Angular hall sensors.

Interested in building your own home grown cockpit?

Moderators: Uros, OXO

Post Reply
mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Angular hall sensors.

Post by mhs » Sat Sep 05, 2020 4:09 pm

Having found most commercial stick/rudder units either expensive or poor sensitivity I constructed some using Allegro a1301
linear hall sensors and BU0836A interface. I had to purchase some small ball bearings and magnets but the rest was sourced
from the recycle pile at the back of the garage. The hall sensors have proved reliable over about 25000km flying but are difficult
to set up as the magnets are direct drive from the stick and rudder axes so required some experiment with magnet strength
to get full voltage swing with small angular displacements.
It is now time to move on to a full seat simulator rather than just the stock rudder units and I have found the Allegro1330 direct
angle sensors becoming available in experimenter quantities, these can provide full swing over a small angle input and are not
sensitive to magnet strength, they are also flat so can be used to build a small sensor unit. The only problem is they require
programming for the small angle range, commercial programmer is expensive, but it looks to be in range of a DIY unit
driven by a small embedded computer.
Has anyone attempted to use these sensors rather than the a1301/2 linear sensors?
I am about to purchase some and start the experiment.
Image

User avatar
wickid
Posts: 2429
Joined: Mon Dec 04, 2006 7:32 pm
Location: Venlo, NL
Contact:

Re: Angular hall sensors.

Post by wickid » Sat Sep 05, 2020 4:13 pm

I used throttle by wire sensors from BMW cars. Maybe you can get them at a scrapyard. They don't lose calibration at all.
PH-1504, KOE

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors.

Post by mhs » Sat Sep 05, 2020 7:23 pm

The throttle unit probably has a similar sensor, they are becoming widely used in cars which is why the
chip is becoming available. Until recently I could only get them with a large minimum order but now
around 1.50GBP in small quantities. The idea is to glue the magnet onto the control axis bolt and mount
the chip in a small plastic disc which can be rotated to zero the axis. Output is analogue and feeds straight
into a BU0836A . Throttle sensor on my car is CAN bus output and this is now common so modification might
not be easy. Car scrapyard might be good source of wiper or door window motors for force feedback if using
simple load up on airspeed method.
Image

sisu1a
Posts: 337
Joined: Sun May 04, 2008 9:07 pm
Location: Bay Area, California (USA)

Re: Angular hall sensors.

Post by sisu1a » Sat Sep 05, 2020 8:36 pm

mhs wrote:
Sat Sep 05, 2020 4:09 pm
Having found most commercial stick/rudder units either expensive or poor sensitivity I constructed some using Allegro a1301
linear hall sensors and BU0836A interface. I had to purchase some small ball bearings and magnets but the rest was sourced
from the recycle pile at the back of the garage. The hall sensors have proved reliable over about 25000km flying but are difficult
to set up as the magnets are direct drive from the stick and rudder axes so required some experiment with magnet strength
to get full voltage swing with small angular displacements.
It is now time to move on to a full seat simulator rather than just the stock rudder units and I have found the Allegro1330 direct
angle sensors becoming available in experimenter quantities, these can provide full swing over a small angle input and are not
sensitive to magnet strength, they are also flat so can be used to build a small sensor unit. The only problem is they require
programming for the small angle range, commercial programmer is expensive, but it looks to be in range of a DIY unit
driven by a small embedded computer.
Has anyone attempted to use these sensors rather than the a1301/2 linear sensors?
I am about to purchase some and start the experiment.

Do the same thing but switch to Alegro A1324 sensors... it's a shame the 1301 and 1302 were documented by early pioneers and became the goto sensor because they're the poorest performers for joysticks due to the limited ranges of motion available and need for powerful magnets. 1301 has low sensitivity so needs really strong magnets and req 180deg for the swing. 1301 is 2.5mv/Gauss unit

The 1324 is the same form factor but 2x more sensitive coming in at 5mv/Gauss. It can use much smaller magnets and also has a 90deg swing which work quite well for joysticks... been using them now since 2012. Here's an album I made about using them to make sealed Hall pots with off shelf components.. I initially started with 1301, then 1302 but have replaced them with superior 1324. I started off following the well known skateboard bearing/bic pen method but refined it to a much smaller form factor https://imgur.com/a/WkQtw

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors.

Post by mhs » Sat Sep 05, 2020 10:22 pm

I am currently using a the a1301 sensor with magnets on the axis bolt, works well as the operating angle is
small enough to be approximately linear (small angle approximation) but the sensor is quite long. I am looking
to build a cockpit with stick between knees so would like a low profile sensor. The a1330 is 3x4mm and around 1mm thick
and will sit at right angle to the axis so should allow a thin stick mechanism. The sensor will also do a linear 360 degree output
for a belt driven pulley air brake or flaps set-up with linear mechanisms which is not possible with the linear sensors.

Pictures of current prototype, must get on with full cockpit but having too much fun flying.
You do not have the required permissions to view the files attached to this post.
Image

User avatar
OXO
Condor Team
Posts: 6360
Joined: Mon Feb 13, 2006 1:08 am
Location: France 42
Contact:

Re: Angular hall sensors.

Post by OXO » Sun Sep 06, 2020 9:52 am

If you used the A1308 instead, it gives an angle output. That would appear to be much more suitable for your purpose.
Chris Wedgwood,
Condor Team

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors.

Post by mhs » Sun Sep 06, 2020 11:41 am

The a1301/2 can be used with a diametric magnet by mounting them parallel to the magnet but the
result is dependent on the chip to magnet separation so output can be unreliable. I had to use the
option of chip edge on between two magnets to get reliable results resulting in a sensor about 2.5cm high.
The a1308 is also a linear sensor so I would expect the same problem.
The advantage of the a1330 is that it does a direct measurement of the magnet angle and is not sensitive
to the field strength so is not sensitive to spacing and can also do a full linear output over 360 degrees but
can be sub-ranged to full output over an angle as small as 12 degrees. With the a1330 geometry I should be
able to stick a thin magnet on the end of a control axis and the chip only adds another 2mm. The 8 pin TSSOP
package only uses three pins and I intent to just solder the wires and probably pot the chip in a circular resin disk.
Image

User avatar
OXO
Condor Team
Posts: 6360
Joined: Mon Feb 13, 2006 1:08 am
Location: France 42
Contact:

Re: Angular hall sensors.

Post by OXO » Sun Sep 06, 2020 12:12 pm

Ahh, typo, I meant to say A1330.

https://www.allegromicro.com/en/product ... sensor-ics


Thats what comes from not reading the whole thread :?
Chris Wedgwood,
Condor Team

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors.

Post by mhs » Sun Sep 06, 2020 1:50 pm

The A1330 was the one that I was proposing to use, they ship from USA by courier so shipping
is expensive, unless you qualify for free shipping. I have ordered 25, and have some magnets on order,
so I should have enough to cover experimental error while attempting to program them with a home
made programmer, there seems to be sufficient information in the data sheet. I will post results when available.
Image

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors A1330

Post by mhs » Sat Sep 26, 2020 8:07 pm

I have received the Allegro 1330 sensors, the chip is very small it is impossible to directly
connect wires so the chip must be mounted on a breakout board. Sourcing suitable magnets
was not easy and I recommend getting the ones from the chip vendor as they are matched to the
chip requirements. I have mounted some of the chips and wired with 3 pin socket wire obtainable from
RC model suppliers, these will plug directly on to the BU0836A pins. I have constructed a test rig to
rotate the magnet over the chip to check the set-up works, check the magnet gap tolerance and any
magnet shunting effect of the A2 stainless axis bolts. There is some magnetic attraction to the A2 stainless
so I will probably use brass bolts in final assembly. As supplied the chips provide a linear voltage ramp from
0.25 to 4.75 over a 360 rotation (5v supply), next step is to programme them for short stroke. Programming
hardware and software is under construction using an Arduino. Potting the chip into a mounting is also
under consideration.
chip_mounted.jpg
chip_board.jpg
The background grid squares are 5mm.
You do not have the required permissions to view the files attached to this post.
Image

Allen
Posts: 42
Joined: Tue Dec 04, 2018 10:42 pm

Re: Angular hall sensors.

Post by Allen » Thu Oct 15, 2020 6:01 pm

Re. the use of angular sensors:

I gave up trying to use these. All the stick and rudder functions end up in a linear motion in a glider, and so I used the 'Magnetopots' from DigiKey:

https://www.digikey.co.uk/en/product-hi ... ntiometers

Very, very easy to use, a wide range of lengths to suit the linear actuation - the whole length can be used, giving the full resolution of the 10K ohm potentiometer (or whatever you choose). I used a bent bit of spring steel to operate the internal magnet.

These have been used for over a year on our 5 projector screen, two seat fuselage, touch screen operated (no keyboard or mouse) simulator at Essex Gliding Club.

Allen

P.S. We had to build the fuselage as well - no volunteers to crash their two-seat glider for us. :roll: :roll:

mhs
Posts: 41
Joined: Thu Nov 28, 2019 8:52 pm

Re: Angular hall sensors.

Post by mhs » Thu Oct 29, 2020 1:58 pm

I have previously had problems with mechanical potentiometers (but have not tried the membrane pots) due
to wear with most movement over a small range, also mechanical backlash. I suppose a linear sensor makes
sense for a two seater with the controls linked but for just a stick/rudder I have found the easy mechanical solution
is a rotation sensor on the pivot axis. I do not have access to extensive workshop facilities so I make everything out
of plywood using just hand tools. The linear film potentiometers are probably good for flaps/air brake linear actuators
but would have to add extra linkage to convert stick/rudder to linear. My current version of simulator is just separate
stick and rudder units (pictures in this post), air brake is on a quadrant mechanism and flaps are just up/down switches. I am currently
contemplating a home build cockpit (in plywood, and will not look much like a glider) with seat and detachable rudder unit
(for reducing storage space), single screen (with head tracker) on a small table over rudder unit and computer alongside so that keyboard
is available until I get everything built in, I do not have space for a large set up.
Haptic feedback is under experimental development using magnetic motor unit similar to disk drive arm actuator, this will drive
directly onto the stick and possibly rudder.
Image

Post Reply