Hopefully this is the video.
Frequency hopping experiments on the NRF24
-
AndyS
- Posts: 42
- Joined: 12 Oct 2018, 19:03
- Location: Biggleswade
-
AndyS
- Posts: 42
- Joined: 12 Oct 2018, 19:03
- Location: Biggleswade
Re: Frequency hopping experiments on the NRF24
Getting near to the problem. Difficult to video and adjust pots. But when the led is flashing at a slow rate the servo movement is jumpy. When the led flashing rate is faster the servo movement is ok.
I found this out by swapping the Nrf module on the Tx. The Tx Nrf is powered through an adaptor suppling the 3v. I connected a Nrf module with no 100uf capacitor across the vcc and ground and that is when I got a fast led and smooth servo. When I removed the capacitor from the Nrf module that was giving the slow rate it was working Ok.
Another issue just emerged is when I connect a Nrf module with amp and antenna the rx led rate is again slow. I do have another one so will try it.
perhaps the 100uf is not the right one, pic attached.
two videos attached, one with slow led rate and one with faster.
PS - UPDATE- I have replaced the two white servos with better quality ones which have a smoother action.
Tried my second NRF24 with amp and antenna and it does not work at all. Either two duff purchases or I need to create an independent power supply for them.
UPDATE 2- just tried another low power nrf24 with a capacitor across the supply and it works ok, so maybe not a capacitor problem.
I found this out by swapping the Nrf module on the Tx. The Tx Nrf is powered through an adaptor suppling the 3v. I connected a Nrf module with no 100uf capacitor across the vcc and ground and that is when I got a fast led and smooth servo. When I removed the capacitor from the Nrf module that was giving the slow rate it was working Ok.
Another issue just emerged is when I connect a Nrf module with amp and antenna the rx led rate is again slow. I do have another one so will try it.
perhaps the 100uf is not the right one, pic attached.
two videos attached, one with slow led rate and one with faster.
PS - UPDATE- I have replaced the two white servos with better quality ones which have a smoother action.
Tried my second NRF24 with amp and antenna and it does not work at all. Either two duff purchases or I need to create an independent power supply for them.
UPDATE 2- just tried another low power nrf24 with a capacitor across the supply and it works ok, so maybe not a capacitor problem.
- Attachments
-
-
Phil_G
- Posts: 606
- Joined: 15 Feb 2018, 23:32
- Contact:
Re: Frequency hopping experiments on the NRF24
Hard to tell whats happening without the traffic LED Andy. They can swamp - try grasping the tx aerial with your hand to shield it a bit, or try a few feet of spacing. We really need to see an LED 
I power the tx regulator directly from the switched battery pos rather than from 5v.
They can swamp and drop packets at very close range.
On all mine I use a 100n in parallel with a 100µF or 220µF, on the tx its across the 3v3,
On the RF-Nano receivers I have the same across the 5v, and on the 8mhz Promini+separate NRF receivers I have it across 3v3. See photos throughout the thread, heres your own photo with caps added:
I have had 2 or 3 duff or suspect NRF boards, one works but not at 250k, one didnt work at all, on one tx module the aerial physically wouldnt fit.
I'm coming to trust them more but I always give them a thorough soak test spread over several days before using them in a model or a tranny conversion
after that they seem perfectly reliable, I've not had a single failure 'in service'
Cheers
Phil
PS We all enjoy a video and as we all know, "if there's no video, it didn't happen"
We want to encourage the posting of videos but some dont have a Yooch account, and videos as attachments are a significant overhead on Ron's hosting allocation. For this purpose I've created a Youtube channel "Mode-Zero videos" where any of the mods can transfer these video attachments, so they're not stored on the forum nor locally on your PC and no theres need to download the file. They're much more pleasant to watch on a proper video hosting platform, no stuttering and you dont waste any of your own disk space with a local copy.
If you dont have your own Yooch account just post the file as an attachment, and one of us will transfer it later. All the videos are 'unlisted' so no-one will find them other than through the mode-zero threads, they're not publicly displayed, nor is the channel.
Pretty much any format is ok be it Mov, Avi, mp4, wmv etc
Keep posting videos!
That cap should be ok Andy but if fitting/removing the cap definitely affects the rate of packets received then maybe its a duff cap, do you have another to try? Unless fitting/removing the cap is disturbing some wiring or something? You're adding it across the 3v3 supply on the NRF itself aren't you?
I power the tx regulator directly from the switched battery pos rather than from 5v.
They can swamp and drop packets at very close range.
On all mine I use a 100n in parallel with a 100µF or 220µF, on the tx its across the 3v3,
On the RF-Nano receivers I have the same across the 5v, and on the 8mhz Promini+separate NRF receivers I have it across 3v3. See photos throughout the thread, heres your own photo with caps added:
I have had 2 or 3 duff or suspect NRF boards, one works but not at 250k, one didnt work at all, on one tx module the aerial physically wouldnt fit.
I'm coming to trust them more but I always give them a thorough soak test spread over several days before using them in a model or a tranny conversion
after that they seem perfectly reliable, I've not had a single failure 'in service' Cheers
Phil
PS We all enjoy a video and as we all know, "if there's no video, it didn't happen"
We want to encourage the posting of videos but some dont have a Yooch account, and videos as attachments are a significant overhead on Ron's hosting allocation. For this purpose I've created a Youtube channel "Mode-Zero videos" where any of the mods can transfer these video attachments, so they're not stored on the forum nor locally on your PC and no theres need to download the file. They're much more pleasant to watch on a proper video hosting platform, no stuttering and you dont waste any of your own disk space with a local copy.
If you dont have your own Yooch account just post the file as an attachment, and one of us will transfer it later. All the videos are 'unlisted' so no-one will find them other than through the mode-zero threads, they're not publicly displayed, nor is the channel.
Pretty much any format is ok be it Mov, Avi, mp4, wmv etc
Keep posting videos!
-
AndyS
- Posts: 42
- Joined: 12 Oct 2018, 19:03
- Location: Biggleswade
Re: Frequency hopping experiments on the NRF24
Thanks for your reply phil, I had noted that you also use a 100n in parallel so will try that.
-
AndyS
- Posts: 42
- Joined: 12 Oct 2018, 19:03
- Location: Biggleswade
Re: Frequency hopping experiments on the NRF24
OK
Just found that I have a batch of 100 x 100nf capacitors
Just found that I have a batch of 100 x 100nf capacitors
-
MaxZ
- Posts: 330
- Joined: 31 Jan 2019, 11:48
- Location: Boskoop, Netherlands
Re: Frequency hopping experiments on the NRF24
All this reminds me: I finished my transmitter but I forgot to add the capacitor. I hope I can still get it in, it is quite cramped in there, more so than I anticipated..............
But I did get it in and closed up
Cheers,
Max.
Max.
-
Tobe
- Posts: 667
- Joined: 16 Feb 2018, 06:19
- Location: Varberg or Stockholm, Sweden
Re: Frequency hopping experiments on the NRF24
Have you been in touch with those tribes that know how to shrink heads to fit all that stuff in that tiny box? Very impressive and very well done Max
One of the reasons I cut and bend my own boxes if I'm not recycling an old transmitter
One of the reasons I cut and bend my own boxes if I'm not recycling an old transmitter
Cheers,
Tobe
Tobe
-
MaxZ
- Posts: 330
- Joined: 31 Jan 2019, 11:48
- Location: Boskoop, Netherlands
Re: Frequency hopping experiments on the NRF24
Hi all,
First, a basic question; can I program an RF Nano board with usbasp even though pins D11/12/13 are permanently connected to the RF module?
The reason I am asking is because the Arduino IDE on my Mac OSX (High Sierra) has developed a problem in recognizing any usb to ttl adapters, the only one that will work is an UNO tricked into thinking it is an adapter by grounding the reset. The other one I have is a FTDI adapter that does not seem to register as a port (as it used to), and this morning I tried to connect a NRF Nano via usb and again no port seems to be available for selection.
Cheers,
Max.
Edit: I found out the RF Nano has a CH340 chip, and I also found this https://github.com/adrianmihalko/ch340g ... s-x-driver , I will try to (re)install the driver and see if that solves it.
Edit2: No, that didn't solve it
First, a basic question; can I program an RF Nano board with usbasp even though pins D11/12/13 are permanently connected to the RF module?
The reason I am asking is because the Arduino IDE on my Mac OSX (High Sierra) has developed a problem in recognizing any usb to ttl adapters, the only one that will work is an UNO tricked into thinking it is an adapter by grounding the reset. The other one I have is a FTDI adapter that does not seem to register as a port (as it used to), and this morning I tried to connect a NRF Nano via usb and again no port seems to be available for selection.
Cheers,
Max.
Edit: I found out the RF Nano has a CH340 chip, and I also found this https://github.com/adrianmihalko/ch340g ... s-x-driver , I will try to (re)install the driver and see if that solves it.
Edit2: No, that didn't solve it
Last edited by MaxZ on 04 Jul 2020, 10:55, edited 1 time in total.
-
Phil_G
- Posts: 606
- Joined: 15 Feb 2018, 23:32
- Contact:
Re: Frequency hopping experiments on the NRF24
@AndyS
Today I put together another transmitter, and this time across the 3v3 pins used a yellow 100µF tantalum bead like Andys. On testing, it worked, but for the first time I had a lot of packets lost, the receiver traffic LED was very slow and there were was an occasional failsafe. I didnt have another module to try but it looked very much like the old 3v3 power problem again. I took the bead off and mounted an SMT 100µF in parallel with a 100n, directly across the 3v3 pins (pin 1 & 2) of the NRF module. It completely cured the problem to the extent that I had to check. I replaced the parallel combo with the bead, and lost loads of packets again.
I wanted to post this so to be absolutely certain I swapped the caps over three more times and its absolutely clear that with the bead alone, the power was not stable, and with the parallel SMT cap + 100n it was rock solid.
This is reminiscent of a problem both me and Ron separately experienced some time ago with LP2950 5v regulators oscillating. If you check the datasheets, many LDO regulators have a critical spec for the output capacitor ESR value, outside which the datasheet warns of instability. Consequently all my PIC encoders have a TAP226 controlled ESR tantalum on the LDO output. The AMS1117 datasheet however gives no specific esr requirement, only that the cap forms part of the compensation loop.
Despite this I am almost totally convinced that this is a similar problem. Until today I've used the SMT + 100n combination and had no trouble whatsoever. Maybe its RF induced into the DC, or ripple caused by the pulsed-current-burst nature of packet systems - this was a 'proper' NRF regulator board BUT these reg boards sell for 25 pence, who knows if its a full-spec AMS1117 or a chinese ripoff.
Its probably not the bead itself but the 100n that makes all the difference.
The first time I use just a bead alone, I get the exact same symptom as Andy.
Replace the bead with the SMT+100n as in the pic below, and its rock solid.
Swap back and I get packets lost. This is endlessly repeatable.
You may find just the tantalum alone has been ok, but I'd definitely suggest adding a 100n in parallel.
On the transmitters I've found its best to take the battery neg and switched pos wires first to the RF module regulator and then on to the processor, and to avoid taking the RF module neg 'through' a Nano - ie not to take battery neg to one Nano GND and use the other Nano GND to power the module.
The 3v3 regulator (for the RF module) I always power from switched battery, not from the Arduino 5v.
I use an adapter rather than the six-way ICSP pads underneath:
viewtopic.php?f=41&t=661&p=5124&hilit=adapter#p5124
Today I put together another transmitter, and this time across the 3v3 pins used a yellow 100µF tantalum bead like Andys. On testing, it worked, but for the first time I had a lot of packets lost, the receiver traffic LED was very slow and there were was an occasional failsafe. I didnt have another module to try but it looked very much like the old 3v3 power problem again. I took the bead off and mounted an SMT 100µF in parallel with a 100n, directly across the 3v3 pins (pin 1 & 2) of the NRF module. It completely cured the problem to the extent that I had to check. I replaced the parallel combo with the bead, and lost loads of packets again.
I wanted to post this so to be absolutely certain I swapped the caps over three more times and its absolutely clear that with the bead alone, the power was not stable, and with the parallel SMT cap + 100n it was rock solid.
This is reminiscent of a problem both me and Ron separately experienced some time ago with LP2950 5v regulators oscillating. If you check the datasheets, many LDO regulators have a critical spec for the output capacitor ESR value, outside which the datasheet warns of instability. Consequently all my PIC encoders have a TAP226 controlled ESR tantalum on the LDO output. The AMS1117 datasheet however gives no specific esr requirement, only that the cap forms part of the compensation loop.
Despite this I am almost totally convinced that this is a similar problem. Until today I've used the SMT + 100n combination and had no trouble whatsoever. Maybe its RF induced into the DC, or ripple caused by the pulsed-current-burst nature of packet systems - this was a 'proper' NRF regulator board BUT these reg boards sell for 25 pence, who knows if its a full-spec AMS1117 or a chinese ripoff.
Its probably not the bead itself but the 100n that makes all the difference.
The first time I use just a bead alone, I get the exact same symptom as Andy.
Replace the bead with the SMT+100n as in the pic below, and its rock solid.
Swap back and I get packets lost. This is endlessly repeatable.
You may find just the tantalum alone has been ok, but I'd definitely suggest adding a 100n in parallel.
On the transmitters I've found its best to take the battery neg and switched pos wires first to the RF module regulator and then on to the processor, and to avoid taking the RF module neg 'through' a Nano - ie not to take battery neg to one Nano GND and use the other Nano GND to power the module.
The 3v3 regulator (for the RF module) I always power from switched battery, not from the Arduino 5v.
Absolutely yes Max. On all my RF-Nanos I've blown the old bootloader via SPI as I dont want to risk updating the IDE and the old IDE wont talk to the RF-Nano serial port. Except on the red Tstar ones, they talk to the old IDE.
I use an adapter rather than the six-way ICSP pads underneath:
viewtopic.php?f=41&t=661&p=5124&hilit=adapter#p5124
-
MaxZ
- Posts: 330
- Joined: 31 Jan 2019, 11:48
- Location: Boskoop, Netherlands
Re: Frequency hopping experiments on the NRF24
Thanks Phil.
Well, programming with usbasp seems to work fine. I loaded some sketch, then loaded Blinky to see if it went ok.
But I am still bugged by the serial ports not working, at least not with the Arduino IDE. Mind you, the usb device shows up in the Hardware overview, there are no outstanding kernel extension approvals, so....?
Max
Well, programming with usbasp seems to work fine. I loaded some sketch, then loaded Blinky to see if it went ok.
But I am still bugged by the serial ports not working, at least not with the Arduino IDE. Mind you, the usb device shows up in the Hardware overview, there are no outstanding kernel extension approvals, so....?
Max