Possible actuator problem

Waggly tails
AllenD
Posts: 11
Joined: 21 Mar 2019, 10:32

Possible actuator problem

Post by AllenD »

While I was installing and sitting up my Fidget, I notices that once in a while the actuator would stop, always at full rudder travel, either full right or full left, for about a second. I know have five flights on my Fidget and it has happened on three of the five flights. the airplane will suddenly roll hard to one side, get about 90 degrees from level, then the actuator will start again. The first time, on the second flight, I was on approach and only about ten feet up, the next two time I was 80 to 100 feet up. Does anyone have an idea of what is going on. I would hate to have it happen on landing or takeoff just a couple off the ground.
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Possible actuator problem

Post by Tobe »

Check PM Allen
Cheers,

Tobe
JohnH
Posts: 30
Joined: 18 Jun 2019, 03:48
Location: Eden, MD USA

Re: Possible actuator problem

Post by JohnH »

I am having the same problem. My Rand actuator will go momentarily hard to the counter clockwise position every 5 seconds or so. I am using a converted a Controlaire GG transmitter with Phil's GG encoder and I am using it with Phil's GG recoder with throttle control for the Rand Servo. I have not yet successfully trouble shot the problem. Can anyone help?
Many thanks,
John Haffner
John H.
"Here's another fine mess you've gotten me into."
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Possible actuator problem

Post by Tobe »

John, please check PM. Quick question, at what voltage do you run the actuator?

Cheers

Tobe
Cheers,

Tobe
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Possible actuator problem

Post by Tobe »

A few notes about the used "Recoders" either Pic or Digispark.

In my opinion the "glitch" is not caused by interference from actuator but by a voltage drop that causes the Pic/DigiSpark to reset, the "Glitches" have been reported both from original Rand GG-1 as the replica but not from the Adams or Adams clones.

Basically this new generation of "Recoders" that allow us to connect Galloping Ghost actuators of any kind directly to modern receivers has to be seen as a software defined apparatus i.e. a processor runs a program that control a H-bridge but if there is a disturbance like a voltage current drop under certain levels the processor reboots and voila we have glitch!
The key issue here is to have a reliable power supply that will give enough headroom to allow for current surges as when the actuator reverses (pulsing) to prevent the PIC/Digispark from dropping out during current surges.

I generally run all my test from a battery pack, 4.8 v NiMh or an old fashion power supply w/o any protection except a regular 5 A fuse i.e. no short circuit protection, my Dad used to call that power supply for "The smoker".
Last night I collected several different speed controls with built in BEC and some "pure" voltage regulator and I spent some times in the lab and was able to verify the following.

I connected a scope to monitor pulse and voltage and found out that there are current surges such as when the actuator reverses (pulsing) and/or the back current from the motor when the spring i dragging the motor to center position and/or a stiff linkage which trigger the short circuit, reverse voltage and/or the over voltage protection on most BEC's and the processor "reset" and voila a glitch, as a regular battery pack would have just a very short voltage drop in the range of a few 1/10 of volt w/o any affect on the recoder. It doesn't matter if it's a Pic or Digispark recoder.

No BEC is the same and some works better then others, you just have to try it out but if in doubt I would recommend a 4 cell NiMh 700 mAh minimum capacity to start with.

I would caution from using LiFe mainly because the 6.6 V 2S has too low a voltage to give enough headroom to allow for current surges; meaning it can drop the voltage enough during peak current demands to cause a glitch.

This suggestion comes from Jay Mendoza:

I accidentally found that an old Rand LR-3 I was using to test my recoders had a 100 ohm resistor across the motor terminals. It did not glitch at all so I tried the same with my Tobes and viola! The glitch that had been plaguing me went away :D.


When I tested this set up I could clearly see that there were occasional voltage drop but not of such entity it would trigger the protection on the BECs thus no reset of Pic/DigiSpark however the current to the actuator would increase with a few mA
Last edited by Tobe on 30 Oct 2019, 18:58, edited 1 time in total.
Cheers,

Tobe
User avatar
Mike_K
Posts: 669
Joined: 16 Feb 2018, 06:35
Location: Hertfordshire

Re: Possible actuator problem

Post by Mike_K »

Hi Tobe

I too have suffered from the "glitching" problem with a Tobe Rand, but couldn't believe it was the Arduino resetting that was causing the problem. Therefore I've done a few tests this afternoon and can confirm that you're correct. What I did was as follows:

I programmed an ATtiny85 with Phil's recoder, and put a 5-second delay in the setup, so if it reset, the "glitch" would last at least 5 seconds. I had to wait for ages for a "glitch", but when it "glitched" it was at least 5 seconds before things got going again (previous "glitches" were only momentary). Interestingly, when it glitched, my Spektrum Rx started flashing, indicating either lost frames or a brownout.

Next, I tried separate 4S NiMH batteries for the Rx and Rand (the second battery was connected directly to the DRV8838 drive side) and I never encountered the problem. I also tried Jay's suggested resistor but used a 120R as it was the nearest I had. Again no problem.

So it looks like all we have to do is put a 100R resistor across the motor terminals and it should sort the problem.

Finally, the problem never happens when the Rand is disconnected from the control surfaces, And in my case, the control surfaces are far too stiff, I'm about to replace the mylar hinges with sown hinges. So maybe stiff control linkages play their part too?

MIke

PS what frequency has everybody been programming the ATtiny85 / Digispark, 8MHz or 16MHz?
AllenD
Posts: 11
Joined: 21 Mar 2019, 10:32

Re: Possible actuator problem

Post by AllenD »

Tobe, as suggested in our conversations, I have a 4 cell nimh battery on the way to check to see if the glitch will go away with my actuator, however I'm going to use a 350 mil pack as I'm going to have trouble enough fitting the battery and the ESC in my Fidget and I'd rather keep the weight down as low as possible. If you believe that soldering a 100 ohm resistor across the motor terminals will help, I don't mind trying that also, through I will try the fixes one at a time to try to determine which fix is really helping.

Allen
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Possible actuator problem

Post by Tobe »

Good to hear your comments Mike, hope to find you well.
Spectrum Rx are known to be more voltage sensitive than old fashion Rx and in the old days we used to add a quite large capacitor to avoid voltage surges that could reset the Spectrum Rx.
I mention this as it's my understanding also that who is flying with old equipment on 35 have not been prone to glitches as much as those flying on 2.4 Ghz.
I'm going from now on, wise of this have the option of the add on of a 100 ohm resistor across the motor terminals

I did also checked the current when having a load on the actuator and found out that if I doubled the load the current increased up to 4 times so smoothness in the linkage sure is a key factor

Cheers Mike and all the best in hope to see you again at next Ponte

Tobe
Last edited by Tobe on 30 Oct 2019, 18:23, edited 2 times in total.
Cheers,

Tobe
Tobe
Posts: 665
Joined: 16 Feb 2018, 06:19
Location: Varberg or Stockholm, Sweden

Re: Possible actuator problem

Post by Tobe »

Allen,
I sure would recommend you to first try the 100 ohm resistor not least because the GG actuator are voltage sensitive which means as voltage drops you get less throws and you will have to correct a pitch down with up trims. The voltage sensitiveness is a well know issue for all GG actuator and not typical just for the GG-Tobe. A battery pack will always suffer a voltage drop as it's discharged
Cheers,

Tobe
AllenD
Posts: 11
Joined: 21 Mar 2019, 10:32

Re: Possible actuator problem

Post by AllenD »

Ok Tobe, I'll give the resister a try.I wouldn't think voltage drop would be a problem as my ESC has a regulated 5 volt output and most flights the flight battery is still around 70 percent when I land after a six minute flight, so the ESC shouldn't have any trouble maintaining 5 volts at that point.
Post Reply