Anonator

  • *
  • 36
  • UAAV experimenting
strange faults with Revo
« on: October 11, 2019, 02:56:41 am »
Plz look at this log file, as I cannot work out why faults occur in system health and no gyro scope log. Maybe a bad Revo? but what part?
Ubuntu 18.04LTS.  F450 Quad.  3 Fixed wing planes.  Revo and OmniF4 flight controllers.

Re: strange faults with Revo
« Reply #1 on: October 11, 2019, 10:50:14 pm »
A new board or suspected crash damage on a previously working board?

Crash damage is fixable, but you could spend a lot of time finding and fixing it if it isn't an obvious problem like broken board or trace.  Also, I do recall some reports of boards where the small square gyro/accel chip popped off in a crash.  I would use a magnifying glass or better to look for cracked/broken traces and missing or damaged IC chips.

If it's a new board and it doesn't work correctly with new firmware and EraseSettings, it is broken.  Personally, I usually look for solder bridges with a magnifying glass in this case.  The two boards I bought that were bad had some IC pins shorted together with solder.

After these simple checks, then unless you simply enjoy fixing boards, if it's not simple you are better off buying a new one.  Even if you can figure it out, buying a new IC chip can cost $15 or so by the time you pay shipping.  DigiKey wants $12.44 plus shipping for a single MPU6000 gyro/accel chip.  You can buy a CC3D shipped for less than that.  Add your time and decide if you want to fix or buy.

Like a car, it can always be fixed, but at what price?

I don't know about other developers, but it would save a little work if you would attach a screen capture of the whole expanded branch of System->DataObjects->SystemAlarms that would look something like the file I attach here.  Of course with the Revo plugged in and the alarms showing.  Two images is OK if it isn't easy to get it all on one image.

Anonator

  • *
  • 36
  • UAAV experimenting
Re: strange faults with Revo
« Reply #2 on: October 12, 2019, 03:58:33 am »
Thx Cliff,

Revo from bangood about a year ago, no crash damage and inspected with magnifier. Looks ok.  Seems to give (system event) UAVO callback error first then Stabilise error. Occasionally a GPS error (black with cross thu). Error lasts less than 1 sec but can cause glitch in flight.  No pattern to it, sometimes goes for 5 min no issue.

Anyone had this?

Using data and control thu oplink.  ppm from futaba skysport to flex port.  usb cable from oplink to laptop for telemetry.
Ubuntu 18.04LTS.  F450 Quad.  3 Fixed wing planes.  Revo and OmniF4 flight controllers.

jdl

  • ***
  • 246
Re: strange faults with Revo
« Reply #3 on: October 12, 2019, 11:22:59 am »
I've once had similar issues. MPU6000 chip occasionally going "dead" for a fraction of a second, causing STAB and ATTI alarms. All the other time except these glitches the gyro and accel data was coming perfectly stable. Finally a longer MPU6000 "hang" (for a second or two) caused the copter mad and crashed (flying in rate mode).
Discussed in the forum and f5soh pointed me to look for a hardware deffect in the MPU6000. I was quite enthusiastic these days and managed to get a new chip and replaced it. Issues stopped and this Revo board is still in use. But I'll not probably try to do this again :) Without proper equipment it is a matter of luck to succeed.

https://forum.librepilot.org/index.php?topic=2449.msg17375#msg17375


Btw, xt-xinte sells Revo boards at very nice price:
https://www.xt-xinte.com/OpenPiolot-CC3D-Revolution-Flight-Controller-Revo-Support-Auto-Return-DIY-Mini-250-Quadcopter-Multirotor-Heli-Fixed-Wing-p140119.html

I've bought 5 or 6 Revos from them till now, last one few days ago. They look (and are packaged) absolutely the same as these from Banggood (maybe same manуfacturer) but are lot cheaper. All I've bought from xt-xinte work flawlessly, the one with dying MPU6000 was from Banggood. Not saying Banggood sells bad ones, my deffective Revo was just bad luck.

Anonator

  • *
  • 36
  • UAAV experimenting
Re: strange faults with Revo
« Reply #4 on: October 13, 2019, 02:14:49 am »
Thx jdl,

Investigating the gyros,  just to windy to fly atm.  Will report bk.
Ubuntu 18.04LTS.  F450 Quad.  3 Fixed wing planes.  Revo and OmniF4 flight controllers.

Anonator

  • *
  • 36
  • UAAV experimenting
Re: strange faults with Revo
« Reply #5 on: October 15, 2019, 09:22:18 am »
Turned out to be an LG33 with C8 issue so removed cap and added more decouple on output... Better but not good enuff :/  Don't have a 22pf SM cap.

Ordered new Revo from www.xt-xinte.com/.  I will report how it goes.  Almost every usual supplier discontinued Revo.

I have a Revo Mini with cable attached "oplink mini air". It has good sensors.  Sanded off RF chip type.  Glitches with rx packet errors at ground station.  Is it possible to config the revo mini to use oplink connected to main/flex port instead, data and control.  I could not get it to work (known good oplinks) trying all the obvious.
Ubuntu 18.04LTS.  F450 Quad.  3 Fixed wing planes.  Revo and OmniF4 flight controllers.

Re: strange faults with Revo
« Reply #6 on: October 15, 2019, 10:06:56 am »
I have a Revo Mini with cable attached "oplink mini air". It has good sensors.  Sanded off RF chip type.  Glitches with rx packet errors at ground station.  Is it possible to config the revo mini to use oplink connected to main/flex port instead, data and control.  I could not get it to work (known good oplinks) trying all the obvious.

Check the size of the X2 label on the RFM22B (RF board) that is part of the OpLink.  Lately there are bad RFM22B boards around where one frequency in the hopping list is wrong.  Bad will talk perfectly to bad since they both change to the same (wrong) frequency, but good and bad do not mix.  If you try to mix them, then at close range you will get regular RSSI drop and at anything farther, you will get dropped connection.

Also, I bought a pair of the OpLink Ground/Air at a good price and because I wanted the SMA RF connector type.  This pair had this problem.  I wound up replacing the RFM22B and the problem went away.  See this thread to understand how to see whether you have good or bad RFM22B on your OpLink:
https://forum.librepilot.org/index.php?topic=4197.0

Yes, you can configure Revo class FCs to use a separate OpLink connected to Main/Flexi (at least).

Anonator

  • *
  • 36
  • UAAV experimenting
Re: strange faults with Revo
« Reply #7 on: October 21, 2019, 07:28:23 am »
Thanks Cliff,

You guessed correct, large X2.

Ordered new RFM22B and will see if my old eyes still up to the solder job ;)
« Last Edit: October 21, 2019, 07:34:54 am by Anonator »
Ubuntu 18.04LTS.  F450 Quad.  3 Fixed wing planes.  Revo and OmniF4 flight controllers.

Re: strange faults with Revo
« Reply #8 on: October 24, 2019, 12:28:26 am »
Just FYI:  ALL e RC (www.allerc.com) has the Revo for $29.00, same day shipping from Phoenix, AZ.

Re: strange faults with Revo
« Reply #9 on: October 25, 2019, 08:20:23 am »
Any idea how much shipping costs?  I didn't see an "estimate shipping" button, at least without filling out the whole "become a customer" stuff.

Re: strange faults with Revo
« Reply #10 on: October 28, 2019, 09:01:46 pm »
Standard shipping to any continental USA address is $5.49.  Free shipping on orders over $100.

jdl

  • ***
  • 246
Re: strange faults with Revo
« Reply #11 on: November 05, 2019, 10:33:39 am »
Yesterday I received my last ordered Revolution FC & Oplink Mini from xt-xinte. Thoroughly tested them and I'm happy to report they seem to be good quality units. RFM22B modules are from the "good" ones, small X2 label, chip markings not grinded. All sensors work fine, baro altitude also reported correctly, normally deviations are within 0.5m, max 0.8m.

Boards came with firmware 15.05. Well known plastic cases, rubber antennas. Old stock, obviously. Sadly, it won't last forever...