LibrePilot Forum
Development => Hardware => Topic started by: isk on April 13, 2017, 07:04:15 pm
-
Hi, first post here. I'm trying to get OPLink running between a Revo and a Mini without much success
I set everything according to the docs with the Revo as receiver and the mini as coordinator and connected to a PC running GCS 16.09. Flash erased both and installed firmware, set everything in the config, and binding was successfull and I get wireless telemetry. The only problem is, the connection is only active when the devices are about 10cm apart, any further apart and the signal drops.
I always have the stock antennas connected on power up and only set the power > 0 when testing. I tried with two separate minis but same problem. Is this a problem with the OPLink antenna on the Revo? Or have I missed something blindingly obvious?
Thanks, Ian
-
It is likely the OPLink. Where did you get you OPLink from? There have been problems with clone OPLinks having short on the antenna, which (of course) greatly reduces the power/range.
If you have a multimeter, please test the resistance between the center pin of the antenna connector and ground (all of the other pins on the connector are ground). If it's very close to zero, you have a bad OPLink.
-
Hi Brian, I got them both from RadioC, the product link is https://www.radioc.co.uk/ProductDetails.asp?ProductCode=3116-1. I checked the resistance as suggested and get 0.6 ohms and 0.8 ohms respectively. I guess that's pretty close to 0? Thanks
-
Yes. That's not good.
The problem is that there is insufficient clearance between the center pin of the antenna and the inner ground plane. If the manufacturer tolerances are not good enough, the antenna pin shorts.
We were actually able to track down one manufacturer and told him about this, and he said he would fix it, but there could be other manufacturers, or these could be from the older batch.
In any case, the only fix is to de-solder the antenna connector and drill out the center hole with a very small drill bit to break the short. You will not be able to put the antenna connector back on after this, but you can solder a wire antenna directly onto the RFM22B module. I have done that on a couple of moudles, but you probably will want to talk to the re-seller and see if you can send them back.
-
Thanks again Brian, I did a bit of Google research and see now that this is a common problem. I emailed the reseller but don't expect a response until next week. Frankly, for the price and limited availability of these things, I might have a crack at the mod as you've suggested. Soldering a wire antenna onto the RFM22B seems achievable but I wouldn't know which wire to use or pins to solder. Can I chop and wire the MMCX end of the SMA pigtail that came with it?
Failing that, is there a reputable supplier selling working modules out there?
Thanks.
-
Take a look here:
https://forum.librepilot.org/index.php?topic=312.msg18007#msg18007
And a few post before for cutting pcb trace.
-
I don't know of any supplier that is known to have good ones. I haven't bought any in a while.
Any small gauge standard wire will work. I like using something like 24 gauge silicon wire, which is flexible. You just need to solder it onto the first pin of the rfm22b next to the antenna connector, and cut it to ~173mm in length.
-
Thank you both for the advice and guidance. I went ahead and made the mod with promising results. I'll do a full range test tomorrow but for now I get a really good telemetry signal at about 30m. I'll mod the other one later. I have attached an image of my work. Cheers :D
-
For anyone else looking for OpLinks, about 2 months ago I bought one of these and it was perfect (all working and frequency correct). Of course they printed my review because it was favorable:
http://www.xt-xinte.com/F16086.html
I also bought a second of these (OpLink with case with SMA) and the USB connector had to be resoldered this time (perfect a year or two ago). I like the SMA connectors. I have watched this link for a while, and the review numbers don't add up. I now have two sets, and thus two "OpLink Air" that I can't use except if I ever burn an RFM22B (I haven't ever so far).
http://www.xt-xinte.com/F17009.html
They also have Sparky2's for about $39 shipped. I bought 2 years ago and 2 months ago and both needed power jumper resoldering, and rf connector solder blob removal. Good baros. Again review numbers are complete lies, they have a 1 star, a 3 star and a 5 star review averaging out to 4.5 Go figure (standard overseas lies).
http://www.xt-xinte.com/F16821.html
These Sparky2's (about $50 shipped) are the real authentic black ones that aren't sold any more. Probably the last of the authentic stock. They appeared after the official seller (hobbiesfly) stopped selling them. I bought 2, two months ago. Perfect in all regards.
http://www.ebay.com/itm/Taulabs-Sparky-2-0-32bit-F4-MCU-Based-Flight-Controller-compatiable-Librepilot-/172353561802
I guess I have been lucky because all the RFM22B stuff (10?) I have bought in the last couple years (most recent purchases 2 months ago) has been good (no shorts and correct frequency) except for the one USB connector that had to be resoldered. (I have an uncrashed authentic Nano that needed USB resoldered when it just popped off one day. I blame ROHS solder.)
-
Hi guys. A question about the range of signal reception. With the classic antennas that are sold that distance of reception should be expected based on the transmission powers?
-
All max ranges assume 100mw setting. 1/4 the power goes 1/2 as far (e.g. 25mw goes half as far as 100mw), so you can figure out ranges for lower power levels. 1/64th the power goes 1/8th as far.
Any one with similar or different experiences please post. :)
There are reports of 7km on other omnidirectional antennas. This was probably under ideal conditions and should not be expected. I've gotten easily 2km on other good antennas (at 100mw ... that translates to about 250m at 1.6mW) without doing any of the extended range things:
- mounting the base station antenna on a pole
- directional base station antenna
- flying at a high angle above base
but doing the minimum:
- both antennas vertical
- aircraft always high enough to be in line of sight
Beware that simply banking the aircraft can cause loss of reception when at long range. This is an antenna polarization issue. Two antennas like this | | work well but like this | _ do not work well.
The wire dipole antennas that come with OpLinks and Revos are good antennas if configured and mounted correctly. I would put their performance as about as good as you will get. These wire dipoles should be mounted like ---| with the dashes being the coax coming out sideways and the two wires being vertical (one up, one down). A ferrite bead on the coax at the antenna wire junction probably helps.
Here is a post I did testing many antennas:
https://forum.librepilot.org/index.php?topic=4199.msg28510;topicseen#msg28510
My guess is that anyone who has decent antennas mounted/held correctly can hope for 2km LOS when both antennas are vertical but your RF environment may be different (noisy): You really really should either:
- have a working failsafe RTB
- fly with base antenna held in a non-optimal way (low or pointing toward the aircraft) so that when it goes out of range you can just raise the transmitter above your head with antenna vertical and know you can get control back.
You should also fly in attitude mode. Loss of FPV vision or RC control when banking in Rate mode will cause a crash.
Careful base station antenna mounting (high and clear of obstruction) and you might get 4km, but you should assume it will fail at any time until proven otherwise. Again, RTB should be considered a requirement.
Experts can replace the RFM22B with an RFM23BP (requires soldering rework station, expert soldering, electronics knowledge to add an additional clean 5.5V or 6V power supply for only the RFM23BP, possibly additional licensing depending on your country) which runs at basically 1000mW only until you measure/verify the lower power settings (which for the RFM23BP are undefined). There is a reason those lower power levels are undefined (the amplifier design doesn't throttle well/cleanly).
Be careful of switching power supplies, such as switching BECs. They can produce a lot of RF noise and greatly affect range. I had an issue before, so now I run linear BECs or at least test. The 3rd harmonic of 433 is 1299, which is really close to 1280. 433 will affect 1280 FPV reception on the ground (but 1280 shouldn't hurt 433).
Test test test. Use min power for tests. Do not set it on the ground for a test. Keep metal away. Turn things off (fpv cameras sometimes have switching regulators) or use linear power supplies for the test to see what range you can get with accessories on/off and using linear/switching power supplies. I get hundreds of meters on 1.6mW. When you get that your antennas are good and your noise level is good.
-
great you are a mine of information! I read that to check the quality of the oplink module, just measure with a Ohm Etro the resistance between the central and the ground of the antenna connector. is it enough and safe?
-
Some outputs are shorted out.
Measure with an Ohm meter. It should show as an open circuit (no connection) because it is capacitively coupled. There are reports of successful repairs where a trace was cut before the short and a cable with connector was soldered on, center conductor soldered upstream of the cut and cable shield soldered to ground.
-
hi
If I change rfm22 module to rfm23b how I can change transmit power via gcs such as 500mw or 1000mw
thanks
-
RFM22 - RFM23 comparison :
(https://forum.librepilot.org/index.php?action=dlattach;topic=3331.0;attach=7923)
(https://forum.librepilot.org/index.php?action=dlattach;topic=3331.0;attach=7927)
https://www.rapidtables.com/convert/power/dBm_to_mW.html
-
IMPORTANT
Good omnidirectional antennas will get you hundreds of meters on 1.25mw. When you have hundreds of meters at 1.25mw, and have increased to 100mw and still need more, then consider the RFM23BP. See post on 433mhz antennas tested.
https://forum.librepilot.org/index.php?topic=4199
There are 8 non-zero power levels controlled by 3 bits in the OpLink configuration. They are given names like 1.25mw and 100mw, but those are really just names and those two really mean "lowest power" and "highest power", so if you change rfm22b to rfm23bp and set it to 100mw you actually get 1000mw.
Several things to be aware of:
- 1W OpLink needs a 5V-6.00V power supply. You must power the RFM23BP with 6V to get a full 1000mW but the rest of the OpLink needs to be powered by the OpLink 3.3V regulator. That means you can't just connect the RFM23BP power (5V-6.00V) connection where the RFM22B power (3.3V) was connected.
- Due to the design of the RF section of the RFM23BP, the power level does not modulate very well. That is why lower power levels on RFM23BP are not defined in the datasheet. You can reduce the power setting, but don't expect it to follow any nice linear scheme. I also don't know if the RF splatters when you do this. I would suggest RF filters and an RF power meter if you you intend to run reduced power levels. Really... don't be an RF hog and splatter on legal users. ;)
https://www.ebay.com/itm/RF-Power-Meter-27-2400-5800-mhz-arduino-ad8319-5-8Ghz-2-4Ghz-antenna-tool-FPV/253921388053
https://www.readymaderc.com/products/details/low-pass-filter-for-433mhz-sma-antennas
- The RFM23BP pinout is the same as RFM22B even though it looks like one of the signals is inverted
- The RFM23BP is physically longer than the RFM22B, so you will probably be soldering one side and running wires for the other side.
- For just control, you can have a modded OpLink on the ground and a stock OpLink in the air, but for bi-directional communication, you need both ground and air OpLinks (or Revo or Sparky2) modified.
- Someone was making 1W OpLinks a year or more ago. Kind of expensive, but worth it if you aren't expert at de-soldering, don't have the equipment, or don't consider the tedious upgrade to be a fun part of the hobby.
-
Few things to add:
Some useful info here:
http://www.itluxembourg.lu/site/modding-the-orangerx-1w-modules-for-full-power/ (http://www.itluxembourg.lu/site/modding-the-orangerx-1w-modules-for-full-power/)
http://www.itluxembourg.lu/site/2-5w-narpat-mod/ (http://www.itluxembourg.lu/site/2-5w-narpat-mod/)
Consider powering an *unmodified* RFM23BP module with slightly less than 6V. I use a Shottky diode to drop the voltage down to 5.8V that still gives 1W RF output and does not stress its onboard 3.3V LDO regulator .
I've found info in rcgroups that tests/measurements on RFM23BP modules conform their stated spec and do not emit more than expected/desired in harmonics when working in 1W mode. No tests made with RFM23BP modified for higher than 1W RF output.
I just did a series of measurements how the output of a modified for 1W OpLink (RFM22B replaced with RFM23BP and separately powered with 5.8V instead of 3.3V) changes according different Max Power settings in GCS:
100 - 30.1dBm 1.0W
50 - 30.1dBm 1.0W
25 - 30.0dBm 1.0W
12.6 - 30.0dBm 1.0W
6.3 - 30.0dBm 1.0W
3.16 - 29.8dBm 962mW
1.6 - 29.7dBm 940mW
1.25 - 29.2dBm 836mW
0 - 0.0dbM 0.0mW
So just set Max Power to 100. RF output power can be modified by changing power supply voltage of RFM23BP in the range 3.3V - 6V (or higher - up to 10V, if you dare to try "Narpat" mod or similar).
If you run a modified OpLink on the ground and a stock OpLink in the air, the one on the UAV hugely benefits from the fact its antenna is high above the ground and not affected by it, so the difference in link quality of control packets and back telemetry is not that noticeable.
P.S. A small comment on the "Narpat" mod: I believe the useful changes are just adding a second R25 transistor in parallel with the existing one and cutting the trace/adding 100nF capacitor to provide a separate power supply for RF MOSFET.
-
When I first built one, I only did power output testing using a simple field strength meter. My testing showed the first few GCS settings were all about the same, but that power did reduce significantly for the lower values.
I have since bought an ExpertRC power meter and I just now tested the RFM23BP OpLink power vs. GCS configured power level. RFM23BP is powered with a 3A Hobbywing switching BEC that has a jumper for 5V or 6V. I currently have it set for 5V (I thought I had it set for 6V) and measures 5.13V at the RFM23BP under load.
I have a second power option because my circuit has a diode from USB power to RFM23BP power so I can run it with just USB and get a measured 3.95V (long USB cable and diode both reduce the voltage) at the RFM23BP under load.
There are some things that don't add up, like at the 100mw setting, the 3.95V power is less than half of the 5.13V power, but at the 1.6mw setting, the 3.95V power is more than half of the 5.13V power. I suspect that the 3.95V voltage increases at reduced loads.
GCS Power Power
Power at at
Level 3.95V 5.13V
-----------------------
100mw 407mw 1016mw
50mw 380mw 1007mw
25mw 380mw 989mw
12.6mw 380mw 955mw
6.3mw 331mw 851mw
3.16mw 245mw 566mw
1.6mw 112mw 214mw
1.25mw 68mw 129mw
The actual values could be a bit lower than shown in this table. My research makes me think it should be less than 1W for 5.13V. The important thing is that for this batch of RMF23BP in this circuit, the actual power levels produced are nearly identically high for several of the top programmed power levels but the actual power levels drop significantly for programmed power levels below those top levels.
Note that this table is quite a bit different than JDL's. There is a reason that the lower power levels are undefined in the datasheet and this may be part of it. On the other hand, JDL discussed some Narpat mods. My RFM23BP is stock and came from eBay for about $7 a few years ago. Moral of the story is that you probably need to measure your own circuit. You probably want to do that anyway to make sure your output power is as expected.
-
so,
when I change Tx out power 100mw (via GCS) rfm23 module is setup to 30dbm (1watt) (answer is yes :) )
and
when I change ground tx to rfm23 module, air module must change to same rfm23b
-
when I change ground tx to rfm23 module, air module must change to same rfm23b
Not obligatory. Unless you plan to fly with control and telemetry far away (more than 10km in RF clear environment) with plain omnidirectional antennas.
-
thanks
only my flight plan whit in 2-3 km (so no need change to air side rf module)
-
thanks
only my flight plan whit in 2-3 km (so no need change to air side rf module)
For such short distances you really do not need more than standard 100mW OpLink on both sides. On the other hand, extra transmitter power can help in very noisy environment, like near radio towers. Although, flying in such areas is generally not a good idea. :)
-
You should be able to get around 2km on 100mw with good (omnidirectional) antennas and clear Line Of Sight.
Set up and test Return To Base close by. Very important to have it set in Flight Mode Switch and in Input->FailsafeSettings. Test close by, by switching off ground OpLink/Transmitter (not just 2.4GHz transmitter if you use e.g. a 2.4GHz to 433MHz "relay box"!! ) !
Next flight you can fly far out to find out how far it goes.
Edit: @JDL beat me to it. :)
-
thanks
only my flight plan whit in 2-3 km (so no need change to air side rf module)
For such short distances you really do not need more than standard 100mW OpLink on both sides. On the other hand, extra transmitter power can help in very noisy environment, like near radio towers. Although, flying in such areas is generally not a good idea. :)
thanks for nicely helps
other thinks;
how I can fly POI flight mode whit revo whit flight mode switch
I was reading some poi flight mode whit revo but not clearly
-
I have never used POI mode. This is where I would start:
https://librepilot.atlassian.net/wiki/spaces/LPDOC/pages/84672520/Autonomous+POI+setup
To ask a question about a different subject, it is best to start a new thread in the correct section of the forum. That way, it can be searched and seen by people interested in that topic. So for further discussion about this new topic, please start a new thread. :)
-
thanks you are fully right
new post about POI opened
https://forum.librepilot.org/index.php?topic=4509.msg30693#msg30693
-
You should be able to get around 2km on 100mw with good (omnidirectional) antennas and clear Line Of Sight.
Set up and test Return To Base close by. Very important to have it set in Flight Mode Switch and in Input->FailsafeSettings. Test close by, by switching off ground OpLink/Transmitter (not just 2.4GHz transmitter if you use e.g. a 2.4GHz to 433MHz "relay box"!! ) !
Next flight you can fly far out to find out how far it goes.
Edit: @JDL beat me to it. :)
is also setting baud rate to 56000 for telemetry and control
because when I downgrade baud rate to 38400 (for long range tx power) I got master EVENT alarm
is it normal
-
It is normal and we sometimes remember to advise to do this. :) More information needs higher data rate.
-
It is normal and we sometimes remember to advise to do this. :) More information needs higher data rate.
does it affect flight or stabilization
-
For me at least, if the data rate (com speed) is set too low it basically doesn't work at all or at least poorly enough that it is obvious there is a problem. When your link includes Control you would see noticeable control lag or no control. For data (telemetry) you might notice data dropouts or complete lack of data.
I suppose there are configurations tweaked to have just a little too much control/data for the rate that would work intermittently with some dropouts or delays.
Even with rate set correctly, when you fly far away and start to drop packets, you have basically the same effect as a rate set too low.
Note that for long range flying, where dropouts (control or FPV videol) and lags are possible, you really need to consider using Attitude mode (or really any GPS flight mode). Imagine loss of FPV video. You give a roll command and then don't know if you are upside down or sideways or level. Better to hold the stick for a given attitude than a given roll rate.
-
thanks again
can i benefit this link openpilot osd code
https://github.com/cdaller/Arduino-Openpilot-UAVTalk
because I need only RSSı,batt voltage ,satellite count data