DSM telemetry and Dev. 5.0.0 = TX crash ??

I flew hundreds of hours with Deviation 4.0.1 and several models without one problem (this was with Walkera receivers in DEVO protocol and as long as telemetry was off)

These days I'm trying Deviation 5.0.0 on my Devo 10.
I set up a cheap EPO glider with an HK OrangeRx receiver with telemetry using DSMX protocol, and activated the telemetry for this model.
Today after about 4 hours in total of flying this model, I lost control as it was at about 250 meters of distance, while the glider was coming toward me.
Output power was set to 100 mW.
I tested the receiver at 1 mW one week ago, and it was good for 120 meters. So with 100 mW, a distance of 250 meters should be no problem.
The telemetry showed 0 frame losses and holds.
I was just in a wide turn to align myself on the landing runway, and when I wanted to go straight again, the plane simply continued the turn and landed in a corn field, luckily with no damage.
I crossed the field a first time, without powering the transmitter off.
From time to time, I activated the throttle, in case I hear the model. But nothing happened.
Then I switched the transmitter off and on again, and crossed the field a second time.
And when I activated the throttle I heard the motor of the glider running !

So I think before I switched it off and on again the Transmitter was "frozen". (but timers went still running)
The error.txt file in the transmitter is empty.

Looks like there are still problems with the DSM telemetry (or DSM even without telemetry ??)

Meanwhile I will continue to use Walkera receivers in DEVO protocol (with telemetry off), I'm more confident this will work.

Roland

Q1 for Roland: How long have you been in the air before this happend? 2-5 minutes? 10-30 minutes? Any longer?

Q2: Did you get any telemetry out of range warnings (Devo 10 inverting FlightLog fields black)?

Q3: Are you sure that your receiver antennas are oriented perfectly and are not in the near of lipo/motor/servo/ESC or that the S-BEC creates noise on the receiver?
You can easily check that if Fades A = max 0xFF = 255 (not that good).

---

Well, 0F / 0H is good, but can you really rely on this OrangeRX receiver?
Any LED blinking? Any problem with BEC power? Well, as you could re-connect / re-bind on transmitter power-off -> maybe not?
Can you even really rely on the OrangeRX telemetry implemention??? I guess not...

Too bad you hit power-off before you reached your plane and you could check working telemetry (either fields black = telemetry out of range) and running FlightLog numbers.
Telemetry screen should have re-activated itself (if once out of range) and not lock out the transmitting of frames (stick movements).


Roland, personally probably I would not waste my time testing around with OrangeRX on initial Deviation releases and new protocol code (changes) as in German forums these (reversed-engineered) telemetry receivers are highly discussed (also on the telemetry send back stuff).

---

Suggestions:
1) Try with a genuine Spektrum receiver with DataPort + TM1000
2) Try with enabled FlightLog logging within Deviation

3) Re-test if you really run into a firmware TX<->RX deadlock scenario (with new 4b3dc91 code from 29 Jul 2015), so maybe you run into a rare case where telemetry receive code locked down transmit part???
-> I have NEVER encountered that on:
- nightly-build v4.0.1-fcd0669
- nightly-build v4.0.1-92e1705 (0xFF patched by me)

Somehow it is weird that it worked for 4 hours and it never happend on detailed ground range tests...

-> I once had this "deadlock" case on one (very) older test-build from Indigo with my 200QX on table.
At that time it was completely firmware / HF-module deadlock, not on any official nighlies.
But Indigo was re-engineering at that time the whole DSM protocol with new buffers, new error checking code, etc. So pretty much (almost) everything new...so yes, that could potentially happen. Until now I could not re-test latest test-build

4) Re-test with nightly-build v4.0.1-fcd0669
-> nightly-build fcd0669 misses newer DSM protocol changes Vlad committed 4b3dc91 for DSM + DEVO
www.deviationtx.com/forum/protocol-devel...port?start=400#52587
www.deviationtx.com/forum/protocol-devel...port?start=400#46358

AR8000 + TM1000 and AR6210/AR600 before worked quite good on my side with this fcd0669 nightly-build with enabled DSM telemetry and checking FlightLog: www.deviationtx.com/forum/protocol-devel...port?start=400#45377


As to Walkera DEVO receivers:
Be warned that starting nightly-build commit version 4b3dc91 from 29 Jul 2015 (which is included in 5.0 release) you can also see some DEVO protocol code changes.
Same facts may apply (depending on protocol code changes) to Devo telemetry as well as?!


I am not sure how many pilots besides Vlad could actually detail test DEVO + DSM protocol changes nightly-build version >=4b3dc91 from 29 Jul 2015 + 5.0 with TM1000 enabled telemetry. I could not so far.

On the other side: The code changes have been committed by Vlad 1 year before.
So I am almost sure that already more than only one Deviation pilot using DSM + telemetry was testing this nightly build code?!?

---

IMHO only the core protocol developers may be able to comment how critical the changes from 4b3dc91 are compared to before or Indigos test-build code (RX<->TX improvements, register error checking).
They where writing back and forth in the two DSM telemetry support and Devo telemetry threads a while ago where Indigo was into active DSM development.
Kinda hard to keep the overview if you are not into CYRF register and error handling programming


Any core developer might want to respond or answer my open questions in the linked DSM telemetry thread?


Maybe best would be if some pilots could give Indigos latest DSM test-build (with crash proof EPP planes + quads) a go and see how well these latest register error checking goes on possibly RX<->TX deadlocks.
And then get back to compare Indigos DSM code changes to Vlad's code changes from nightly 4b3dc91, which made it into release 5.0

You may also have to step into CYRF register programming to understand the two threads what these guys were trying to do or improve

Just some ideas....

Thomas

Hello Thomas, and thank you for taking the time for your long answer !

Thomas.Heiss wrote: Q1 for Roland: How long have you been in the air before this happend? 2-5 minutes? 10-30 minutes? Any longer?

About 10 minutes (total time in this setup about 4 hours)

Q2: Did you get any telemetry out of range warnings (Devo 10 inverting FlightLog fields black)?

Some fades (about 30-50 depending on flight distance and duration), but no losses and holds.
Telemetry losses I don't really know while flying, but after crash yes it was lost (corn field "shielding" the transmission I think)

Q3: Are you sure that your receiver antennas are oriented perfectly and are not in the near of lipo/motor/servo/ESC or that the S-BEC creates noise on the receiver?
You can easily check that if Fades A = max 0xFF = 255 (not that good).

Antennas are perpendicular each other and under the EPO wings, about 15 cm from ESC and battery.
And when I lost control the motor was off (it's a glider) so no interference from the drive train...

---

Well, 0F / 0H is good, but can you really rely on this OrangeRX receiver?
Any LED blinking? Any problem with BEC power? Well, as you could re-connect / re-bind on transmitter power-off -> maybe not?
Can you even really rely on the OrangeRX telemetry implemention??? I guess not...

Can I rely on the Orange Rx ? that's a good question !
The red led on the Rx blinked (sequence of 1 blinking) when I found the plane in the corn, this means the Rx registered one hold, wich I think was when the TX "froze" (stopped to transmit)
This led is interesting, it emits a sequence of flashes with the number of holds encountered.
When you land your plane and control this (of course before switching off the TX) you see if you had one or more holds.
In the 4 hours of flight before my problem, this led had never blinked after a flight, even at double the distance I had my problem.
This makes my think it's not at all a range issue.

...

Roland, personally probably I would not waste my time testing around with OrangeRX on initial Deviation releases and new protocol code (changes) as in German forums these (reversed-engineered) telemetry receivers are highly discussed (also on the telemetry send back stuff).

Thomas, can you give me an internet address to this forum (I speak and read german currently)

---

Suggestions:
1) Try with a genuine Spektrum receiver with DataPort + TM1000
2) Try with enabled FlightLog logging within Deviation

3) Re-test if you really run into a firmware TX<->RX deadlock scenario (with new 4b3dc91 code from 29 Jul 2015), so maybe you run into a rare case where telemetry receive code locked down transmit part???

Err, I don't understand why to test with an old nighty when I have 5.0.0 installed on the TX ?

-> I have NEVER encountered that on:
- nightly-build v4.0.1-fcd0669
- nightly-build v4.0.1-92e1705 (0xFF patched by me)

Somehow it is weird that it worked for 4 hours and it never happend on detailed ground range tests...

Yes all worked well, on ground and in the air, without losses and holds, so I'm thinking (but no clue for that) that in fact my transmitter locked.

...

4) Re-test with nightly-build v4.0.1-fcd0669
-> nightly-build fcd0669 misses newer DSM protocol changes Vlad committed 4b3dc91 for DSM + DEVO
www.deviationtx.com/forum/protocol-devel...port?start=400#52587
www.deviationtx.com/forum/protocol-devel...port?start=400#46358

AR8000 + TM1000 and AR6210/AR600 before worked quite good on my side with this fcd0669 nightly-build with enabled DSM telemetry and checking FlightLog: www.deviationtx.com/forum/protocol-devel...port?start=400#45377

Having no Spektrum receivers and flying at places with big fields and forest (slope in the Schwarzwald !) I don't want to risk to loose a plane again (I feel really lucky to have recovered this one !) and will continue to fly with Walkera receivers and DEVO protocol without telemetry.
I have a battery telemetry from HK wich works well : www.hobbyking.com/hobbyking/store/%5F%5F...%5F5Mhz%5FFM%5F.html
I can live with this.
This is not to critize DeviationTX wich I think is fantastic, but I want to fly with peace of mind instead of thinking all the time "when will the next TX lock come" !!! and risk to injure a mate on the flying field or "landing" in a car winshield.

As to Walkera DEVO receivers:
Be warned that starting nightly-build commit version 4b3dc91 from 29 Jul 2015 (which is included in 5.0 release) you can also see some DEVO protocol code changes.
Same facts may apply (depending on protocol code changes) to Devo telemetry as well as?!

I'm flying some planes with DeviationTX 5.0.0 and DEVO protocol for about 15 hours now, so far no problems (telemetry off of course)
For my big glider that I REALLY don't want to lose I still have my "old" black Devo10 with DeviationTX 4.0.1, DEVO protocol with no telemetry.

...

Just some ideas....

Thomas

Once more, thanks Thomas for your long answer.

MfG,

Roland

Rollmops67 wrote:

---

I'm flying some planes with DeviationTX 5.0.0 and DEVO protocol for about 15 hours now, so far no problems (telemetry off of course)
For my big glider that I REALLY don't want to lose I still have my "old" black Devo10 with DeviationTX 4.0.1, DEVO protocol with no telemetry.

.............Roland

I am sorry to jump into this topic...,but
I have been posting this question several times but got no answer.

Why is it not possible to turm the telemetry off for DSMX on a 7e ?

As far as I remember this was different till v4.0.1-92e1705. This feature was reintroduced by Indigo and again
it disappeared some times later ...(..?) From June to August 2015 I used Indigo´s firmware and it was ok.
After an update to a more recent Nightly I had several short failsafe moments (40-120 m) and two crashes that let me
switch to Frsky in the end.
Range with 1mW was 124m but all this was not reliable even at 100mW.

I really would appreciate to have this option back.

Forums for OrangeRX telemetry:
www.rclineforum.de
www.rc-heli.de

Err, I don't understand why to test with an old nighty when I have 5.0.0 installed on the TX ?

Because this / my nightly build was well tested by me on (450er) 6-8min. heli flying with genuine Spektrum receivers
However I need to get a EPP glider / warmliner for long distance tests.

AND: Because this nightly-build does not yet include the DSM protocol changes 4b3dc91 from 29 Jul 2015.
I just would like to find out if any of these code blocks could be the reason.

Please note: NO, I am NOT sure if any of those code changes on 4b3dc91 could bring any potential deadlock with and how Indigos latest test-build was going well on this.

However I would not fly a very expensive plane anyways without multiple weeks / months of good Deviation version testing results.

Unfortunately when a new EPP glider, 3D plane or race quad is up and running for me it won't be much sooner than end of Sept/Oct - so any detail version / range testing probably will have to occur next year on my side

Thomas.Heiss wrote: Forums for OrangeRX telemetry:
www.rclineforum.de
www.rc-heli.de

Err, I don't understand why to test with an old nighty when I have 5.0.0 installed on the TX ?

Because this / my nightly build was well tested by me on (450er) 6-8min. heli flying with genuine Spektrum receivers
However I need to get a EPP glider / warmliner for long distance tests.

AND: Because this nightly-build does not yet include the DSM protocol changes 4b3dc91 from 29 Jul 2015.
I just would like to find out if any of these code blocks could be the reason.

Ah OK I understand now.
But I feel not very excited to test this, because to be sure that it works (or not) it can take 10th of hours of flight (I got one probleme after 4 hours, it could also occur after 10 or 20 hours)
Concerning my glider in DSMX protocol wich crashed, I will continue to fly it with the Orange Rx receiver, but with telemetry off, just to see if there are still problems or not.

...

However I would not fly a very expensive plane anyways without multiple weeks / months of good Deviation version testing results.

Yes I agree, that's why I wait to switch to Dev. 5.0.0 with my big glider.
I will wait until I have flown about 30 hours with cheaper models (this with DEVO protocol)...

...

Roland

www.deviationtx.com/forum/builds/6306-ni...v4-0-1-fcd0669#53115

Rollmops67 wrote: But I feel not very excited to test this, because to be sure that it works (or not) it can take 10th of hours of flight (I got one probleme after 4 hours, it could also occur after 10 or 20 hours)

Concerning my glider in DSMX protocol which crashed, I will continue to fly it with the Orange Rx receiver, but with telemetry off, just to see if there are still problems or not.

Let make a short summary.

A) You are turning off telemetry (avoiding any whatever potential deadlock scenario between receive and send)
You go back to DEVO receivers (where the RX+TX telemetry protocol code needs to be synced too...anyways..)
C) You stick to not so big / expensive gliders for the moment
D) You agree that detail-tests are not finished in some short days (total of hours) tests

I agree on all these points.

However my Indigo older test-build TX<->RX deadlock could easily be replicated with my 200QX on table. Didn't take too long.
So maybe 2-4 weeks and you now what is going on if you fly 3-5days/week after you installed a new release.


> > I will continue to fly it with the Orange Rx receiver

If your glider is very crash proof and you will not put it into highest tree in your fly area - go and make your OrangeRX tests
I understood that this fly area / green thing might be a real problem for you in risking models.
A better crash proof thing is IMHO a 3-4mm CFK quad which you can bomb into ground.
OrangeRX telemetry is 3rd party - reverse engineered...nothing Spektrum originally. I would not trust it too much.


The end result still could be that the OrangeRX receiver or antenna is corrupted
and
it is not Deviation firmware and newer protocol code block faults.

---

Somehow I feel the need to start a dedicated poll thread
- how many people
- and how long

they have been using
- V4.0.1-nightly builds >= 4b3dc91 since it's commit in Summer 2015
- or release 5.0
- or 5.0-nightly-build
- with DEVO or DSM telemetry turned on and/or off.

I want to see no text in this new thread at all but real multiple selections, concrete numbers, bar graphs just like real HeliFreak polls.


As previously said: The protocol code change is not new and ~1 year is a longer time.
I would expect that it would have popped off much earlier.
And Vlad probably made his own tests with his equipment before and after commit for a while too.
But what do I know with my grounded RC stuff and flying fcd0669 build version.

vlad_vy wrote: I don't like new GUI for Devo 6s, 8s and 12S too. Now I use build v5.0.0 and don't want to spend my time for nightly builds.

So how are you doing on your 5.0 release build Vlad?

What receivers are you using? Oranges? Lemons? Genuine Spektrum? With telemetry ON or OFF?

Any deadlocks, and loss of signal, anything bad which could explain what happend to Roland (besides I would start at the receiver side)?

Crash-proof 3mm CF Quad with electronics is still not ready
EPO glider ASW17 better should not crash because of any DSM code changes in nightly build 4b3dc91 / release 5.0.0

Any newer experiences for DSM + telemetry?

DSMx + telemetry (TM1000) works quite fine on devo10-v4.0.1-fcd0669 so far for me (without latest DSM code changes) but with bogus (non-filtered) telemetry out-of-range FlightLog data: www.deviationtx.com/forum/protocol-devel...d-field-out-of-range

Thomas

Just wanted to let you know that I had my first 3 successful flights with the EPO glider on

- Deviation release 5.0.0 (DSMx)
- activated telemetry
- Spektrum AR6210 (+SAT)
- TM1000

Ground range test @100µw and 1mw output power looked quite promising on release 5.0.0.

First test: I walked >~30-35m away and still could steer the rudder @100µw with no problems - even the TM1000/telemetry already was out of range.
I have to redo some detail ground tests to get the exact meters and write them down.

Second test: I re-tested ground range @1mw >~60-65m distance (and a little bit above) and this was good too (like Spektrum DX8G1 up to ~60m), no telemetry out of range (black/inverted monitor fields), no FrameLosses, no Holds.

Had three more flights on devo10-v4.0.1-fcd0669 (BitBucket) before which is also proven to be working fine on Blade 4503D on AR8000 (with SAT) +TM1000.

Rollmops67 wrote: Today after about 4 hours in total of flying this model, I lost control as it was at about 250 meters of distance, while the glider was coming toward me.
Output power was set to 100 mW.
I tested the receiver at 1 mW one week ago, and it was good for 120 meters. So with 100 mW, a distance of 250 meters should be no problem.
The telemetry showed 0 frame losses and holds.

I have setup a permanent timer which calculates CH1 (THR) motor running time.
I better should switch it to real flying time.

Can you guys give me a hint how I would have to do this? Any switch where I activate and de-activate a 3rd permanent timer?
How do guys measure real flying time and not motor running time?
And tricks / trips on timer configuration?

Thomas

Thomas.Heiss wrote: I have setup a permanent timer which calculates CH1 (THR) motor running time.
I better should switch it to real flying time.

Can you guys give me a hint how I would have to do this? Any switch where I activate and de-activate a 3rd permanent timer?
How do guys measure real flying time and not motor running time?
And tricks / trips on timer configuration?

Thomas

Hello Thomas.

i asked the same question to me when I began flying gliders.
I even asked on this forum.
There seems no other solution than to activate the timer with a switch.
In the beginning I often forgot it, but now it's OK.

I have this setup on my devo 10 for electric gliders with DSM protocol :

---

[timer1]
src=ELE DR1
resetsrc=GEAR0
[timer2]
type=permanent
src=ELE DR1
val=90828285
[timer3]
src=Ch1
resetsrc=GEAR0

---

Some remarks :
Timer1 is the duration of actual flight, ELE DR1 is the switch I take (and sometime forget) to begin and stop the counting, GEAR0 resets the timer.
In my setup GEAR0 also is used as throttle hold. So when I set throttle hold after each flight, the timer is reseted.

Timer2 is for total flight duration of the plane (permanent timer), like timer 1 it is activated by ELE DR1.

Timer3 is for the run time of the motor during actual flight. Same remarks apply as for timer1, but it's triggered when Ch1 (throttle output) has a value bigger than about -90 (you cant change this value)

MfG, Roland

P.S. No much flying these days, I was in holidays.

Next 5.0.0 custom build to test for Vlad's new CYRF reset and reliability of RX-telemetry side: www.deviationtx.com/forum/builds/6407-de...s?limitstart=0#54569

Corresponding info DSMx protocol error / enhancement thread: www.deviationtx.com/forum/protocol-devel...m2-x-protocol-errors

I just installed Vlad's last .DFU file (from 12/11/2016) on my DEVO 10, I will try to resume flying with my EPO glider these days.
I allready did some tests in my street with an orange RX (1° Celsius this evening, brrrr), I can walk about 200 meters before telemetry stops to work (black boxes in telemetry screen), but as I come back toward the Rx telemetry works again.
Will have to do long time tests, as with the stock 5.0.0 firmware, it took me 4 hours of flight to have a TX lock...

Anyway, thank you Vlad and Thomas for your work !

Roland

For your information :
Today I was flying my test-glider with Vlad's last .DFU file from 12/11/2016 on my Devo 10 (the white one with the 5dB gain antenna)
I had 4 flights with a total time of 90 minutes.
Telemetry was enabled, receiver is an OrangeRX R720X V2 (with integrated telemetry), protocol was DSMX.
I flew as far as I could, about 400-500 meters with my 1,5 meter glider, this was the limit before it became difficult to see.
At each fly (about 22 minutes) I had between 150 and 200 fades from each of the 2 antennas (A and B ), about 100-150 frame losses (F) and no Holds (H)
On the Devo I had the telemetry screen, of course I wasn't able to look at it all the time, but each time I looked at it the telemetry worked, I saw no blacked out boxes.

So far, so good.
Will try this again for some other hours.

Regards, Roland

Good evening Vlad !
Yes, I've enabled the Orange Rx option in the Protocol submenu.
My 2 antennas are perpendicular each to the other (there are not big differences in fades between antenna A and B ), but the battery wires are going rather near of the receiver (difficult to change, there is no much place in the glider)

Regards, Roland

Thomas.Heiss wrote: Suggestions:
1) Try with a genuine Spektrum receiver (e.g AR6210) with DataPort + TM1000

What are your FrameLoss + Fades values on a good Spektrum receiver, e.g AR6210?
How far is your receiver and antennas from lipos / ESC power cable?

You could even try a newer AR660T and get rid of TM1000: www.horizonhobby.com/product/radios/radi...-receiver-spmar6600t


Even for genuine Spektrum AR6210 valid:
Fades A = 255 in my in-air test as the ESC/lipo power cable is TOO NEAR the receiver because of this smaller glider fuselage.

But you can get a Spektrum satellite 90cm extension cable and mount the satellite at the backside of your HLW/SLW and not in your fuselage by itself.

Most of the times I got for about 15-20minutes glider flight time max ~8-10 FrameLosses, and I am flying higher / more far away @100mw output power.
Having the Spektrum satellite at the glider backside (outside of the fuselage, far away from any cables, lipos, electronics, etc.) does really help to have good Fades B and low FrameLosses.

Thomas.Heiss wrote: What are your FrameLoss + Fades values on a good Spektrum receiver, e.g AR6210?
How far is your receiver and antennas from lipos / ESC power cable?

Hallo Thomas.
I have no other DSM receiver, the DEVO is my only Transmitter and all my other planes have Walkera Receivers with DEVO protocol.
So I can't compare.
My antennas are about 6 to 10 cm from the lipo pack power cable (difficult to get it better, it's a little glider with a little place inside)

When I flew yesterday I had my Handy in my jean's pocket, maybe I have to try it without next time !

I don't plan to make tests with satellites, when this is really necessary to have a good reception in DSMX, I will stay with Walkera receivers, wich gave me 0 problems in 5 years of flying. (albeit there is no way to know if there are frame losses too with them !)

MfG, Roland

If Spektrum I would use something with true MultiLink technology (2+ internalreceivers): www.spektrumrc.com/Technology/Multilink.aspx
or
receiver with two antenna diversity + Sat (external).

e.g AR600 has only one receiver, two antennas with diversity, if there are Fades they result in FrameLosses.
AR6210 has two receivers, Fades A (internal) + Fades B (SAT).
Not sure about AR610: It is probably described on www.rc-network.de if one vs two receivers, right here it is: www.rc-network.de/forum/showthread.php/5...ewfull=1#post3810564


Sorry Roland, I personally do not have currently the overview what Spektrum receivers use two real internal receivers (MultiLink).

There is e.g also AR6255 (carbon, longer antennas), or the newer T series RX.

Maybe the word "MultiLink" is given on some Spektrum receivers for specs / manuals??
There should be better a search / filter option on the Spektrumrc.com site!