Operating at 472 Mhz instead of 433 Mhz

[oden]

In Australia we have laws that limit (unlicensed operators) the power output of 433Mhz transmitters to 25mw EIRP (means includes the gains from antenna).

Wikipedia reference
http://en.wikipedia.org/wiki/First-pers ... ansmission

How every we can push 100 mW within the range of 472.0125 MHz - 472.1125 MHz hence I am interested in researching how a Flytron and Orange TX & RX can be modified to operate at 472 Mhz (as I have this hardware).

Have noticed that the code for openlrsng contains the declaration RFMXX_868 for using modified hardware operating at different frequency instead of the usual 433 Mhz, hence the idea seems possible. Although a google search shows no results on this topic so I am not able to find any details on how to do this modification, assume this involves replacing the RF module with a different one.. but which?

Any suggestions on this?

[Rob Thomson]

Have you looked at the Rlink? By royal way?

You can set the frequency to anything within quite a large range.

Not sure if it goes as high as you want!

Sent from my GT-I9300 using Tapatalk

[tilmanb]

OpenLRS uses the RFM23BP HopeRF module
http://www.hoperf.com/rf/fsk_module/RFM23BP.htm

As you can see, it has a very wide frequency range. However, as far s I know there might be hardware mods required. Least of all the correct antenna lengths. Possibly some RF filter changes.

The RF module re programmed using Arduino code. I guess you can take the RFMXX_868 modifications and follow down the same way.
But keep in mind, the 868Mhz regulation required very low duty cycle times. Perhaps your band has similar restricions. (Meaning, you can't just open a carrier on a frequency and stay there. You are only allowd to ctively transmit for short bursts)

[oden]

Have you looked at the Rlink? By royal way?

No, i had not seen this product before your post. 410-510Mhz is the frequency range which is perfect for me but unfortunately the low power mode is 400mW which is way over the power limit.

Thanks for the suggestion

[oden]

So people have done this before for 868Mhz, are there any details on what hardware was needed to be changed. Should I troll though the commit history to see how made this code change and try to contact them?

As you can see, it has a very wide frequency range

So according to the website these units will cover the following frequencies 433/470/868/915MHZ and it looks like there is no need to replace the HopeRF module.

However, as far s I know there might be hardware mods required. Least of all the correct antenna lengths. Possibly some RF filter changes.

Good point, I will have to read the dock very carefully.

But keep in mind, the 868Mhz regulation required very low duty cycle times. Perhaps your band has similar restricions.

Was not aware there might be a restriction here, thanks Will look into it

[oden]

Reading the documentation on transmitter module shows there is no hardware changes needed to support different frequencies (other then the antenna). The schematics from the Flytron site confirms no additional components are between the module and the antenna.

Decided to test this out using the stock 433 Mhz antenna, knowing its range will be limited but will be fine for testing purposes.

Reviewing the code, the only section that is different for the 868 Mhz implementation is the declaration of the frequencies

Code: Select all

binding.h
#if (defined RFMXX_868)
#  define MIN_RFM_FREQUENCY 848000000
#  define MAX_RFM_FREQUENCY 888000000
#  define DEFAULT_CARRIER_FREQUENCY 868000000  // Hz  (ch 0)
#  define BINDING_FREQUENCY 868000000 // Hz  

Googling DEFAULT_CARRIER_FREQUENCY seemed to indicate this is also know as "Arithmetic mean" which can be calculated here HERE

Which lead me to adding this into the code:

Code: Select all

binding.h
#elif (defined RFMXX_472)
#  define MIN_RFM_FREQUENCY 472012500
#  define MAX_RFM_FREQUENCY 472112500 # http://www.comlaw.gov.au/Details/F2011C00543   
#  define DEFAULT_CARRIER_FREQUENCY 472062500  // Hz  (ch 0) Arithmetic mean 
#  define BINDING_FREQUENCY 472062500 // Hz     

openLRSng.ino
//### Module type selection (only for modified HW)
//#define RFMXX_868
//#define RFMXX_915
#define RFMXX_472

Code compiled, uploaded to TX and RX but they would not bind.

Of cause the web configuration had issues as it did not know of the frequency changes.

Attempted to get the serial CLI to fire up with both the stock and modified firmware. Able to receive boot message of the firmware version and hardware type but could not enter the menu or receive any other messages.

Anyone have any suggestions?

[Kilrah]

Reading the documentation on transmitter module shows there is no hardware changes needed to support different frequencies (other then the antenna). The schematics from the Flytron site confirms no additional components are between the module and the antenna.

There are no other compenents there, but the module is a different version. The RFM22 module is sold in 4 different versions depending on band, and that's where components change. So you should replace the whole RF module, providing you can find the one you want as it seems to only be stocked in the "common" ones in most places.

Hard to find exactly what freq range is "supported" by each of them too.

[oden]

he RFM22 module is sold in 4 different versions depending on band

THANK YOU! Did not pick up that massive point from the product manual. Makes perfect sense!

Interesting that 470Mhz is not an option on the unit you have shown.

providing you can find the one you want as it seems to only be stocked in the "common" ones in most places.

Usually use http://octopart.com to locate parts, amazed that nothing is listed. Can not seem to locate any units from the usual suppliers.

Any suggestions on where to go?

[Kilrah]

Nope, no idea. When I looked, any place I found them would only stock the 433 and 915MHz versions.
Anyway to operate at 470MHz your best bet would be to try and configure a 433MHz module up there, and see what you get as it's the closest. That would be software only, and might work or not.

[oden]

Ok the 470 Mhz is a 'custom' order from HopeRF, they don't normally manufacture them. They have a 'large' minimum order quantity, no point even asking what large is.

your best bet would be to try and configure a 433MHz module up there

Did test the 433 module with 470 frequency settings, it was just dead air.

It appears that 915Mhz would also be an option but needs to be configured with 20 hops so I will give that a go.

I would like to do a detailed how to guide once its up and running, nice to contribute something.

[Harro]

Oden,

Any further progress with this? I am also in Aus and am facing similar problems trying to get a wireless link up and running without breaking any regulations. The problem I have is I need the link to see over a small hill so the lower frequency spectrum is best for me however the limited power allowed is causing problems.
I am using the RFM-23BP and have been doing some experimenting. The following is what I have found that may be of assistance.
I setup the 434Mhz version to Tx and Rx at 472Mhz and surprisingly there was no noticeable decrease in received power (Measured using the RSSI register, I get a received power of -10 dBm at both 434 and 472 Mhz with units about a metre away and non tuned antennas (78mm 1/4 wave, setup for the 915Mhz testing)). When the 434Mhz version is set to run at 915Mhz the power is significantly decreased (received power of ~-70dBm with the more suitable antenna). All of the above measurements were taken with the device set at the minimum power setting (The data sheet doesn't specify what this is but through some basic testing I am fairly certain it is less than 100mW)
In terms of setup there is a great spread sheet on the hopeRF website. It is for the RFM23B but it will give you all of the required register settings to setup frequencies, packets, BW etc.

Good luck,

Harro

[Kilrah]

FWIW This is the power setting - measured power table I measured on my 1W Orange module:

7-469mW
6-497
5-408
4-426
3-385
2-365
1-279

This is with that module's 3.3V supply, but as you can see the power setting makes very little difference.

[oden]

Very interesting. I had given up on the idea of getting it running on 472Mhz as the 472 module is not available for small orders. Ended up ordered 2x RFM23B 1watt 915Mhz modules from Ebay, unfortunately these were the only units available as they have a different form factor so it won't be a simple swap out. They arrived the other day but not time to play ATM

Very surprised that you were able to get a result from setting the frequency to 472 and 915Mhz. I attempted 472 via custom code changes and the config tool settings. Both methods rendered the device deaf, would not even do spectrum analysis.

Can I ask how, or what you did? Would like to repeat your tests.

[Harro]

Can I ask how, or what you did? Would like to repeat your tests.

Hi Oden,

I just stumbled across this forum while googling this sort of thing so I'm not sure what you are using/doing and how compatible this is with what i'm doing (I.e I'm not sure what openlrsng is).

I am programming the RFM23BP directly via the SPI ports of a PIC18F microcontroller. This gives me full control over the module and allows me to TX/RX my own code (I.e Whatever I want). Changing from 433 to 472 Mhz is not a simple matter of changing the frequency band select register, there are a bunch of register changes involved however the spreadsheet mentioned takes care of all of these for you and makes it pretty simple . The code I am using to set the registers is as follows:

char config1[2] = {0x3B, 0x40}; //Start reg = 0x1C

char config2[6] = {0x7D, 0x00, 0x83, 0x12, 0x04, 0x1B}; //Start reg = 0x20

char config2a = 0x10; //Start reg = 0x2A

char config3[3] = {0x19, 0x38, 0x27}; //Start reg = 0x2C

char config4 = 0xAC; //start reg = 0x30

char config5[21] = {0x00, 0x02, 0x08, 0x22, 0x2D, 0xD4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00,
0x00, 0x00, 0xFF, 0xFF, 0xFF, 0xFF}; //Start reg = 0x32

char config6[2] = {0x10, 0x62}; //Start reg = 0x6E

char config7[3] = {0x2A, 0x23, 0x03}; //Start reg = 0x70

char config8[3] = {0x57, 0x33, 0x80}; //Start reg = 0x75

This is at a 2kbs data rate (For max range), a narrow BW (For min noise), Manchester encoding and packet handling enabled.

I have since replaced the antenna with a 474Mhz 1/4 wave and this now maxs out the RSSI whenever at close range.

Like I said, i'm not sure what you're doing and if any of this even relates but feel free to ask any more questions. I can send you my code if desired, it is programmed in C using XC8.

I have done some more testing regarding the power output and as Kilrah said it will not be possible to get these below 100mW so technically they are against the regs. I'm changing to the RFM22b module which has a TX power of 100mW.

It is legal to Tx up to 1watt in the 915Mhz band so this may be the best option for you.

Regards,

Harro

[RightRudder]

I am programming the RFM23BP directly via the SPI ports of a PIC18F microcontroller.
Harro

Hi Harro

I just saw this. I have been playing around with a PIC18F2553 together with some ANT transceivers and have an application which would need three transceivers arranged in a star network. The ANT devices are only 1mW and quite low bandwidth which didn't work out so well. Would your HopeRF module be suitable? I'd be interested in having a look at your code as a starting point for that if you don't mind to share.

Joe

[Harro]

Joe,

PM sent.

Harro

[oden]

Have finally received some RFM23B 915Mhz units, it appears they can operate at a wider frequency then the RFM22B. Setting to 433Mhz seems to operate even when this is a 915Mhz unit, have confirmed this by the spectrum analyser detecting other devices on their correct frequencies.

Mounting the RFM23B onto a RFM22B foot print was fun

Once the RF module was replaced it was a simple matter of compiling the code with the correct frequency set and off she went.

[RightRudder]

Interesting. One word of warning though, just because you can pick up a signal from a similarly configured second unit using the spectrum display does not mean everything is right. Usually there will be some output filtering or at least tuning (often the network does both functions) network components between the chip and the antenna. These components are typically not very wide band, certainly not over an octave of frequency shift. There will certainly still be some signal output but the antenna will not likely be matched to the amp and harmonic suppression would be poor. (you would see the harmonics on the SA display). Running the transmit in such a way may result in a lot of reflected power which will heat up the chip and do bad things and it also means the transmitted power would be low, affecting range. You might want to look into this more closely!

Joe

[jhsa]

433 Mega Hertz, is the frequency the module transmits.

João

[Kilrah]

That was spam

[jhsa]

Who approved it then?

I guess I was spammed then

João