Searching \ for '[EE]: DIY force sensitive joystick' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=diy+force+sensitive
Search entire site for: 'DIY force sensitive joystick'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: DIY force sensitive joystick'
2015\07\31@214728 by Mike Hord

picon face
Hi all-

A complaint I've long had about video game joysticks is the relatively high
deflection
they have during use. I'm not much of a gamer, but lately I've been sucked
into the
re-release of the old Star Wars X-Wing series games.

I'd *like* to try and DIY a low-deflection force sensitive (not The Force,
mind you)
joystick, but I'm kind of having a hard time figuring out how to do this.
There are
commercial options but they are $200+.

I've toyed with the idea of strain gauges but they tend to be both pricey
and finicky.
I'm kind of trying to wrap my head around a way to use a force sensitive
resistor <https://www.sparkfun.com/products/9376> to
get a measure of the force being applied, then some other means to
calculate the
vector.

Any ideas? Bonus points if it uses mainly parts from SparkFun.com; I can
use it as
an excuse to write a tutorial or blog post and get the parts for free. ;-)

Mike
-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.


'[EE]: DIY force sensitive joystick'
2015\08\01@044635 by RussellMc
face picon face
You can get some extremely good strain gauges in digital kitchen scales -
and some very poor ones. Odds are even the poor ones will be good enough.
By mechanically biasing one you get two axis (and may be able to use it
over a useful range without biasing.

I have seen various kitchen scales with poor accuracy, poor linearity and
quite often poor temperature sensitivity.

Quite some years ago (10+?) a local supermarket chain endlined some scales
with 2kg max reading and 1 gram resolution.
A test showed that a sample was able to resolve added or removed 2g test
weights (coins) anywhere across the range and that accuracy and resolution
allowed it to be used as a coin counter for as many hundreds of coins as I
has to hand. ie extremely good for what they were. Temperature shift was
essentially zero when heated to well over 50C - again, amazingly good.
AFAIR they have an extra gauge apart from the main bridge quad for
temperature compensation.
They sold for under $10 each as I recall. I and a friend bought the dozen
or so available and I've been 'using them up' ever since. I'll be sad when
I destroy the last one.

If a cheap looking low priced scale can achieve the above performance 10+
years ago a few hours spent in a few Kitchenware departments may reveal a
modern equivalent. (You'll need a source of warm air for likely candidates)..

You may be able to access the amplified analog signal in such devices, but
even if not adding an instrumentation amplifier per channel is affordable.




Russell
-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@053546 by Peter

picon face
Precedent: https://en.wikipedia.org/wiki/Pointing_stick
(there are others but 2 orders of magnitude more money, typically LVDT or
resolver based)


-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@092824 by alan.b.pearce

face picon face
> If a cheap looking low priced scale can achieve the above performance 10+
> years ago a few hours spent in a few Kitchenware departments may reveal a
> modern equivalent. (You'll need a source of warm air for likely candidates).

They will have been made accurate as they are used for trade, and so will be tested periodically by Weights and Measures (not sure which govt. department they come under in NZ these days), hence the accuracy and linearity. Because they are used for trade purposes they have to be accurate to ensure the customer isn't short changed.

Kitchen scales are a case of 'near enough is good enough' as recipes don't require high accuracy measurement of ingredients, so their accuracy and linearity doesn't need to be so good.

However for the purposes of a force joystick kitchen scale sensors should be good enough. I would have thought they would be repeatable enough when used within the sensors range that they would serve the purpose.

However for the original joystick with too much travel, is there any possibility of lengthening the distance between the hinge point of the joystick, and the point on the joystick that applies pressure to the sensor? Without knowing how these things measure the force I am just pondering the 'how to' possibilities.





-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@130036 by Eoin Ross

flavicon
face
Perhaps this might work at the core? http://www.ti.com/product/ldc1614/description
They have a tool on that page for designing a coil on a circuit board.

*------------------------------
The LDC1612 and LDC1614 are 2- and 4-channel, 28-bit inductance to digital converters (LDCs) for inductive sensing solutions. With multiple channels and support for remote sensing, the LDC1612 and LDC1614 enable the performance and reliability benefits of inductive sensing to be realized at minimal cost and power. The products are easy to use, only requiring that the sensor frequency be within 1 kHz and 10 MHz to begin sensing. The wide 1 kHz to 10 MHz sensor frequency range also enables use of very small PCB coils, further reducing sensing solution cost and size.
*------------------------------

On 31/07/2015 9:47 p.m., Mike Hord wrote:
{Quote hidden}

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@141910 by Mike Hord

picon face
General feedback on things that have passed thus far:
1. RE: repurposing a scale. I hadn't considered finding a strain gauge in a
scale.
I expect mainly to find load cells in them BUT now am considering that a
load cell
is just a well-characterized set of strain gauges pre-attached to a bar. I
bet I can
do something with that- and we sell an assortment of load cells and
accessories. <www.sparkfun.com/search/results?term=load+cell>
2. Pointing stick seems equally valid, and maybe I'll try to source
something to
see what I can figure out by disassembly. Also, maybe a good product for us
to
start selling?
3. RE: lengthening the moment arm. Part of the problem I'm having is
finding a
means to actuate against the pressure sensor without destroying it. Also,
most
existing joysticks use either some kind of variation on optical mouse,
optical
encoder, or potentiometer to measure displacement. Force is rarely
considered.
4. Recalibrating the range is possible, but it's going to make delicate
control
even harder. Also, the Alliance upgrade is a thing I intend to use
eventually.
5. Inductance or capacitance based is a good idea.

So, I'm going to experiment with load cells, for now; it'll give me some
excuse to
go into the machine shop!

Mike

On Fri, Jul 31, 2015 at 7:47 PM, Mike Hord <spam_OUTmike.hordTakeThisOuTspamgmail.com> wrote:

{Quote hidden}

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@211505 by Brooke Clarke

flavicon
face
Hi Mike:

Yesterday I was looking at the Stanford on line classes and found this:
Introduction to Haptics (Self-Paced) <online.stanford.edu/course/introduction-haptics-self-paced>
It has to do with touch sensitive control.

They have a couple of 1-dimensional control kits (one where you glue Acrylic plates together and another that's 3D printed) where a DC motor provides feedback using a custom board from Seeed Studio and a computer or Arduino.
http://hapkit.stanford.edu/
www.seeedstudio.com/depot/hapkit-p-1622.html
The kit includes a Spark Fun pressure sensor as well as an X-Y magnetic sensor to read back the motor shaft position.

You could improve on this be making a 2-axis version.  Not you can program the compliance on the control from a brick wall to very springy.
Also can add feedback like used in commercial jets where the vibrate the control stick to indicate a serious condition.

Touch, engineered: Allison Okamura at TEDxStanford <https://www.youtube.com/watch?v=pPKqW3tdWCQ>
TEDxManhattanBeach - Paulo Blikstein - One Fabrication Lab per School: the FabLab@School project <https://www.youtube.com/watch?v=ylhfpDAniqM>- Big Education Implications
that leads to:
Transformative Learning Technologies Lab <https://tltl.stanford.edu/>& FabLab@School <https://tltl.stanford.edu/project/fablabschool>

Mail_Attachment --
Have Fun,

Brooke Clarke
http://www.PRC68.com
www.end2partygovernment.com/2012Issues.html
http://www.prc68.com/I/DietNutrition.html

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\01@211640 by RussellMc

face picon face

On 2 August 2015 at 01:28, <.....alan.b.pearceKILLspamspam@spam@stfc.ac.uk> wrote:

> > If a cheap looking low priced scale can achieve the above performance 10+
> > years ago a few hours spent in a few Kitchenware departments may reveal a
> > modern equivalent. (You'll need a source of warm air for likely
> candidates).
>
> They will have been made accurate as they are used for trade, and so will
> be tested periodically by Weights and Measures (not sure which govt.
> department they come under in NZ these days), hence the accuracy and
> linearity. Because they are used for trade purposes they have to be
> accurate to ensure the customer isn't short changed.
>
> Kitchen scales are a case of 'near enough is good enough' as recipes don't
> require high accuracy measurement of ingredients, so their accuracy and
> linearity doesn't need to be so good.
>

​The unit I described WAS a "kitchen scale". Typical flat white plastic
type, sloping front panel with on/off and tare buttons. grams/oz switch on
back. 4 x AA cells.
Performance was (and is) well be​yond what I'd have expected - and well
beyond many others. Temperature compensation was a suprise and a joy.

Probably more suitable for trade than many "real" ones, but not intended to
be.


               Russell

Similar to many of these:

www.google.com/search?site=&tbm=isch&source=hp&biw=1280&bih=893&q=kitchen+scale+digital&oq=kitchen+scale+digital&gs_l=img.3..0j0i8i30l6j0i24.3254.9298.0.10023.21.17.0.4.4.0.251.3274.1j0j14.15.0....0...1ac.1.64.img..2.19.3291.XxC69xMFyBQ#tbm=isch&q=kitchen+scale+digital+round+white
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist

2015\08\02@121031 by Sean Breheny

face picon face
Russell,

What do you mean by "By mechanically biasing one you get two axis"? How can
a single analog signal represent two axes? (I can think of lots of possible
trickery where you encode a digital signal, representing two degrees of
freedom, as an analog signal - but such is inherently non-analog and I
don't think it's what you meant)

Sean


On Sat, Aug 1, 2015 at 4:45 AM, RussellMc <apptechnzspamKILLspamgmail.com> wrote:

{Quote hidden}

-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

2015\08\02@194022 by RussellMc

face picon face

On 3 August 2015 at 04:10, Sean Breheny <.....shb7KILLspamspam.....cornell.edu> wrote:

> What do you mean by "By mechanically biasing one you get two axis"?


​Sorry - excessively sloppy terminology on my part.
If somebody else had written that I would have taken it to mean what you
thought, so I don't know why I used that word.

What I meant was that the strain gauge is usually set up for positive
(downwards) load operation and while the mecahniical aspect will probably
work with negative (upwards) loads, the electronics may not deal well with
inverted polarity SO by applying a net positive bias you can then get +/-
variations around the zero point.

So I should have said something like "2 of the 4 possible directions"
(poor) or +/- X or Y axis operation (better).
ie 2 load cells needed.


Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist

2015\08\03@042123 by alan.b.pearce

face picon face
> Perhaps this might work at the core?
> http://www.ti.com/product/ldc1614/description
> They have a tool on that page for designing a coil on a circuit board.
>
> *------------------------------
> The LDC1612 and LDC1614 are 2- and 4-channel, 28-bit inductance to digital
> converters (LDCs) for inductive sensing solutions. With multiple channels and
> support for remote sensing, the LDC1612 and LDC1614 enable the
> performance and reliability benefits of inductive sensing to be realized at
> minimal cost and power. The products are easy to use, only requiring that
> the sensor frequency be within 1 kHz and 10 MHz to begin sensing. The wide
> 1 kHz to 10 MHz sensor frequency range also enables use of very small PCB
> coils, further reducing sensing solution cost and size.
> *------------------------------

Hey, those look like nifty devices, might have a use for them.

Thanks for the pointer.



-- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
.

More... (looser matching)
- Last day of these posts
- In 2015 , 2016 only
- Today
- New search...