openrcforums

Would it be possible to fit these new gimbals to a 9XR Pro, and if so what modifications would be needed.

Dave

Since Taranis normal gimbals have been fitted, then these may well fit with minor mechanical fiddling. Electrically they look to be compatible.

Mike.

Thanks Mike, as soon as they are in stock I will order a pair and give it a go. I will report back after I fit them.

Dave

Take care to check the supply and ground connections. The pots on the existing gimbals are OK if you get the power and ground swapped, but I suspect the hall effect devices would get damaged if that happens. (And don't rely on the existing wire colours, they seem to be a bit "random").

Mike.

Looking at the new M9 Hall Effect Gimbal for the Taranis - X9D and Taranis Plus - X9DP radios,

Specifications:
· Supply Voltage: DC 3.0~±0.5V
· Sensitivity: 2.50mV/G
· Linear Output Voltage Range: 0.1~(VCC +0.1)V
· Quiescent Output (TA=25°C, B=0G): 0.5 x VCC
· Noise (no load): ≤ 40 mVpp

The specification of Supply Voltage: DC 3.0 ~±0.5 Volts is to low to be used with the FlySky and Turnigy 9x radio's,
as they use 5 volts, two volts higher.
Could we lower the 5V down to 3V so these M9 Hall Effect Gimbal could be used ?
See more info here: (http://alofthobbies.com/m9-gimbal.html) and here (http://openrcforums.com/forum/viewtopic ... 44#p123344)
and here (http://openrcforums.com/forum/viewtopic ... 16#p122893) and (http://openrcforums.com/forum/viewtopic.php?f=95&t=9977)
and (http://openrcforums.com/forum/viewtopic ... ls#p132195).

Bob B.

Note that the SKY board, AR9X board, 9Xtreme and 9XR-PRO are all 3.3V based and only supply 3.3V to the gimbal pots.

Mike.

Thanks Mike, I was thinking (trying to remember) that these other 9x radios use (have) the lower Gimbal voltage.
So this new M9 Hall Effect Gimbals will work with these (SKY-Board, AR9X-Board, 9Xtreme and 9XR-PRO) 9x radios. (COOL! )

If someone wanted to upgrade to the M9 Hall Effect Gimbals in the FlySky and Turnigy 9x radios,
what could they use or need to use to lower the 5Vs down to 3Vs ?

Would we need something for each of the 4 Gimbal Stick circuit ?

A single 3.3V regulator should be OK to power all 4 gimbal circuits. There is a downside, the analog input to the processor will still handle up to 5V, so you wont be using all available 10 bits of the conversion (0-1023), only 0-682, so less resolution.
The other radios mentioned not only work from 3.3V, but also have a 12-bit converter.

Mike.

Thanks Mike,
With the FlySky and Turnigy 9x radios with there 5V Gimbal circuits would not be a good upgrade,
because we would be loosing a high amount of resolution (341, is a lost of about 1/3 of the resolution).
10 bits down to 6.6 bits, would this, in actually = 6 bits only or can it use the .6 bits ?

Edit: MikeB reply about this: (viewtopic.php?f=122&t=9683#p123482).

Bob B.

Add Edit Feb 26, 2917 M9 Gimbals will work with the SKY9x, AR9X, 9Xtreme and 9XR-PRO radios:

MikeB wrote: ↑Fri Jan 20, 2017 10:33 pmNote that the SKY board, AR9X board, 9Xtreme and 9XR-PRO are all 3.3V based and only supply 3.3V to the gimbal pots.
Mike.

Also make sure when installing M9 Hall effect Gimbals, that you know which wires are which, as you could have Magic Smoke (A BIG PROBLEM).

MikeB wrote:Take care to check the supply and ground connections.
The pots on the existing gimbals are OK if you get the power and ground swapped,
but I suspect the hall effect devices would get damaged if that happens.
(And don't rely on the existing wire colors, they seem to be a bit "random").
Mike.

Mike, what about powering the processor from a 3.3V regulator? Could the Atmega handle 16Mhz at that voltage?

João

Not per datasheet, 4.5V min.

Thanks Kilrah.. It could be tested though. The m328 we use on other projects, also according to the datasheet, is not able to oscillate at 16Mhz when powered by 3.3V, but it does work. The multi module and the oXs are proof of that. The chips are different though.
I have a couple 9x boards that I could power from 3.3V and check if they work ok Might have to bypass the 3.3V regulator for the test, and feed 3.3V to a 5V pad.

João

I wouldn't trust running at 3.3V. I don't like running the '328 at 16MHz on 3.3V. While it works, it is not guaranteed. Go flying on a cold day and you could have problems!

Mike.

Hmm, I have a couple DIY Frsky compatible receivers with m328 running on 3.3v at 16Mhz. I also have loads of cold days here at the moment. I think I could test this

João

But you will still only be testing a couple of samples. Keeping to the spec. means all processors will work OK across temperature, voltage and manufacturing variations.

Mike.

Mike, you are absolutely right. One chip working out of specs doesn't mean all will work. The m328p seems to be handling really well though. But this thread is not about that chip, it is about the 9x chip(s).
So, the only option would be to use an amplifier circuit on each stick? I remember you have designed such a circuit a few years ago..

I have some spare 9x boards that were repaired. I can still go ahead and test the 3.3V if people find it interesting, even if just for the sake of it, and if it is not a waste of good time that I could spend doing something else

João

I was thinking about it, too. Losing 30% of resolution isn't an option at all, in my opinion. So 9x users will have to use opamps between the sensors and the mainboard. But then, given the price of AR9x, it is probably better to replace the mainboard rather than put four new opamps...

It's not that I try to advertise AR9x, but this board gives so much more advantages over the stock mainboard! I would certainly recommend this upgrade to anyone looking to install M9 gimbals in a 9x.

bob195558 wrote:10 bits down to 6.6 bits, would this, in actually = 6 bits only or can it use the .6 bits ?

You are not down to 6.6 bits more like 9.3 bits.

We use some over sampling with averaging to turn the 10 bits into 11 bits already.
I just tried moving a stick on a Taranis (12 bits analog), and I'm struggling to get the output to change by less than 0.3%.
I wouldn't be surprised to find a resolution of 680 steps on a stick is actually OK.
When I put a 'scope on the PPM output of my Hitec Eclipse 7, I found the pulse width changed in steps of 2.5uS. Since full movement of the stick in one direction only changed the pulse width by 400uS, this was only 320 steps across the full range of stick movement.

Mike.

wow Mike, the 9x has better resolution than the eclipse 7?

João

As I understand, sticks resolution is challenged more often by so-called "pinch flyers". They prefer longer levers, and the longer the lever is - the easier it is to "feel" the lack of resolution.

I read a discussion some time ago with someone complaining about sticks resolution after he moved from thumb to pinch flying on his 9x (he was a heli guy, if I'm not mistaken). In fact, he wasn't really complaining, but he said he could easily move the stick by just just 1 step of resolution (0.2%, if I got the math right) if he wanted to. Can't find that post right now though...

He must have put his radio at his feet, and extended the sticks by about a meter

João

andrewju wrote:I was thinking about it, too. Losing 30% of resolution isn't an option at all, in my opinion. So 9x users will have to use opamps between the sensors and the mainboard.

DC amplifiers introduce new problems such as temperature and offset drift. I would not do this.

Has anybody tested how much resolution is actually lost if you use a voltage divider to step down the 5V to 3.3V?
I wonder if the loss would be even noticeable.

Stefan

This thread is for the 9XR-PRO, are you using a 9X or 9XR with and Atmel AVR processor?

Mike.

Hi Mike,
I'm building my own transmitter based on a small AVR dev-board with an ATmega128a.

Stefan

Little suggestion here

Look at ar9x.net and check the ArUni board if you want to build your own transistor.
You would be able to have voice, telemetry, haptic, etc, right away.
You would also run ersky9x instead of er9x. The ArUni board has much more flash, eeprom memory and RAM. The AVR chips at now quite obsolete.. I do still have a radio with an m2561 though

João

EDIT: Ahh, and you wouldn't need any voltage divider on the gimbals because the ArUni already works at 3.3V..

Hi Joao,
thanks!
I already had a look at the ArUni, but I am building a One-Hand-Transmitter (much like the Becker-fm s48mc), so the ArUni board is way to large

Stefan

To bring this thread back up after more than a year:

Yes it is possible to use these gimbals in a T9X without loosing any resolution. You just need a linear 3.3V converter. You need to run all gimbals and potentiometers at 3.3V and supply the AREF pin of the Atmega with 3.3V. This way you get the full resolution of the ADC.

I just tested this by plugging a receiver into a FC and looking at the RX-Output in Betaflight. No matter how hard I tried I couldn't see a difference between my core9x-prototyp with M9s and my taranis x9d (with M9s). (Of course this isn't a scientific method)