Incorrect Deviation

[Brian Patterson]

I am currently using the RFM22 board on an Arduino Uno for transmit only.  I am using the Direct FIFO mode since I don't need the preamble, sync words, etc etc.  While testing at 413MHz with a deviation setting of 100KHz I have noticed that my FSK modulation is running at approximately 145KHz and 40KHz.  I need the modulated signal to be closer to 94KHz and 44 KHz.  I have tried several dozen values for the deviation and none of them gets me close plus I have experimented with almost every setting I can think of to try and get this gap narrowed down.

I am praying that somebody out there might be able to clue me in as to how I can accomplish this. (trying to control a home automation device is my goal).  Just FYI - here are a few more deviation settings I tried and the results of each (all at 8.34 kbps which is about where I know I need to be):

• 32KHZ -->  85KHz/20KHz
• 25KHz --> 76KHz/23KHz
• 75KHz --> 125KHz/9KHz
• 125KHz --> 180KHz/69KHz

I am also not sure if this is related *but* in the past when dealing with FSK - my experience had always been that a "1" bit would result in the high frequency and a "0" would result in the lower frequency.  What I am seeing is complete opposite.  "0"'s are generating the higher frequency and "1" bits are resulting in lower.  It's not that big of a deal since I just flipped the bits on everything I am sending to get the appropriate signal but now I just need to fix the darn deviation issue.

ANY help at all would be greatly appreciated! 

Brian P

[Mike McCauley]

Hi,

you might wan to check the RF22 register calculator referred to in the doc.
That might allow you to construct your own modemconfig register settings to
achieve what you want.

Cheers.

On Wednesday, June 18, 2014 11:34:25 AM Brian Patterson wrote:
> I am currently using the RFM22 board on an Arduino Uno for transmit only.
> I am using the Direct FIFO mode since I don't need the preamble, sync
> words, etc etc. While testing at 413MHz with a deviation setting of 100KHz
> I have noticed that my FSK modulation is running at approximately 145KHz
> and 40KHz. I need the modulated signal to be closer to 94KHz and 44 KHz.
> I have tried several dozen values for the deviation and none of them gets
> me close plus I have experimented with almost every setting I can think of
> to try and get this gap narrowed down.
>
> I am praying that somebody out there might be able to clue me in as to how
> I can accomplish this. (trying to control a home automation device is my
> goal). Just FYI - here are a few more deviation settings I tried and the
> results of each (all at 8.34 kbps which is about where I know I need to be):
>

> - 32KHZ --> 85KHz/20KHz
> - 25KHz --> 76KHz/23KHz
> - 75KHz --> 125KHz/9KHz
> - 125KHz --> 180KHz/69KHz

>
>
> I am also not sure if this is related *but* in the past when dealing with
> FSK - my experience had always been that a "1" bit would result in the high
> frequency and a "0" would result in the lower frequency. What I am seeing
> is complete opposite. "0"'s are generating the higher frequency and "1"
> bits are resulting in lower. It's not that big of a deal since I just
> flipped the bits on everything I am sending to get the appropriate signal
> but now I just need to fix the darn deviation issue.
>
> ANY help at all would be greatly appreciated!
>
> Brian P

--
Mike McCauley VK4AMM mi...@airspayce.com
Airspayce Pty Ltd 9 Bulbul Place Currumbin Waters QLD 4223 Australia
http://www.airspayce.com
Phone +61 7 5598-7474

[Brian Patterson]

That is what I have been using to construct my modem config.  Now matter what I try I haven't been able to get the deviation down to where  need it.  Granted I am no whiz when it comes to RF so I assumed perhaps some setting in there could help me do what I need to do (oversampling stuff, clock settings, etc...)..

[Mike McCauley]

So this sounds like an issue to do with the RF22 chip, rather than the
RadioHead library?

How are you measuring the deviation?

Cheers.

On Wednesday, June 18, 2014 02:24:01 PM Brian Patterson wrote:
> That is what I have been using to construct my modem config. Now matter
> what I try I haven't been able to get the deviation down to where need
> it. Granted I am no whiz when it comes to RF so I assumed perhaps some
> setting in there could help me do what I need to do (oversampling stuff,
> clock settings, etc...)..
>
> On Wednesday, June 18, 2014 1:34:25 PM UTC-5, Brian Patterson wrote:
> > I am currently using the RFM22 board on an Arduino Uno for transmit only.
> > I am using the Direct FIFO mode since I don't need the preamble, sync
> > words, etc etc. While testing at 413MHz with a deviation setting of
> > 100KHz
> > I have noticed that my FSK modulation is running at approximately 145KHz
> > and 40KHz. I need the modulated signal to be closer to 94KHz and 44 KHz.
> > I have tried several dozen values for the deviation and none of them gets
> > me close plus I have experimented with almost every setting I can think of
> > to try and get this gap narrowed down.
> >
> > I am praying that somebody out there might be able to clue me in as to how
> > I can accomplish this. (trying to control a home automation device is my
> > goal). Just FYI - here are a few more deviation settings I tried and the
> >
> > results of each (all at 8.34 kbps which is about where I know I need to
be):

> > - 32KHZ --> 85KHz/20KHz
> > - 25KHz --> 76KHz/23KHz
> > - 75KHz --> 125KHz/9KHz

> > - 125KHz --> 180KHz/69KHz

> >
> > I am also not sure if this is related *but* in the past when dealing with
> > FSK - my experience had always been that a "1" bit would result in the
> > high
> > frequency and a "0" would result in the lower frequency. What I am seeing
> > is complete opposite. "0"'s are generating the higher frequency and "1"
> > bits are resulting in lower. It's not that big of a deal since I just
> > flipped the bits on everything I am sending to get the appropriate signal
> > but now I just need to fix the darn deviation issue.
> >
> > ANY help at all would be greatly appreciated!
> >
> > Brian P

[Brian Patterson]

Excellent question.  Either it's a chip issue - in which case I would have expected to hear others complain about it.  Or even more likely - I am just not using the library right - missing a setting, wanting a setting that isn't standard and thus need to "tweek" something....?  There are quite a few settings out there that I just don't get (i.e. clock recovery oversampling rate, Gear Shift, Timing loop gain, etc)

I checked the deviation with an oscilloscope.

[Mike McCauley]

Hi,

On Wednesday, June 18, 2014 02:46:38 PM Brian Patterson wrote:
> Excellent question. Either it's a chip issue - in which case I would have
> expected to hear others complain about it. Or even more likely - I am just
> not using the library right - missing a setting, wanting a setting that
> isn't standard and thus need to "tweek" something....? There are quite a
> few settings out there that I just don't get (i.e. clock recovery
> oversampling rate, Gear Shift, Timing loop gain, etc)

All the radio internal settings are encapsulated in the modemconfig. The
builting modemconfig parameters included in the library are derived from the
calculator. So I would seriouslyt expect them to be right (modulo the
manufacturers correct information). There are no libray specific paramters that
should be relevant to deviation.

Cheers.