EDIT, Additional answers
The example is a part of the SDK, you can download it here. The latest SDK version to support the nRF51 series is 12.3.0. The gazell example code can be found in the \examples\proprietary_rf\gzll\gzll_ack_payload subfolder.
I think we would need to see the design to properly evaluate it. If you would like for us at Nordic to evaluate your schematic design and/or layout you can either upload it to your original post here on DevZone, or if your design is confidential you can create a mypage case at nordicsemi.com.
When frequency hopping the timeslots per channel option affects the number of retransmits before it changes frequency. If many of your channels are filled with noise from other sources this can affect your range. The nRF51's radio has a sensitivity of -96 dBm (0.1% BER) at 250kbps, but if there is a lot of noise present at the receiver you will need a stronger signal, or shorter range, to achieve the same BER.
How much power do you measure? Do you have a picture of the measured spectrum?
If I completely cover the antenna I have around 10-13 meters range, also through a glass door.
If you still have poor range with the SDK examples there might be a problem with the module you are using.
Best regards
Jørn Frøysa
Hi Jorn,
First of all, thank you very much for your fast response.
2)We are using this Module as I already said on both sides, on the one hand in an Remote and on the other side we are using an RF PCB in our Control box where also this Modul from Xungtong (NRF51822) is located.
We already made some approvements especially by the Remote(copper around antenna, bigger hole in main pcb) to get the output power better. It helps a little bit but we are not satisfied with that.
We also made a lot of tests with this frequency hopping (jump between at least 15 channels) which we need that we can certify our HS also for the american market. This makes the range even worse than with an fixed channel.
a little bit better is the performance with 250k Baudrate.
We also made some measurements regarding output power in our EMC chamber and matched this with an stand alone PTR5518 (only supplied, sending all the time without main HS PCB) and we can't see any differences on the output power when we compare it to our Design.
As I already said, the problem is when something is in between, can you maybe test with your Setup these two nRF51DK's which you talked in the comment above, what happens when you shield it with your hands, we are loosing our connection when we are 2-3 meters away from the other module.
I hope you understand my description.
I'm looking forward to hearing from you, thank you upfront.
Best Regards Tanja
I have updated my answer with additional information.
Answers to you additional informations below.
