Hello everyone,

I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.

I'm using Rabbit BL4S200.

Thank You!

------------------------------------



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

I do not have the RIO but am confused by the question. It is a bit vague in details. What are the characteristics of the signal you are trying to capture - analog, discrete. By "bouncing", it this a switch input ? is the input trigger ringing ? is there a periodic signal that is interfering or beating with the desired signal?

kevin
kc6pob

--- On Mon, 5/25/09, jeanyves.garneau wrote:

From: jeanyves.garneau
Subject: [rabbit-semi] Digital signal filter with the RIO chip
To: r...@yahoogroups.com
Date: Monday, May 25, 2009, 12:31 PM

Hello everyone,

I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.

I'm using Rabbit BL4S200.

Thank You!



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Hi Kevin,

Thank you for your reply. The signal is digital (0-5 V) and come from photocell, which is the board moving on the conveyor and trig sometimes because wood dust in the factory. We have to track the encoder position for the real board not dust. We think to use the RIO because it's the I/O found on the BL4S200 board. The RIO can be programed to decode encoder pulses from 2 inputs (A,B from one I/O block) and programed to edge detection (or capture) of photocell from another input (Another I/O block). Having these signals, we want to apply filter to found the beginning end ending of board with good precision.

Jean-Yves

________________________________
From: kevin asato
To: r...@yahoogroups.com
Sent: Tuesday, May 26, 2009 1:45:09 AM
Subject: Re: [rabbit-semi] Digital signal filter with the RIO chip

I do not have the RIO but am confused by the question. It is a bit vague in details. What are the characteristics of the signal you are trying to capture - analog, discrete. By "bouncing", it this a switch input ? is the input trigger ringing ? is there a periodic signal that is interfering or beating with the desired signal?

kevin
kc6pob

--- On Mon, 5/25/09, jeanyves.garneau wrote:
From: jeanyves.garneau
Subject: [rabbit-semi] Digital signal filter with the RIO chip
To: rabbit-semi@ yahoogroups. com
Date: Monday, May 25, 2009, 12:31 PM
Hello everyone,

I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.

I'm using Rabbit BL4S200.

Thank You!



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

I'm wondering if the problem is simple as cable length.
When I have noise in the system, I usually use the amplifier like
this one (from Us digital)
http://www.usdigital.com/products/interfaces/encoder/cable-drivers/pc4/
George G

Jean-Yves Garneau wrote:
> Hi Kevin,
>
> Thank you for your reply. The signal is digital (0-5 V) and come
> from photocell, which is the board moving on the conveyor and trig
> sometimes because wood dust in the factory. We have to track the
> encoder position for the real board not dust. We think to use the RIO
> because it's the I/O found on the BL4S200 board. The RIO can be
> programed to decode encoder pulses from 2 inputs (A,B from one I/O
> block) and programed to edge detection (or capture) of photocell from
> another input (Another I/O block). Having these signals, we want to
> apply filter to found the beginning end ending of board with good
> precision.
>
> Jean-Yves
>
> ------------------------------------------------------------------------
> *From:* kevin asato
> *To:* r...@yahoogroups.com
> *Sent:* Tuesday, May 26, 2009 1:45:09 AM
> *Subject:* Re: [rabbit-semi] Digital signal filter with the RIO chip
>
> I do not have the RIO but am confused by the question. It is a bit
> vague in details. What are the characteristics of the signal you are
> trying to capture - analog, discrete. By "bouncing", it this a switch
> input ? is the input trigger ringing ? is there a periodic signal that
> is interfering or beating with the desired signal?
>
> kevin
> kc6pob
>
> --- On *Mon, 5/25/09, jeanyves.garneau / > >/* wrote:
> From: jeanyves.garneau
> Subject: [rabbit-semi] Digital signal filter with the RIO chip
> To: rabbit-semi@ yahoogroups. com
> Date: Monday, May 25, 2009, 12:31 PM
>
> Hello everyone,
>
> I'm searching for the best method to apply digital filter to
> remove bouncing with the help of RIO chip. I need filter at the
> signal beginning and signal ending with different values.
>
> I'm using Rabbit BL4S200.
>
> Thank You!
------------------------------------



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

If the photocell is picking up a beam cut by the board it is easy to loook for the beam being cut and check that again in a series of bursts. Dust will not have much effect on the trailing edge allowing the beam to hit the photocell again.

It is a different problem if the source and detector are side-by-side and you are picking up a reflected source beam by the detector. In that case, dust is a bigger problem.

Another question is what are you doing? Are you counting ?? tossing off the conveyor belt ?? controlling sawing machinery ?? The solution will depend on what you want to do.

bob moore

________________________________

From: r...@yahoogroups.com on behalf of Rabbit
Sent: Tue 5/26/2009 10:38 AM
To: r...@yahoogroups.com
Subject: Re: [rabbit-semi] Digital signal filter with the RIO chip

I'm wondering if the problem is simple as cable length.
When I have noise in the system, I usually use the amplifier like
this one (from Us digital)
http://www.usdigital.com/products/interfaces/encoder/cable-drivers/pc4/
George G

Jean-Yves Garneau wrote:
> Hi Kevin,
>
> Thank you for your reply. The signal is digital (0-5 V) and come
> from photocell, which is the board moving on the conveyor and trig
> sometimes because wood dust in the factory. We have to track the
> encoder position for the real board not dust. We think to use the RIO
> because it's the I/O found on the BL4S200 board. The RIO can be
> programed to decode encoder pulses from 2 inputs (A,B from one I/O
> block) and programed to edge detection (or capture) of photocell from
> another input (Another I/O block). Having these signals, we want to
> apply filter to found the beginning end ending of board with good
> precision.
>
> Jean-Yves
>
> ------------------------------------------------------------------------
> *From:* kevin asato
> *To:* r...@yahoogroups.com
> *Sent:* Tuesday, May 26, 2009 1:45:09 AM
> *Subject:* Re: [rabbit-semi] Digital signal filter with the RIO chip
>
> I do not have the RIO but am confused by the question. It is a bit
> vague in details. What are the characteristics of the signal you are
> trying to capture - analog, discrete. By "bouncing", it this a switch
> input ? is the input trigger ringing ? is there a periodic signal that
> is interfering or beating with the desired signal?
>
> kevin
> kc6pob
>
> --- On *Mon, 5/25/09, jeanyves.garneau / > >>/* wrote:
> From: jeanyves.garneau
> Subject: [rabbit-semi] Digital signal filter with the RIO chip
> To: rabbit-semi@ yahoogroups. com
> Date: Monday, May 25, 2009, 12:31 PM
>
> Hello everyone,
>
> I'm searching for the best method to apply digital filter to
> remove bouncing with the help of RIO chip. I need filter at the
> signal beginning and signal ending with different values.
>
> I'm using Rabbit BL4S200.
>
> Thank You!
------------------------------------



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Hi Bob,

Yes, the source and detector are side by side. This is by design. And dust it's a real problem! The objective is to measure with good precision the length of the board (in encoder pulse), the board curve in height, printing cutting mark on board, controlling camera, etc.

The RIO decode quadrature in hardware and count pulses, and maybe (not sure RIO is able to do that) latch the decoder count in the begin and end register on edge detection of another pin on the same RIO block. If not, we can enable interrupt on edge detection and get the decoder count in the interrupt but can take up to 100 clocks before reading the counter, so the counter is offset.

Thank you!

Jean-Yves

________________________________
From: "Moore, Robert"
To: r...@yahoogroups.com
Sent: Tuesday, May 26, 2009 11:19:15 AM
Subject: RE: [rabbit-semi] Digital signal filter with the RIO chip

If the photocell is picking up a beam cut by the board it is easy to loook for the beam being cut and check that again in a series of bursts. Dust will not have much effect on the trailing edge allowing the beam to hit the photocell again.

It is a different problem if the source and detector are side-by-side and you are picking up a reflected source beam by the detector. In that case, dust is a bigger problem.

Another question is what are you doing? Are you counting ?? tossing off the conveyor belt ?? controlling sawing machinery ?? The solution will depend on what you want to do.

bob moore

____________ _________ _________ __

From: rabbit-semi@ yahoogroups. com on behalf of Rabbit
Sent: Tue 5/26/2009 10:38 AM
To: rabbit-semi@ yahoogroups. com
Subject: Re: [rabbit-semi] Digital signal filter with the RIO chip

I'm wondering if the problem is simple as cable length.
When I have noise in the system, I usually use the amplifier like
this one (from Us digital)
http://www.usdigital.com/products/interfaces/encoder/cable-drivers/pc4/
George G

Jean-Yves Garneau wrote:
> Hi Kevin,
>
> Thank you for your reply. The signal is digital (0-5 V) and come
> from photocell, which is the board moving on the conveyor and trig
> sometimes because wood dust in the factory. We have to track the
> encoder position for the real board not dust. We think to use the RIO
> because it's the I/O found on the BL4S200 board. The RIO can be
> programed to decode encoder pulses from 2 inputs (A,B from one I/O
> block) and programed to edge detection (or capture) of photocell from
> another input (Another I/O block). Having these signals, we want to
> apply filter to found the beginning end ending of board with good
> precision.
>
> Jean-Yves
>
> ------------ --------- --------- --------- --------- --------- -
> *From:* kevin asato
> *To:* rabbit-semi@ yahoogroups. com
> *Sent:* Tuesday, May 26, 2009 1:45:09 AM
> *Subject:* Re: [rabbit-semi] Digital signal filter with the RIO chip
>
> I do not have the RIO but am confused by the question. It is a bit
> vague in details. What are the characteristics of the signal you are
> trying to capture - analog, discrete. By "bouncing", it this a switch
> input ? is the input trigger ringing ? is there a periodic signal that
> is interfering or beating with the desired signal?
>
> kevin
> kc6pob
>
> --- On *Mon, 5/25/09, jeanyves.garneau / > >>/* wrote:
> From: jeanyves.garneau
> Subject: [rabbit-semi] Digital signal filter with the RIO chip
> To: rabbit-semi@ yahoogroups. com
> Date: Monday, May 25, 2009, 12:31 PM
>
> Hello everyone,
>
> I'm searching for the best method to apply digital filter to
> remove bouncing with the help of RIO chip. I need filter at the
> signal beginning and signal ending with different values.
>
> I'm using Rabbit BL4S200.
>
> Thank You!

------------ --------- --------- ------

Yahoo! Groups Links



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

--- In r...@yahoogroups.com, "jeanyves.garneau" wrote:
>
> Hello everyone,
>
> I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.
>
> I'm using Rabbit BL4S200.
>
> Thank You!
>

In the quadrature decoder mode the digital filters on the inputs will
not pass any pulses that are less than three sample clocks wide.
The prescaler can be used to make the sample clock whatever width is
convenient. Tie off one quadrature input and use the other to generate
an interrupt on both incr and decr. The counter in another channel
can be used for timing.

Monte

------------------------------------



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Hi Monte,

I need two decoder channels for detect count up and count down. The digital filter is not for the encoder but for the photocell.

Thank you.

Jean-Yves

________________________________
From: monte_dalrymple
To: r...@yahoogroups.com
Sent: Wednesday, May 27, 2009 3:19:33 PM
Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip

--- In rabbit-semi@ yahoogroups. com, "jeanyves.garneau" wrote:
>
> Hello everyone,
>
> I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.
>
> I'm using Rabbit BL4S200.
>
> Thank You!
>

In the quadrature decoder mode the digital filters on the inputs will
not pass any pulses that are less than three sample clocks wide.
The prescaler can be used to make the sample clock whatever width is
convenient. Tie off one quadrature input and use the other to generate
an interrupt on both incr and decr. The counter in another channel
can be used for timing.

Monte



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Jean-Yves,

The point I was trying to make was that you could filter the "dirty"
signal from the photocell by using a RIO channel in an unconventional
way. The digital filters that are present in the quadrature decode
mode are ideal for this filtering, and then you can use the incr and
decr signal, or pattern match, or rollover condition, to create a
status signal that is a clean version of the photocell output, for
output on a RIO port pin. Or perhpas I am not understanding the
requirement.

Monte

--- In r...@yahoogroups.com, Jean-Yves Garneau wrote:
>
> Hi Monte,
>
> I need two decoder channels for detect count up and count down. The digital filter is not for the encoder but for the photocell.
>
> Thank you.
>
> Jean-Yves
> ________________________________
> From: monte_dalrymple
> To: r...@yahoogroups.com
> Sent: Wednesday, May 27, 2009 3:19:33 PM
> Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip
>
> --- In rabbit-semi@ yahoogroups. com, "jeanyves.garneau" wrote:
> >
> > Hello everyone,
> >
> > I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.
> >
> > I'm using Rabbit BL4S200.
> >
> > Thank You!
> > In the quadrature decoder mode the digital filters on the inputs will
> not pass any pulses that are less than three sample clocks wide.
> The prescaler can be used to make the sample clock whatever width is
> convenient. Tie off one quadrature input and use the other to generate
> an interrupt on both incr and decr. The counter in another channel
> can be used for timing.
>
> Monte
>

------------------------------------



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Hi Monte,

I'm very sorry because i have not understood your first answer. I have not realized this unconventional way for the decode mode. But is not the same for all inputs mode? I thought each input pin is sampled at clock rate, exact? If the clock rate is low, the glitch is not detected?

I'm going to study this possibilities...thank you!

Jean-Yves

________________________________
From: monte_dalrymple
To: r...@yahoogroups.com
Sent: Thursday, May 28, 2009 12:46:45 PM
Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip

Jean-Yves,

The point I was trying to make was that you could filter the "dirty"
signal from the photocell by using a RIO channel in an unconventional
way. The digital filters that are present in the quadrature decode
mode are ideal for this filtering, and then you can use the incr and
decr signal, or pattern match, or rollover condition, to create a
status signal that is a clean version of the photocell output, for
output on a RIO port pin. Or perhpas I am not understanding the
requirement.

Monte

--- In rabbit-semi@ yahoogroups. com, Jean-Yves Garneau wrote:
>
> Hi Monte,
>
> I need two decoder channels for detect count up and count down. The digital filter is not for the encoder but for the photocell.
>
> Thank you.
>
> Jean-Yves
> ____________ _________ _________ __
> From: monte_dalrymple
> To: rabbit-semi@ yahoogroups. com
> Sent: Wednesday, May 27, 2009 3:19:33 PM
> Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip
>
> --- In rabbit-semi@ yahoogroups. com, "jeanyves.garneau" wrote:
> >
> > Hello everyone,
> >
> > I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.
> >
> > I'm using Rabbit BL4S200.
> >
> > Thank You!
> > In the quadrature decoder mode the digital filters on the inputs will
> not pass any pulses that are less than three sample clocks wide.
> The prescaler can be used to make the sample clock whatever width is
> convenient. Tie off one quadrature input and use the other to generate
> an interrupt on both incr and decr. The counter in another channel
> can be used for timing.
>
> Monte
>



(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )

Hi Jean-Yves,

I had originally thought that the inputs were filtered all of the
time, but when I went beck and looked at the source code I realized
that I only did the multi-cycle filtering in the quadrature decode
mode. Hence my comment.

In all other modes the input is sampled by the sample clock, so it
will remove any glitches that are not coincident with the sample
clock. But the quad decode mode adds in the extra filter function,
requiring three successive samples before passing the signal on.

Monte
--- In r...@yahoogroups.com, Jean-Yves Garneau wrote:
>
> Hi Monte,
>
> I'm very sorry because i have not understood your first answer. I have not realized this unconventional way for the decode mode. But is not the same for all inputs mode? I thought each input pin is sampled at clock rate, exact? If the clock rate is low, the glitch is not detected?
>
> I'm going to study this possibilities...thank you!
>
> Jean-Yves
> ________________________________
> From: monte_dalrymple
> To: r...@yahoogroups.com
> Sent: Thursday, May 28, 2009 12:46:45 PM
> Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip
>
> Jean-Yves,
>
> The point I was trying to make was that you could filter the "dirty"
> signal from the photocell by using a RIO channel in an unconventional
> way. The digital filters that are present in the quadrature decode
> mode are ideal for this filtering, and then you can use the incr and
> decr signal, or pattern match, or rollover condition, to create a
> status signal that is a clean version of the photocell output, for
> output on a RIO port pin. Or perhpas I am not understanding the
> requirement.
>
> Monte
>
> --- In rabbit-semi@ yahoogroups. com, Jean-Yves Garneau wrote:
> >
> > Hi Monte,
> >
> > I need two decoder channels for detect count up and count down. The digital filter is not for the encoder but for the photocell.
> >
> > Thank you.
> >
> > Jean-Yves
> >
> >
> >
> >
> > ____________ _________ _________ __
> > From: monte_dalrymple
> > To: rabbit-semi@ yahoogroups. com
> > Sent: Wednesday, May 27, 2009 3:19:33 PM
> > Subject: [rabbit-semi] Re: Digital signal filter with the RIO chip
> >
> >
> >
> >
> >
> > --- In rabbit-semi@ yahoogroups. com, "jeanyves.garneau" wrote:
> > >
> > > Hello everyone,
> > >
> > > I'm searching for the best method to apply digital filter to remove bouncing with the help of RIO chip. I need filter at the signal beginning and signal ending with different values.
> > >
> > > I'm using Rabbit BL4S200.
> > >
> > > Thank You!
> > >
> >
> > In the quadrature decoder mode the digital filters on the inputs will
> > not pass any pulses that are less than three sample clocks wide.
> > The prescaler can be used to make the sample clock whatever width is
> > convenient. Tie off one quadrature input and use the other to generate
> > an interrupt on both incr and decr. The counter in another channel
> > can be used for timing.
> >
> > Monte
>

------------------------------------

______________________________
controlSUITE™ software. Comprehensive. Intuitive. Optimized.
Real-world software for real-time control. Details Here!


(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )