Frequency drifting on nRF24L01+

[Peter]

Hello Nordic Semi,

We are using the nRF24L01+ for a remote switch application. We have a small transmitter button (nRF24L01+ layout as per datasheet with micro controller attached and CR2450 lithium coin cell battery) and a base unit in 100% receive mode with some additional outputs. The system is set to operate at 2482MHz (channel 82).

During testing, we have found a transmitter voltage range where the transmission is not received correctly by the receiver. Above or below this voltage range, the transmission is processed correctly. If we watch the transmitter system voltage on an oscilloscope, the transmit pulses can be observed during this "dead voltage range" so we can see that the RF chip is transmitting (or at least attempting to). It can also be observed that if 2 receivers are placed close by; unit 1 may acknowledge the transmission and the other does not. Next time, only unit 2 may acknowledge.

Here is some more information from our R&D engineer:

"When it is around 2.2 volts looks like it is transmitting but no reception then works OK when you take it lower.. (a) Definitely no packet received by the Receiver. It continues polling so status doesn’t flag a received packet of any description. (b) Tried adding delays, changing timings in the transmitter TX data area. Make sure everything was settled and also making it go through as quick as I could etc. No difference. (c) Tweaked the xtal cap values and even changed xtal. No difference. (d) Changed the transmitted power level and “bingo”. At the lowest power level it will transmit through the whole range of voltage. Middle two power levels were improvements but not perfect. Off course our range drops off. So while I thought the xtal was changing frequency at the suspect voltages I think the transmitter or PLL is being pulled off frequency or dramatically reduced in power . I tweaked the antenna matching capacitors and no difference. Fitted larger the Rx antenna no difference. Didn’t try to tweak the matching inductors."

Are we observing a known issue here?

Any suggestions as to a root cause and solution?

Regards, Peter

[Håkon Alseth]

I cannot say that this is an issue we've seen before.

1. Can you verify that the L01+ is transmitting within the "dead voltage range"? This can be done by scoping VDD_PA pin (goes high for each on-air TX)

2. Can you please post the nRF chip markings? Should be something like:

   NRF M 24L01+ 1313AA

3. Does this happen with other boards as well? Or is it only one of them?

[Peter]

I will test the VDD_PA pin and then provide some feedback.

[Peter]

We can confirm that the VDD_PA pin is being asserted and therefore it is transmitting. This happens on several boards with different chip markings.

[Peter]

If I setup 5 receivers with flashing LED's to indicate a correctly received packet, I can see that receivers will sometimes not acknowledge transmissions evenly. These "dropped packets" are seemingly random amongst the 5 receivers. This seems to indicate that the tuning of the transmission changes sometimes with each transmission (within seconds of each-other)? Could this be a result of the RF tuning circuitry?