2.4GHz/868MHz/900 MHz radio chips / modules for radio control

[CrashingDutchman]

This thread is started to discuss the various chips/modules that are available for radio control. It is a continuation of this thread that was getting off topic: http://www.rcgroups.com/forums/showthread.php?t=748694#post8328742

[CrashingDutchman]

rmteo (assuming you are joining this thread),

Do you know the exact difference between the chips and the various available modules? I assume the modules contain a chip that works like the other chips we found, but have an amplifier circuit inside them?

I am not an RF expert at all, but do you know if it is even possible to amplify the signal somewhere between the output of the chip and the antenna?

CD

[darkith]

FYI,
I built a 900mhz system around the Integration/RFM chips, could be tuned to another band as well.
http://darkith.dyndns.org/~darkith/html/diyuhf.shtml
Even have a newer version with better PCBs and better code...just need a little motivation to package and post it. Have flown it for a few years now. Found the chip nice and easy to work with on the RF side, minimal components, etc.

Cheers,
David



My webserver should be accessible again.

D.

[CrashingDutchman]

Just checked you site and it works... You can find it here:
http://darkith.dyndns.org/~darkith/html/diyuhf.shtml (http://darkith.dyndns.org/%7Edarkith/html/diyuhf.shtml)

edit: lol - you posted the link in your quote... didn't see that

[village_idiot]

Yes, many of these chips are designed to have an external power amp, that's why Maxim makes two versions. Same with the Integration chips. Generally the modules already have any PA in them, but not always.

[village_idiot]

Just read your post about the TI chip, more important then listen before transmit is the (optional) forward error correction!

[rmteo]

rmteo (assuming you are joining this thread),

Do you know the exact difference between the chips and the various available modules? I assume the modules contain a chip that works like the other chips we found, but have an amplifier circuit inside them?

I am not an RF expert at all, but do you know if it is even possible to amplify the signal somewhere between the output of the chip and the antenna?

CD

Modules simply make it easier for you to work with the RF chips by combining them a with microcontroller, antenna matching circuits and usually an output power amplifier as well - but not always. The microcontroller is used to set up the operating parameters of the RF chip and to communicate with your application. Lets take the XBee modules - used by XPS - as an example. The are 2 versions, the standard and the Pro, see pictures below.

The main advantage to using modules is that they usually have agency (such as FCC, ETSI etc.) approvals/compliance/certification. This may not be of consequence in a diy situation but in a commercial one, it will be the responsibility of the developer to obtain the proper papers. The other side of the coin is that the modules are typically 5-10 times the price of the chip solution (excluding development costs and other overheads).

Most of the RF chips have an output power of around 0-10dBm - 1-10mW. You can amplify the output to whatever you need by incorporating a power amplifier. Using a power amp will also require 2 RF switches to switch out the PA in receive mode.

[rmteo]

Just read your post about the TI chip, more important then listen before transmit is the (optional) forward error correction!

What is interesting about Chipcon is that they have nearly identical 2.4GHz (CC2500) and 413/868/915MHz (CC1100) versions - although the same can be said for Nordic as well. :D

[GavinBLDC]

RF chips, RF modules and RF modules with micros.
Depends on what level of capabilty of the users.
RF modules for guys who want to get up and running faster.
Lots of IP required to DIY modules.
PC world ie wireless USB is pushing technology, modelers can reap the benefits of small size and weight.

Low volume then Sparkfun, Parallax, Digikey etc have the chips and modules.
Some Chip suppliers offer free samples, but more effort is needed to use them.

Toys like PiccoZ, TwinX etc give lowest cost.

Ease of code writing then AVR or PIC based solutions or what ever user is most familar with.

Smallest size/weight then all in single RF chip and Micro solutions are best.
TI/Chipcon, Nordic, Cypress WUSB, RFM

Two chip solutions are RF plus micro of user choice.

Many transciever chips can be used, allows for multi channels.
RF can be simple OOSK, FHSS, DSSS.
Freqs of use 433, 868(Europe) 900MHz, 2.4GHz.
27Mhz, 40MHz common for toys.

Need longer range then PA is required for most transmitters, this has issues with freq allocations in each country.

Gavin

[TheSteve]

The regular xbee (non-pro) has also recently received a hardware change agency approval update. The MCU and RF chip have been combined into a single chip by Motorola. It could make for a simple solution.(this is what I believe XPS is experimenting with for the micro 4 channel receiver Jim once spoke of)

[CrashingDutchman]

Unfortunately, I don't have much time on my hands, but it would be a nice winter project to make a tx/rx combo with a good range, maybe as a joint effort.

Anybody interested? It would be nice if we could make something where you can just drop in a different chip for 444MHz, 868MHz, 900MHz or 2.4GHz.

Any suggestions?

[village_idiot]

You can hit the first 3 with a single chip, and really I think it would be better to shoot for the lower frequencies when possible. Problem is that 900Mhz and 868Mhz don't have much space available for channels.

[rmteo]

The 900MHz band (which spans 902-928MHz in the US) is not too bad. You can easily get 50+ channels for FHSS which will allow the higher output levels per 15.247. Now, the European 868MHz band is another thing altogether - much more limited in space and with more restrictions regarding power and duty cycle. This is where the LBT might be needed.

[GavinBLDC]

If data rate is fast enough in 868MHz band then could use Ch ID addressing.
That's assuming all digital radio, 8 bit address would give 256 channels.
FHSS or DSSS would be better.

Does 433MHz have a higher allowed power in Europe? Then longer range?

[CrashingDutchman]

I did a quick search during my lunch brake and found this on the Dutch Telecom Agency website:

http://farm3.static.flickr.com/2382/1541988508_1e6a415944_o.jpg
I am not sure how much different the above is for other countries.

The website of the Dutch Telecom Agency can be found here: http://www.agentschap-telecom.nl/nfr/tpl/main_nfr_uk.html

[village_idiot]

Lower frequency for a given power equals longer range. A very rough statement, but generically accurate. It also means lower bit rates too.

[GavinBLDC]

SRD, short range device? 434MHz walkie talkies?
Can get these quite cheap, best to add a micro with digital CRC etc to avoid getting shot down by the neighbours kids:)

[rmteo]

It is just terminology. SRD (Short-Range Devices) is used in the EU while in the US they are referred to as ISM (Industrial, Scientific and Medical).

[AnthonyRC]

Unfortunately, I don't have much time on my hands, but it would be a nice winter project to make a tx/rx combo with a good range, maybe as a joint effort.

Anybody interested? It would be nice if we could make something where you can just drop in a different chip for 444MHz, 868MHz, 900MHz or 2.4GHz.

Any suggestions?
Not sure how many readers of this thread visit the FPV forum, but this may be of interest:
http://www.rcgroups.com/forums/showthread.php?t=728111

[rmteo]

Thank you for the link, Anthony. Very interesting stuff there.

[CrashingDutchman]

For me, the downside of Anthony's project are the 'expensive' Aerocom modules (allthough they are value for money - I realize that).

[AnthonyRC]

For me, the downside of Anthony's project are the 'expensive' Aerocom modules (allthough they are value for money - I realize that).
It is true, they are (relatively) expensive, but since a little more than 1/2 of the expense is firmly planted on the ground I'm not so worried :-)

The use of the modules does bring advantages though, module interchangabiliy between EU and US versions, and not having to worry about getting the design approved if I decided to sell them.

Incidentally, these modules seem to use the CC1010 internally.

[fmkit]

Not sure how many readers of this thread visit the FPV forum, but this may be of interest:
http://www.rcgroups.com/forums/showthread.php?t=728111

I use UHF only, miles of range and cost nearly nothing - $4.50 per radio + $1 microcontroller


http://www.rcgroups.com/forums/showthread.php?t=754288

[AnthonyRC]

I use UHF only, miles of range and cost nearly nothing - $4.50 per radio + $1 microcontroller


http://www.rcgroups.com/forums/showthread.php?t=754288
That is pretty funny, I did the same thing a couple of years ago for downlinking telemetry from a UAV, worked fairly well!
I hit a couple of problems with the radios that I had, one being that they would only transmit for a relatively short period of time before returning to Rx mode. To work around it you had to hold a button while applying power, was a bit of a pain.
I was also a bit worried about the legality of using this band (and its EU equivalent) for data.

[fmkit]

That is pretty funny, I did the same thing a couple of years ago for downlinking telemetry from a UAV, worked fairly well!
I hit a couple of problems with the radios that I had, one being that they would only transmit for a relatively short period of time before returning to Rx mode. To work around it you had to hold a button while applying power, was a bit of a pain.
I was also a bit worried about the legality of using this band (and its EU equivalent) for data.

Most radios don't have TX timeout, for example Midland and Uniden. They also allow overmodulating when PLL chip should signal CPU "out of lock"
Motorola does time out but the problem easily solved by releasing TX for about 80mS every minute. I made it full duplex to get telemetry real time and shortened control frames to 9ms for improve response compared to traditional 20ms.
Well, cause so many pointed at legality of FRS license free radios let's just forget about FRS. I know nothing.

[rmteo]

High power (up to 100mW) stuff for 2.4Ghz and 900Mhz.
CC2591 is a cost-effective and high performance RF Front End for low-power and low-voltage 2.4-GHz wireless applications.
CC2591 is a range extender for all existing and future 2.4-GHz low-power RF transceivers, transmitters and System-on-Chip products from Texas Instruments.
CC2591 increases the link budget by providing a power amplifier for increased output power, and an LNA with low noise figure for improved receiver sensitivity.
CC2591 provides a small size, high output power RF design with its 4x4-mm QFN-16 package.
CC2591 contains PA, LNA, switches, RF-matching, and balun for simple design of high performance wireless applications.
http://focus.ti.com/docs/prod/folders/print/cc2591.html

Integration's IA443x series of transceivers are members of the new EZRadioPRO™ product family. While retaining all the attractive features of the earlier products such as high integration, low cost, flexibility, low BOM, and easy design-in, these parts are targeted to more sophisticated applications and offer several enhanced parameters and features, including continuous frequency coverage from 240-960MHz, and output power up to +20dBm. Also included are built-in features like wake-up timer, low battery detector, transmit and receive data FIFOs, power-on reset circuit, and general purpose digital I/Os.
http://www.integration.com/prd_ia44303132.php.

[CrashingDutchman]

Are you working on something?

[rmteo]

Are you working on something?
Yes, but not R/C related.

[CrashingDutchman]

Are you going to use an amplifier for longer range?

[rmteo]

My primary focus is on the 900MHz ISM band and the IA443x is interesting for me as it has a power amplifier (and support circuitry) built-in.

The CC2591 is a good way to increase power for 2.4Ghz transceivers.

You can increase power by adding a power amplifier, Rx/Tx switch, balun etc., but the CC2591 has all of this integrated into one convenient package.

[Gordito Volador]

Are you working on something?

Are you kidding? :D Rene is the personification of the original "Mad Scientist" from what I have been observing. Please correct me if I am wrong Rene. ;)

Regards, Bill

[rmteo]

I wouldn't go that far. :) But thanks for the compliment (it is a compliment, right?). :p

[Gordito Volador]

I wouldn't go that far. :) But thanks for the compliment (it is a compliment, right?). :p

Yes it is a compliment. :o

[mrdunk]

anyone got any more module recommendations for this post?
i'm looking for a transceiver with relatively high bandwidth and over 1000m range.
must be available in the EU.

been digging around on DigiKey all day and come up with the following contenders.
all have a range of around 1000m or more.
all are in the 2.4GHz range. (anyone fond a similar product in other frequencies?)
prices in Euro.

NEC ZMXM-400-1
price: 22.41
Output Power: +20dBm
Sensitivity: -92dBm
Data Rate: 250kbps
interface: UART
link: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=ZMXM-401-1-ND

Artaflex Inc AWA24S
price: 24.63
Output Power: +20dBm
Sensitivity: -92dBm
Data Rate: upto 1MHz (depending on error checking and packet format)
interface: 4MHz SPI
link: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=748-1003-ND

Unigen Corp UGWJ4USHN33
price: 14.24
Output Power: +23dBm
Sensitivity: -97dBm
Data Rate: up to 1MHz (depending on error checking and packet format)
interface: 4MHz SPI
link: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=784-1013-ND

i was originally thinking of using the Xbee-PRO but having found these alternatives i'm not convinced the Xbee is the best option.
the NEC is very similar to the Xbee but cheeper.
the other two are also cheaper and more flexible (providing you don't mind the SPI interface.)

on paper the Unigen module looks like a clear winner to me.
the SPI bus is easy to interface to any microcontroller, the transmission range is in excess of what i'm aiming for and at that price i can crash it without breaking the bank...

what does everyone else use for medium and long range serial radio?


dunk.

[tclark]

dunk-
I bought one of the Unigen modules you listed, as well as another SPI interface one without the power amp (UGWG4USHN33A) and a USB interface one (UGQLUE4USE050A). These modules use the same Cypress radio chip as the Spectrum series of radios use, as do the Artaflex modules. Unfortunately, I have not yet found the time to play with them. I am interested to see what you come up with. I got them from Future Electronics, before Digi carried them. Digi is charging more for them, but they are easier to buy from.

-Tracy

[CrashingDutchman]

A friend just mailed me that he has very good results with a CC1101 and 250mW amplifier on 868 MHz. His connection was still very strong at >1000 meter (3000 feet) with almost no LOS (many buildings between TX and RX).

He used 2 home-made antenna's from this site: http://273k.net/gsm/designing-and-building-a-gsm-antenna/monopole/

[mrdunk]

so i got a few of the Unigen Corp UGWJ4USHN33 (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=784-1013-ND) modules in the post today.

it comes with a mini coaxial connector (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=H9161-ND) which is a little annoying as the cables that mate to that connector are almost as expensive as the modules them selves.

at the receiver end i plan to remove the connector entirely and just solder on a patch antenna to reduce weight but i want proper 2.4GHz antennas on the modules at the transmitter end.

fortunately after a little research (and when that failed, taking apart people in my office's laptops) it turns out this connector is widely used on Intel networking cards.
this includes the cards in most of the IBM T-series laptops.
i think there is a pile of broken IBM laptops in one of my work sites so i should have that hurdle under control.

i'll let you guys know how it turns out.


dunk.

[mrdunk]

so i'm making some progress with the UGWJ4USHN33 modules.
they are basically just a Cypress CYRF6936 with an RF amplifier on there so the Cypress CYRF69xx datasheets are actually a whole lot more use than the Unigen one.

the SPI interface works fine with my chosen AVR microcontrollers implementation of the SPI bus which simplifies things.

it's non trivial getting a set of startup parameters which work for these modules but now i have that part down i don't expect too many more problems.

now i have some TX power level code working i'm nearly ready to do some range tests.
i'll keep you posted.


dunk.

[village_idiot]

Hmmmm..... Could be Spektrum compatible if you work it correctly.

[CrashingDutchman]

Interesting...

[beanoman]

Hey, I kept forgetting to post.. I've been running Cypress 6935 modules with atmega88 controllers for a few months. I have mine working pretty well, but only on the lowest speed, they won't work for me if I try to set them to the other rates. Because of that I might switch to the 6936 modules but I haven't really decided yet.. Have you had any problems with switching the data rate?

Anyways, my current plan(changes pretty quick) is to use the smallest 6936 modules, if I switch to them, for a basic 4 channel receiver that works with standard stuff.. Do you think you'll be able to get some interesting features like telemetry and stuff into yours? I don't think I'll ever have enough time, the 6935 test setups have been sitting next to my computer for the last two months.

[mrdunk]

yup.
as it happens i just got it working to the point i expect to be able to fly with it although i'll be on the road for the next few days so it will have to go untested until then.

control of 3 servos with loads of spare bandwidth.
i'm only using the return signal for flow control at the moment but it will be trivial to add return data.

Have you had any problems with switching the data rate?
i didn't play with all the rates but mine work for GFSK and 8DR modes.
when i get a bit of time i'll post the config i'm using for these.

there's a clip of it in action here:
http://www.youtube.com/watch?v=HI3KNy9GKB8


dunk.

[ZAGNUT]

so i got a few of the Unigen Corp UGWJ4USHN33 (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=784-1013-ND) modules in the post today.

it comes with a mini coaxial connector (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=H9161-ND) which is a little annoying as the cables that mate to that connector are almost as expensive as the modules them selves.

at the receiver end i plan to remove the connector entirely and just solder on a patch antenna to reduce weight but i want proper 2.4GHz antennas on the modules at the transmitter end.

fortunately after a little research (and when that failed, taking apart people in my office's laptops) it turns out this connector is widely used on Intel networking cards.
this includes the cards in most of the IBM T-series laptops.
i think there is a pile of broken IBM laptops in one of my work sites so i should have that hurdle under control.

i'll let you guys know how it turns out.


dunk.

those little U.FL connectors and cables are used in just about all wifi equipment. laptops will have either two or three of these cables inside, each with its own connector.

lots of routers and desktop wireless cards use the same mini-pci cards made for laptops, sometimes under a metal shield. in these you will many times find a short coax with the U.FL on one end and an RP-SMA on the other end, perfect for an antenna set up on the Tx end.


dave

[beanoman]

Looks pretty good! Progress is slow over here. I got the new 6935 modules and other parts in last week, but I only have one board partially assembled, although there's more on it now. I hope to finish the pair and have some results this weekend, but things always end up taking a lot longer than they should :D

edit- Also, you can find those connectors on ebay(usually as adapters to a different size connector) for pretty cheap, I think..

[cstratton]

dunk-
I bought one of the Unigen modules you listed, as well as another SPI interface one without the power amp (UGWG4USHN33A) and a USB interface one (UGQLUE4USE050A).

Could I ask if you've had any lucky getting the USB one going?

If it's easy to work with, that seems like it could be great for a PC-based test tool... but programming USB micros can also be a real pain...

[tclark]

I have not had much time to play with it, and Unigen did not make life easy by providing nice documentation. I could not get any info on how to communicate with the micro on the USB module; they only had a demo application (and no source code for that). I too had hoped that the USB module would be 'easy to work with', and I'd bet that it would be with any reasonable documentation. Sadly, I have not found any.

-Tracy

[cstratton]

Sounds like one is left to figuring it out from the data sheet for the cypress USB microcontroller on there - and probably writing firmware for it. Which is possible of course, it's just a Project with a capitol "P" rather than simply plug it in and call this api to access it, make yourself a data logger and see what the off the shelf radio sets are actually saying to each other kind of thing...

Maybe a seperate FTDI usb-serial chip of the sort that can do SPI as an option to the usual async would be easier...

[tclark]

Writing new firmware for the micro in the USB module is the only way I can see to make it useful, but that is a big project (even if I already had programming tools and experience with the Cypress line of micros, which I don't). I also have the 'plain' modules with the SPI interface, and I need to cobble together some hardware to communicate with them. That is a less big project, but still bigger than I have gotten time for. Maybe over the holiday break.

-Tracy

[cstratton]

I would look at the FTDI USB to serial chips, I think those to SPI. If you want to go slow and still have a parallel port that works, too.

Will probably be ordering unigen modules, bare chips, and some atmel microcontrollers next week, but while picking up heli parts I was like- oh, look, buy a gyro sensor for $33 (which is a good price) and get a free cypress chip already on a board - in other words I ordered a replacement 4-in-1 for my walkera just to play with on the lab bench.

[mrdunk]

following up on my earlier post about my experimentation with the Unigen UGWJ4US modules,
i started a thread here: http://www.rcgroups.com/forums/showthread.php?t=971053 with a link to the project documentation of my success.

dunk.