Mavic pro battery & antenna mod!!!!!!!!!!!!!!!

[CyberNate]

Cybernate, I have no doubt you witnessed that, but I would like to see the setup and check the calibration of the testing equipment. DJI publishes on the mavic datasheet 26 EIRP.

EIRP = Radiated power output from antenna ( Transmitter (dBm) - Cable_Loss (dB) - Connector_Loss (dB) + Antenna_Gain (dBi) )

Making a safe assumption that the controller has 0.5 db loss in cable/connectors and 3.8 dBi antenna gain, that would equate to the radio transmitter at 22.7 dBm, or 186 mW.

45mW is only around 1/4 of the FCC allowed EIRP for this type of setup. It is possible that DJI may dynamically adjust transmit power when the aircraft moves away from the remote. I know other technologies do this, but I am purely speculating trying to understand this low measured power output.

I didn't do the testing. The guy that did simply hooked up his meter to left side of controller and turned on controller and Mavic. It read 45mW in FCC mode and 9mW for CE mode. There is also a "special" version of DJI Go 4 that boosts the output to over 150mW, but I've heard it's not stable, resulting in overheated controller and potential disconnects and damage to controller.

You may be right, that the output is dynamic. I've also wondered about that.

 

[beanbubba]

Well I'm going off the measurement that @JakeMaxxUAV made in this post:
NanoSync Antenna for DJI Mavic Coming Soon

I'm not sure what he used for input on the Sunhams. I've seen the output on the controller measured at 45mW for FCC power level, while Jake measured 170mW off Sunhams output.

OK, I got to the bottom of this. Sunhams used some seriously deceptive marketing for their units. What they call a 3 Watt amp is actually 15dBm, or 32mW gain (see attached data sheet).

So, using Jake's spectrum analyzer readout of 170mW, then subtract the boosters 32mW, that leaves left 138mW coming from the controller.

138mW is 21.4dBm and I had predicted 22dBm in all my other posts based off of the published EIRP of 26.

These boosters are complete crap compared to readily available commercial versions at nearly the same price point that will produce 250mW of actual gain versus Sunhams paltry 32mW.

So in summary, Suhams markets a 3000mW amp that effectively produces 32mW gain.

By switching back and forth from dBm to mW in their data sheet descriptions, only those of us that do this for a living would see this deception trying to convert numbers from a log scale to a linear scale.

15dBm=32mW
20dBM=100mW
35dBm=~3000mW

BUT, 32mW + 100mW DOES NOT EQUAL 3000mW !!!!!

This is the most deceptive marketing I have ever seen and I am surprised nobody has called them out.

So Cybernate, I have reverse engineered all the equations except for one. I can't find the antenna gain for your Rexuav CP antenna. Does anyone know what it is? Once I get this we can positively do correct RF equations for any combination of equipment and stop this guessing game along with the vendors not publishing their RF specifications.

We have enough data points now and once I get all my test gear, I will publish the results.

 

[sprtbkrydr]

It is supposed to be 9.5 dbi gain for the RexUAV antenna. I was told this by the owner of RexUAV himself.

 

[beanbubba]

It is supposed to be 9.5 dbi gain for the RexUAV antenna. I was told this by the owner of RexUAV himself.

So I plugged in the 9.5 dBi antenna estimate and the 15dB Sunham amp gain into this RF link budget. I am predicting that Cybernate had 3 signal bars at 10 miles. These link budgets do carry a plus/minus 6dB margin of error.

You can probably expect 1.0 to 1.5 signal bar increase on the mavic from the Sunham boosters.

 

[Brojon]

Now you've said this, I stand corrected... it can't be a Yagi as there would also need to be a reflector behind the dipole and at least one radiator in front of the dipole. There are only 2 linear parts to this antenna. I may get one to have a
play with. Once fitted I'd then have the screw connectors in place so would be open to other antenna options.

In fact a Yagi is technically a travelling wave design - the spacing and dimensions of the reflector and the director elements is specific to this end - the ARRL used to have design notes on making a yagi. Wonder if I still have my ARRL Antennas book...

 

[CyberNate]

Wow! I think you're about right on the bars. It varies depending on interference, but once I get to the bay around 3 bars on TX.

Well that's frustrating about Sunhams. I spent a lot to buy 2. Can you recommend a better amp?

 

[beanbubba]

Wow! I think you're about right on the bars. It varies depending on interference, but once I get to the bay around 3 bars on TX.

Well that's frustrating about Sunhams. I spent a lot to buy 2. Can you recommend a better amp?

Hey, Cybernate

The Sunhams will serve their purpose just fine for this application. I was a little reactive to their marketing numbers game, but your link budget will take you out to 20 miles in a perfect environment (probably 15 miles in reality).

There is the 6dB rule in RF math that says you can double the distance of a link and incur a 6dB penalty. You have plenty of link budget left to double the 10 miles out to 20 miles and still be on the outer edge of the mavic remote functionality. I doubt this aircraft platform will ever exceed this.

If you do ever want go past this range, I can hook you up with hardware to meet any opportunity.

 

[beanbubba]

It is supposed to be 9.5 dbi gain for the RexUAV antenna. I was told this by the owner of RexUAV himself.

Hey sportbiker, the rexuav website publishes ZERO specifications. Thanks for calling the owner and getting this info.

 

[sprtbkrydr]

Hey sportbiker, the rexuav website publishes ZERO specifications. Thanks for calling the owner and getting this info.

Yeah no problem. He has a listing on Ebay also and he advertises them as being 9.5dbi there as well. I know one thing for sure the antenna just works for me.

 

[CyberNate]

So I plugged in the 9.5 dBi antenna estimate and the 15dB Sunham amp gain into this RF link budget. I am predicting that Cybernate had 3 signal bars at 10 miles. These link budgets do carry a plus/minus 6dB margin of error.

You can probably expect 1.0 to 1.5 signal bar increase on the mavic from the Sunham boosters.

View attachment 16488

Hey, Cybernate

The Sunhams will serve their purpose just fine for this application. I was a little reactive to their marketing numbers game, but your link budget will take you out to 20 miles in a perfect environment (probably 15 miles in reality).

There is the 6dB rule in RF math that says you can double the distance of a link and incur a 6dB penalty. You have plenty of link budget left to double the 10 miles out to 20 miles and still be on the outer edge of the mavic remote functionality. I doubt this aircraft platform will ever exceed this.

If you do ever want go past this range, I can hook you up with hardware to meet any opportunity.

I guess 20 miles of range should do it ;-). Thank you for the detailed calculations!

 

[sprtbkrydr]

I guess 20 miles of range should do it

Hey Nate I'm gonna have to get you to show me how to pop the top on this bird of mine so I can go ahead and do the permanent battery mod. I'm not to happy with the clip.

 

[Hyprmtr]

I didn't do the testing. The guy that did simply hooked up his meter to left side of controller and turned on controller and Mavic. It read 45mW in FCC mode and 9mW for CE mode. There is also a "special" version of DJI Go 4 that boosts the output to over 150mW, but I've heard it's not stable, resulting in overheated controller and potential disconnects and damage to controller.

You may be right, that the output is dynamic. I've also wondered about that.

I was the person that did the testing. CE power was +/-9mw and FCC was +/-45mw. Feeding FCC power to a 2W booster measured just under 400mw. Most of these bidirectional booster accept input powers up to 20dbm(100mw). The modified APK app that has been configured for 150mw would be too much for the external amplifier and might be too much for the internal SMD amplifier. I believe the external boosted FCC power of just less than 400mw is going to be pretty close to the maximum power that the boosted downlink Rx(+/-11db) can handle without getting scrambled by the boosted uplink. The next step is going to be increasing the downlink power on the copter itself. That is in progress.

I attached photos of the measured powers.

Ken

 

[beanbubba]

I was the person that did the testing. CE power was +/-9mw and FCC was +/-45mw. Feeding FCC power to a 2W booster measured just under 400mw. Most of these bidirectional booster accept input powers up to 20dbm(100mw). The modified APK app that has been configured for 150mw would be too much for the external amplifier and might be too much for the internal SMD amplifier. I believe the external boosted FCC power of just less than 400mw is going to be pretty close to the maximum power that the boosted downlink Rx(+/-11db) can handle without getting scrambled by the boosted uplink. The next step is going to be increasing the downlink power on the copter itself. That is in progress.

I attached photos of the measured powers.

Ken

Ken, pleasure to meet you.

In my experience I have yet to see a device that couldn't handle 200mW on the receiver sensitivity (even though their doc stated 100mW). Just curious if you have tested this.

Question, your measurement device appears to have an attenuator connected. If so, did you account for that in your measurements?

Thank you.

 

[Hyprmtr]

Ken, pleasure to meet you.

In my experience I have yet to see a device that couldn't handle 200mW on the receiver sensitivity (even though their doc stated 100mW). Just curious if you have tested this.

Question, your measurement device appears to have an attenuator connected. If so, did you account for that in your measurements?

Thank you.

Hi there,


Yes, of course that is an attenuator. The attenuation is configured in the tester. It is fully adjustable, even fine attenuation. We do not know each other's experience or backgrounds so any questions to the similar will be alright by me.

The accuracy of this little meter is +/-0.45dB. I have used this meter a lot over the past couple of years. I've tested 433mhz 2.4ghz, 1.3ghz, and many 5.8ghz and compared the measurements with larger more expensive meters and it is well within similar ranges. Once all the local fpv guys find out you have a meter, they are always curious on what their Vtxs are really putting out.

I can review the manual and see what the manufacturer suggests max input power is without attenuation. The attenuation is configured in the meter so direct connection shouldn't be much change if any at all.

The only explanation that we can up with as far as DJI's claim of just under 400mw FCC is that the power is variable under low RSSI conditions. I attempted to duplicate a low RSSI condition all the way to disconnect and the power stayed the same throughout. Now, this was with the copter in a faraday cage so, it may very well need to be actually flying for the power to increase if it does at all. I'm going to have my meter next to me next time I do a long flight. I will see if the power goes up once I start to lose signal and failsafe triggers. I will have the meter and the transceiver in a fixed position so there is no chance of any movement from either one. I will post results as soon as this is completed.

Thank you for you interest.


Ken

 

[beanbubba]

Hi there,


Yes, of course that is an attenuator. The attenuation is configured in the tester. It is fully adjustable, even fine attenuation. We do not know each other's experience or backgrounds so any questions to the similar will be alright by me.

The accuracy of this little meter is +/-0.45dB. I have used this meter a lot over the past couple of years. I've tested 433mhz 2.4ghz, 1.3ghz, and many 5.8ghz and compared the measurements with larger more expensive meters and it is well within similar ranges. Once all the local fpv guys find out you have a meter, they are always curious on what their Vtxs are really putting out.

I can review the manual and see what the manufacturer suggests max input power is without attenuation. The attenuation is configured in the meter so direct connection shouldn't be much change if any at all.

The only explanation that we can up with as far as DJI's claim of just under 400mw FCC is that the power is variable under low RSSI conditions. I attempted to duplicate a low RSSI condition all the way to disconnect and the power stayed the same throughout. Now, this was with the copter in a faraday cage so, it may very well need to be actually flying for the power to increase if it does at all. I'm going to have my meter next to me next time I do a long flight. I will see if the power goes up once I start to lose signal and failsafe triggers. I will have the meter and the transceiver in a fixed position so there is no chance of any movement from either one. I will post results as soon as this is completed.

Thank you for you interest.


Ken

Glad to know you Ken and we speak the same RF language. I will be performing my own tests/validation this week with my spectrum analyzer. I came to the same conclusion that DJI was using variable output (but this is pure speculation until I can prove it).

 

[Hyprmtr]

Glad to know you Ken and we speak the same RF language. I will be performing my own tests/validation this week with my spectrum analyzer. I came to the same conclusion that DJI was using variable output (but this is pure speculation until I can prove it).

That sound wonderful. I will be patiently waiting for the outcome of your tests. It is nice to talk to people with the same interests and with similar motivation to get true and accurate results. There is so much misinformation online and it gets thicker by the day. It will be a few days before I can do any more tests of my own. I will share the information as soon as it is verified.


Ken

 

[beanbubba]

That sound wonderful. I will be patiently waiting for the outcome of your tests. It is nice to talk to people with the same interests and with similar motivation to get true and accurate results. There is so much misinformation online and it gets thicker by the day. It will be a few days before I can do any more tests of my own. I will share the information as soon as it is verified.
Ken

I did a coupe quick tests today with my spectrum analyzer and made a few educated guesses. Much more testing is needed, but I did see a solid and repeatable 10dB difference (supporting the theory of DJI adjusting TX power dynamically) from the remote TX when the aircraft was nearly out of range:



And this is the remote with aircraft nearby:



There is strong evidence that the remote TX is using spread spectrum as opposed to the fixed channel video stream sent from the aircraft:



I have several more tests planned that will be conclusive, but I had to order an attenuator and U.FL cables on Amazon and they are scheduled to arrive Friday. Then I can cable directly to the remote and aircraft U.FL ports on the main circuit boards.

Hyprmtr, until then I can not make a determination to compare your output findings because the tests today were using an external antenna (apples and oranges as they say).

 

[Hyprmtr]

Hello,

Thank you for sharing your results. I'm curious why I didn't get any evidence of power change at all when I moved the copter nearly out of range? I'm still have the speculation it needs to be in flight before the variable power activates, but the contradicts what you found unless yours was flying? Do you mind sharing the firmware version your copter and Go4 app is on? My copter and controller is still on .400 and I'm using the 4.1.3 app. I will be doing this particular test over the weekend for sure. The weather looks to be good for the weekend but very hot.

I been procrastinating on getting the RF Explorer 6g combo. I think I would really enjoy the PC interface it has to offer unlike what I have now. I've been interested in radio for a couple years now so I think this hobby will keep me happy for a good while. It should be a worth while purchase.


Ken

 

[beanbubba]

Hello,

Thank you for sharing your results. I'm curious why I didn't get any evidence of power change at all when I moved the copter nearly out of range? I'm still have the speculation it needs to be in flight before the variable power activates, but the contradicts what you found unless yours was flying? Do you mind sharing the firmware version your copter and Go4 app is on? My copter and controller is still on .400 and I'm using the 4.1.3 app. I will be doing this particular test over the weekend for sure. The weather looks to be good for the weekend but very hot.

I been procrastinating on getting the RF Explorer 6g combo. I think I would really enjoy the PC interface it has to offer unlike what I have now. I've been interested in radio for a couple years now so I think this hobby will keep me happy for a good while. It should be a worth while purchase.


Ken

I was flying and flew it completely out of range. As of today, I am on the most current software for both aircraft and iPad.

 

[beanbubba]

I was flying and flew it completely out of range. As of today, I am on the most current software for both aircraft and iPad.

Ohhh, forgot to mention... I put the aircraft inside my microwave oven across the room from the remote (poor mans faraday cage, but very effective for 2.4 microwave). I didn't lose signal completely but down to one bar, and I didn't see a TX increase from the remote. Only saw the increase on a real flight.