No Telemetry after 30 sec

[Seaquest]

Hello,

I've got a CC ice75, Jlog 2.6 and a JR XG14 with RG731

When I load the Jlog with the "26-CC_J-126.14.bin" the telemetry appears on my screen. But after approximately 30 sec the data disappears from the screen..

When I load the Jlog with the "26-CCop_J-126.13.bin" the telemetry appears on screen and keeps working, but I get the "engine twitch".

Led on the Jlog blinks green.

This is in the config.txt (with the 26-CC_J-126.14.bin firmware)

9600,0,0,6,0,10,93,14,153,0,1,0,16,22,2,85,0,1,0,0 ,0,0,0,0,0,0,7,0,0,6,16,16,16,16,16,16,16,16,16,16 ,16,16,16,16,16,16,132,126,0,0,39,43,15,20,1

Any thoughts?

Thanx in advance.

Erik

[dl7uae]

Strange that you have the twitching with the "own pulse" version and not w/ the other.. Not that I mistook it in the downloader, mixed up "op" with "non-op"...

ESC data acquisition is independent of telemetry. So this is a pure telemetry issue.

When does it happen? Is it only time dependent or is there anywhere some higher current drawing what could cause that effect by a ground loop / interference?

Sure that the transmitter is not getting too close to the telemetry transceiver during testing? I myself observed such effects from time to time, but they are based purely on the JR RF path.

[Seaquest]

Yes, I was surprised it twitched with that version.

It happens on the table at home. Everytime when I connect the battery.

Switch on the transmitter, connect battery, initialization from ESC. At that moment I see data. After approximately 30 sec no more data.

I tried to connect en hover it outside on the street but still no data.

Got the transmitter away up to 2 meters. Same result. Strange enough it doesn't happen with the other firmware.....

I can reproduce it evertime. Even after reflash from the firmware.

I don't know if it helps, but here a video.

Telemetry (0 min 43 sec)

Erik

[dl7uae]

I see... Will test both firmwares tomorrow, - mean today..

[Seaquest]

Don't work too hard today, Tom.

It's weekend. (although I'm working today too)

[dl7uae]

I have anyway no free time.

I see what you mean. However, it only happens randomly (sometime) with me, with the version w/o "ownpulse" only.
That's the reason why I have not noticed it when testing.

Well.. Castle with JR telemetry (besides Futaba s.bus2) is particularly hard. With JR this is due to a relatively high interrupt frequency from 250kBd serial line speed.

So that was caused by a racing condition. JLog got stuck in tx mode on COM (JR telemetry line, 1-wire).
Fixed it: 26-CC_J-126.16.bin

-----

"Own pulse": Another unfavorable condition comes from the JR receiver, at least from a RG831B what I have here: instable/jumping pulse frequency. So it can happen, if JLog comes to read "on the edge", that one or two measurements (pulse length) are wrong in between.

My thinking was: Oh no, don't use a hazard filter, that adds latency.
Now I changed my mind: BS! A 1-stage hazard filter adds 20ms latency on throttle change at maximum.
I added a 1-stage hazard filter. On top I set in transmitter: travel of throttle pulse length, LOW, rised from 100 to 110% (even better to use 120%). That makes no noticeable difference in usage.
Result: Almost never a twitch happens.
New Firmware: 26-CCop_J-126.15.bin

Positive side effect: Both firmwares have the mod for the HV²BEC already - one with a new version of the main chip, - automatic address detection, different handling of the temperature value.

Test conditions at my end:
- Castle ICE2 HV 120 on a motor test stand
- receiver: RG831B
- transmitter: XG8
- JLog 2.6
- logic analyzer, sources, brain, fingers, patience

[dl7uae]

Oops, concerning the pulse frequency behavior, I think I mixed-up FrSky's receiver "X8R" with JR's transmitter "XG8" and so the JR receiver.

I never measured on a RG831B, at least I cannot remember yet.

[Seaquest]

Thanks a lot for the superfast help, Tom.

I'm not able to test today. But I'll let you know tomorrow.

If there are specific things you want to know, that might help you in the future, and I can test it let me know!

Erik

[Seaquest]

Goodmorning.

Telemetry seems to work on both firmwares.

What I did notice is that the voltage showing is a bit low. Fully charged 6s 25.2V but I get a 24.7V reading. Can I change that somehow?

In the Owmpulse firmware I can't find the 1 stage hazard filter. It's not showing in JLC 7.0.3.2.

[dl7uae]

This is Castle.., always a bit too low.

No means in JLog to correct it. Wouldn't make much sense because...

Some Castle ESC have also the problem w/ low end current measurement. Current rising, voltage rising. Huhh?!

[Seaquest]

The voltage just caught my attention. No problem.

And the 1 stage hazard filter. Should it show in JLC 7.0.3.2?

[dl7uae]

Should it show in JLC 7.0.3.2?

No, it shouldn't. Stage 1 is standard, always on. Can just be changed to 2 or 3.

[Seaquest]

Tom,

The telemetry works. But I get twitches while flying. The tail reacts like static hits but I can hear the engine pulsate.

When I disconnect the jlog, the tail is stable.

I'm using the own pulse firmware at this moment.

Erik.

[dl7uae]

Glitches, hmm..

With JLog logging (logging enabled and SD card inserted) or w/o?

Did you try the non-ownpulse firmware also?

If you do not mind: I appoint you to a beta tester of the S32 (aka JLog3).
Only, problem is that currently the configurator is not finished that way that you could do the setup by yourself.

[Seaquest]

Sd card inserted. I don't know if logging is enabled. It's not a full version. Logging from the castle was enabled I think. But I have to check.

I will try the non own pulse.

I don't mind. ☺