Searching \ for '[OT] Quadrature Encoders' 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/io/sensors.htm?key=quadrature
Search entire site for: 'Quadrature Encoders'.

Exact match. Not showing close matches.
PICList Thread
'[OT] Quadrature Encoders'
2000\05\08@204558 by Nicholas Irias

flavicon
face
I'm interested in comments on quad encoder solutions.  For my application, I
have already decided to use HCTL-2000 decoders rather than decoding in
software, because the PIC that will be reading shaft speed will need to read
2 encoders at high speed and will have additional tasks to attend to.

The tricky part is the encoder itself.  I have a 1/4" shaft, spinning at a
maximum of 8000 rpm.  I also want to be able to sample meaningfull results
every 10 mSecs.  With minimum shaft speed around 180 rpm, this requires
around 100 holes (400 state positions) per revolution to ensure that the
10mSec count totals can provide at least 1:10 speed resolution on the low
shaft speeds.

I have considered making my own encoder, but machining an encoder disk with
100 holes is a little tricky.  And I'm guessing that a poorly designed or
constructed encoder would have problems with stray light reaching the
photodiodes at the wrong times.

The only packaged solution I have found that is up to the task is an E5M
from USDigital, at $69 in single quantities.  Are there any cheaper and/or
better encoder solutions that I ought to be looking at?

thanks

2000\05\09@091227 by Randy A.

picon face
Nicholas:

I would very much doubt that you will find anything much cheaper.  I am a
field engineer in the industrial woodworking machine industry and specialize
in CNC machinery.  I use quadrature encoders on virtually all of the
equipment that I service and design,  those encoders are designed for shaft
speeds less than what you are requiring however they do have a higher
resolution usually around 500 up to 2500 pulses per revolution and they run
from $250 up to over $1000 each so the price from U.S. Digita sounds pretty
good to me.  Make certain that the encoder can handle the 8000 rpm as that is
quite a bit higher than even most industrial encoders will handle.  And, you
are correct about making your own.  You will need some good equipment and in
order to have an accurate encoder it needs to be made with precision.

Regards,
Randy A.

2000\05\09@121315 by Ricardo Seixas

picon face
<x-flowed>Nicholas,

<SNIP>
>The tricky part is the encoder itself.  I have a 1/4" shaft, spinning at a
>maximum of 8000 rpm.  I also want to be able to sample meaningfull results
>every 10 mSecs.  With minimum shaft speed around 180 rpm, this requires
>around 100 holes (400 state positions) per revolution to ensure that the
>10mSec count totals can provide at least 1:10 speed resolution on the low
>shaft speeds.

Please clarify.
Do you need to measure speed only ?

>I have considered making my own encoder, but machining an encoder disk with
>100 holes is a little tricky.  And I'm guessing that a poorly designed or
>constructed encoder would have problems with stray light reaching the
>photodiodes at the wrong times.

I think 100 holes is pretty easy to do with a typeseter (Agfa for example).
I've already done discs with 1800 slots with this method.
Then you cut the disc and fix it between a washer and a hub.
If you need I can send you a PostScript file that generates the slots.
The photodiodes positioning is the problem here.

>The only packaged solution I have found that is up to the task is an E5M
>from USDigital, at $69 in single quantities.  Are there any cheaper and/or
>better encoder solutions that I ought to be looking at?

Maybe you can buy only the discs and HEDS modules ($39) from USDigital and
build your own case and hub.
I've already done a lot of research on this and never found a solution
cheaper than
USDigital (for big resolutions).
A little big thing is to ensure that the disc is firmly fixed on the shaft
so, it won't slip
when there's a big change in rotational speed, this effect can be found on
USDigital S2 series
where the washer is glued on the shaft and the disc can spin easily, and it
becomes worst when
you use index. (I have a big problem to track down this thing once)


Hope this helps




-----------------------------------
Ricardo Seixas
spam_OUTrseixasTakeThisOuTspampobox.com
-----------------------------------

</x-flowed>

2000\05\09@171329 by Giles

picon face
I don't think I can help the the RPM you are dealing with, but..
I am wanting to make an encoder using film.  I have at my disposal f 1/8mil
photo plotter.  Does anyone have access to a DXF or drawing of an encoder
wheel?

Best regards,
Giles

{Original Message removed}

2000\05\09@175857 by Nicholas Irias

flavicon
face
-----Original Message-----
From: Ricardo Seixas <.....rseixasKILLspamspam@spam@POBOX.COM>
To: PICLISTspamKILLspamMITVMA.MIT.EDU <.....PICLISTKILLspamspam.....MITVMA.MIT.EDU>
Date: Tuesday, May 09, 2000 9:25 AM
Subject: Re: [OT] Quadrature Encoders


{Quote hidden}

I need speed and direction at low (<500rpm) speed.  When turning at high
speeds, direction is always known.

2000\05\09@223357 by James R. Cunningham

flavicon
face
Hi guys,

I have a somewhat similar problem.  I need to drive an Astroflight 60 motor
(turning a ball screw drive) from 0-9000 rpm and back to 0 in 0.4 seconds, and
then do the same thing with the opposite rotation during the next 0.4 second.
For me, correcting to the appropriate position at each time increment based on
an error signal is the important thing.  My intent is to use the ball screw to
drive the flapping wing plunge amplitude of a 16 foot span pterodactyl replica.
Any thoughts about position measurement techniques?  I was tentatively thinking
about using a rotation encoder to set a digital pot for use as feedback for an
H-bridge servo driver.

Thanks,
Jim


Nicholas Irias wrote:

> {Original Message removed}

2000\05\10@050924 by Ricardo Seixas
picon face
Nicholas,

There's a small dual-channel Omron module that may fit your application
but, I think you need a disc > 50mm in diameter to use 100 slots with that
module.
I don't remember the part number but DigiKey sells it, if you need more
information
I can track the part nr from the DigiKey catalog.
Also you can use as decoder the LS7183 or LS7184 (USDigital sells) it plugs
directly to the encoder
and do the X4 conversion so the only thing you need is to measure the
frequency
directly with the PIC with a very small overhead using TMR, the direction
is also provided by the chip.


{Quote hidden}

-----------------------------------
Ricardo Seixas
EraseMErseixasspam_OUTspamTakeThisOuTpobox.com
-----------------------------------

2000\05\10@050940 by Ricardo Seixas

picon face
Giles,

I've writed a Postscript program (resolution independent) that generates
the disc, you change the
external/internal diameter and the number of slots and it draws directly
onto the printing
device and, you can fit several inside an A4 sheet.
But I guess if you convert from PS to DXF you will lose resolution.

{Quote hidden}

-----------------------------------
Ricardo Seixas
rseixasspamspam_OUTpobox.com
-----------------------------------

2000\05\10@092743 by James R. Cunningham

flavicon
face
Curiosity strikes.  Why does converting from PS to DXF lose resolution?

Jim

Ricardo Seixas wrote:

> But I guess if you convert from PS to DXF you will lose resolution.

2000\05\10@114243 by Nicholas Irias

flavicon
face
The pterodactyl makes my project seem quite simple.  Is it PIC controlled?
Is there a project web page ?  If so, please provide the url.

-Nicholas


{Original Message removed}

2000\05\10@114449 by Ricardo Seixas

picon face
<x-flowed>Jim

>Curiosity strikes.  Why does converting from PS to DXF lose resolution?


In this particular case the resolution is determined by the printer, you
just copy
the file to the printer and it uses the maximum resolution to generate the
encoder.
I don't know a converter from PS to DXF so, if you convert from PS to an
intermediate format
then to DXF maybe you lose resolution.

> > But I guess if you convert from PS to DXF you will lose resolution.



-----------------------------------
Ricardo Seixas
@spam@rseixasKILLspamspampobox.com
-----------------------------------

</x-flowed>

2000\05\10@121851 by Giles

picon face
Post Script uses vector boundaries with properties assigned to those
boundaries.
DXF is exclusively vector boundary.
When converting, you must assign arc resolution to the vectors, because you
are kinda converting true arcs to segmented vectors.
I would love to play with the conversion and could probably do it for you.
We have some software that helps with this type of problem.

Best regards,
Giles

P.S.  If you want to send me the PS and DXF, e-mail them to
KILLspamgilesKILLspamspamonecorpsource.com


----- Original Message -----
From: James R. Cunningham <RemoveMEjrcceaTakeThisOuTspamBELLSOUTH.NET>
To: <spamBeGonePICLISTspamBeGonespamMITVMA.MIT.EDU>
Sent: Wednesday, May 10, 2000 9:41 AM
Subject: Re: [PICLIST] [OT] Quadrature Encoders


> Curiosity strikes.  Why does converting from PS to DXF lose resolution?
>
> Jim
>
> Ricardo Seixas wrote:
>
> > But I guess if you convert from PS to DXF you will lose resolution.
>

2000\05\10@132859 by James R. Cunningham

flavicon
face
Probably 50 Mh Scenix -- perhaps several of them.  No web page yet.  My actual
project involves the flight mechanics and biomechanics of two late Cretaceous
azhdarchid pterodactlys, Quetzalcoatlus species, and Quetzalcoatlus northropi.
The flying replica project is actually secondary, just an effort to see how well
I can implement what we've learned about the little critters in the last couple
of years.  Qsp had a wingspan of about 16 feet and a weight of about 44 pounds.
Qn had a wingspan of about 36 feet and a weight of about 330 pounds.  New full
scale skeletal replicas of the big animal based on some of our work have
recently been completed, and are hanging at the TMM in Austin, TX; the Carnegie
Museum, and a couple of other locations.  I may do a web page later -- haven't
decided yet. I'll send you a rather poor quality JPEG of the skeletal replica of
the big animal if you wish.

Jim

Nicholas Irias wrote:

> The pterodactyl makes my project seem quite simple.  Is it PIC controlled?
> Is there a project web page ?  If so, please provide the url.

2000\05\10@133912 by Severson, Rob

flavicon
face
Please check your facts. The late Cretaceous azhdarchid pterodactlys were
based on the Intel 8080. The Scenix would come much later.

;-)

2000\05\10@142856 by James R. Cunningham

flavicon
face
You just made my day.  Thanks.  And if I ever learn to type 'pterodactyls'
corectly, I'll be dangerous.  But then, spelling wasn't standardised during
the late K.

Jim

Severson, Rob wrote:

> Please check your facts. The late Cretaceous azhdarchid pterodactlys were
> based on the Intel 8080. The Scenix would come much later.
>
> ;-)

More... (looser matching)
- Last day of these posts
- In 2000 , 2001 only
- Today
- New search...