Thread Tools
May 27, 2019, 02:25 PM
Registered User
Short additional information:

Same problem / error messages when I try to use the FirmwareUpdater with the A3S2 hardware (USB adapter and gyro) !?

I'm totally confused...
Sign up now
to remove ads between posts
May 28, 2019, 03:15 AM
Registered User

Problem with FirmwareUpdate solved


Folks,

after hours working on it, I could solve the problem:

There were several old instances of the CP210x drivers on my WIN 10 computer.

By means of regedit I deleted rigorously ALL occurrences of "CP210" and "silab".

Next, just insert the USB adapter provided by HobbyEagle into an USB slot. Do not attach a device! Do not install the USB driver manually, but wait until Windows has managed to install the appropiate driver by itself. It installed version 10.1.7.2399 as of 26.03.2019.

Afterwards, I started my laptop new, and everything worked fine. IF the installation was successful, the blue LED in the A3S3 USb adapter should be on permanently, IF you have selected the correct COM port in Updater.

Some other issue I don't know the real importance of: I'm using my laptop on two different docking stations (@home and @work). Now, when using the dockingstation at work, it appeared that the COM port 6 the driver was assigend to in my home is already used by some other device! Now it is COM port 12.

Hope all this helps somebody with the same problem^^.

Cheers

Torsten
Aug 25, 2019, 11:02 AM
Registered User
deadbug's Avatar

Firmware Bug - AUTO LEVEL STILL NOT FIXED


The following is copied from my letter to Motion RC regarding the problem with the A3S3 Auto-Level:

I just finished my 20th flight on my F-35 with the Hobby Eagle A3 Super III (A3S3) and wanted to get my latest and final test results to you for coordination with Hobby Eagle tech support to get to the bottom of the problem with the Auto-Level mode.
First of all to recap from Day 1:

1. When I decided to purchase the HE A3S3, I really wanted a reliable gyro that not only would stabilize the F-35, but one that would have a reliable, rapid and accurate auto-level or recovery mode as this model can be very difficult to keep in sight and easy to lose orientation.
I received the gyro and RX from MRC before the airplane arrived, so I used my time to upgrade the firmware to v 2.2 (the latest) and to program it carefully per the instruction manual and MRC videos. You can check my programming with the returned A3S3 under RMA item that was put in the mail today.

2. The maiden flight was conducted entirely “gyro off”. Second and subsequent flights have been with gyro “on” with sampling and testing Auto-Level Mode” on every flight. The 3-switch positions are: Gyro Off; Gyro On and Auto-Level. The basic gains are set at 30% - Ailerons; 40% Elevator and 40% - Rudder. Master Remote Gain is set to 100%. Auto Level Gain set to 75%. Level calibration was done and set to 5 degrees nose up.

3. The normal gyro mode is perfectly stable – period. The only negative is that in the normal gyro on mode, roll rate is dampened a lot compared to gyro “off” mode. i.e.—full stick gives a verrrry slooooow roll rate until I turn the gyro off.

4. Auto level testing results were interesting to say the least…every possible unusual attitude was tested on the ground with the airplane powered and the data cable connected to my PC and the A3S3 with the software in my PC running and connected. All normal gyro functions were tested with correct control inputs. When placed in Auto Level mode, the control response to each scenario seemed absolutely proper. A positive roll to the correct wings level upright position and then elevator as appropriate to recover from the climb or descent.

5. Inflight testing was a failure…Roll commands for the most part were correct. Pitch commands during the entire recovery were incorrect and many times during the A/L commanded “recovery”, I had to dramatically intervene to prevent a loss of my model. Several times I was uncomfortably close to not being able to recover. Nose high/wings level attitude—then Auto Level: airplane usually rolls to wings level correctly, but then very slowly, if at all…goes to a shallow descent. Nose low, wings level—aircraft does not even attempt to level off and continues to increase the dive angle more and more till I intervene.
Nose high/wings level/INVERTED – pitches the nose even higher while inverted and is slow to level wings. Wings level/INVERTED…usually rolls quickly to level upright.

6. After discussions with you, I tried Trainer Mode with identical results. Failure of A/L.

7. Swapped out new replacement A3S3 to see if it was a faulty unit. Test flew—same problems with A/L. Failure of A/l.

8. As a final test, I rolled back the f/w to v2.1—even flashed it twice. Recalibrated the accelerometers just to be safe (3 times). Reinstalled in aircraft. Test flew today. Yep—same unacceptable performance of the gyro in Auto Level. Failure.

There is a problem with the program in this gyro. <snip...>. I think at this point HE needs to become involved and fix their product.
Aug 25, 2019, 10:32 PM
Registered User
Go4it's Avatar
Quote:
Originally Posted by deadbug
The following is copied from my letter to Motion RC regarding the problem with the A3S3 Auto-Level:

I just finished my 20th flight on my F-35 with the Hobby Eagle A3 Super III (A3S3) and wanted to get my latest and final test results to you for coordination with Hobby Eagle tech support to get to the bottom of the problem with the Auto-Level mode.
First of all to recap from Day 1:

1. When I decided to purchase the HE A3S3, I really wanted a reliable gyro that not only would stabilize the F-35, but one that would have a reliable, rapid and accurate auto-level or recovery mode as this model can be very difficult to keep in sight and easy to lose orientation.
I received the gyro and RX from MRC before the airplane arrived, so I used my time to upgrade the firmware to v 2.2 (the latest) and to program it carefully per the instruction manual and MRC videos. You can check my programming with the returned A3S3 under RMA item that was put in the mail today.

2. The maiden flight was conducted entirely “gyro off”. Second and subsequent flights have been with gyro “on” with sampling and testing Auto-Level Mode” on every flight. The 3-switch positions are: Gyro Off; Gyro On and Auto-Level. The basic gains are set at 30% - Ailerons; 40% Elevator and 40% - Rudder. Master Remote Gain is set to 100%. Auto Level Gain set to 75%. Level calibration was done and set to 5 degrees nose up.

3. The normal gyro mode is perfectly stable – period. The only negative is that in the normal gyro on mode, roll rate is dampened a lot compared to gyro “off” mode. i.e.—full stick gives a verrrry slooooow roll rate until I turn the gyro off.

4. Auto level testing results were interesting to say the least…every possible unusual attitude was tested on the ground with the airplane powered and the data cable connected to my PC and the A3S3 with the software in my PC running and connected. All normal gyro functions were tested with correct control inputs. When placed in Auto Level mode, the control response to each scenario seemed absolutely proper. A positive roll to the correct wings level upright position and then elevator as appropriate to recover from the climb or descent.

5. Inflight testing was a failure…Roll commands for the most part were correct. Pitch commands during the entire recovery were incorrect and many times during the A/L commanded “recovery”, I had to dramatically intervene to prevent a loss of my model. Several times I was uncomfortably close to not being able to recover. Nose high/wings level attitude—then Auto Level: airplane usually rolls to wings level correctly, but then very slowly, if at all…goes to a shallow descent. Nose low, wings level—aircraft does not even attempt to level off and continues to increase the dive angle more and more till I intervene.
Nose high/wings level/INVERTED – pitches the nose even higher while inverted and is slow to level wings. Wings level/INVERTED…usually rolls quickly to level upright.

6. After discussions with you, I tried Trainer Mode with identical results. Failure of A/L.

7. Swapped out new replacement A3S3 to see if it was a faulty unit. Test flew—same problems with A/L. Failure of A/l.

8. As a final test, I rolled back the f/w to v2.1—even flashed it twice. Recalibrated the accelerometers just to be safe (3 times). Reinstalled in aircraft. Test flew today. Yep—same unacceptable performance of the gyro in Auto Level. Failure.

There is a problem with the program in this gyro. <snip...>. I think at this point HE needs to become involved and fix their product.

Have you tried trainer mode to see if you get the same results as in auto self-level mode?
Aug 26, 2019, 04:11 AM
Blue Skies
hifinsword's Avatar
Quote:
Originally Posted by Go4it
Have you tried trainer mode to see if you get the same results as in auto self-level mode?
See deadbug's item #6, Trainer mode tried.
Aug 26, 2019, 07:20 PM
Registered User
Go4it's Avatar
Quote:
Originally Posted by hifinsword
See deadbug's item #6, Trainer mode tried.
My bad for missing number 6.
Anyhow on the last 6- A3S3gyros, I have had the same problem when trying to use Self-level mode or trainer mode. I have my transmitter set for using 6 different modes. I fly warbirds and jets not really into 3D. But with that said I do use 3D mode to fly High alpha on a few jets. Low and slow.
I, like you mentioned, have tried several times to set auto level but can never trust it. And if I do try this mode, I fly very high to save the airplane. If I have a plane that I really do want to use auto self-level mode, I use an A3SII gyro. It seems to always work from me.
Aug 26, 2019, 08:36 PM
Registered User
deadbug's Avatar
Thanks Go4it. I just wish Hobby Eagle would fix this issue...it really bothers me that things have been going on so long with a flagship product, and nothing is being done to fix this.
DB
Aug 26, 2019, 09:08 PM
Registered User
Go4it's Avatar
Quote:
Originally Posted by deadbug
Thanks Go4it. I just wish Hobby Eagle would fix this issue...it really bothers me that things have been going on so long with a flagship product, and nothing is being done to fix this.
DB
The factory started this thread but I do not know if they still monitor it as they did at the beginning.
You might want to contact PeterVRC Australia, VIC, Melbourne to see if he can shed some light on it. I don't fly as much as I used to.
Aug 28, 2019, 05:08 AM
Registered User
Hi Kallend - this is timboh49 from Perth Australia

I have just crashed my T-28 - it had an A3 Super 3 V2.0 fitted and I was amazed to read in this forum that it behaved exactly as you described - i.e. flying normally then crazy pitching - it was relatively stable in the roll axis - I was unable to regain control and the plane crashed!

Did you get any response or resolution from Hobby Eagle on this problem? I am looking for answers! Can you help?

I have previously experienced this same problem on a Super II - I totalled a Freewing A-10 - I upgraded to the Super 3 's hoping that they would fix this jinx.

In both planes I (foolishly) did not have failsafes set and I was starting to think it may have been caused by poor or lack of signal strength - I was using a Flysky FS-i10 TX and FS-iA10B RX using AFHDS 2A protocol with telemetry - it's meant to be a very good radio!


Any ideas or solutions will be very much appreciated

Cheers Tim
Aug 28, 2019, 06:32 AM
Blue Skies
hifinsword's Avatar
I had thought about getting the A3 S3 V2 but it seems the SW still isn't ready for prime time. Too many unexplained crashes being reported. I had a problem with the A3 S2 and it not liking throws over a certain amount. Maybe that is causing some of the crashes? In the A3 S2, if you are over the throw limit, the controls will revert to neutral after a second or two of max throws. It could still be the same in the A3 S3 V2?
Don
Aug 28, 2019, 06:42 PM
Registered User
What do you mean by 'over the throw limit' ??

The INPUT values??
Aug 28, 2019, 07:28 PM
Registered User
mshagg2's Avatar
Presumably we need to set up some headroom in the travel limits which is above the maximum servo travel achievable with stick inputs, so that the gyro can apply some additional input above and beyond this if it deems this to be necessary for a corrective movement, without triggering the upper travel limit (which don suggests prompted a momentary return to neutral on the A3S2).

A difficult hypothesis to test on the bench.
Aug 28, 2019, 08:07 PM
Registered User
I couldn't make sense of that post.....

RC systems have a 'normal' range. eg the 800 -2200uS maximum. But there is no solid standard.
The more standard range taken as '100%' is 900 - 2100uS.

Any device taking an input should deal with this 'variable' range possibility.
In Analogue servos it is difficult if not impossible to 'control' an Input - so to deal with these variables they design it to allow and cope with the maximums possible, and even allowing a bit more to be sure. But in their case it is easy enough to just allow that greater range and travel.

In a digital system you can 'control' an Input - by setting a limit that you truncate the value. eg Nothing under 900uS does any more than hit 900uS in their internal outputs. It is capped/clipped.. And the same for 2100uS.... nothing above that does any more than set 2100uS internally.
But if they DO NOT do that, then they could encounter issues when higher values than expected, and programmed for, are input to it....

It is a pretty standard thing to set those limits!! Standard programming.
If they did NOT do that, then there is either a bug, or bad programming.....(bad programmer)

Once you have such an 'overrun' case occur, who knows what the following programming will do when it doesn't handle the overrun. Anything from 'nothing' of note that matters, to some major bug/glitch sending it off seriously astray.
The need to be doing this limting has been known 'forever', and I doubt HobbyEagle don't know its importance and possible detriment also. Though maybe they overlooked it.. or made a 'mistake'....

I have not checked this aspect out. I do tend to use the 'wider' ranges of 800 - 2200uS in Channels, and I have my OpenTX using that by default. So any Channel I use is going to have that 800-2200uS range... and I have never reduced it for any Flight Controller.
I at least know the Eagle Tree Guardian does clip it at 900-2100uS..... because they outline that in their Manual. I expect that HobbyEagle also do.....
I recall looking into the A3S2 PC setup program and the Limits areas long ago, but I don't recall seeing anything causing any issue in the defaults I was using/inputting to it.

I have three (more?) A3S3's, but only one in use flying (one other 'dormant' in a plane and I never use it - so far). The one being used only uses the Trainer portion of Angle Mode control (not Auto Level), because I want the fixed Angle Limits. It is also has 3D-lock Mode set up and I think that acted odd - but I have not tested it out in detail as it is not a jet you want acting 'weird'!!
I need to dig out a 'simple' expendable prop plane to try these A3S3 things out in...... but my best candidate for that needs a bit of 'fixing' first!
Aug 28, 2019, 08:34 PM
Registered User
mshagg2's Avatar
Apologies, probably poor paraphrasing on my part. The reference is to servo travel limits (clips) in the gyro (comparable to the outputs page on OpenTX, rather than inputs).

From the manual (p24):
Travel limits can be varied from 0% to 150% in each direction of the servos. The default setting is 130%. It’s used to protect the linkages, for example, if you have set the stick rate to 100% for aileron, after superimposing the correction of the gyro, the final amount of the movement of the aileron servo(s) may probably exceed 100%, which could cause damage to the linkages and pushrods. But after setting a certain limits the total movement of the servos will be limited in a safe range. The travel limit setting is not available for the pass-through channels.
So, what it's saying here is - the gyro might (will) attempt to move the servo above and beyond what you are inputting via the stick movements, if it deems this necessary. If the maximum travel limit set in the A3S3 software aligns with the maximum throw you can achieve with stick input, then the gyro doesnt have any additional travel it can apply, because it will already be up against the travel limit. Hence, you need some headroom in those travel limits above your maximum stick inputs.

If i'm reading his comment correctly, Don suggested in a prior model that a similar scenario resulted in the gyro resetting servo positions to neutral momentarily.
Aug 28, 2019, 10:02 PM
Registered User
It doesn't really work the way it seems you are thinking....

Your Inputs are now 'commands' not direct servo control anymore (in some Modes). And they do processing (maths etc) to create an Output for the Servos - which has no direct relationship to your TX values that were sent to the A3. Yes related, but not directly linked/related.
So now THEIR system (A3) then needs to know the servo's and aircraft's physical limitations. It is the exact same thing as your TX output limits does.... but with any devices the FINAL Output device needs to be set up to get that right.

So.... if you send in 900 -2100uS.... they do their maths and create whatever..... THEIR Limit Settings are set (need to be set) so they only Output what the Servo/Aircraft can deal with. eg possibly that same 900-2100uS range maximums.

If your INPUT sends over the 'normal' values, they SHOULD be Clipping them anyway.....
In a way you could view it as 'There is some sensible person in there that watches and makes sure it is all done 'right' (this is the programming/algorithms) - dealing with all forms of inputs, and creating correct outputs, based off the output parameter limitations they asked you to set up. (they do the input stuff themselves - no need for parameters set on that).

This Limits setting is what stops it over-driving your servo/aircraft set up.
This part has nothing to do with them creating 'bad actions' due to INPUTS being over-range. Which is what has been suggested to be causing the weird AutoLevel cases.
Our only control over that..... IF is is occurring due to over-range INPUTS.... is make sure that your TX OUTPUTS do not go over range - via the Output Limits. In OpenTX that would mean Output limited to 90% or less - I am not sure the exact value that will create a 900-2100uS range.

If it really is the case, I would expect that even in STATIC testing, any over-range inputs would make it go nuts.
But it sounds like it is occurring on NEUTRAL TX Sticks anyway..... expecting Auto Level, but it corrects incorrectly.
This means it is more likely an "Internal Maths"/programming issue/bug.
(plus how it is UNLIKELY that H.E. don't know how to deal with INPUTS, and their over-range cases, correctly. I expect they DO.)


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion HOBBY EAGLE New A3Super3 known as A3S3 questions and answers Go4it Electric Plane Talk 670 Aug 14, 2019 01:31 PM
Question What Hertz set for Hobbyeagle A3-L gyro JIMA Radios 0 Mar 14, 2017 11:44 AM
Idea Hobbyeagle Super A3 Gyro for Tail Sitters PaulB VTOLs 1 Oct 20, 2016 04:05 PM
Question Question, Suggestion, Kudos, Complaint, and Bug! FourDollaRacing Site Suggestions / Complaints 2 Apr 14, 2016 06:27 PM
Discussion Hobbyeagle A3 Pro SE gyro for aircraft 4stripes Electric Ducted Fan Jet Talk 10 Aug 02, 2015 06:33 PM