Tactile feedback controls

[Paul Hovnanian P.E.]

Also call ed haptics, I believe.

Anyone have any good app notes (URLs) on things like rotary controls
where tuning nulls can be used to produce 'detents' in the knobs
operation? I need some quick and dirty rules of thumb covering hardware,
software and other consideration$ for some preliminary development and
estimates

--
Paul Hovnanian mailtoaul@Hovnanian.com
------------------------------------------------------------------
Sleep is for wimps. Happy, healthy, well-rested wimps, but wimps
nonetheless.

 

[martin griffith]

On 31 May 2005 04:29:25 -0700, in sci.electronics.design "Tim Shoppa"
<shoppa@trailing-edge.com> wrote:

Anyone have any good app notes (URLs)
on things like rotary controls where
tuning nulls can be used to produce 'detents'
in the knobs operation?

You mean not fixed detents, but electrically controllable detents?

Interesting concept, if so. I haven't seen it in use anywhere!

One implementation I could think of is a stepper motor with phases
energized to "hold". Combine with a shaft encoder and some smarts, of
course, to turn off the hold except at the electronic detents.

Tim.
I think I remember a pro video machine, years ago that had a magnetic

clutch on the jog/shuttle knob that changed the "feel" according to
mode.

May be a (magnetised) ball that could be pushed into an idented disc


martin

 

[Luhan Monat]

Tim Shoppa wrote:

Anyone have any good app notes (URLs)
on things like rotary controls where
tuning nulls can be used to produce 'detents'
in the knobs operation?


You mean not fixed detents, but electrically controllable detents?

Interesting concept, if so. I haven't seen it in use anywhere!

One implementation I could think of is a stepper motor with phases
energized to "hold". Combine with a shaft encoder and some smarts, of
course, to turn off the hold except at the electronic detents.

Tim.

I've used stepper motors as shaft encoders. Power up one winding and
look for pulses on the other one. Changing the current thru the powered
winding could change the feel.

--
Luhan Monat: luhanis(at)yahoo(dot)com
http://members.cox.net/berniekm
"Any sufficiently advanced magick is
indistinguishable from technology."

 

[Paul Hovnanian P.E.]

Tim Shoppa wrote:

Anyone have any good app notes (URLs)
on things like rotary controls where
tuning nulls can be used to produce 'detents'
in the knobs operation?

You mean not fixed detents, but electrically controllable detents?

Interesting concept, if so. I haven't seen it in use anywhere!

Its called 'Haptics'. Google found me some good R&D stuff, but I'm
looking for some quick and dirty rules of thumb to get an idea what sort
of angular resolution, frequency response, etc.one needs to give a knob
a good 'feel'.

One implementation I could think of is a stepper motor with phases
energized to "hold". Combine with a shaft encoder and some smarts, of
course, to turn off the hold except at the electronic detents.

Typical stepper motors (directly coupled to a knob) might not have
sufficient angular resolution for good feel. I can feel them cogging, so
unless some form of microstepping was used, they might be too rough.

Basically, I need some quick guidelines on which parameters are
important and where corners can be cut to save a few $$.

Tim.

--
Paul Hovnanian mailtoaul@Hovnanian.com
------------------------------------------------------------------
Q: Why do mountain climbers rope themselves together?
A: To prevent the sensible ones from going home.

 

[Joerg]

Hello Paul,

Its called 'Haptics'. Google found me some good R&D stuff, but I'm
looking for some quick and dirty rules of thumb to get an idea what sort
of angular resolution, frequency response, etc.one needs to give a knob
a good 'feel'.

Quite frankly, the marketing folks will be calling the shots, usually.
This happened every time we finished a design. "This knob is too stiff,
that one needs feedback" etc. I ended up marrying on of the marketeers
that I met during such a session ;-)

Typical stepper motors (directly coupled to a knob) might not have
sufficient angular resolution for good feel. I can feel them cogging, so
unless some form of microstepping was used, they might be too rough.

I have seen some with 4096 (regular) steps but yes, they all cogged.

Basically, I need some quick guidelines on which parameters are
important and where corners can be cut to save a few $$.

Talk to a marketeer. At the end of the day your solution is only as good
as the customer says it is.

If it has to be rock bottom in cost I guess you'll need a plain old
clutch and a DC driven solenoid, possibly spring-loading the whole
thing. Solenoids are available cheaply for automotive use (may be too
big though) but the clutch will most likely have to be a custom part. If
space is an issue and you have to brake from the side that could be done
via a bell crank.

For very long lifetimes look at spring loaded pool pump seals for clutch
action. They have very long lasting friction surfaces but you have to
make sure it is ok to operate under dry conditions with no stuff flaking
off etc.

Regards, Joerg

http://www.analogconsultants.com

 

[Matthias Melcher]

I agree. It really depends on who this thing is built for. If you want to
simulate the endoscopic instruments for heart surgery, you'll need an
immense resulution. On the other end of the spectrum are thos Logitech mice
that would 'click' or 'hop' when the user moved across a window border -
1-bit haptics ;-)

The simple way is to start some motion (magnet, motor with weight,
vibration) when something happens with the device. This is not really force
feedback, but works in low dosage.

If you want real force feedback, you must measure the position of the
manipulator and the force that is applied to it (many only measure the
position!), and make some kind of actuator react as fast as possible. The
human sense for touch is very hard to fool. So cutting corners is limited to
what feels sufficient for you (or your client).

"Joerg" <notthisjoergsch@removethispacbell.net> schrieb im Newsbeitrag
news:ZI8ne.926$wy1.225@newssvr13.news.prodigy.com...

Hello Paul,

Its called 'Haptics'. Google found me some good R&D stuff, but I'm
looking for some quick and dirty rules of thumb to get an idea what sort
of angular resolution, frequency response, etc.one needs to give a knob
a good 'feel'.

 

[Pooh Bear]

"Paul Hovnanian P.E." wrote:

Also call ed haptics, I believe.

Anyone have any good app notes (URLs) on things like rotary controls
where tuning nulls can be used to produce 'detents' in the knobs
operation? I need some quick and dirty rules of thumb covering hardware,
software and other consideration$ for some preliminary development and
estimates

ALPS - no doubt among a few others - have made DC motor controlled pots
mainly for remote control audio applications. Less cogging than the stepper
approach.

Wouldn't be too tricky to energise the motor so as provide resistance to
movement ( feedback from the servo track ) when you found your null point.

Graham

 

[Zak]

Paul Hovnanian P.E. wrote:

Also call ed haptics, I believe.

Anyone have any good app notes (URLs) on things like rotary controls
where tuning nulls can be used to produce 'detents' in the knobs
operation? I need some quick and dirty rules of thumb covering hardware,
software and other consideration$ for some preliminary development and
estimates

What software? Join the knob to a DC motor, and energize the motor
according to tuning.

The old FM discriminator curve:

/|
------/ | /-------
|/

seems what should be the controlling element. And if there is not enough
friction, it might become an autotune feature. Be sure to use a
non-cogging motor.


Thomas

 

[Joerg]

Hello Paul,

It might pay to check out higher end joysticks. I guess high end in this
category means over $20 or so.

I have never seen one myself but was told by a kid that there are
tactile feedback joysticks. IOW the stick force changes according to
some software parameter that the video game spits out. The same is true
for aircraft controls but that would be very expensive stuff.

Regards, Joerg

http://www.analogconsultants.com