Frequency hopping experiments on the NRF24

[tiptipflyer]

First class work Tobe. Love the mount and your Tx design incl. the sticks.
Frank

[Shaun]

It's not real if there's no video, so here's the video.



Flown in my 180g indoor model with the G45 Space Commander tx mentioned earlier.

viewtopic.php?f=42&t=971&p=8405#p8384

Cheers,

Shaun

[Shaun]

Not being very adept at designing 3D widgets, I used this approach for my NRF24 converted Tx.
The antenna extension leads are easy to make (assuming you have the correct parts) and very reliable as they are soldered up. I have a small stock of various SMA & RPSMA male and female ends left over from my prolific Drone building days. This also takes the strain off the thin NRF PCB's if the antenna gets knocked, but I do like Tobes holder- pity I don't have a 3D printer anymore.

Shaun

NRF Antenna conn.jpg (22.24 KiB) Viewed 51952 times

[Tobe]

will have to put a few in the mail for you and Phil!

[AndyS]

I am having a lot of fun with this NRF24 project.

I have some (hopefully not too dumb) questions.

First I made the 2.4 GHz scanner with an oled display which works fine and is a good test for my NRF Tx.

The NRF Tx is a arduino nano and the low power NRF24 module which seems to work Ok with a visible display on the scanner. The nano was programed through the ide using the USBASP programmer.

The Rx is a RF Nano which was programmed through the USB connector presumably with a bootloader.

When the Rx is switched on the servos initially twitch but no movement from the Tx pots.

I have scoped the Rx servo signal which shows a pulse output but no variation in width to the Tx.

Is there a point in the Tx nano where I can scope the ppm pulses thus ensuring that Tx pulse train is working?

Secondly is programming the Rx nano via the bootloader ok or is there a way to use the USBASP?

I am using 5 volt NiMh packs to power both Tx and Rx, for the Rx connected to the 5v pin.

I have included some pics.

[Phil_G]

Hi Andy, thanks for giving it a try
A few pointers for a start:

First of all, have you fitted the 5v LED between A0 and A1 on the receiver? that will tell you if packets are being received. Using a 5v LED means no need for a resistor - much neater.

With the tx switched off, the rx should be in failsafe, so the LED should be solid and you should see 1100uS on the throttle and 1500uS on the other channels.

The number one source of problems is power, the NRF needs those decoupling caps. And forgetting to set CE and CSN to the correct pins (10 & 9 resp) though if your tx is working then you have that right.

Be careful with a 4-cell NiMh on the Vcc receiver pin, as explained in the thread, the NRF section is 5v tolerant, but only to 5.5v iirc, a 4-cell can exceed that. I use a 1S lithium.

Is there a point in the Tx nano where I can scope the ppm pulses?
There is no pulse train Andy, its entirely digital, no PPM anywhere.

Your band scanner - is that this project being displayed on the screen? I see only two carriers, did you spread the 16 FHSS channels throughout the band? It should fill the display.

I'm not a fan of breadboards, I dont use them because of the hours they can waste trying to find faults!
Try to simplify the layout - put links on A6 and A7, thats a start, looking at the pics your NRF wiring looks ok but I cant make out your transmitters 3v3 supply - thats crucial

Thinking about it, the RF-Nano receiver is so simple its unlikely to be the problem, other than power, so I'd be going over the tx wiring again. That OLED screen defo looks wrong.

Did you do the calibration? ground D8, switch on, waggle all the sticks, centre all the sticks, release D8. Dont switch off at any point during this process.

The code has been updated a couple of times, make sure yours is the latest

The receiver LED can be just a normal one with a 330 ohm in series Andy.
Long leg to A0, short leg via a 330R to A1 - its not a pullup though, its in series.
5v LEDS are just easier & neater as they have the resistor inside.

With A6 and A7 floating on the tx you probably wont get any movement from the sticks, easiest is
to ground them with Spektrum bindplugs.

The receiver LED will give the best clue as to whats happening - it should blink rapidly.
Each blink is 20 packets received, 'on' for 10 packets, 'off' for 10 packets.

Cheers
Phil

[AndyS]

Phil Thanks for your quick reply.
I have not got a LED that I am sure is 5v. I have led's of various colours but not sure what their voltage is. So unable to use the LED method to find out if the Rx is in failsafe. I will have to work a bit more on this. Will it be OK to use an LED with a pullup resistor?

I have some 100uf capacitors so will fit one to the Rx power rail.

The Tx low power NRF24 is supplied through the power adaptor, but I will fit a capacitor across the 3v supply.

TX has the CE and CSN correct. Breadboard agreed is not the best but my way of seeing how it is connected and adding some control pots.

4 cell NiMh is giving 5.25 volts but have moved Rx supply from 5v to Vin.

Pity about not seeing the Tx pulses on a scope, but understand it is all code.

The 2.4GHz was showing just my house WiFi, Tx was off. Tx shows ouputs across the band.

Good point about the A6 an A7, not good practise leaving them floating.

[AndyS]

All good stuff Phil.

I had sussed out a led series circuit and will apply to my rx.

I have connected A6 and A7 to gnd.

I have also found that my single channel switch on the breadboard was not wired to switch, hence the pots never calibrated.

Not tested all this yet. At 71 of age and after a lifetime of working on radars I find I am still learning. I read somewhere that electronics now is all software

PS The plan of the bench is for a Keil Kraft Talon, converted to RC and electric power. All pushed to one side while I divert with arduino modules.

[Tobe]

I sure second Phil on the breadbord, we call it the dreadbord here as they are not reliable when you have a lot of wirering and I have to many hours troubleshooting.
I go for the iron! Components are quite cheap today so i just buy a few extra and for the critical parts I use sockets.

[AndyS]

Finally managed to get a working Tx and Rx.

But the servo response is a bit jittery.

I would appreciate any pointers where the problem is.

Hopefully the video will attach as well.