FrSky 2.4ghz - looking good so far!
An index compiled by Renatoa, many thanks for that!
Beta test: http://www.frsky-rc.com/download.asp?id=23
Alternate manuals, wiki:
The most comprehensive: http://www.eflightwiki.com/eflightwi...rSky_Telemetry
Dan Thompson Receiver Chart - all you want to know before buying a Frsky receiver
First RC world headsup, Bruce review:
Martin (PLMS) spectrum analyser RF tests:
kneedrag video: Frsky on 9X V1 radio
V2 first report of RF levels changing:
PPM input issue
Start of discussions, followup on... many pages
Chase Wu first solution for PPM input protection:
Other relevant posts on PPM input issue
Alexcmag, 2874m on Sao Paulo, Brazil
Daemon, 7.88km with 5dB antenna:
20km with patch and booster
MPX Royal EVO:
JR XP7202 (with issues):
Hitec Optic 6 Sport:
Polk Hobbies Tracker III (with issues):
DX7, DX8, DX6i, DSX9, DX6i
Hardware issues and reports:
VD5M bind wiring color issue
Blazen 1024 and Futaba 9Z supposed (connector) multiple issue, warning followup on many pages:
D8FR replacement with V2 without any warn issue:
CPPM firmware, reflashing how-to:
3rd party developments:
Green LED tap
Beep codes meaning
Bind board alternative for V8HT and DHT modules
************************************************** ************************************************** *********************************************
Also posted in the GiantCod store forum courtesey of Rob Carpenter
I'm in possession of a set of the FrSky 2.4 GHz modular gear – thanks Rob!
The FrSky arrives as a transmitter module, an eight channel receiver and a longer than usual aerial. My first impressions were that it's a step up from Corona, having the rx aerials attached with micro plugs, just like the big boys Futaba and Spektrum.
The large tx aerial was a bit of a surprise, there was some comment on RCGroups website about, the manufacturer stating that the longer (5dB) aerial should give just as good range as the shorter (3dB?) aerials. Something to do with radiation patterns. It plugged into my FF7 and fired up straight away.
The instructions are reasonably clear for chinese translations with minimum Chinglish phrasing. In all it took about 30 seconds from opening the box to binding the rx. Just like the Corona gear, you hold down the button on the back of the tx module while switching on, and the module LED starts flashing to show it's in bind mode. Now power up the rx while pressing down it's bind button. It binds so quickly I didn't realize anything had happened first time around. A red and green flash as you switch on shows that the rx has found a tx and mated. Switch everything off, then back on (the traditional TX first way) and the rx locks on within a second or so. Being used to Corona and Radiolink, to me this was amazingly fast.
I tried plugging in a variety of servos, the only ones where any stepping was in the least visible were the TowerPoo ones. Futaba, DYS and Emax all worked smoothly enough.
I tried various settings and mixes, but I couldn't replicate the glitching
that the early Coronas had shown when used with my FF7 at certain settings for ch 5. As an aside, I had a new Corona module off Rob a couple of weeks ago, and this doesn't glitch at particular settings of channels 5, 6 or 8 either – looks like they've fixed the software in the TX modules.
I can swap between the two Corona and the FrSky modules, and the 'old' Corona plays up every time, the other two have behaved perfectly so far.
The receiver is a shade narrower than the Corona 8, and generally more
conveniently designed than the RadioLink ones, but then I like end
connectors for the servos rather than having them poking up in the air. The bind button and indicator LED are safely recessed in front of the case, but they are still easy to access. Alignment slots for Futaba style plugs are molded into the case, but JR style plugs go in easily enough as well.
I ran it in a small electric powered wing (indoors!) with the aerials
exiting the rx case about 1 cm from the motor, and the esc alongside the
Rx, no problems of any kind, but then I wouldn't really expect any.
The failsafe set up is a matter of holding the controls where you want them
set and pressing the bind button on the rx. Two flashes of the LED show that it's recorded the positions.
Switching of the tx gives about 1 second delay before the failsafe kicks in - I had 5 servos hooked up at the time and all went to their assigned settings. After switching the TX back on it took the rx something like 2
seconds to re-aquire the signal and return to normal service.
The instructions say when switching on to wait for a green light on the TX before turning on the RX – wrong, if you wait for green, you will be waiting a long time. It's an orangey/pinky light for normal service.
Press the TX bind button for 3 or 4 seconds once the Tx is on, and it goes
into range check mode, now it shows a green light. It claims to cut
output to 10%, which should give a ground range of 60 metres. I'm roping
in a clubmate to check that out tomorrow, we will also try it alongside a
variety of other 2.4 sets just for a laugh. It totally ignores my Corona and RadioLink sets unless I touch the 'other' tx aerial against the FrSky rx aerial – no signal swamping there then.
Last night I had a play with a variable power supply. The rx was hooked up to one SD200, an SD200 clone and a good old fashioned Futaba 148. I started with a 6 volt supply, and slowly dropped it down, At 3.1v the SD200 clone gave up the fight and the SD200 started getting flaky while the 148 still followed the stick movements, just very slowly. At dead on 3.0v, the LED on the rx went out and the system froze. The rx came back to life and got an instant lock somewhere between 3.05 and 3.1 volts. I know Xjet's Review site records the Corona as coping down to 2.5 v, but as the servos were dying at 3v, I think that last 0.5v is academic. If your power supply is dropping to these sort of voltages you've got serious problems beyond the rx.
All I want now is a dry day to try it for real, but I'm quietly confident it will work just fine. As long as it plays ok with other 2.4 sets en mass that is. I was so impressed by my initial impression that I've gone and bought a second hand FF8 from the classifieds to give it a permanent home.
To be continued when I can get to the field - not the best time for an assessment, January, last week snow, today rain. Never mind, I'll have to get inventive.
Last edited by BrumBob; Feb 11, 2014 at 10:00 AM.
Turned out nice again.
Ground range in range check mode turned out to be 170 paces - call it 150 yards. The instruction sheet gives the full power output as 60mw, and the range-check mode as 10% of the normal output with a recommended minimum operating range in check mode of 60 metres for what it's worth. It wasn't an ideal day, the clear sky and low sun meant the model became hard to orient much sooner than if it had been a little overcast. I'd got the receiver in a 100" old school Sunrise soarer converted to electric power. I set the failsafe for a left turn and power off, and tried to avoid turning left on the climb. I admitted defeat when I could only tell which way the model was turning when the wings started fluttering and flashing in the sun. We estimated height around 800-1000 feet. Thinking back to the days when I launched models using a 150m surgical bungee, I voted for the bigger number. Either way, it was far more than I would need in any normal situation.
There are only 2 flaws I've found so far:
1) the instructions say the transmitter module light should show green in normal use, when it's actually an orangey-pink colour.
2) The transmitter module socket molding could do with a chamfered edge to guide the transmitter pins in as it didn't want to drop straight in the Robbe tx. This was cured by taking the sharp edge off the molding with a scalpel blade. It's not a good picture, but it shows the offending item after trimming the rhs edges as they face you.
As far as I'm concerned based on playing around with this for a couple of days, it's next home will be in an OS61 powered Katana, something I wouldn't even consider with RadioLink and it's single aerial, and I can never feel totally at ease with Corona in view of it's 'idiosyncrasies'. Admittedly I've not had any problems using the Corona gear in my Acrowot, but I really don't like flying larger models without a failsafe.
Last edited by BrumBob; Jan 17, 2010 at 02:01 PM.
I was also given a FrSky set to test for GiantCod, thanks Rob!
Not owning a Futaba modular radio, I had to ponder how best to use my Frisky module. I'd intended building a new encoder board for it and fitting it into the case of my old 27mhz DigiFleet - but as BrumBob has been flying it already, I though I'd better do something quick!
A quick poke with the scope around 6 different transmitters revealed just one that would supply the same PPM format & polarity to the module as a Futaba set would - an old Irvine Sanwa that belonged to my late Dad. Futabas feed the module with positive going pulses at full battery level, just like the Sanwa.
How to connect it up? A standard header just fell into the socket - no good at all, so I emailed BrumBob to measure his Futaba pins for me - much longer pins were needed to connect to the module socket, and a section of a wire-wrap IC socket proved ideal. So all connected up, mounted on the back panel and bound as per BrumBobs instructions above (I didnt get any with mine) it all burst into life. I rediscovered one long-forgotten trait of the old Sanwa - throttle is channel 1, 2 is aileron, 3 is elevator and 4 is rudder. Weird, but no problem.
On the bench it seems very linear. I cycled the tx power to check the re-lock time - almost instant. Tried the failsafe, works fine. Low power for range-checking is invoked by holding the tx button for a sec as BrumBob says, but note also that you can cancel it & return to full power with another press. This makes it quite feasible to 'try' low power in flight.
I'll add more comments soon but here are a few poor quality photos to show the current set up. I'll replace these with better ones taken in daylight.
Note the proper ufl connectors on the receiver antennas, making them really easy to replace.
Somehow I'm feeling very confident about this set, dont know why, it just seems to sit on a much higher tier than the other cheap sets.
Time will tell, but its looking good.
Hello,every one. I am Chase From FrSky.
Thanks for your attention.
I would like to explain one point about the range check of our 2.4GHz unit.
Actually, the power of the transmitter module will be reduced to ab. 1/100-th of the nominal value, and the effective distance will be shortened to ab. 1/10-th of the nomial value based on our test.
Secondly, It's really an orange light for normal service as BrumBob mentioned.
We will correct the instruction manual, thanks to BrumBob
Chase, you might want to look into the alignment of the transmitter socket. It's only a fraction of a mm, but with no chamfer or slope on the molding to guide the pins into place it can stop the module going in at all.
I've no idea how that 1/100th of the signal (0.6 mW?) for range check compares with other manufacturers, but it's a long walk when you are looking over your shoulder to see what's happening at the model.
Yes, RobertWing understand correctly my meaning about our range check.
Regarding the module chamfer, thanks for your suggestion which is very helpful to us. We will improving our module(as your suggestion to add chamfer) to guide the pin into place easily.
What I do know is that 150 yards was impressive, and by that reasoning should give 1500 yards - nearly a mile - ground range! Unfortunately the only Spektrum module we had to hand on the day for comparison was faulty, giving just 2M range.
Chase - that's encouraging that you are taking action on the molding shape so promptly. Glad I could be of assistance.
It seems that there are some chinese 2.4Ghz modules around used by many manufacturar. It looks like that Frysky, Corona, Flydream, WOWRC are all using excactly the same chines RF modules.
Could you confirm that the chipset used consist of a cc2500 Texas Instruments, and a T212 RF amplifier?
The cc2500 can be used for frequency hopping and DSS, likely each manufacturar implements it's own software around.
Could you also make a close up of the antenna diversity on the RX board?
It is indeed the usual CC2500 tranceiver, at the front end the antennas connect to an ES02 change-over switch followed by an unmarked 16-pin RDK chip which I guess is an RF bidirectional amp (T212 equiv?), then another ES02 before the CC2500. I guess the 2500 selects which of the two antennas to use based on either digital link quality or analogue signal strength. I dont know why the second switch has been implemented - input and output switching suggests the board was intended to be used as a tranceiver.
I expect this will raise the question once again of "does antenna switching constitute diversity". My own view is yes, the unit as a whole has a choice of signal paths via one or other of the antennas. YMMV, as they say over there.
Last edited by phil_g; Jan 19, 2010 at 05:37 PM.
We would like to note the pin annotations via a little sticker or note it in the instruction manual.
What's your opinion, I look forward to your thoughts on my comments.
Ok Confirmed it is indeed same RF hardware.
The RDAT212 have a single 50 Ohm antenna output and contain LNA for recption and PA for transmission, also it has some internal tx/rx switch.
The RDAES02 are antenna switch, I do not understand why there are two of them, also I am not sure if the switch are controlled by RSSI.
Would be interesting to find the datasheet of these RF modules used.
Btw Could you test reception range by removing one or the other antenna and check if you find any range difference?
Last edited by fernandez; Jan 20, 2010 at 07:25 AM.
I would be interesting to look with a scope on the ctx & crx pins of the front-end ES02 to see if the 2500 switches antennas only when the link is bad, or if it constantly switches all the time - in other words, at close range with a good signal would crx/ctx be static, or a squarewave. (or if it doesnt switch at all, like Corona!)
Xjets FrSky review is excellent but unfortunately he didnt include this particular test.
EDIT: I asked Rob from Giantcod (our importer) to query this with FrSky. This is their response:
Giantcod Presales Enquiry Posted On: 22 Jan 2010 04:11 AM
I just asked them...
Response is below:
'jonathan: It does switch when signal lost at one anttena.'
I've not had chance (weather!) to try the system in practise yet...
Last edited by phil_g; Jan 22, 2010 at 06:03 AM.
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|Mini-Review||The video for FrSky 2.4Ghz||Chase Wu||Radios||0||Dec 16, 2009 02:18 AM|
|Discussion||Realflight G5 Combat so far so good!||splitpilot||Simulators||0||Nov 22, 2009 07:36 AM|
|Discussion||So far... So good||rikks||Mini Helis||35||Feb 17, 2006 01:42 PM|
|Aerofly Deluxe so far so good!||Gary Morris||Simulators||5||Aug 19, 2005 08:23 AM|
|Schulze 18.46K v7.02 - so far, so good!||Jason M||Electric Heli Talk||4||Aug 11, 2002 02:25 AM|