pass caps for the nRF24L01

I am trying to consolidate this issue so I apologise if I am raking over old coals.

I understand

  1. there can be issues with the power supply for the nRF24L01 on the Mega 2560, but apparently not with a Uno.

  2. It pays to keep the power wires as short as possible.

  3. adding extra bypass capacitors is a good precaution.

Somewhere else it was suggested to use a 10mf tant and a 100n. I've got those and I guess I could wire one between the 3v3 and gnd pins, but I propose to mount both on a proto board immediately adjacent to the nrf24. With the current arrangement, the path to one cap is considerably longer than the other.

My question for the 2.4 GHz gurus is, if the two caps are merited in the way I understand, which one should have the shorter path?

I'm looking at setting up the Mega proto board like below.

proto mega centre.pdf (102 KB)

Put both capacitors on the VCC line as close to the radio as possible (the radio connector is good) The capacitors have little to do @ 2.4 GHz but a great deal to do with how clean the digital internals are and this is the real issue. I've never bothered to put an O'scope on it... I believe religiously that the parts are NECESSARY NOT OPTIONAL PLACES FOR EXTRA UNUSED CAPACITORS. I have in 50 years of "Messing" with this stuff proved often enough that By-pass capacitors a wise investment and a great time saver when trying to fix a new circuit. They are Never "Optional".

Bob

Docedison:
Put both capacitors on the VCC line as close to the radio as possible (the radio connector is good) The parts are NECESSARY NOT OPTIONAL PLACES FOR EXTRA UNUSED CAPACITORS. I

I'm not sure what you are saying here - not helped by not actually possessing the transceivers yet. Is there a vacant spot on the board to accommodate these? Is there a rule for which cap has the priority?

Currently, the distances from the 3v3 pin to the holes on the board are 25 and 6mm. No doubt I can put one between 3v3 and Gnd on the 8-pin block.

Thanks

Basic electronics makes the point that it is a good idea to bypass any power used by a non trivial device with suitable components and in 50 or so years I have NEVER found the Opposite to be true. It was a suggestion not my own as a matter of fact but one I've seen before in reading about the nRF24Lxx trancievers in several threads and one very much common sense suggestion.

Bob

OK. I take it that you are not actually familiar with these transceivers. I'll wait for mine to turn up in the mail.

Nick_Pyner:
OK. I take it that you are not actually familiar with these transceivers. I'll wait for mine to turn up in the mail.

What does the datasheet say about this stuff?

In the case of the datasheet from Nordic Semiconductor - very little. The word "bypass" occurs once and I don't think it is relevant to the end user - me. This is indeed why I asked the question in the first place.

To answer one of your question, the smaller cap needs to sit closer to the radio. These capacitors are needed to filter out noise on the power supply side and have not much to do with the 2.4GHz side as Docedison mentioned.

That does not need to be on a spec sheet as adding decoupling caps is very common practice ( as Docedison also mentioned) and one of the very first things I learned ( the hard way :slight_smile: when I started with electronics 20+ years ago. That is not specific to these transceivers!

Nick_Pyner:
In the case of the datasheet from Nordic Semiconductor - very little. The word "bypass" occurs once and I don't think it is relevant to the end user - me. This is indeed why I asked the question in the first place.

Actually, if you look at the schematic for the nrf2401 "module", you'll see they have
5 bypass caps on the 3 Vdd pins, which is kind of interesting.

Headroom:
To answer one of your question, the smaller cap needs to sit closer to the radio. T

Thank you.

oric_dan:
Actually, if you look at the schematic for the nrf2401 "module", you'll see they have
5 bypass caps on the 3 Vdd pins, which is kind of interesting.

Yes, I've seen all that, hence my question. The board from Yourduino looks fully populated. We shall see.

Darn, I recall having seen such info for one of the RF modules, but have so many 1000s
of PDF files on my HD, I can't find the specific one. If you've seen the DS_nRF24L01.pdf
file from Nordic Semiconductor, they show several actual pcb layouts on pages 70..72.
The bypass caps are smt and almost touching the RF chip.

Like HR said, if that can't be done, and if possibly using non-smt parts, I would expect
it's best to place the smallest value caps closest to the Vcc pins, since those are squelching
the highest frequencies, and this is where it's most important to be keep the pcb copper
trace inductances as small as possible.