Thread Tools
Aug 13, 2021, 06:04 PM
Registered User
Also misbehaving Guardian...

I did the normal install and adjustments and everything looked good.
Then I switched the Guardian on and things went nuts to where the elevator was reversed and the right aileron was suddenly coupled to the elevator and reversed as well.

I then turned the entire system off, turned it all back on and it was normal again until I turned the Guardian on again.

I went through all the settings and everything checked out.

A similar situation happened on a different plane with a different guardian as well.

I'm afraid I can't trust my Guardians anymore so I guess I need to learn how to fly better.
Sign up now
to remove ads between posts
Aug 16, 2021, 01:36 AM
Rick
Some might find this interesting. This log file was recorded during an uneventful flight with a Durafly Tundra with a stock ESC and a Guardian on board. In the first part of the flight the G was off. The red line is the mode channel and shows when it was switched to 2D. The change in the receiver voltage is pretty clear. When a stabilizer is active the rapid corrections going to the servos can create transient loads that significantly impact the BEC with a resulting voltage drop. Even though the voltage dipped nearly .4V at times it was not enough to cause brownout or any other problem, this time.

The log probably needs a little context explanation. This was logged with the Rx Snoop. The voltage is sampled at 1ms intervals to capture short transient dips characteristic of servos loads. The points in the log file are recorded with each sbus frame, which in this case is every 7ms. The value logged is the lowest value sampled during that interval. This is quite different from the infrequent samples and averaged values typical of transmitter logs.
Aug 16, 2021, 02:13 AM
ancora imparo
jj604's Avatar
Rick, that Rx snoop gadget of yours is really useful.
I have a TattleTale but the continuous logs matched to other events by the Snoop is invaluable.
I am tempted to make one.

John
Aug 16, 2021, 04:29 AM
Great Southern Land
Berkie's Avatar
Quote:
Originally Posted by plasoneal
Well, a little more info. I intentionally reversed the aileron servos last night and checked the impact. Indeed, the aileron throw was still correct but the flap activation was reversed. Still scratching my head a bit on that one. But the servo centering was fine.

Net, I'm still not sure why (2 outings ago) I experienced both a flap reversal AND center position offset. It is however looking less like a failure of the G itself. 2 out of the 3 issues now have plausible explanations. I'll certainly keep my eyes on this very closely going forward though.

So what would you advise choochoo?
Would a good ubec remove the risk?
Aug 16, 2021, 09:50 AM
Registered User
Galand's Avatar
A UBEC is the answer.
It would be interesting to see the voltage drop when you add the simultaneous draw from the flaps and the landing gear.
Rick, would there be a way to visualize the amp draw?
Dirk
Aug 16, 2021, 06:00 PM
Rick
Quote:
Originally Posted by Berkie
So what would you advise choochoo?
Would a good ubec remove the risk?
I don't have anything to add to what's already been said. It sounds like basic operator error.

It's really easy to toggle the wrong number of times and set the G for elevons. IMO that setting should be inside the stick menu so it isn't so easily reached by accident. I discussed that with ET a few years ago and they didn't want to make the change. From time to time someone here on RCG asks why their controls suddenly went bonkers, that's usually it.

With aileron servos installed in the usual way mirrored, swapping the plugs will cause the ailerons to work normally but the flaperons are reversed. If there is a trim offset in the installation, reversing the plugs will reverse the offset as well. This is normal operation on all installations with or without a G.

A good ubec is always a good idea but doesn't seem to be related to your problems.

Personally, I'm not a fan of flaperons on powered planes. On competition sailplanes it makes sense to be able to trim the camber across the wing for optimal performance in different flight modes, but these planes rarely have a G, it's probably not even legal for competition. When I first got a capable programmable radio I put flaperons on everything because I could. Eventually it became clear that not one plane benefited in any observable way other than "that's cool". On the other hand deploying flaperons on landing destroyed roll control, which was very observable.
Aug 16, 2021, 06:44 PM
Rick
Quote:
Originally Posted by Galand
A UBEC is the answer.
It would be interesting to see the voltage drop when you add the simultaneous draw from the flaps and the landing gear.
Rick, would there be a way to visualize the amp draw?
Dirk
I tried checking the flap impact (no retractable gear). There were some other issues at the time that prevented getting a clear picture but the general look didn't seem to have much impact on that plane.

It's probably fair to say that in this scenario the current visualization is the inverse of the voltage but to get an exact value reading is problematic.

The common way to measure current is to measure voltage drop across a known resistor. That's how most ammeters work. It might be possible with some re-wiring, to insert a known resistor in the circuit, say .1Ω, then measure the voltage across that. I'm reluctant, however. At 5A this resistor would lower the voltage available by half a volt. The AD converter is designed to measure up to 3.3V . Measuring to only .5V may provide too little resolution while jeopardizing the reliability of the radio link.

For in-flight measurements I'm happy just being able to see the impact of transient events on the rx voltage. As usual though, more information raises more questions. Like; how quickly will a rx go into brownout? It's fairly easy to measure the voltage at which a given rx or other device quits working but if the voltage drops below that level for just a millisecond is that enough to trigger a re-boot? Don't know.

Now we are getting really off topic for the Guardian thread.
Aug 16, 2021, 06:51 PM
Rick
Quote:
Originally Posted by jj604
Rick, that Rx snoop gadget of yours is really useful.
I have a TattleTale but the continuous logs matched to other events by the Snoop is invaluable.
I am tempted to make one.

John
I have a Tattletale also and it's a pretty cool little device but it provides only a single data point and that data is presented rather vaguely. When the Snoop was conceived, voltage wasn't considered, but when it occurred to me that it might be easy to add, I had to do it. V3 will add the real time clock and a few other improvements by the way.
Aug 16, 2021, 07:28 PM
ancora imparo
jj604's Avatar
Although we are a bit off topic, it is in fact a critical discussion for folks using stabilisers as it is almost certainly the biggest cause of failures. And the Guardian thread is where most of the intelligent stabilizer discussion has happened.

So if I may pursue it further.

One simple way to get a true measure of the total instantaneous current drawn from the receiver supply would be to put one of these in the power lead. It looks like a sense resistor and transistor amplifier but it is cheap and neat mechanically and light enough (<4g). It is a 0.5mOhm 1% resistor in line with a voltage tap point on the PC board and an XT30 connector integrated. Looks like the voltage at the sense point is amplified and could be monitored using the RxSnoop I would think? As far as I can gather these XT30 units output an analog voltage of 3.3V at 80 Amps. Might be sensitive enough. At 5 Amps the output would be about 200mV.

The insertion loss of a 0.5mOhm resistor amounts to only 2.5mV on a 5Amp current draw.

LINK

Worth considering? Be simpler than starting from scratch.

John

Quote:
Originally Posted by choochoo22
I tried checking the flap impact (no retractable gear). There were some other issues at the time that prevented getting a clear picture but the general look didn't seem to have much impact on that plane.

It's probably fair to say that in this scenario the current visualization is the inverse of the voltage but to get an exact value reading is problematic.

The common way to measure current is to measure voltage drop across a known resistor. That's how most ammeters work. It might be possible with some re-wiring, to insert a known resistor in the circuit, say .1Ω, then measure the voltage across that. I'm reluctant, however. At 5A this resistor would lower the voltage available by half a volt. The AD converter is designed to measure up to 3.3V . Measuring to only .5V may provide too little resolution while jeopardizing the reliability of the radio link.

For in-flight measurements I'm happy just being able to see the impact of transient events on the rx voltage. As usual though, more information raises more questions. Like; how quickly will a rx go into brownout? It's fairly easy to measure the voltage at which a given rx or other device quits working but if the voltage drops below that level for just a millisecond is that enough to trigger a re-boot? Don't know.

Now we are getting really off topic for the Guardian thread.
Aug 16, 2021, 09:45 PM
Rick
Is it really critical to know the current? A rx doesn't reboot because the servos are drawing too much current. The rx doesn't have any visibility to how much current the servos are drawing, it just passes through the rx on bus bars, usually. Unless the bus bars melt the rx is good with any amount of current. The rx reboots because the rx supply voltage dropped below a critical level and the rx components can't function. This may have been caused by the servo current draw but it was the voltage dip that caused the brownout.

The current data might tell you if your BEC is performing well but that is a different question. It's the voltage that indicates how close you might be to a brownout.

Or am I wrong here? You're the engineer.
Aug 16, 2021, 10:51 PM
ancora imparo
jj604's Avatar
Fair point. And I am not really disputing it.

However matching the actual current draw to what is actually happening gives you the best information for understanding what the cause of the problem might be.

It is true that the voltage dip is an ersatz equivalent for current as we might assume that the voltage dip is proportional to current due to the internal resistance of the battery plus its wires and connectors but there are quite number of assumptions behind that which may or may not be true. Personally I would prefer to know what the actual current is. It just gives you more primary info.
For example if the current shoots up if and only if one particular stab axis is active it tells you that gain/mechanical linkage combo might be worth looking at.

It's just my old preference for eliminating as many assumptions and uncontrolled variables as possible.
It's the excessive current that is the actual cause of any problems - it is good to measure that if we can do it easily.
On the other hand the excessive current is just the root cause. The voltage dip is the final result of a number of causes and it is the voltage dip that crashes your plane in the end.

Both numbers have meaning, but in analysing the reason for a problem IMO the closer you can get to the root cause the better.
On the other hand, in determining how close you are to disaster the voltage is the critical one.

John

Quote:
Originally Posted by choochoo22
Is it really critical to know the current? A rx doesn't reboot because the servos are drawing too much current. The rx doesn't have any visibility to how much current the servos are drawing, it just passes through the rx on bus bars, usually. Unless the bus bars melt the rx is good with any amount of current. The rx reboots because the rx supply voltage dropped below a critical level and the rx components can't function. This may have been caused by the servo current draw but it was the voltage dip that caused the brownout.

The current data might tell you if your BEC is performing well but that is a different question. It's the voltage that indicates how close you might be to a brownout.

Or am I wrong here? You're the engineer.
Sep 04, 2021, 10:35 AM
Registered User
Jack Reacher's Avatar

question


This may be silly but can a guardian be mounted upside down? Arrow facing forward but pots pointing down
Sep 04, 2021, 12:03 PM
Registered User
Quote:
Originally Posted by Jack Reacher
This may be silly but can a guardian be mounted upside down? Arrow facing forward but pots pointing down
Yes.
I have several mounted that way.
Sep 04, 2021, 05:59 PM
Registered User
Quote:
Originally Posted by clearprop88
Yes.
I have several mounted that way.

Yes. I also have several mounted upside down.
Latest blog entry: Nexa OV-10
Oct 10, 2021, 10:47 AM
Mach One
captain MoMo's Avatar

3D HH and Rudder control question


I was setting up an AR8360 Spektrum RX and saw a note saying that head hold should be avoided for rudder or if applied it should be done with caution. Then I started thinking that most of my Guardian setups for my EDFs that use 3D HH have their gain set high 90% and other than the delta wing jets I do not see the tail dragging - so I don't bother applying rudder on turns using roll only. On the delta wings I apply a slight rudder to keep the tail with the body and for it to not lag behind.

Question is - what is the proper way of setting up the rudder in 3DHH mode? How does the G compensate rudder? If you bank to the right without rudder applied, does the rudder fight the roll right? I was thinking about adding a mix which seems to be the simplest way to correct this, but how much and will this work ?


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Can I control Guardian stabilizer with DX6i? njwxyz Eagle Tree Systems 9 Jun 02, 2014 12:17 PM
Discussion GPS 2D Plots incomplete, 3D OK jackerbes Eagle Tree Systems 1 Oct 23, 2011 09:42 PM
Discussion Premier 1st generation, Gen 2 Vector unit, 2D or 3D motor unit lanternfly RC Paragliders and Kites 0 Oct 05, 2011 09:42 PM
Data 2d or 3d wing ribs? Marty7n Beginner Training Area (Aircraft-Electric) 56 Sep 22, 2011 04:06 PM
Discussion 3View to 3D to 2d techasist CAD/CAM 0 Mar 27, 2011 12:19 AM