thUndeadMod v3.4 - openLRS S**T just got real :)

[Coyote]

I can't remember off hand but I will post it up tomorrow when I get in, I will check the voltage level too to be sure

[raptor_flyer]

thUndeadMod v 3.6 Beta

- Config GUI bugs fixed
- added channels in GUI
- 459 band spacing fix
- overall hopping and recovery algorithm improvement
- RSSI fixed (PIN 8)
- Stress Tested for 4h - no dropouts / random rssi encountered.

NEEDS TESTED!
- 459 range (should be better)
- Serial PPM

------
WHo wants to beta test PM me and I'll give u the password.

Cheers,
Mihai.

Mihai,

From the testing I've done there still seems to be something strange going on with Frequency Hopping on 459.

See the following scan from Dennis OpenLRS Spectrum Analyzer





This is with the default hopping setup :-

//###### HOPPING CHANNELS #######
#define CHNO 6
static unsigned char hop_list[10] = {7,4,5,6,0,1,9,3,2,8};

There are 3 nice peaks between 459000 and 459500 but the other 3 peaks are up
at 461000 to 461400 and at reduced power.
[Test was done with TX on reduced power and RX about 2 m away on a dining room table]
[Also before anyone asks, Yes I did check for these 461 to 461.4 peaks with the OpenLRS TX switched off]

I can't see anything wrong with your code since you're setting the Hopping step to 5 (x 10KHz)
and the maximum hop step (6 channels) is 7 (7 x 50Khz = 350Khz) so the max hopping frequency should be 459350Khz.

I wonder is there's some "undocumented feature" of the  RFM22B or Si4431 occurring here.

Has anyhone else tried using the OpenLRS Spectrum Analyzer on thUndeadMod v 3.6 Beta ?

[Coyote]

It use to do that on the previous version as well

[papa_lazerous]

Ive been trying the beta but not with the spektrum analyser software yet.  I need to get out and do a range test but its been raining all day :(

[raptor_flyer]

It use to do that on the previous version as well

Yes I recall seeing it hopped outside the 458500 to 459500 band in previous versions but
I thought Mihai said this was a known problem he's tried to fix. Based on the info in the
Si4431 datasheet I can't see anything wrong with his code.

What is concerning now is the 3 hops above 460500Khz are at a much lower power
level for some bizarre reason.  I really need someone else to try the same test to make sure
there's not something weird with my measurements or setup.

[ghost2212]

Did 925m today using DIY OpenLRS hardware and thundeads v3.6 code on 435. RSSI was around 20-30% according to the Dragon OSD but this hasnt been calibrated recently and I wasn't holding the transmitter aerial straight. Transmitter aerial was a Diamond SRH-771 with a half wave bazooka dipole on the reciever.

[papa_lazerous]

Well I have some EPIC news  :o

I did a range test this afternoon with very surprising results. I tested at 459MHz on the new 3.6 software running @100mW using a V2 Tx and a V2 Rx.  On the Tx I had a Diamond SRH-771 antenna (not ideal for 459MHz but ok)  and on the Rx I have a home made dipole tuned to 459MHz.

What I did was make a little test rig up, well it was a box with the antenna taped vertically to keep the antenna's pointing the same to stop any issues with polarisation. a BEC a servo and my trusty Fluke meter for RSSI.



So I took the otherhalfs boy with me on this test, I got him to sit at one end of the beach with instructions to keep the antenna pointing straight up and I would be back in 10 minutes (I should point out that he's 14 so its ok to abandon him at the beach lol)

So I set off in the car and lost signal pretty quick, the reason is 2 fold.  Firstly the car is pretty much a crude farraday cage so not much is going to be recieved, and secondly there is a heavy duty concrete sea wall between the road and where the boy was sitting.  So I made a few stops along the way to get out of the car and check the signal.  Heres the pics and the range on each.



Ok so 1.65Km I drove up the slip way over the sea wall hung my arm out the window and got instant signal with 1.65v RSSI no glitches so back on the road....

I drive a bit further to the next break in the sea wall where I can stop the car.....



Ok 2.23Km this time and I am impressed, but I still have 0.9v RSSI and no glitching so I carry on :)



This is the bit you have been waiting for 3Km on 459!!! :)  When I got out the car I was getting some glitching and losing some packets, but I held the test rig up in the air and I had signal again that was perfectly fine to fly on.  the RSSI was around 0.6v I didnt go any further than that as I felt I was at the limit of the range.

Now here's the small print lol  I was 100% definately on 459MHz hopping on 6 channels I confirmed this by using the spektrum analyser before I went out.  So no question of me being on the correct frequency, I was running standard OpenLRS hardware with no mods or boosters, so I was outputting 100mW nothing more.  By testing over the sea like I did and with the tide being in I will have been experiencing a reduction in range due to multipathing, its also quite a poluted RF environment as I am on the edge of town plus the starting point of the range test is the coastguard office with a massive antenna, I know they use UHF/VHF and radar from there.  So I have to assume its afecting me.  Out in the countryside with a wide open area to test in I expect that you will get more range, perhaps 25% more.  Another thing is that the range in the air is ALWAYS allot more than on the ground, so you can safely expect 3 to 4 times the range when flying.  we are talking 10km easy!!

A BIG THANKS TO MIHAI!! :party:

[thundead]

Yaaaay Awesome range test and result! Hopefully someone else will do a range test on 459 and I can get the fw out of beta.
Big thumbs up for papa_lazerous! I'm officially calling you my main α β γ Δ tester muhaha :D

[ghost2212]

Coyote, Have you had a chance to check the 9x voltage yet?

[papa_lazerous]

Coyote, Have you had a chance to check the 9x voltage yet?

I am failing to understand the correlation between the voltage on the 9x and if we have a good firmware?

To put your mind at rest and get you back on topic I had a look at my 9x, the voltage outputted from the 9x radio for the module to connect to is equal to pack voltage!

This is NOT a concern, the OpenLRS Tx module has built in voltage regs that are designed to take upto 16VDC input, and we are getting nowhere near that. here is a link to the schematic http://www.flytron.com/pdf/openlrs-m2-module.pdf

Now if you use a Rx as a Tx then you would naturally use a BEC, or take 5v directly from inside the radio (this can be done but for the cost of a bec its hardly worth it)

[ghost2212]

Coyote, Have you had a chance to check the 9x voltage yet?

I am failing to understand the correlation between the voltage on the 9x and if we have a good firmware?

To put your mind at rest and get you back on topic I had a look at my 9x, the voltage outputted from the 9x radio for the module to connect to is equal to pack voltage!

This is NOT a concern, the OpenLRS Tx module has built in voltage regs that are designed to take upto 16VDC input, and we are getting nowhere near that. here is a link to the schematic http://www.flytron.com/pdf/openlrs-m2-module.pdf

Now if you use a Rx as a Tx then you would naturally use a BEC, or take 5v directly from inside the radio (this can be done but for the cost of a bec its hardly worth it)

This is Directly related to whether we have a working firmware or not! Coyote was using an RX as a TX(if you read his post you'll see that) He says he was powering it off 5v but if he is wrong then this would cause the voltage regulator to overheat and the RSSI to drop to 0 and oscillate as the voltage regulator came back on and then off again.

I know how the voltage regulator reacts to overvoltage because I had the same problem powering it with too higher voltage.

He also states that the RX was powered off of the 5v just as the TX module was. But where the TX module plugs in there is no 5v so not sure how this is possible.

Mihai advised people not to use the firmware that Coyote was using as he suspected there may be a serious bug that caused the crash. If it was not the firmware that was the problem then there may have never been a problem with the firmware on the 435 band at least.

Hope this clears your misunderstandings up

Brett

[thundead]

Hey Brett,
Something might have been wrong with the fw because when I changed the pin layout I might have messed up one of the three timers running on the arduino. All should be fixed in the 3.6 fw as I reverted back to the old pinout (rssi on pin 8 and now Serial PPM on in 9) and further optimized timing, hopping and recovery algorithms. I also changed some stuff for the 459 band and apparently works good now :)

using the rx as tx without a BEC or with a BEC on 6v is a big no no as you said beacuse of the overheating problems.

Coyote, Have you had a chance to check the 9x voltage yet?

I am failing to understand the correlation between the voltage on the 9x and if we have a good firmware?

To put your mind at rest and get you back on topic I had a look at my 9x, the voltage outputted from the 9x radio for the module to connect to is equal to pack voltage!

This is NOT a concern, the OpenLRS Tx module has built in voltage regs that are designed to take upto 16VDC input, and we are getting nowhere near that. here is a link to the schematic http://www.flytron.com/pdf/openlrs-m2-module.pdf

Now if you use a Rx as a Tx then you would naturally use a BEC, or take 5v directly from inside the radio (this can be done but for the cost of a bec its hardly worth it)

This is Directly related to whether we have a working firmware or not! Coyote was using an RX as a TX(if you read his post you'll see that) He says he was powering it off 5v but if he is wrong then this would cause the voltage regulator to overheat and the RSSI to drop to 0 and oscillate as the voltage regulator came back on and then off again.

I know how the voltage regulator reacts to overvoltage because I had the same problem powering it with too higher voltage.

He also states that the RX was powered off of the 5v just as the TX module was. But where the TX module plugs in there is no 5v so not sure how this is possible.

Mihai advised people not to use the firmware that Coyote was using as he suspected there may be a serious bug that caused the crash. If it was not the firmware that was the problem then there may have never been a problem with the firmware on the 435 band at least.

Hope this clears your misunderstandings up

Brett

[papa_lazerous]

Coyote, Have you had a chance to check the 9x voltage yet?

I am failing to understand the correlation between the voltage on the 9x and if we have a good firmware?

To put your mind at rest and get you back on topic I had a look at my 9x, the voltage outputted from the 9x radio for the module to connect to is equal to pack voltage!

This is NOT a concern, the OpenLRS Tx module has built in voltage regs that are designed to take upto 16VDC input, and we are getting nowhere near that. here is a link to the schematic http://www.flytron.com/pdf/openlrs-m2-module.pdf

Now if you use a Rx as a Tx then you would naturally use a BEC, or take 5v directly from inside the radio (this can be done but for the cost of a bec its hardly worth it)

This is Directly related to whether we have a working firmware or not! Coyote was using an RX as a TX(if you read his post you'll see that) He says he was powering it off 5v but if he is wrong then this would cause the voltage regulator to overheat and the RSSI to drop to 0 and oscillate as the voltage regulator came back on and then off again.

I know how the voltage regulator reacts to overvoltage because I had the same problem powering it with too higher voltage.

He also states that the RX was powered off of the 5v just as the TX module was. But where the TX module plugs in there is no 5v so not sure how this is possible.

Mihai advised people not to use the firmware that Coyote was using as he suspected there may be a serious bug that caused the crash. If it was not the firmware that was the problem then there may have never been a problem with the firmware on the 435 band at least.

Hope this clears your misunderstandings up

Brett

I see what you are getting at Brett,  the FW had an issue as Mihai just stated. Coyote had initially been running a Tx as a Tx so no issue there with voltages but still the same problem,  a problem I managed to duplicate also.  Running a Rx as a Tx I assume there was a BEC being used.  Either way there was a FW bug regardless of coyote's wiring

[Coyote]

Yeah my Tx as a tx was fine for lots and lots of range tests and was left on for hours at a time as I went back and forward 8km a time before 3.5 FW

[raptor_flyer]

Hey Brett,
Something might have been wrong with the fw because when I changed the pin layout I might have messed up one of the three timers running on the arduino. All should be fixed in the 3.6 fw as I reverted back to the old pinout (rssi on pin 8 and now Serial PPM on in 9) and further optimized timing, hopping and recovery algorithms. I also changed some stuff for the 459 band and apparently works good now :)

Ah ha, I didn't realise the significance of moving the RSSI out back to pin 8.  I've been modifying the setup in your 3.6B code to put it back on pin 1 so I didn't have to rewire the plane.  :-[  I will rewire the plane so I have RSSI out on pin 8 and use your code unmodified before we try any further flight tests.

Thanks,

Steve.