Quadrature encoder hardware.

[Jai]

I have a couple of projects in mind that require fairly precise tracking of encoders, but not fast updating, if you see what I mean. Over, say, a 15 second move, I really don't care about microsecond/micrometer precision during the move. But at the end and start of the move, I want precise feedback, and I want to know that the end position is precise in comparison to the start.

So rather than waste an interrupt pin or endless cycles of a microcontroller tracking the encoder when I don't need it, I'm looking at a hardware solution like this:
http://www.lsicsi.com/pdfs/Data_Sheets/LS7266R1.pdf

Does anyone here have experience with these chips for tracking rotary encoders?
Does anyone here know where to get the things? I may just have search fail today, but neither Mouser nor Digikey seem to have these.
Digikey have a quadrature encoder chip, but it's $11.50 a piece (http://www.digikey.com/product-search/en/integrated-circuits-ics/interface-specialized/2556698?k=quadrature), at which point I could happily devote an Atmega to both counting the encoder *and* running the motor. Or use one of the LPC ARM chips, which have dedicated QE interfaces and more than enough smarts.

Is that the way of the future?

[Andy Gelme]

hi Jai,

On 2013-07-22 13:30 , Jai wrote:
> at which point I could happily devote an Atmega to both counting the
> encoder *and* running the motor. Or use one of the LPC ARM chips,
> which have dedicated QE interfaces and more than enough smarts.
>
> Is that the way of the future?

Due to volume sales, it is often cheaper to dedicate a general purpose
microcontroller to specific tasks.

Angus recently pointed out some ARM Cortex-M0 32-bit microcontrollers
for $1.60 !

For measuring rotation, another solution (original pointed out by Bernd)
is to use AMS encoders ...


http://www.ams.com/eng/Products/Magnetic-Position-Sensors/Rotary-Magnetic-Position-Sensors

AMS ship free samples that arrive in 3 business days (from Austria).

--
-O- cheers = /\ /\/ /) `/ =
--O -- http://www.geekscape.org --
OOO -- an...@geekscape.org -- http://twitter.com/geekscape --

[Jai]

Were those the LPC810 ? I can't see if they have a dedicated QE interface (the biger LPCs certainly do), but yes, at that price even doing it in software is probably a more efficient solution than paying $11 for a specialist chip.

I have an LPCXpresso board on order, I'll have a play with the ARM stuff, and I think you may be right, that may be the way to go. Chers! 

--
You received this message because you are subscribed to a topic in the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/connected-community-hackerspace/vo7ZVFtJvgc/unsubscribe.
To unsubscribe from this group and all of its topics, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/51ECAC7C.6060202%40geekscape.org.

For more options, visit https://groups.google.com/groups/opt_out.

[Angus Gratton]

Hi Jai & Andy,

On Mon, Jul 22, 2013 at 02:10:53PM +1000, Jai wrote:
> Were those the LPC810 ? I can't see if they have a dedicated QE interface

Yes, those were the ones. The biggest problem with them at the moment
is that they seem to be made of unobtainium! The restock order Mouser
said was due last week was belatedly postponed until September, and
they're out of stock for months at Digikey as well. :(

As you say though, almost any microcontroller can probably do what you
want in software - from a <$2 ATTiny85 on up.

> On 22 July 2013 13:52, Andy Gelme <an...@geekscape.org> wrote:
> > > at which point I could happily devote an Atmega to both counting the
> > > encoder *and* running the motor. Or use one of the LPC ARM chips,
> > > which have dedicated QE interfaces and more than enough smarts.
> > >
> > > Is that the way of the future?

I think you're right. Nearly every design that needs to read shaft
encoders will need a microcontroller for some other function anyhow,
and they're cheap and comparatively powerful. So why add costs for a
dedicated IC?

I don't know for sure what your other options are, though. Would be
interested to hear what you end up going with.

- Angus

[Luke Weston]

There are also different tradeoffs you may make, depending on exactly what sort of rotary encoder it is and what it's used for.

For example a rotary encoder for a user interface dial is different from a rotary encoder used for precise feedback in a mechatronic servo system - if it's the former you may be able to get away without needing reliable real-time detection of every single step every time.

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.

To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/20130722073630.GA9985%40ex2.

For more options, visit https://groups.google.com/groups/opt_out.

--
This email is intended only for the personal and confidential use of the human(s) named above. If intercepted by an extraterrestrial civilization, all opinions expressed in this email are my own and do not necessarily reflect the opinion of mankind as a whole.

[Jai]

On 22 July 2013 18:28, Luke Weston <reindeer...@gmail.com> wrote:

There are also different tradeoffs you may make, depending on exactly what sort of rotary encoder it is and what it's used for.

For example a rotary encoder for a user interface dial is different from a rotary encoder used for precise feedback in a mechatronic servo system - if it's the former you may be able to get away without needing reliable real-time detection of every single step every time.

And those tradeoffs are always the interesting bit. It's somewhere between the two - I work in theatre automation, and I'm building a mini version of the kind of systems we use - to make it useful I suppose I should build a small scale model theatre, but that's a longer term project :) So the state of the art is still largely PLC based from people like Seimens and Allen-Bradley, and different systems will split control slightly differently, and hand different bits off to different systems. But broadly speaking, the PC-based console (operator UI) hands cue information off to a PLC, which then calculates move information for axes, sends that to individual drive controllers, and co-ordinates the lot. So an individual drive controller may care very much about individual counts, because they will have a tolerance for how many counts the actual encoder value can drift from the expected encoder value. But the master PLC may care much less about encoder counts, because it actually cares about millimetres of piece travel - so long as the piece is within a millimetre of expected target, it's happy, and a millimetre of travel may be hundreds of counts. So *something* has to count each and every tick, and care, but it doesn't have to be every component.

As Angus points out, though, whatever cares about the individual count also has to know about, and be able to compare, the expected count. The other tradeoff here is between offloading tasks to dedicated hardware, and then comms between that hardware - there's no point having a precise counter and a precise calculator and a precise co-ordinator if by the time they all compare notes, they become less precise than a single chip with twice the power and a dedicated QE interface.

So many tradeoffs, so much to play with. It becomes vastly more interesting again when you allow for a little uncertainty in the initial position, and calculate 3D inverse kinematics for multiple motors at once in real-time. Traditionally, cues are combinations of single axes with pre-plotted moves, but the 21st century is an exciting time to be playing with this stuff.

[Bernd Wachs]

Hello Jai,

Here are some sites in relation to Quadrature encoder:

1)   http://circuitelec.blogspot.com.au/2009/03/rotary-encoder-circuit.html

2)   http://profmason.com/?p=775&doing_wp_cron=1374572132.9559929370880126953125

Regards

Bernd

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/ad5d7c2b-97dd-45d6-b5dd-46e1d2d54b30%40googlegroups.com.

[Jai]

Ah, thanks, Bernd. I do have some of those dual flip flops lying around, and offloading the direction would be a good start. I'll look at that, cheers. But I think the LPC solution might be the way to go.

--
You received this message because you are subscribed to a topic in the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/connected-community-hackerspace/vo7ZVFtJvgc/unsubscribe.
To unsubscribe from this group and all of its topics, send an email to connected-community-h...@googlegroups.com.

To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/842B14F3-9146-43CD-8C94-DDE2F5135FE8%40gmail.com.

[tubular]

The only place that seems to have LPC810's is adafruit, in a starter kit 

http://www.adafruit.com/products/1336

Given it includes the same $10/$7 debug cable we use with the Pi, its not bad value.  $9.10/kit with cchs discount. 

I think we'll be placing another hackerspace order once the 5v 10A PSUs come back into stock 

regards

Lachlan

[Jai]

I think I already have one of those cables - I might email adafruit and see if they'll consider doing a pack of the chips.

--
You received this message because you are subscribed to a topic in the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/connected-community-hackerspace/vo7ZVFtJvgc/unsubscribe.
To unsubscribe from this group and all of its topics, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/d8a2d107-f41a-49ca-9831-bf42f6b3362b%40googlegroups.com.

[Bernd Wachs]

G'Day Jai,

Sorry about the short reply last time.

 I am making up a board that fits in the back of my  VW T3 ( first of the water cooled )  Kombi's.  During my exploring / investigations on how they operate i worked that  with they way the magnets are placed i could put in a little box with 2 Hall effect  sensors and apply the Quadrature sensor principle.

 I am going to use 2 Op Amps as comparators, because the output waveform is in the shape of a sine wave, and feed the signal into 2 D Flip Flops and then into my Arduino.  The direction pin will go into one of the Digital inputs which determines the direction of wheel travel. The pulses will be connected will be fed into pin 5 on my UNO board which has a 16 Bit counter ( counts up to 65535) connected. I read the counter continuously determining if it has changed. If it has changed the distance counters are incremented.

You can use the edges of the pulses to determine position.

The other option if you want to determine the angular position of a shaft at any point in given time is to use a sensor such as   AS5408 which can measure fractions of a Degree

Regards

Bernd

On 22/07/2013, at 13:30 , Jai wrote:

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.

To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/connected-community-hackerspace/ad5d7c2b-97dd-45d6-b5dd-46e1d2d54b30%40googlegroups.com.

[Nick Fryer]

Hi guys

I use an encoder counter shield for an Arduino mega 2560 on my robot,  to track the wheel rotations.

Here's a link to their site. http://www.robogaia.com/two-axis-encoder-counter-mega-shield-version-2.html

Nick