Second email! As generally happens when I'm about to travel, I'm
struck with all sorts of ideas I want to try out, and no time to do
them. For now this will remain a thought experiment, but I hope to do
some of the design while traveling, then build a prototype when I
return.
I've had an idea for a project that would use motorized faders. These
are linear (slide) potentiometers with a DC motor attached. I've
started reading about these, and apparently driving them can be a bit
tricky. They have two resistance tracks, originally one was used to
control your signal (audio or otherwise), and the second was used as
positional feedback for your servo control. In essence these were
linear servos.
I'd like to drive mine digitally, as I'm going to be connecting the
unit to an ADC input anyways. What's tricky is that there's a
mechanical component to the design that needs to be taken into
consideration, otherwise you stand the chance of having the fader
yo-yo around as it "homes in" on the correct value it should be at.
This is due to inertia and friction. In addition, you need to be
careful about not slamming the unit into the end stops repeatedly, or
trying to drive the fader if it's stuck (lest you burn out the motor).
I'm thinking about an H bridge to drive the motor, and software to
help with the positioning. I will also need to come up with some sort
of capacitive touch sensor, as the unit has a touch path from the
slider, and I'd like to know when it's being touched so as to disable
the motor drive.
I've never done any PID work, and I'm wondering if that's the correct
route to go for the control. I've done some searching, but haven't
come up with much for this particular application. Does anyone have
any tips, or links to other documentation? I've been sent an
application note from the North American rep for Penny and Giles, one
of the most respected fader manufacturers. I'm working through it, but
it's mainly based on analog control schemes. Still, useful stuff.
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
I've not used _linear_ motorized pots, but the rotary ones I have used were
quite slow. Inertia and friction were basically non-issues.
The other detail is they had slip clutches so the knob could be moved
without the motor, this also acted to slip when the motor "hit" the end
stop.
Yes, an H-bridge is a necessary driver. You can get these as a driver chip
(TI's DRVxxx series comes to mind) or do it with discrete transistors.
For my money, PID is complete overkill for this application. i drive lots
of resistive feedback linear actuators with full ON-OFF control and a
deadband of +-2 ADC counts, never seen a hunting problem with that.
> Hi all.
>
> Second email! As generally happens when I'm about to travel, I'm
> struck with all sorts of ideas I want to try out, and no time to do
> them. For now this will remain a thought experiment, but I hope to do
> some of the design while traveling, then build a prototype when I
> return.
>
> I've had an idea for a project that would use motorized faders. These
> are linear (slide) potentiometers with a DC motor attached. I've
> started reading about these, and apparently driving them can be a bit
> tricky. They have two resistance tracks, originally one was used to
> control your signal (audio or otherwise), and the second was used as
> positional feedback for your servo control. In essence these were
> linear servos.
>
> I'd like to drive mine digitally, as I'm going to be connecting the
> unit to an ADC input anyways. What's tricky is that there's a
> mechanical component to the design that needs to be taken into
> consideration, otherwise you stand the chance of having the fader
> yo-yo around as it "homes in" on the correct value it should be at.
> This is due to inertia and friction. In addition, you need to be
> careful about not slamming the unit into the end stops repeatedly, or
> trying to drive the fader if it's stuck (lest you burn out the motor).
>
> I'm thinking about an H bridge to drive the motor, and software to
> help with the positioning. I will also need to come up with some sort
> of capacitive touch sensor, as the unit has a touch path from the
> slider, and I'd like to know when it's being touched so as to disable
> the motor drive.
>
> I've never done any PID work, and I'm wondering if that's the correct
> route to go for the control. I've done some searching, but haven't
> come up with much for this particular application. Does anyone have
> any tips, or links to other documentation? I've been sent an
> application note from the North American rep for Penny and Giles, one
> of the most respected fader manufacturers. I'm working through it, but
> it's mainly based on analog control schemes. Still, useful stuff.
>
> Thanks!
>
> Josh
> --
> A common mistake that people make when trying to design something
> completely foolproof is to underestimate the ingenuity of complete
> fools.
> -Douglas Adams
>
> quite slow. Inertia and friction were basically non-issues.
> The other detail is they had slip clutches so the knob could be moved
> without the motor, this also acted to slip when the motor "hit" the end
> stop.
>
> Yes, an H-bridge is a necessary driver. You can get these as a driver chip
> (TI's DRVxxx series comes to mind) or do it with discrete transistors.
>
> For my money, PID is complete overkill for this application. i drive lots
> of resistive feedback linear actuators with full ON-OFF control and a
> deadband of +-2 ADC counts, never seen a hunting problem with that.
>
> Good luck.
> -Denny
>
>
> On Tue, May 29, 2012 at 10:10 AM, Josh Koffman <joshybearKILLspamgmail.com>
>
>> Hi all.
>>
>> Second email! As generally happens when I'm about to travel, I'm
>> struck with all sorts of ideas I want to try out, and no time to do
>> them. For now this will remain a thought experiment, but I hope to do
>> some of the design while traveling, then build a prototype when I
>> return.
>>
>> I've had an idea for a project that would use motorized faders. These
>> are linear (slide) potentiometers with a DC motor attached. I've
>> started reading about these, and apparently driving them can be a bit
>> tricky. They have two resistance tracks, originally one was used to
>> control your signal (audio or otherwise), and the second was used as
>> positional feedback for your servo control. In essence these were
>> linear servos.
>>
>> I'd like to drive mine digitally, as I'm going to be connecting the
>> unit to an ADC input anyways. What's tricky is that there's a
>> mechanical component to the design that needs to be taken into
>> consideration, otherwise you stand the chance of having the fader
>> yo-yo around as it "homes in" on the correct value it should be at.
>> This is due to inertia and friction. In addition, you need to be
>> careful about not slamming the unit into the end stops repeatedly, or
>> trying to drive the fader if it's stuck (lest you burn out the motor).
>>
>> I'm thinking about an H bridge to drive the motor, and software to
>> help with the positioning. I will also need to come up with some sort
>> of capacitive touch sensor, as the unit has a touch path from the
>> slider, and I'd like to know when it's being touched so as to disable
>> the motor drive.
>>
>> I've never done any PID work, and I'm wondering if that's the correct
>> route to go for the control. I've done some searching, but haven't
>> come up with much for this particular application. Does anyone have
>> any tips, or links to other documentation? I've been sent an
>> application note from the North American rep for Penny and Giles, one
>> of the most respected fader manufacturers. I'm working through it, but
>> it's mainly based on analog control schemes. Still, useful stuff.
>>
>> Thanks!
>>
>> Josh
>> --
>> A common mistake that people make when trying to design something
>> completely foolproof is to underestimate the ingenuity of complete
>> fools.
>> -Douglas Adams
>>
part 1 820 bytes content-type:text/plain; charset="iso-8859-1" (decoded quoted-printable)
Josh,
I don't think you'll have any problem with the pot yo-yoing or
hunting. It will be very slow compared with the processing speed
of the PIC and has no significant mechanical momentum to carry
it on past a set point. The reduction gearbox means a pretty much
instant stop
Attached is what I use for a rotary pot controlled manually or
with IR. The wiper goes to an 18F, but could also go to the 12F
for feedback. It's a little slower than spec because of the voltage
drops across the transistors, but that's not too important in my
application
Might not apply to your pot, but I notice in the Penny And Giles
tech notes for the PGFM9000 series that the motor should NOT
be driven against the end stops, so you'll need to address that.
Rotary pots have a clutch
Joe
part 2 3367 bytes content-type:image/gif; name="ir_h-bridge_675.gif" (decode) part 3 181 bytes content-type:text/plain; name="ATT00001.txt" (decoded base64)
-- http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
mailman.mit.edu/mailman/listinfo/piclist
On Tue, May 29, 2012 at 2:13 PM, Denny Esterline <.....desterlineKILLspam.....gmail.com> wrote:
> I've not used _linear_ motorized pots, but the rotary ones I have used were
> quite slow. Inertia and friction were basically non-issues.
> The other detail is they had slip clutches so the knob could be moved
> without the motor, this also acted to slip when the motor "hit" the end
> stop.
These ones can zip along fairly quick, but from what I'm now hearing,
inertia and friction might not be as big a problem as I was initially
thinking. I think I'll try to whip up a quick circuit at some point
and just try it to see what happens.
> Yes, an H-bridge is a necessary driver. You can get these as a driver chip
> (TI's DRVxxx series comes to mind) or do it with discrete transistors.
I wasn't able to locate much that was relevant in the DRVxxx series.
Admittedly TI's website seemed to be acting a bit weird so perhaps I
missed something. Do you have any more exact reference that I can
search for on there?
> For my money, PID is complete overkill for this application. i drive lots
> of resistive feedback linear actuators with full ON-OFF control and a
> deadband of +-2 ADC counts, never seen a hunting problem with that.
>
> Good luck.
> -Denny
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
On Tue, May 29, 2012 at 3:15 PM, Vicent Colomar Prats
<EraseMEvicentecolomarspam_OUTTakeThisOuTgmail.com> wrote:
> As a driver for small motors you can use mcp14e4 or similars. And a small
> pic to communicate with.
Hm, this part looks quite nice for this application. The MCP14E5 is
even a dual driver, one inverting, one non inverting. Seems that I
could tie the two inputs and two enables together, and have a two
control line solution without any external signal inverting. The only
area I'm a little hazy on is if I'm able to run the motors on a higher
voltage than the PIC this way. According to the MCP14E5 datasheet,
logic level 1 on the input is minimum 2.4V, typical 1.5V, which
doesn't make any sense. It does seem as though I could run the chip
off say 9V, and the PIC off 5V and not require level translation, but
I'm not entirely sure.
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
On Tue, May 29, 2012 at 8:08 PM, IVP <joecolquittspam_OUTclear.net.nz> wrote:
> Josh,
>
> I don't think you'll have any problem with the pot yo-yoing or
> hunting. It will be very slow compared with the processing speed
> of the PIC and has no significant mechanical momentum to carry
> it on past a set point. The reduction gearbox means a pretty much
> instant stop
Sadly these guys don't have gearboxes, at least, I'm pretty sure they
don't. They can move rather rapidly, but I could just limit the speed
all the time rather than feeding it full juice at every go. We'll see.
> Attached is what I use for a rotary pot controlled manually or
> with IR. The wiper goes to an 18F, but could also go to the 12F
> for feedback. It's a little slower than spec because of the voltage
> drops across the transistors, but that's not too important in my
> application
Thank you!
> Might not apply to your pot, but I notice in the Penny And Giles
> tech notes for the PGFM9000 series that the motor should NOT
> be driven against the end stops, so you'll need to address that.
> Rotary pots have a clutch
Yep, and while I'm a bit worried about that, I think I should be ok.
The servo track of the linear pot means that I'm always getting
feedback, so unless something gets stuck on, I shouldn't be in a state
of trying to drive it too far. One of the application notes also
suggests having a timeout on the action equal to double the time to
travel the total length. This means that if the fader is stuck, you
aren't trying to dump current in the motor indefinitely. And coupled
with the feedback by the linear pot, it means that errors can be
recognized by the PIC and indicated with an LED or something.
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
I tried it with a two cell LiPo (8,4V at max charge) directly for the motor
and 5V for dspic, and it works fine. So, I think it can also do the job
with 9V, but look for absolute max. in datasheet, I do not remember what
the value it was.
> On Tue, May 29, 2012 at 3:15 PM, Vicent Colomar Prats
> <KILLspamvicentecolomarKILLspamgmail.com> wrote:
> > As a driver for small motors you can use mcp14e4 or similars. And a small
> > pic to communicate with.
>
> Hm, this part looks quite nice for this application. The MCP14E5 is
> even a dual driver, one inverting, one non inverting. Seems that I
> could tie the two inputs and two enables together, and have a two
> control line solution without any external signal inverting. The only
> area I'm a little hazy on is if I'm able to run the motors on a higher
> voltage than the PIC this way. According to the MCP14E5 datasheet,
> logic level 1 on the input is minimum 2.4V, typical 1.5V, which
> doesn't make any sense. It does seem as though I could run the chip
> off say 9V, and the PIC off 5V and not require level translation, but
> I'm not entirely sure.
>
> Thanks!
>
> Josh
> --
> A common mistake that people make when trying to design something
> completely foolproof is to underestimate the ingenuity of complete
> fools.
> -Douglas Adams
>
Was where I first seen them.
Note, I've not used them, just thought they were interesting little
beasties.
I used to use the TI TPIC107b parts, but TI obsoleted them and caused me
much grief. Now unless there's a compelling reason not to, I try to design
with discrete transistors/fets instead - many, many more second source
options.
I think Allegro micro has some suitable H-bridge chips in your current /
voltage range as well.
-Denny
> On Tue, May 29, 2012 at 2:13 PM, Denny Esterline <spamBeGonedesterlinespamBeGonegmail.com>
> wrote:
> > I've not used _linear_ motorized pots, but the rotary ones I have used
> were
> > quite slow. Inertia and friction were basically non-issues.
> > The other detail is they had slip clutches so the knob could be moved
> > without the motor, this also acted to slip when the motor "hit" the end
> > stop.
>
> These ones can zip along fairly quick, but from what I'm now hearing,
> inertia and friction might not be as big a problem as I was initially
> thinking. I think I'll try to whip up a quick circuit at some point
> and just try it to see what happens.
>
> > Yes, an H-bridge is a necessary driver. You can get these as a driver
> chip
> > (TI's DRVxxx series comes to mind) or do it with discrete transistors.
>
> I wasn't able to locate much that was relevant in the DRVxxx series.
> Admittedly TI's website seemed to be acting a bit weird so perhaps I
> missed something. Do you have any more exact reference that I can
> search for on there?
>
> > For my money, PID is complete overkill for this application. i drive lots
> > of resistive feedback linear actuators with full ON-OFF control and a
> > deadband of +-2 ADC counts, never seen a hunting problem with that.
> >
> > Good luck.
> > -Denny
>
> Thanks!
>
> Josh
> --
> A common mistake that people make when trying to design something
> completely foolproof is to underestimate the ingenuity of complete
> fools.
> -Douglas Adams
>
> Sadly these guys don't have gearboxes, at least, I'm pretty sure
> they don't. They can move rather rapidly, but I could just limit
> the speed all the time rather than feeding it full juice at every go
On audio boards for example they'll snap to memorised positions
pretty quickly, as you say. If you don't want them to move that
fast then you can run the motor at a lower voltage and/or PWM
If PWM, you'd have the option of full speed until within 'braking
distance' of the destination, rather like driving a car. And braking
could mean either coast to a stop or applying oppositie drive, ie
slam it in revers
Admittedly off-topic, but not by too much:
Audio line mixer, signal path is all analog, the audio is never digitised.
Controls are all digital (MIDI).
There are NO POTS, NO SLIDERS, no chance of getting "scratchy" with age.
And regarding how fast they change level, these can "jump" directly to the
setting without traversing the distance in between.
This uses Burr-Brown audio chips, and an 18F452 (I think) to manange the
controls.
>> Sadly these guys don't have gearboxes, at least, I'm pretty sure
>> they don't. They can move rather rapidly, but I could just limit
>> the speed all the time rather than feeding it full juice at every go
>
> On audio boards for example they'll snap to memorised positions
> pretty quickly, as you say. If you don't want them to move that
> fast then you can run the motor at a lower voltage and/or PWM
>
> If PWM, you'd have the option of full speed until within 'braking
> distance' of the destination, rather like driving a car. And braking
> could mean either coast to a stop or applying oppositie drive, ie
> slam it in reverse
By the way, if anyone has seen conductive plastic or metal fader caps
for sale somewhere, I'd appreciate a steer. I have found straight
plastic ones at Mouser, but I'm looking for a supplier of ones that
will allow me to sense a finger on the fader. I know I can order them
as spare parts from audio console manufacturers, but I'd prefer to
find a farther upstream source.
As for the status of my project, I've thrown together a quick demo
board (well, an add on to an existing board), but I won't get to try
it out for at least a month. Ah well!
Thanks for the help!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
On Thu, May 31, 2012 at 8:24 PM, Lyle Hazelwood <RemoveMElylehazeTakeThisOuTgmail.com> wrote:
> Admittedly off-topic, but not by too much:
> Audio line mixer, signal path is all analog, the audio is never digitised..
> Controls are all digital (MIDI).
> There are NO POTS, NO SLIDERS, no chance of getting "scratchy" with age.
>
> And regarding how fast they change level, these can "jump" directly to the
> setting without traversing the distance in between.
Looks cool Lyle! Unfortunately my project has nothing to do with
audio, it's more about using the faders as a control interface.
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
> By the way, if anyone has seen conductive plastic or metal fader caps
> for sale somewhere, I'd appreciate a steer. I have found straight
> plastic ones at Mouser, but I'm looking for a supplier of ones that
> will allow me to sense a finger on the fader. I know I can order them
> as spare parts from audio console manufacturers, but I'd prefer to
> find a farther upstream source.
>
Come to think of it, if anyone has any examples of using a PIC to detect
touch in this way, I'd love to see it. I was looking at cap sense but I
don't think it's quite the right solution. I believe there's an Arduino
library for doing touch sense that doesn't require too many external
components, ideally I'd be looking for something similar. I have found one
at Bytecraft (http://www.bytecraft.com/Touch_Sensitive_Switch) but I'm
always on the lookout for more!
Thanks!
Josh
-- A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adam
> If PWM, you'd have the option of full speed until within 'braking
> distance' of the destination, rather like driving a car. And braking
> could mean either coast to a stop or applying oppositie drive, ie
> slam it in reverse
>
>
>
Braking could also mean braking the motor by grounding (or connecting to
Vcc for that matter) both leads via the H bridge, which would help eat up
any inertial energy, without needing to be precise, like reversing it to a
stop.
You could also use this to damp/brake the sliders when they are manually
moved (I.E. it could be free moving, or damped/braked). For a soundboard
application, you could possibly brake the slider if the channel is
clipping, etc. The user could still move it, but would have to do so
intentionally, just as an example.
gmail.com> wrote:
