27Mhz - Radio Ideas

I have used the concept of "primary" core & "secondary" auxiliary control parameters in packet message structures before in other projects. It is a neat solution to servicing both needs and keeping latency down on the primary parameters.

I encode the packet type identifier into the packet ID sequence field achieving "two birds" with one stone...

Currently I am evaluating the interrupt driven capture of the ppm stream from the RC TX (from the trainer port) and what impact this has on the integrity of the flow of the data to the RF TX pin. I want to make certain that the system is not compromised by the processor servicing the interrupt and being unable to service the Manchester encoded TX stream in a timely manner. Maybe it's as simple as enable interrupts, capture ppm stream, turn off interrupts, transmit 27MHz packet, rinse and repeat as any additional ppm stream received will not be used until next 27MHz transmission...

On the hit list to include in the project

• Telemetry to allow evalution of what is going on inside the sub  - especially RSSI, dropped packets, chechsum errors
  
• Some form of dynamic antenna tuning (ATU) - lots of ideas to trial from my SWL days 50 years ago (how time flies). Servo controlled air spaced cap(s) - or suggestion from SubJohn to use a Varicap
  
• Failsafes - low voltage, lost comms, some sort of intelligence around compromised packet sequence (>x), checksum errors (>y)
  
• Review size of primary control parameters (8,10,16 bit or mix) with regards to efficieny vs. control resolution
  

I am playing with the ideas of using either a direct WiFi RF link back to shore for telemetry realising that the associated antenna (whip) may interfere (+ or -) with the 27mhz RX or perhaps using a "light" comms link from the sub to a model boat of the water which then relays the telemetry back to shore via RF

I have also wondered about the merits of having a "repeater" model boat communicating with the sub. Second (first controls boat manoeuvres) 2.4GHz link to boat, boat repeats transmission to sub on 27MHz |458MHz | xxMHz using onboard TX with antenna slung underneath hull...

Regards
Jonathan

C-3PO wrote:

I am playing with the ideas of using either a direct WiFi RF link back to shore for telemetry realising that the associated antenna (whip) may interfere (+ or -) with the 27mhz RX or perhaps using a "light" comms link from the sub to a model boat of the water which then relays the telemetry back to shore via RF

Regards
Jonathan

ExpressLrs is based on ESP32 chips that have wifi built in, the idea is they use the wifi interface and do all the configuration on a webpage served by the chip then it powers off the wifi when using the RC. but it could also be used as a telemetry link using wifi or ESP-NOW. it would be ideal for submarines then you can either change settings or flash it over wifi using OTA updates without removing it.

esp32 ... got that T shirt  :) that's what I would/do use for telemetry.

Indeed I am only using a Nano for the 27MHz side of the project due to compiling/development speed - I would migrate to esp32

Both the projects below are based on esp32 with web pages served directly from the board - I got bogged down and bored with writing code to store/retrieve multiple parameters to eprom.

This is a project of mine still in development - initially for a leveller with on shore control of settings and live feedback rather than having to gain access to sub's innards



and here is another project of mine for controlling turrets on my HMS Rodney, insert screen is web pages on iPad, control of turrets goes via esp32 and 458Mhz to boat





Did some transmission speed tests today - 10 x 8 byte packets a seconds - 100% copy!! - I am sure I can make that go wrong...

Regards
Jonathan

C-3PO wrote:esp32 ... got that T shirt  :) that's what I would/do use for telemetry

Regards
Jonathan

very cool!

I'm currently making a esp32 model boat controller, gyro stabilization, gps guidance, scanning lidar proximity avoidance system. I was using inav and another controller for the lidar but the more I got into it the more I realized it would be better to just make my own from scratch and get the esp32 to do everything inc espnow for rc and control.
https://github.com/geofrancis/Easyboat-controller

https://www.youtube.com/watch?v=jHB3Qs5lelg


I'm working on adding a miniature 24ghz radar to it to replace the lidar.

C-3PO wrote:
Currently I am evaluating the interrupt driven capture of the ppm stream from the RC TX (from the trainer port) and what impact this has on the integrity of the flow of the data to the RF TX pin. I want to make certain that the system is not compromised by the processor servicing the interrupt and being unable to service the Manchester encoded TX stream in a timely manner. Maybe it's as simple as enable interrupts, capture ppm stream, turn off interrupts, transmit 27MHz packet, rinse and repeat as any additional ppm stream received will not be used until next 27MHz transmission...

I am using flysky protocal for all my projects these days rather than PPM, Its 14 channels (13+1 for RSSI), it operates at 115200 non inverted serial so you can send it over any serial link without the inverted funky serial speeds that you need with SBUS or timing that ppm requires.  Any opentx radio or flysky receiver should put out ADHDS 2A protocol.

this is the library im using, you can use the no timer option if you want to save on timers. I think the timer is only needed if you want to guarantee receiving every packet. but since were sending at low speeds it wont matter.

https://github.com/bmellink/IBusBM

this works using softserial on esp32 and 8266 if your short on hardware serial ports but it can have performance issues

https://github.com/Kunnuck/FlySkyiBusESP8266


this is the code im using for generating Ibus

https://en.ovcharov.me/2021/05/04/can-esp01-be-used-as-a-receiver-for-drone/

I wanted to "pull" the Manchester library code back into a sketch so I could get a better feel for what was actually happening.

So far so good - the TX side has been transplanted and is working well - I can see some tidying up opportunities that will reduce the code and make it more simple to understand - a plus for me :)

I feel better prepared now for playing with packet structures and checksums.

Running test data - approx 10 x 8 byte packets a second



Regards
Jonathan

Part Two of pulling the Manchester library code back into a sketch (the receiver part) now completed and working.

Took longer than expected but isn't that always the way with Arduino if it doesn't fire first time...


Regards
Jonathan

Radio Control PWM range

I am not intenting to down scale (map) PWM to 0-255 range and leave it uncompressed.

Common PWM range is 1000-2000 microseconds with 1500 being neutral.

I seem to remember Robbe F14 RX split out 800 - 2200 microseconds with 1500 being neutral.

Some jazzy servos seem to accept 500-2500 microseconds ( https://wiki.lynxmotion.com/info/wiki/lynxmotion/view/lynxmotion-smart-servo/lss-radio-control-pwm/)

I think I am going to move forwards with 500-2500 microseconds as the PWM range as it fits nicely into an 11bit (2048) number - e.g. 11 bit number range 0 to 2047 plus static offset of 500 allows pwm range of 500 - 2547

Am I missing or overlooking something?

Thankyou
Jonathan

C-3PO wrote:Radio Control PWM range

I am not intenting to down scale (map) PWM to 0-255 range and leave it uncompressed.

Common PWM range is 1000-2000 microseconds with 1500 being neutral.

I seem to remember Robbe F14 RX split out 800 - 2200 microseconds with 1500 being neutral.

Some jazzy servos seem to accept 500-2500 microseconds ( https://wiki.lynxmotion.com/info/wiki/lynxmotion/view/lynxmotion-smart-servo/lss-radio-control-pwm/)

I think I am going to move forwards with 500-2500 microseconds as the PWM range as it fits nicely into an 11bit (2048) number - e.g. 11 bit number range 0 to 2047 plus static offset of 500 allows pwm range of 500 - 2547

Am I missing or overlooking something?

Thankyou
Jonathan

Generally i have found that on most servos 1000-2000us range works out about 90 degrees of travel on the servo. If you want 180 degrees you have to use 500-2500. Most motor controllers work in the 1000-2000 range. the default Arduino library maps 0-180 to 1000-2000us so you need to use the microsecond range to get the full travel on the servo. I had to do it with both the servo that yaws my camera and the servo that turns my lidar as i wanted 180 degrees of travel on them.

whilst it looks as if the conversion is a bit easier it also uses more data bits, especially if a number of servos (or speed controllers) are being used and do you need the 11-bit resoltuion? assuming a 90deg servo then 11-bits gives use a resoltion of 90/2^11 = 0.043deg, 8-bits gives you 90/2^8 =0.35deg.
Obvioiusly if you need 180deg range then that changes, but the comparison is still 2048 positions vs. 256 positions.
It depends on you control/software, in arduino its just something like
servo = int(float(Hi) - float(Lo))/float(N) * float(n)) +Lo
where
n = input value
N = max input value (2047 for 11 bits, 255 for 8 bits)
Lo, Hi are the min & max values for the servo.

I must admit I used the PCA9685 based servo controller that provides all the servo drivers I need with a simple I2C interface rather than drive directly from the arduino with all the connectors as well.

the reason robbe/futaba transmitters went from 800-2200 was to support their multi-modules... basic range was between 1000-2000ms, but their mutli-modules (prop or switch) sent a "sync" signal that was less than 900ms to indicate the end of 8 multiplexed channels of data. if you had another way of sending that sync signal to your receiver, you could use a 10 bit value (0-1023) add 1000 to it for your basic 1000-2000 at the servo. or just drop the support for robbe multi-prop/switch

Thank for all for your input - food for thought.

I seem to remember a while back I had an issue of granularity using 8 bit control for an ESC - however it could have been my clunky coding.

It's an easy switch so I will incorporate 8 & 11 bit as options - I will be stuffing the unused bits in transmitted bytes with control data ( Packet ID | Checksum | simple 2 & 3 state channels) so at the moment my worst case scenario will be 9 bytes per packet which I have running on the bench so far without any problems...

It will be an interesting experiment to see which of the two options has the least dropped packets / Checksum errors.



Simon - I also avoid using Arduino to drive servos/ESC's wherever possible, I have used the PCA9685 based servo controller also. A friend of mine used one in his Speedline Lifeboat - when the Arduino lost contact with the PCA9685, due to it having it's own clock and running pwm independantly etc, his boat shot out of the lake and over the Tarmac - he eventually came up with a nifty failesafe shutdown.

If you have not used before - you might want to take a look at a Hobbytronics i2c servo controller - the servo speed control comes in very handy and of course there are several Pololu boards.

https://www.hobbytronics.co.uk//index.php?route=product/search&search=servo%20controller


Documentation gives you a flavour of functionality - only pain is it's not on a breakout board...

https://www.hobbytronics.co.uk/datasheets/slave_servo_pro_datasheet_v01.pdf

Regards
Jonathan

I have been playing with Arduino for coming up to 10 years... I think I have experienced many of the foibles of the topic...

and then....

The RX side of this project I have running currently in it's most simple form - the Radiometrix RX connected an Arduino Nano.

I have a USB lead from my laptop connected to the Nano.

There are 3 other connections:
1. Nano Ground  - RadioMetrix RX GND
2. Nano Pin 5 - Radiometrix RX Data
3. Nano 3.3v - Radiometrix RX - RX Enable

A few days back I suddenly could not upload programs from the Arduino IDE to the RX Nano - the leds would flicker to start and then the upload would just hang. So I presumed the Nano had somehow died, it could not have been shorted as it's mounted on a piece of breadboard in the clear with nothing else near it. Did all the obvious - new USB cable, etc etc

I removed and replaced old Nano with a new one. Earlier this afternoon - some 48 hours or so from the initial incident the new Nano died - same symptoms - initial flicker of leds then hanging on upload.

So now I am a little confused - I have fried several Arduino's over ther last few years usually accompanied with the fragrant smoke - but nothing like this.

A few hours later I wondered if I could have "bricked" them so I reloaded the bootloader via an Arduino Uno - and to my surprise they have both come back to life.

The Nano's are good quality clones - I bought 2 packs of 20 boards a few years back resulting in a unit price of less than £2.00 - I have used these boards in many projects without any problems.

So how is this possible?

David - have you had any issues?

Regards
Jonathan

C-3PO wrote:I have been playing with Arduino for coming up to 10 years... I think I have experienced many of the foibles of the topic...

and then....

The RX side of this project I have running currently in it's most simple form - the Radiometrix RX connected an Arduino Nano.

I have a USB lead from my laptop connected to the Nano.

There are 3 other connections:
1. Nano Ground  - RadioMetrix RX GND
2. Nano Pin 5 - Radiometrix RX Data
3. Nano 3.3v - Radiometrix RX - RX Enable

A few days back I suddenly could not upload programs from the Arduino IDE to the RX Nano - the leds would flicker to start and then the upload would just hang. So I presumed the Nano had somehow died, it could not have been shorted as it's mounted on a piece of breadboard in the clear with nothing else near it. Did all the obvious - new USB cable, etc etc

I removed and replaced old Nano with a new one. Earlier this afternoon - some 48 hours or so from the initial incident the new Nano died - same symptoms - initial flicker of leds then hanging on upload.

So now I am a little confused - I have fried several Arduino's over ther last few years usually accompanied with the fragrant smoke - but nothing like this.

A few hours later I wondered if I could have "bricked" them so I reloaded the bootloader via an Arduino Uno - and to my surprise they have both come back to life.

The Nano's are good quality clones - I bought 2 packs of 20 boards a few years back resulting in a unit price of less than £2.00 - I have used these boards in many projects without any problems.

So how is this possible?

David - have you had any issues?

Regards
Jonathan

I have only ever see corruption on 328 chips when they are browned out. something about their memory doesnt like low voltages. there is a fuse you can set for brownout detection to stop it happening. its normally only an issue trying to run them at 16mhz on 3.3v as your right on the limit and technically out of spec, as they are rated for 8mhz at 3.3v and 16mhz on 5v.

http://www.itluxembourg.lu/site/hobbyking-orangerx-433mhz-modules-faq/fuse-settings/

geofrancis wrote:

I have only ever see corruption on 328 chips when they are browned out. something about their memory doesnt like low voltages. there is a fuse you can set for brownout detection to stop it happening. its normally only an issue trying to run them at 16mhz on 3.3v as your right on the limit and technically out of spec, as they are rated for 8mhz at 3.3v and 16mhz on 5v.

http://www.itluxembourg.lu/site/hobbyking-orangerx-433mhz-modules-faq/fuse-settings/

Good call - I did not consider low voltage - that would a plausible reason - thank you.

I had jumped to the illogical conclusion that hammering pin 5 however unlikely and impossible had somehow contributed to the event as I could not see what else could be a factor...

I have reconfigured my desktop workarea and am now powering the Nano's from a powered USB hub

I hope my sanity can now be preserved - thank's again

Regards
Jonathan

I HAVE had occasional problems with Nano Clones and I tried using only genuine Nanos for a while but they are expensive. Difficult to pin down the cause.

I currently get Nano clones from:

happybuddhatrading

https://www.ebay.co.uk/itm/325448632453

And they seem OK.

I think that geofrancis has a very good explanation re: brown outs. I gave up BECs on ESCs in submarines long ago and I now always put an LM2940CT-5 linear, low dropout regulator in with a large storage capacitor (5000 micro F). (I have tried switch mode power supplies but they seem to generate different problems!)

This seems OK but I recently noticed the Arduino data logger I have been using for 27MHz trials keeps on resetting when a sub is sailing. Maybe the LM2940 at 1 amp is not hefty enough.

But yes, power supplies create problems!

David

I have bought some LGT8F328P boards, they are supposed to be register compatible with the atmega328 along with having some extra features like a dac and being able to run at 32mhz but I havent tested them yet they only just arrived.
https://github.com/dbuezas/lgt8fx
https://www.aliexpress.com/item/1005003524081449.html
https://www.aliexpress.com/item/33003833124.html

david f wrote:I HAVE had occasional problems with Nano Clones and I tried using only genuine Nanos for a while but they are expensive. Difficult to pin down the cause.

I think that geofrancis has a very good explanation re: brown outs. I  gave up BECs on ESCs in submarines long ago and I now always put an LM2940CT-5 linear, low dropout regulator in with a large storage capacitor (5000 micro F). (I have tried switch mode power supplies but they seem to generate different problems!)

I have used these for years without issue even shorting them out just makes them get hot. https://www.ebay.co.uk/itm/314518771446

Most electronics run internally on 3.3v even if they are rated for 5v as they have an internal linier regulator so they are pretty tolerant to voltage swings. sAa good idea is to use both switching and linier regulators, the switching regulator does the big drop from battery voltage down to 5v then the linier does the drop from 5v to 3.3v that way most of the work is done by the efficient switching regulator and the fine control is done by the linier regulator.  it keeps the linier regulator temperature down as its only dropping a couple of volts.

C-3PO wrote:
I have reconfigured my desktop workarea and am now powering the Nano's from a powered USB hub

Are you trying to power the radiometrix module from the nanos internal 3.3v regulator? they are only rated for a very low current i doubt it would be able to run a radio module im not surprised the chip browned out.  you should use a separate ams1117 3.3v regulator for the radio module.

For a long time i was buying Teensy LCs... 48mhz 32 bit for $12. considerably more than the chinese clones, but less than the authentic arduinos. and much more powerful. but sadly, due to the "chip shortage", the Teensy LCs are no longer being made. :(

geofrancis wrote:

C-3PO wrote:
I have reconfigured my desktop workarea and am now powering the Nano's from a powered USB hub

Are you trying to power the radiometrix module from the nanos internal 3.3v regulator? they are only rated for a very low current i doubt it would be able to run a radio module im not surprised the chip browned out.  you should use a separate ams1117 3.3v regulator for the radio module.

I never power anything directly from the Arduino boards - aware of the regulator constraints - so that was not the issue.

Nano spec sheets states 3.3v regulator 100ma max - the Radiometrix RX draws 20ma - but it's all irellevant as that's not how it was connected :)

The reference to a connection from the Nano to the Radiometrix "RX enable" pin is simply to bring the Radiometrix 3.3v logic pin high to enable the Radiometrix RX - but I will put this on 5v pin via resistor.

I think the RX Enable pin probably is 5v logic - the Radiometrix documentation is not great - they go into great detail on some things and not others so I played it safe

David - My 40 Nanos @ £1.96 each (about 3 years back) came from happybuddhatrading - purchased several things from them - consistently reliable and my go to supplier (well one of them)

With respect to boards - my favourites are Teensy 2.0 or 4.1 & Esp32 (Adafruit Huzzah)- I have not checked recent prices but I am sure I would get a shock!!!

Regards
Jonathan

C-3PO wrote:

Simon - I also avoid using Arduino to drive servos/ESC's wherever possible, I have used the PCA9685 based servo controller also.

I dont mind driving servos from arduino as generally if the program crashes the servo output will stop. IO on the other hand i never use the arduino directly as I have seen the outputs glitch out and rapidly cycle things so I always use a MCP23017 chip for general io that way the arduino has to be in a functional enough state that it sends commands that are understood before pins get switched.

I am using Arduino Nanos to drive the servos from software in both the LoRa and 27Mhz Radiometrix software:

https://github.com/rdforrest/LoRa-TX-RX-Test-458MHz

https://github.com/rdforrest/27MHz--Radiometrix-module

This was Jonathan's original suggestion and it seems to work well. The servo software took a bit of "shoe-horning in" because of, I think, interrupt clashes.

I think I am right in saying that using less hardware in the RX makes it smaller, which is an important consideration for subs.

David

PS Not much 27MHz development from me at the moment. Been busy with getting ready for Papplewick and trying to find an antenna analyzer for 27Mhz for not too much money! I recently bought a cheap, second hand SWR meter for CB radio but the TX output does not move the meter to full scale - for the calibration setup. Presumably 100 mW TX output is not "man enough" compared with CB rigs!?

Hi David,

I am sure SubJohn could illuminate the topic of antenna tuning.

From my young days as an avid SWL (Short Wave Listener) I remember long bits of wire strung out down the garden.

I used an "tuner" between the antenna and the receiver and simply twiddled until the S meter peaked / the audio was the strongest/clearest.

A couple of air spaced caps and a tuning coil

Along these lines -




I would experiment with an antenna cut to length that suggests resonance underwater at operating frequency taking into account the difference in speed an SDR dongle | ATU perhaps servo controlled | Antenna underwater in a waterproof plastic box underwater and a USB lead back to the surface

What I really trying to get across is to approach the problem from the other end - use a receiver to tune the antenna by getting the best RX signal which would suggest a good antenna resonance...

A bit Heath Robinson but with such low TX output I don't think you will be able to realistically measure SWR etc

Just found this - not read it but probably some interesting stuff

https://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/ar/1987/04/page05/index.html

Regards
Jonathan

That's very interesting, Jonathan.

So the point you are making is that an antenna behaves in the same way for TX and RX? Resonance is resonance.

The trimmer could indeed be servo actuated (for trials at least). USB cable to feed RSSI value back to a laptop or maybe a serial link to my oLED display if it works through (say) a 3 m cable?

I am heading to the NARSA rally at Blackpool this Sunday to maybe glean some ideas and/or gear:

https://narsa.org.uk/

The article in the link you give is, at first glance, the original Australian article which kicked this whole underwater antenna idea off?!

Thanks again,

David