HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Jan 09, 2013, 02:09 PM
Registered User
Columbus Ohio.
Joined Aug 2006
151 Posts
Always the naysayers...lol.
I guess we can respectfully agree to disagree.

Srnet...I am curious about what you found on the RFM22B modules. It sounds like you've got so valuable experience working with these...
RJKIRK is offline Find More Posts by RJKIRK
Reply With Quote
Sign up now
to remove ads between posts
Old Jan 09, 2013, 02:36 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
lol! i didnt actually disagree with anything you said. in fact you and i are possibly the only two on this website sharing the same criticism of hold type failsafes.

what i was saying is you seem to be about 10 years behind in thinking radios still use analog ppm fm/am radio transmissions. and your comment about all planes having rtf capabilty within 3 years also indicates a little out of touch. autopilots and rtl have been around for many years and havent yet climbed out of niche market status. even with cost down to less than $10 including gps like my rtl design they are still virtually restricted to the fpv crowd. kinda like the predictions 50 years ago that every garage would have a helicopter (or space ship) and robot servants. dont take offense but im just suggesting you might be a little out of the loop on current technology. im only trying to help you gain a more current level of understanding.

for in depth grasp of rfm22b capabilities and current developments i might suggest you investigate the openlrs project both here and on melihks flytron forum. some searching on arduino might help too.
dave1993 is offline Find More Posts by dave1993
Last edited by dave1993; Jan 09, 2013 at 02:46 PM. Reason: olrs and ardupilot
Reply With Quote
Old Jan 09, 2013, 05:21 PM
Registered User
Columbus Ohio.
Joined Aug 2006
151 Posts
Thanks Dave.

This is fun! I'm glad you're here.
I've followed a lot of these forums and know that you and JakeStew are a couple of...well.......let's just say.........characters......lol.
I've also seen your 9x new radio thread...very cool!!!

First, I never argued that analog is being employed. Let's just say it's more of a digital "bit banging" approach.
Yes I know about the Open LRS initiative using these Hope modules.
There's also a Google Group called "RF22-Arduino" that provides a library written by an Australian coder.

Yes autopilots have been around for 10 years, but so have Ferraris. If helicopters cost $5000 we may have seen one in every garage by now. It's always the price that's the showstopper(or for personal helicopters...a lack of airspace).

But these sensor chips for autopilots are getting remarkably cheap. Now even cameras are getting GPS. Millions of chips drives down the price.

Yet even today the price is still too much. For autopilot today you really need three pieces: a conventional RC Receiver($10-20 for a cheaply Orange), a separate Autopilot Board consisting of a dedicated MCU and Sensors, and finally, a second complete radio board.

From DIY Drones you're looking at $250 for a combination autopilot board and radio. That plus $20 for a cheapy RC Receiver get's us to $270...ouch!!!

What if you could get all three parts in one board for $25 to $50.
Would you stick to your Orange receivers and watch your plane fly off into the sunset.
Or your Corona receivers which will circle your plane round and round before crashing into the windshield of that Ferrari in the parking lot.

Or would you pony up an extra $25 to get an all in one receiver/telemetry/autopilot...and watch the plane land safely out in a designated patch of flying field...lol...

Oh...and you mentioned Flytron...
That's where I bought my Hope Modules and little red breakout boards so I could do standard size thru-hole attachment.

Also, would you consider PXX a modern protocol? Isn't that latest and greatest FrSky Protocol?
RJKIRK is offline Find More Posts by RJKIRK
Last edited by RJKIRK; Jan 09, 2013 at 05:29 PM.
Reply With Quote
Old Jan 09, 2013, 06:04 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by srnet View Post
The coronas allow you to set the failsafe. I normally trim my plane for a power off gentle glide with a soft left turn.
actually my main concern is with default behaviour not ability or inability to set failsafe. nearly everyone here on rcg insists they are careful to do this but i dont believe it. real world observations convince me most ignore this step and just use the rx as-is. many default to flyaway mode, spektrum kills throttle (ch1 only) but holds the rest, and as i said afaik flysky has by far the best failsafe by killing all channels. this has major implications not just safety but uav/multicopter applications. many call this no failsafe at all but in fact it is the best.

Quote:
Originally Posted by srnet View Post
And if you have at least some altitude you at least have time to warn others that you have a flyaway.
by "flyaway" i was mostly concerned about the plane taking off and disappearing over the horizon when out of radio range so no crowds nearby. not so much a safety issue. this would result from other factors during or right after takeoff.

speaking of corona my 4ch single conversion fm units were well behaved but the synth version just kept going with that stupid hold function. and it could not be changed by the user.
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 09, 2013, 06:32 PM
Registered User
Columbus Ohio.
Joined Aug 2006
151 Posts
Thanks Dave.

I got a hold of a PDF the PXX protocol off the 9xforums.com. PXX is what FrSky uses in its products. Here's the link to the forum discussion on "PXX Work", and if you go down to the 5th post, Erazz (the guy who wrote the er9x firmware) posted a link to the PDF of the PXX protocol.
http://9xforums.com/forum/viewtopic.php?f=5&t=72

I'm going to read through it and create a post that details what I mean about bit-banging, and not utilizing the full functionality of today's chips. Even PXX is a perfect example.

If you look back at the Spektrum introduction of 2.4Ghz. digital communcations, I think it was really a game-changer pushing us away from 72Mhz FM. I think another wave of change is just around the corner, and it will be all tied up in the seamless melding of conventional radios and receivers with telemetry and autopilot.
RJKIRK is offline Find More Posts by RJKIRK
Last edited by RJKIRK; Jan 09, 2013 at 06:47 PM.
Reply With Quote
Old Jan 09, 2013, 07:13 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by RJKIRK View Post
We're stuck in the past with methods akin to holding down and releasing a Morse Code ticker switch in correctly timed patterns. Then sending these square-waves over the air to be picked up by the receiver...which takes this clunky pattern, and tries to decode it, and spit it out down the correct servo wires.
looks like i underestimated your level of experience. sorry about that. but you might see where i got the impression from above that you were referring to old fm/ppm transmissions. and your followup comments pretty much described to a tee how most modern radios already work.

Quote:
Originally Posted by RJKIRK View Post
For autopilot today you really need three pieces: a conventional RC Receiver($10-20 for a cheaply Orange), a separate Autopilot Board consisting of a dedicated MCU and Sensors, and finally, a second complete radio board.
im not sure what you mean by "second complete radio board". rtl basically needs little more than a gps module and low end mcu added to an existing setup. heres an example of one that performed perfectly in several test runs last year:





you can probably make out the $8 dealextreme sirfIII module and atmel tiny avr chip. the one cent gps antenna take a while to lock but ok sfter that.far cry from chris andersons ardupilot fiasco. i am convinced intergrating the autopilot functions into the reciever is a good idea. id love to get rid of that clunky d8r and replace it with one of my tiny new frsky micro 4ch which now have telemetry. im also considering adding gps string parsing and waypoint sequencing to the half gram flysky rx i just got going. the frsky telemetry is nice though.

Quote:
Originally Posted by RJKIRK View Post
Also, would you consider PXX a modern protocol? Isn't that latest and greatest FrSky Protocol?
well... probably.. considering it hasnt arrived yet. but im not so excited by these new frsky protocols and radio. like i was telling jake, im not a big fan of oddball formats and exotic hardware. imo when it comes to buying development boards and getting a head start coding nothing beats generic chinese clones of well established products. ill let others stampede over the bleeding edge. when the technology has stablized and cheapo clones start to show up on ebay and aliexpress i jump in and take advantage. thats my sop.
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 10, 2013, 01:08 AM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,330 Posts
Quote:
Originally Posted by dave1993 View Post
actually my main concern is with default behaviour not ability or inability to set failsafe. nearly everyone here on rcg insists they are careful to do this but i dont believe it. real world observations convince me most ignore this step and just use the rx as-is
I would agree.
srnet is offline Find More Posts by srnet
Reply With Quote
Old Jan 10, 2013, 01:59 AM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,330 Posts
Quote:
Originally Posted by RJKIRK View Post
Srnet...I am curious about what you found on the RFM22B modules. It sounds like you've got so valuable experience working with these...
I found when first bread boarding the device that tuning on the TX at high power would trigger a RFM42\22 internal reset. Mostly the device (Si4432) registers would go to reset defaults or occasionally the device internal processor would crash, with the registers constantly changing. The device needed to be re-intialised in the former case (write out all the config registers again) and shut down and restarted in the case of the later.

Things that would cause the problem, or make it worse were;

1. Poor supply regulation, running direct from a single Lipo with a diode in series with the supply was not good, the change in Vf as current went from a couple of mA to 100mA or so during transmit would trigger reset.
2. Poor layout, breadboards being a prime example.
3. RF feedback, you want the RF to go out through the antenna and not couple back into the wiring or power supply of the RFM22.
4. Poor supply decoupling. Large low ESR capacitors on the supply help, but can be impractical in a small footprint layout.

A good PCB layout helps a lot, use a low dropout regulator such as TCS1262 or MCP1700. Put the regulator as close to the power pins of the RFM22 as you can. I fitted a 1uF ceramic and 150pF NPO ceramic almost directly across the RFM22 power pins.

Keep wiring from the battery short, avoid series diodes if you can.

A battery with a low series resistance, such as a Lipo, is better that one with a relatively high one such as a consumer NiMh.

Use a PCB ground plane.

Dont let an antenna wire dangle near the PCB, the RF will could back into the RFM22.

Build in software checks to see if the RFM22 registers have cleared, especially after turning on TX.

Having said all that, with a decent PCB layout, problems are few and with software strategies to detect and work around the problem you can have a reliable setup. You can also get by with only a 100uf or 220uf supply smoothing capacitor.

A filtered LNA in front of the RFM22 on the RX side will give you about 12dB of signal improvement or about 4 times the range.

And for best performance it is essential to tune the antenna. For a 433mhz you might think the correct length of antenna is 17CM or so. But take a look at the graph, measured field strength in -dBm at 50M as the antenna is trimmed from 30cm down in 1cm steps. Note that the field strength for 29cm is 19dB higher than the calculated length of 17cm.
srnet is offline Find More Posts by srnet
Reply With Quote
Old Jan 10, 2013, 08:13 AM
Registered User
Columbus Ohio.
Joined Aug 2006
151 Posts
Thanks for the info Srnet.

This is the first time I'll be playing with radio modules...so this is all good info. Not knowing much about antennas, I was scratching my head on how accurate the lengths of wire had to be.

As you stated, I guess 17 cm is the recommended lengh for this module. So I was concerned that if the antenna was a little short (16) or a little long (18) what kind of effect that would have. Would you loose 50% of the capacity? From your graph it doesn't sound like I need to be concerned. I can play around with the length and not break things.

And while on the subject of antennas...what kind of wire would you recommend? Will normal copper 22 guage hoopup wire work? Single strand or Stranded. I've got some 16 guage if that's better.

Initially I'm going to build prototypes on a 0.10" perfboard with copper pads on one side. So groundplanes and such are out of the question.

Thanks again!!!
RJKIRK is offline Find More Posts by RJKIRK
Reply With Quote
Old Jan 10, 2013, 09:04 AM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,330 Posts
Quote:
Originally Posted by RJKIRK View Post
Thanks for the info Srnet.

This is the first time I'll be playing with radio modules...so this is all good info. Not knowing much about antennas, I was scratching my head on how accurate the lengths of wire had to be.

As you stated, I guess 17 cm is the recommended lengh for this module. So I was concerned that if the antenna was a little short (16) or a little long (18) what kind of effect that would have. Would you loose 50% of the capacity? From your graph it doesn't sound like I need to be concerned. I can play around with the length and not break things.

And while on the subject of antennas...what kind of wire would you recommend? Will normal copper 22 guage hoopup wire work? Single strand or Stranded. I've got some 16 guage if that's better.

Initially I'm going to build prototypes on a 0.10" perfboard with copper pads on one side. So groundplanes and such are out of the question.

Thanks again!!!
I would not worry about the exact wire length, or sizes, although clearly the length will have a significant effect on range.

A tough flexible wire is probably best, less likely to break.

If you want to do antenna tuning tests (very important !!) you can write code to use one RFM22 as a field strength meter, by reading and displaying the RSSI register repeatedly.
srnet is offline Find More Posts by srnet
Last edited by srnet; Jan 10, 2013 at 09:20 AM.
Reply With Quote
Old Jan 16, 2013, 10:20 PM
Registered User
United States, ID
Joined Sep 2011
589 Posts
I'm not really happy with any of the modules out there. They really don't seem to be that well designed.

I'm looking at doing a board based on the OpenLRS type setup, with an Atmega 328, and based around the SiLabs reference design.

I won't be able to use the "grain of sand" sized SMT components that the Chinese use in HopeRF style modules, but I think that 600-800 sized components (grain of rice sized) shouldn't be that difficult to work with.

If anyone's interested I can start putting some files up.
jakestew is offline Find More Posts by jakestew
Reply With Quote
Old Jan 17, 2013, 03:00 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by jakestew View Post
I'm not really happy with any of the modules out there. They really don't seem to be that well designed.
yeah... not ONE of them will put out 100 watts.

seriously though i would be very interested in seeing such a design. specially with an avr chip like that. that would make the tarduino gang (thanks tc) happy too.
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 17, 2013, 03:15 PM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,330 Posts
Quote:
Originally Posted by jakestew View Post
I'm not really happy with any of the modules out there. They really don't seem to be that well designed
In what way ?

The Hope modules for instance seem to be very capable for their intended application.
srnet is offline Find More Posts by srnet
Reply With Quote
Old Jan 18, 2013, 04:25 AM
Registered User
United States, ID
Joined Sep 2011
589 Posts
There are some limitations with the HopeRF modules IMHO. I'm not a RF expert by any means, but I've looked over most of the SiLabs design and application notes. I think the PA matching network could be done better, especially by using a little more board space. There's a couple other things I'd just do a bit differently.

The range claims on the modules are pretty short compared to what they should/could be. Just kind of makes me think the design is not that well optimized.

With respect to the RFM22BP, if the amp really doesn't have any linearity to speak of then it's not very good for our application. With battery powered devices operated at ranges that should usually only need 50-150mW of power it really doesn't make sense to spam out a whole watt of power all the time.

There's a reference design using one of the SiLabs chips and a RFMD front-end module (PA + LNA) with an integrated antenna switch. Apparently the RFMD chip was specially designed (or maybe just specially marketed) for use with the SiLabs chips. I'd like to try out this combo and see how well it does.

SiLabs also had a similar reference design using a Skyworks front-end chip. It seems to be gone now though (I've still got a copy somewhere), so I'm not sure what the deal is there. Maybe it didn't work all that well.

To save board space and cost I'd love to use a Si1000 design. It only makes sense to throw everything in one chip if possible. But the Si1000 HopeRF module doesn't pin out anything to speak of. It's a shame to waste all the features on that processor for no reason.

It should be pretty easy to make a decent board. There's really not very many external components needed. If I'm going to etch or order a board then I might as well do the whole shebang at once. A module is only going to limit the design and save 10-20 minutes of soldering at most.
jakestew is offline Find More Posts by jakestew
Reply With Quote
Old Jan 18, 2013, 05:17 AM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,330 Posts
Quote:
Originally Posted by jakestew View Post
There are some limitations with the HopeRF modules IMHO. I'm not a RF expert by any means, but I've looked over most of the SiLabs design and application notes. I think the PA matching network could be done better, especially by using a little more board space. There's a couple other things I'd just do a bit differently.

The range claims on the modules are pretty short compared to what they should/could be. Just kind of makes me think the design is not that well optimized.
The output matching was no doubt designed around a particular price point product and with a particular size in mind. Better tolerance components, and restricted frequency range, would no doubt help, but I am sure the Hope guys know that.

For their intended purpose the Hope modules seem well enough designed, and I doubt they had long range RC Model control applications in mind when designing the devices.

I know about the range claims and the RFM22B, at least Hope are not extravagant, and they were somewhat surprised when I told them I had got 40km LOS out of a RFM22B at 100mW.

You can of course take the Si4332 and build your own circuit around it, but its not a device I would choose to use for RC Model control because of that glitch detector (the so called 'smart' reset)

I guess it depends on whether your a glass half empty or half full type of guy .....
srnet is offline Find More Posts by srnet
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Mod my 7CAP or buy a new cheap radio to use FRSKY system? OrlandoP Radios 2 Jul 18, 2011 06:12 PM
For Sale Werks Racing .28 Ceo Mods/Trx 3.3 Engines FT/FS Jamminracer18 Cars - Cars and Parts (FS/W) 1 Apr 18, 2009 12:17 PM
Build Log Project Gatika, Modern Voith Tug yacht boy Scale Boats 198 Feb 19, 2009 05:05 PM
Sold modded mini t arr for sale "cheap" or trade for mamba system jessnlaurie Cars - Cars and Parts (FS/W) 2 Oct 28, 2005 11:03 PM
FORMOSA project! Cheap, over 1:1 power system PunkerTFC Foamies (Kits) 12 Jun 16, 2005 09:45 AM