Taranis - Smart Port Lipo Volt. Sensor Cell Low Bug

Thorvald · Post by **Thorvald** » Sun Jan 05, 2014 8:08 pm

While flying my Raptor 30, I happened to review the telemetry log on my Taranis after the flight and found that the new FrSky SP-FLVS - Smart Port Lipo Voltage Sensor was randomly showing a single cell was down to ~1v.

Here are some quick of samples from it:

Taranis Rev A01
OpenTx r2834

BaroAlt = FrSky FVAS-02H High Precision Vario
A1 = Receiver Voltage
Vfas = FrSky SP-40A - Smart Port 40 Amp Sensor
Cells = FrSky SP-FLVS - Smart Port Lipo Voltage Sensor

Code: Select all

Time	BaroAlt	A1	Vfas	Cells	Cell1	Cell2	Current	Consumption
08:41.5	4.09	6.21	6.3	6.3	3.18	3.18	1.2	36
08:41.7	4.04	6.21	6.3	6.3	3.18	3.18	1.2	36
08:41.9	3.93	6.16	6.3	4.3	1.12	3.18	1.2	37
08:42.1	4.22	6.16	6.3	4.3	1.12	3.18	1.2	37
08:42.3	4.18	6.26	6.3	6.3	3.18	3.18	1.2	37
08:42.5	4.12	6.26	6.3	6.3	3.18	3.18	1.2	37
08:42.8	3.89	6.21	6.3	6.3	3.18	3.18	1.3	37
08:43.0	3.79	6.21	6.3	6.3	3.18	3.18	1.3	37
08:43.2	3.58	6.21	6.3	6.3	3.18	3.18	1.3	37
08:43.4	3.44	6.21	6.3	6.3	3.18	3.18	1.3	37
08:43.6	3.21	6.21	6.3	6.3	3.18	3.18	1.3	37
08:43.8	3.28	6.21	6.3	6.3	3.18	3.18	1.2	37
08:44.0	3.18	6.26	6.3	6.3	3.18	3.18	1.2	37
08:44.2	3.05	6.26	6.3	6.3	3.18	3.18	1.2	37
08:44.5	3.26	6.26	6.3	6.3	3.18	3.18	1.3	37
08:44.7	3.32	6.26	6.3	6.3	3.18	3.18	1.3	37
08:44.9	3.72	6.16	6.3	6.3	3.18	3.18	1.3	38
08:45.1	3.87	6.16	6.3	6.3	3.18	3.18	1.3	38
08:45.3	3.93	6.21	6.3	6.3	3.18	3.18	1.2	38
08:45.5	3.89	6.21	6.3	6.3	3.18	3.18	1.2	38
08:45.7	3.81	6.1	6.3	6.3	3.18	3.18	1.2	38
08:45.9	3.81	6.1	6.3	4.3	1.12	3.18	1.1	38
08:46.1	4.01	6.16	6.3	6.3	3.18	3.18	1.1	38
08:46.4	3.88	6.16	6.3	6.3	3.18	3.18	1.1	38
08:46.6	4.05	6.16	6.3	6.3	3.18	3.18	1.1	38

And a sample with Cell 2 going low:

Code: Select all

Time	BaroAlt	A1	Vfas	Cells	Cell1	Cell2	Current	Consumption
13:42.5	8.66	6.21	6.2	6.2	3.14	3.14	1.4	124
13:42.7	8.66	6.1	6.2	6.2	3.14	3.14	1.4	124
13:43.0	8.45	6.1	6.2	6.2	3.14	3.14	1.4	124
13:43.2	8.45	6.1	6.2	4.2	3.14	1.08	1.4	124
13:43.4	8.45	6.1	6.2	4.2	3.14	1.08	1.4	124
13:43.6	9.91	6.21	6.2	6.2	3.14	3.14	1.4	124
13:43.8	10.15	6.21	6.2	6.2	3.14	3.14	1.5	124
13:44.0	10.15	6.1	6.2	6.2	3.14	3.14	1.5	124
13:44.2	9.85	6.1	6.2	6.2	3.14	3.14	1.5	125
13:44.4	9.77	6.1	6.2	6.2	3.14	3.14	1.1	125
13:44.7	9.58	6.1	6.2	4.2	3.14	1.08	1.1	125
13:44.9	9.38	6.1	6.2	4.2	3.14	1.08	1.4	125
13:45.1	9.42	6.1	6.2	6.2	3.14	3.14	1.4	125
13:45.3	9.48	6.1	6.2	6.2	3.14	3.14	1.4	125
13:45.5	9.8	6.16	6.2	6.2	3.14	3.14	1.2	125
13:45.7	9.54	6.16	6.2	6.2	3.14	3.14	1.2	125
13:45.9	9.79	6.21	6.2	6.2	3.14	3.14	1.2	125
13:46.1	9.79	6.21	6.2	4.2	3.14	1.08	1.2	125
13:46.4	11.3	6.16	6.2	4.2	3.14	1.08	1.2	125
13:46.6	11.3	6.16	6.2	6.2	3.14	3.14	1.2	125
13:46.8	11.3	6.05	6.2	6.2	3.14	3.14	1.2	125
13:47.0	12.83	6.05	6.2	6.2	3.14	3.14	1.2	125
13:47.2	13.05	6.05	6.2	6.2	3.14	3.14	1.2	126
13:47.4	13.05	6.05	6.2	6.2	3.14	3.14	1.1	126

I also just did a static bench test, left Rx powered up with all the sensors and servos power up. Then left Tx on and logging for around 12 minutes. I also setup a custom switch if any cell got below 2v or if total cells voltage got below 4 volts and played two different tracks.

Sure enough, randomly throughout the 12 minutes I would get the tracks playing. Watching the Tx telemetry screen it does show the voltage drop.

However watching the Lipo Voltage Sensor little LCD screen it does NOT show the voltage change/drop.

Anyone got any ideas?

Cheers
Tim

Post by **Kilrah** » Sun Jan 05, 2014 8:20 pm

There was an error in telemetry decoding that could cause a spurious value once in a while. It should be already fixed and available in next release.

Thorvald · Post by **Thorvald** » Sun Jan 05, 2014 8:53 pm

Thanks Kilrah. If you want me to test just let me know.

Cheers
Tim

Post by **Kilrah** » Sun Jan 05, 2014 9:06 pm

We'll be setting up a server that provides nightly builds for testing at some point.
In the meantime you could have a go with this:

opentx-taranis-TEST-r2921.zip: (128.8 KiB) Downloaded 381 times

It's the candidate for next release. Should be OK, but standard disclaimers apply. Thanks for reporting any issue

Thorvald · Post by **Thorvald** » Sun Jan 05, 2014 9:24 pm

Ok just flashed with that firmware and started to run on bench. Within seconds I get Cell2 dropping crazy to 1.2v. The frequency of the occurrence appears to be greatly increased with this firmware. I'll leave it running for a good couple of minutes and then post telemetry logs. It did calm down after the first minute.

EDIT: Second full power down and power up is running much better. 5 minutes in and no warnings or drops (besides any you get when first powering the Rx before it reads the sensors).

Cheers
Tim

Thorvald · Post by **Thorvald** » Sun Jan 05, 2014 10:15 pm

Ok I've now done multiple 10 to 15 minute sessions and not one trigger.

I suspect when I first powered up with the new firmware it was using old cached data???

I'll let it run for a few hours

Thanks!!
Tim

Thorvald · Post by **Thorvald** » Tue Jan 07, 2014 2:44 am

Done a lot more testing and the only time I can get it to show the problem again is right after a power on of the Rx. The low cell (1.x volts) warning and one of the cells will flip-flop between the proper 3v and the erroneous ~1.18v. It will continue to flip-flop for 3 to 5 times approx. within the first 5 to 15 seconds of power up and then be fine.

I'll keep at it to see if I can find a pattern or how to reproduce.

Cheers
Tim

Thorvald · Post by **Thorvald** » Fri Jan 10, 2014 12:48 am

Downloaded the 2923 release today and confirmed the bug is still there, randomly a single cell reading drops to 1.x volts for a split second.

I have the non-smart port version as well so I'll do a test with it.

Tim

Thorvald · Post by **Thorvald** » Fri Jan 10, 2014 2:24 am

Well looks like the Sensor hub connected Lipo Voltage sensor does not suffer from the same issue, I've been testing it over the past 2 hours and it's never flinched. Voltage always rock steady.

So it looks like the issue is specific to the Smart Port Version.

Cheers
Tim

totof1965 · Post by **totof1965** » Sun Jan 26, 2014 3:41 pm

Hi, just received and mounted my FLVSS, and got same issue, before receiving the flvss i've been using a lipo sensor thrue a hub, like you and never have this issue

so I think there's something wrong with the FLVSS

rilot · Post by **rilot** » Mon Feb 03, 2014 12:47 pm

To add another one to the list.
I have two FLVSS and both exhibit exactly the same issue. Voltage display on the little OLED screen is correct but one cell will randomly drop to 1.xV on the Taranis.

totof1965 · Post by **totof1965** » Fri Feb 07, 2014 2:44 pm

It seem that this issue have been fixed in last firmware release

Koffi · Post by **Koffi** » Tue Feb 18, 2014 2:52 pm

Unfortunately, the issue still exists in the latest Firmware 2940.

Maybe we have to wait for the next major update to OpentTx 2.0...

Post by **ShowMaster** » Tue Feb 18, 2014 3:27 pm

I have a new FLVSS to try out but haven't programmed my Taranis to display 3 cells or announce the voltage. I could use help with the programming.
Could you post your settings or a copy of your model memory with it please?
I'll be happy to test mine out and report back.
Thank you
SM

Sent from my iPod touch using Tapatalk

Post by **Kilrah** » Tue Feb 18, 2014 3:39 pm

There seems to STILL have random telemetry corruption issues despite everything we've already done to try and solve the problem (happens on all values, not just the lipo sensor). We think more and more that it might be a problem on the "FrSky side" which is out of our control, i.e. either in the module or reciever.

We'll try to find time to repeat the problem (which seems to never happen when looking for it) and carefully analyze and decode the received stream manually to see if it is correct or not when it happens.

Koffi · Post by **Koffi** » Tue Feb 18, 2014 5:34 pm

Tell me of I can help out with a Log, that contains this issue (have lots of them).

Post by **Kilrah** » Tue Feb 18, 2014 5:52 pm

Not really, as the log would need to be synced with a raw dump of the incoming data. That way we can easily find the glitches in the decoded log, note the timestamp, and see what we received from the module at that time to see if it matches and is correct.

Compiling the firmware with DEBUG=YES enables the raw dump, but it's not enabled in "production" firmware.

totof1965 · Post by **totof1965** » Tue Feb 18, 2014 6:14 pm

Yep still got the issue

kilrah if it can help, send me a custom firmware and i can do the tests for you

mstrens · Post by **mstrens** » Tue Feb 25, 2014 12:10 pm

If you want, I could make a test with an oxs on SPORT.
I can let the oxs generates e.g. a set of fix data in several fields ID (e.g altitude, Vspd, T1, T2, cells,...)

ZOZ · Post by **ZOZ** » Mon Mar 10, 2014 10:08 am

Hi All

My first post on this forum

.

Well, bought the new FLVSS Sensors and thought I would get now the voltage telemetrie right.
On the first flight with my YAK 54 was everythink ok. Set the Cell Alarm to 3.58 V and the Sirene came.
On the second flight I had the Issues already on the ground and had to switch on/off few times and then in the air showed zero Volt and
the Alarm sirene didn't work as well so I've killed the LiPo unfortunatelly.

Using the FW 2940

Cheers

ZoZ

HarWee · Post by **HarWee** » Wed Mar 12, 2014 7:27 am

I’m facing the same problem! Within seconds I get different cells dropping down.
Additionally I have Variometer Sensor on the smart port. This shows wrong values to! The height seems constantly to be more.
My Setup:
X8R Telemetry Receiver (with DJI Naza V2 on the SBUS Port)
FrSky Lipo Sensor Smart Port
Variometer Sensor Smart Port
Taranis X9D

Awap · Post by **Awap** » Sun Mar 16, 2014 7:12 pm

FYI: The firmware update for the SP-FLVSS, found on the FrSky site (under "downloads") solved all for me. FrUSB-FUC-3 needed though.

Post by **Kilrah** » Sun Mar 16, 2014 9:41 pm

Oho, so it was a sensor bug after all!

Koffi · Post by **Koffi** » Mon Mar 17, 2014 4:08 pm

After first test with satisfiyng results I had to notice that there seem still to exist some special Cell voltage configuration that confuse the sensors data telemetry:

Awap · Post by **Awap** » Tue Mar 18, 2014 10:15 pm

Mmm.. for me the new FW works on 3S. But it seems there are still problems on 4S with cell 1 and 2 (I couldn't test this). Somebody mentioned this also at the FrSky BBS:

Board index‹ Products and Services‹ Telemetry Sensors‹ SP-FLVS random voltage drops

I suggest to post your problem on the FrSky BBS, I hope they are able to help you with the 4S problem also.

mstrens · Post by **mstrens** » Fri Mar 21, 2014 10:00 pm

I suspect that there is an occasionnal error in the telemetry fields.
I do not know if it is on the RX, XJT or TX side.

I will try to explain what I have seen.

A user (NeilRogers) tested today a new version of an OXS.
In this version, OXS transmit in the field T2 a data that is the difference of altitude in meter between about 10 sec.
The value is calulated and sent around every 500 msec.
The value should normally be quite stable.
The user logged the telemetry data and the log shows very strange occasional variations.

The log is in attachment.

The value (transmitted as T2 but being the diff of alt between 10sec) changes from:
1) -28, 627, ..., 627, -26
2) -3, 650, ..., 638 , -17
3) -2, 654, ..., 656, 3
4) -33, 626, ..., 629, -25
5) -21, 629, ..., 626, -31
6) -16, 630, ..., 627, -24

If you look at this you see that:
- the wrong values are always around 620-660 (the plane never made 650 vertical meter in 10 sec)
- if you suppose that the wrong value should normally be nearly equal to the previous/next logical value and if you make the calculation (wrong - previous or next) (e.g. in first case 627 - -28 = 655 and 627 - -26 = 653), you get most of the time a value around 654.
Here the results
655 653
653 655
656 653
659 654
650 657
646 651

I asked the user following data
- which Tx are you using (Taranis , ...)
- which release of openTx are you running on TX
- did you make the Frsky upgrade on XJT
- which receiver do you use
- did you make the Frsky upgrade on RX

Are other data useful for further analysis?

As already proposed, I could make myself further test with a version of XOS where I should send a fix value (or a value changing continuously just by 1 or -1 between 2 limits.

openrcforums