Neutral Position Adjustment Accuracy

Post Reply
qqQ
Posts: 3
Joined: Thu Sep 22, 2016 7:24 am
Product: R3
S/N: 206

Neutral Position Adjustment Accuracy

Post by qqQ »

Hello guys, just received my R3. I already had some time to test and its as great as I expected (super accurate, great software, intuitive to use). But yet, there is one thing that doesnt work perfectly and its the NPA. If i activate the SMM, press Trigger1 and then let go the Stick until the beeping is finished, it wont be perfectly centered most of the time, but always offset by about 1 percent in a random direction. Im aware that the output isnt smoothed, so there is always some jitter in it, however it does not jitter around the neutral position but another one.

I use the Sensor in the Standalone version (no TM-board) with a Warthog grip. Firmware is GRI_FW_1_00_0. The problem is more noticeable with lower force settings, however its observable even with max force setting, just not as pronounced as with lower ones.
I already reset Win calibration, loaded the Firmware again and reinstalled the drivers.

Would be interesting to hear if this is normal behaviour or if Im doing something wrong.

Thank you in advance.

EDIT: Seems like the direction isnt totally random, the offset is most times 1 percent to the left (almost always), without tilt.

EDIT2: Further testing results: Did intensive testing with tilted Stick calibration. The Stick miscalibrates to a different position depending on tilt angle (for the same tilt angle its always the same one). The problem is more pronounced on the Roll Axis, although it also exists on the Pitch Axis (tests with 13 Ibs for both Axis to be comparable). The calibrated position does not drift over time (atleast not overnight), the Stick has no problem with the Force sensor not working properly around special positions (I can not notice any difference between different tilt angles, and I applied up to approx 45 degrees), the Centre is just wrongly calibrated. Problem also exists with other firmwares. My Warthog grip is absolutely fine, no wobble whatsoever in it.

I can assure I never exceeded the Never-exceed-force (20 Ibs as far as I know), as Im flying on 5,8 Ibs for Pitch and even that requires considerable effort to reach maximum deflection. On 13 Ibs I flip my table ;)

EDIT3: Tested all firmwares now, in some of them (LCJ for example), miscalibrates to a different point, otherwise same as Gripen.
Dragon
Posts: 702
Joined: Thu Jan 10, 2013 6:03 pm
Product: ---
S/N: 0

Re: Neutral Position Adjustment Accuracy

Post by Dragon »

Hi qqQ

Sorry for the delay. I have read your post and I am not sure about your problem, would be possible, do you take contact with me? to try to understand you and help you?

Thanks in advance
Dragon
qqQ
Posts: 3
Joined: Thu Sep 22, 2016 7:24 am
Product: R3
S/N: 206

Re: Neutral Position Adjustment Accuracy

Post by qqQ »

Quick Necropost on this thread....

Finally found time for some flying again, FSSB R3 still working fine after a one year slumber. Yet, the stick still doesnt calibrate to exact center, but is a bit off after NPA. Cant remember how much it was a year ago, so a quick question to you, Dragon:

In our email conversation we had after I created this thread, you told me that this deviation exists because the FSSB R3 uses a set of discrete values in order to calibrate, thus it is possible that this set does not contain a value which will exactly create sensor readout 32768/32768 when in neutral position. So my question is: How big is the maximum possible calibration error at max force settings approximately?

Mine is off by about -30 in the pitch axis and +100 in the roll axis (on max force setting)...just wanna make sure that this means its fine.

Thank you in advance for your help

Regards, qqQ
Dragon
Posts: 702
Joined: Thu Jan 10, 2013 6:03 pm
Product: ---
S/N: 0

Re: Neutral Position Adjustment Accuracy

Post by Dragon »

Dear qqQ

I have check in a new unit in order to give you a real date from a new unit. every R3 are diferent just as are analog sensors and instrument analog amplifier, but the order must be simililar. But in the unit test the values are Roll 32651 that is -116 and in Pich 32480 that is +287.

This numbers look great than your, and you can think, your unit is better than mine, but this is not real, both are as good as all others, just as the sensibility and accuracy are same, the diferente is only that in your case, the value in your center position, is very close to the discrete value that we have inside the electronic to compare and make the "0" offset.

In my case, the 287 numbers in pitch, are about 0,5% of total deflexion, just when I put my hand in the stick I will apply a bigger amount of force, and I will not feel this small deviation from the center position. In reall word, a deviation of 2,3 5% in the center position is the value you insert when you get the grip with your hand, and this percentage is not noticiable in your arm/hand.

I think would be great some day, make a R3 convencion, in order to talk about all this kind of questions ;-) and fly all together, no?

Best Regards
Dragon
qqQ
Posts: 3
Joined: Thu Sep 22, 2016 7:24 am
Product: R3
S/N: 206

Re: Neutral Position Adjustment Accuracy

Post by qqQ »

Thanks Dragon!

Just wanted to make sure, you know, the thing was expensive... but worth every penny. R3 convention would be fun for sure, but I think flying BMS or similar in a larger group would be more realistic (would have to rust off my skills little bit though, last time in the Viper is a year ago....)
Silvershade
Posts: 12
Joined: Fri Apr 13, 2018 11:54 am
Product: R3L
S/N: 60

Re: Neutral Position Adjustment Accuracy

Post by Silvershade »

Mine is also off -1 percent to the left, so are we saying this is normal?
Dragon
Posts: 702
Joined: Thu Jan 10, 2013 6:03 pm
Product: ---
S/N: 0

Re: Neutral Position Adjustment Accuracy

Post by Dragon »

Hi. as you can calculate, if you use the raw numbers instead the percentage number, 1% is not really 1%, for example 0,500001% is normalized to 1% just as the percentage number are calculated with no decimals.

Anycase, no matter if is 1% or 5% you will not apreciate, just as this numbers are for RAW values, not for the values in the flight simulator software neutral position. and belive me, no body in this planet, would feel the diference if the neutral is positioned just in the middle or 10% around the center, the only important is that this value will be used as 0X 0Y for the simulation software.

If you use Falcon for example, in your setting control windows, every time you press center., your X and Y axles values are used to calculate the neutral position.
If you use any other sim software normally have its own neutral position adjust, and if not, is used the neutral position you calibrate in windows.

What is important and what only your R1, R2 or R3 give you, is the absolutely precise and smooth transition from this neutral position to a light, light roll or pitch adjust with a tiny force in your grip stick, this what you get with the force sensing and is not posible with any other system. This is why you can get more precision and accuracy, is not the stability in the cero or the resolution, and is this because you dont need curves or any other additional mathematic assistance.

Thanks and nice flights
Dragon
Dragon
Posts: 702
Joined: Thu Jan 10, 2013 6:03 pm
Product: ---
S/N: 0

Re: Neutral Position Adjustment Accuracy

Post by Dragon »

I have write this before, but I will explain as many times as any of you will ask for.

As you can see the center position is never just in the center 32767/8 that is because the zero adjust is not done with maths or dead zone algorithm, is done with analog electronic, moving the adjust voltage in discretes values, so, it is imposible adjust the zero just in the 32767/8 value, but is absolutely not necesary, as you will feel your control around zero position will be as good as in your car, or your bike, when you move your wheelsteering, you can control absolutely small movements, thas would be imposible, is we would apply a mechanical or mathematical dead zone.

In others words, dont be afraid if you see that your cero position is 32767 in any is 32767. You R3 is full Ok, and will work perfect.

Anycase and only because the owners ask for, we include in some firmwares, a simulation of friction in the surface control, called BFA (Break Force Adjust) in esence is a simulation of how the bike front wheel work, this will get out you some precision around the 0:0 position but we has try to make this as small as possible.


Regards
Dragon
Post Reply