Searching \ for '[EE]: Maximizing ADC Range' 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=rang
Search entire site for: 'Maximizing ADC Range'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Maximizing ADC Range'
2002\10\22@042013 by Jeethu Rao

flavicon
face
Hi,
       I'm working on improving(actually reimplementing) an existing
analog
       Circuit using an MCU. Its  for a grain color sorter(used for
sorting
       rice) I'm using an AVR 8515 for this. That does not exactly
constitute
       the problem.

       The problem is the application. It is to detect the color of a
singular stream of rice, which passes through an illuminated chamber.
       Two photodiodes with a dichroic beam splitter detect colors of
the
       grains against a        distinctly colored background plate, In
the analog
       ckt, lighter shades     give a higher voltage and darker shades
lower
       the signal voltage. It  works well in analog, with some 14
OPAMPS
       (LM725s and 741s)       For each        photodiode. So, I have
an average signal
       on the photodiodes (for instance, 2.5 volts) with all the lamps
lit
       up. A black grain       gives a lower signal (say 2.2 volts),
while a
       normal grain will not   get as low as that, might get a little
higher
       if it's a little lighter        than other grains (In the coffee

       industry, we call such grains Bleaches, but there's no
       nomenclature or purpose for that in the rice industry), but
       never will the signal rise above 3 volts or so.

       Now, if I set the ADCs ref to 3.0 volts so that it reads 1023
(AVR    AT90S8535 has a 10 bit adc like a PIC16f877), I'll be wasting
2/3rds  of the ADCs range       below 2 volts. And the ADC on 8535
doesn't have a  VRef- input. And even if I use an adc with both Vref+
and  Vref-, The         brightness of the CFLs used in the illuminated
chamber tend to         deteriorate over time, so the mid point will not
always be the same,
       it shifts over time. And another problem is that the ADC
       readings initially will be low, since it takes a little time for
the
       lamps to attain full brightness. The solution In the analog
circuit is
       to compensate the value with a multi turn pot for zero setting.
But
       this is inelegant. My idea is to use two DACs to set the Vref+
and
       Vref- once the mid point has been determined. I'm not sure if
this
       Approach works well. A simpler approach would be to use a higher
res
       ADC (~16 bits), but I need an SPI ADC,(low pin count) and I'm
using
       Microchip MCP3201 (12bits) for this. I guess there is a pure
analog  solution to this        problem , just using OPAMPs. Could you
folks please    enlighten me in this regard ?

Thanks,

Jeethu Rao

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spam_OUTlistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\22@062400 by Michael Rigby-Jones

flavicon
face
{Quote hidden}

This is a simple dynamic range problem, you either cover a large range with
poor resolution, or a smaller range with better resolution.  What you need
to define is what resolution is acceptable for your purpose.  You can get
16bit ADC's with an SPI interface, but they are not at all common and most
of them use a DSP type interface rather than plain SPI.  Take a look at the
Linear Technology LTC1865.  Alternatively, you could consider oversampling
and averaging to give you more effective bits.  You need to sample 4^n times
where n is the number of extra bits you want.

The beam splitter arrangment sounds very simmilar to an arrangement we use
to wavelength lock tuneable lasers in the fibre optic communications
industry.  The difference bwteen the two diodes gives wavelength information
(i.e. colour), and the sum of the two gives us optical power.  If this is
the way your system works, then you can automaticaly compenstae for
differening light levels by using the sum term to scale the difference term.
This would mean no manual reset is required.

Regards

Mike

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam.....mitvma.mit.edu with SET PICList DIGEST in the body


2002\10\22@064058 by mike

flavicon
face
On Tue, 22 Oct 2002 11:22:52 +0100, you wrote:

{Quote hidden}

least 16 useable bits, internal PGA up to 128x, and surprisingly cheap
(About $4-5). Will convert up to 30Hz.
--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email @spam@listservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\22@075613 by Russell McMahon

face
flavicon
face
> Hi,
>  I'm working on improving(actually reimplementing) an existing
>  analog Circuit using an MCU. Its  for a grain color sorter(used for
> sorting rice)

Curious minds (there are a few of them here :-) ) want to know Why?
Is this for bulk rice sorting (I imagine not as the speed of response
required would be immense) or (more likely?) for analytical work on a small
sample?

Companion questions:
How often do you have to "measure" a new rice grain and typically how long
is the reading stable for?
Have you got control of the arrival rate or does the arrival rate of the
grains vary widely?


       Russell McMahon

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email KILLspamlistservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\22@095413 by Matt Pobursky

flavicon
face
On Tue, 22 Oct 2002 13:51:38 +0530, Jeethu Rao wrote: ...snip...
>(LM725s and 741s)       For each        photodiode. So, I have an
>average signal on the photodiodes (for instance, 2.5 volts) with all
>the lamps lit up. A black grain       gives a lower signal (say 2.2
>volts), while a
..snip...
>Now, if I set the ADCs ref to 3.0 volts so that it reads 1023 (AVR
>AT90S8535 has a 10 bit adc like a PIC16f877), I'll be wasting 2/3rds
>of the ADCs range       below 2 volts. And the ADC on 8535 doesn't
>have a  VRef- input. And even if I use an adc with both Vref+ and
>Vref-, The         brightness of the CFLs used in the illuminated
>chamber tend to         deteriorate over time, so the mid point will
>not always be the same, it shifts over time. And another problem is
>that the ADC
..snip...
>Microchip MCP3201 (12bits) for this. I guess there is a pure analog
>solution to this        problem , just using OPAMPs. Could you folks
>please    enlighten me in this regard ?

It seems to me that the simplest solution would be to add one final op-
amp stage that has both gain and a negative offset bias for your sensor
signal. You seem to have plenty of dynamic range for the task at 8
bits, but just need to shift your level down and increase the range of
sensor output.

A low input offset op-amp circuit with a negative offset bias of about
67V and a gain of 3 should give you a much wider range of usable
signal level -- about 0V up to maybe 4.5V.

Matt Pobursky
Maximum Performance Systems

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\22@120819 by Matt Pobursky

flavicon
face
On Tue, 22 Oct 2002 08:53:11 -0500, Matt Pobursky wrote:
>A low input offset op-amp circuit with a negative offset bias of about
>67V and a gain of 3 should give you a much wider range of usable

Hmmm... that should have been 0.67V!

Matt Pobursky
Maximum Performance Systems

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spamBeGonelistservspamBeGonespammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\23@040232 by Jeethu Rao

flavicon
face
> This is a simple dynamic range problem, you either cover a large range
> with
> poor resolution, or a smaller range with better resolution.  What you
need
> to define is what resolution is acceptable for your purpose.  You can
get
> 16bit ADC's with an SPI interface, but they are not at all common and
most
> of them use a DSP type interface rather than plain SPI.  Take a look
at
> the
> Linear Technology LTC1865.  Alternatively, you could consider
oversampling
> and averaging to give you more effective bits.  You need to sample 4^n
> times
> where n is the number of extra bits you want.

[JR] Out here in India, availability of these chips is a big problem.
Last time I checked I couldn't find any 16 bit SPI ADCs in the market.
Maybe I'll
have to go in for parallel ADCs.

> The beam splitter arrangment sounds very simmilar to an arrangement we
use
> to wavelength lock tuneable lasers in the fibre optic communications
> industry.  The difference bwteen the two diodes gives wavelength
> information
> (i.e. colour), and the sum of the two gives us optical power.  If this
is
> the way your system works, then you can automaticaly compenstae for
> differening light levels by using the sum term to scale the difference
> term.
> This would mean no manual reset is required.

[JR] Yes, it's the same beam splitter. A 'plain' beam splitter just
reflects a part of the light falling on it and transmits the rest. This
ratio is determined by the coating. In a dichroic beam splitter, we can
get one made to reflect exactly one wavelength and to transmit another.
Actually, its not necessarily one wavelength, it could be two or more
wavelengths reflected/transmitted. Apart from its application, it looks
pretty good!, I sometimes keep gazing on one for hours together.

Regards,

Jeethu Rao

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-requestEraseMEspamspam_OUTmitvma.mit.edu


2002\10\23@040236 by Jeethu Rao

flavicon
face
Ok for the curious kind, here is the brief overview of the machine.
Graded rice contains a lot of foreign bodies (small twigs, husk, dark
rice, bites tc) So, the job of the color sorting machine is to remove
out all these unacceptable particles from rice.

So, the grains from a silo are fed into a feed hopper, which feeds
A Vibrator (the construction of which is strikingly similar to the
images posted by Thomas Lehmann in the "Re: [EE]: First: Orienter w
electromagnet ViBRATOR" thread, The only major difference is in the
driving circuitry). This falls on a 1.2 meter long chute, where the
grains form into a singular stream. This stream passes through a
illuminated chamber, where
the grains are scanned against a background by the photodiodes through
the
associated optics and the signal goes to the processing circuitry.

The Ejectors, which fire out the unwanted grains by a burst of
compressed
Air is located about 1.5" below the scanning chamber. The electronics is
Way too fast for the mechanical ejectors. So, we introduce a delay
between
The sensing and the pulse output to the ejectors. This is done using a
chain of shift registers and clocked by a precise oscillator (you get
the idea!).

That's the rough Idea. As far as the scan rate is concerned, we need an
output of 35 kgs/hr from one channel (that's the limit on the old
design)
And the whole machine has 10 channels (350 Kgs/Hr). So roughly 400
grains
Pass in a second. So, a scan rate of atleast 800 Hz is required.

The arrival rate can and will be varied, to adjust with the varying
plant capacity and the overall plant throughput required. But every time
the arrival rate is varied, the time delay is to be varied. That's
roughly done on trial and error basis. And the delay also needs to be
varied whenever the
Grade of rice is changed, generally, the higher the grade, the heavier
the rice.

To conclude, this is nothing analytical, Its an industrial machine. So
the operator will generally be a dope with no knowledge of electronics
whatsoever. But the delay and the gain settings on the analog system are
A little hard to master. So I just want to make the system as easy to
operate as possible with ready made presets for all grades of rice.

Jeethu Rao

> {Original Message removed}

2002\10\24@084014 by Morgan Olsson

picon face
Jeethu Rao wrote:
>It is to detect the color of a singular stream of rice

( Curiously: )

Some years ago i made a prototype for a brown bean sorter.
It was to sort out sick beans from fresh in the food indutry, eleven devices could sort the whole crop in Sweden.
It used a light chamber, polarised light and reference backgrounds.
Electronic was only a few analog chips.
Free falling beans from mechanical feeder 20 per second.
I used a pneumatic valve and nozzle to "puff" the beans in different directions.
The valves was the speed limiting issue.

May I ask, curiously, what speed you can operate at, and how you divert the bad grains?

- - - Lesson - - -

I designed it for my customer who was to manufacture it.
However his customer had been watching expensive solutions using video processing and fast computers, so they did not believe in our sorting machine although we demonstrated it worked good, fast and reliably.  -That´s what you get for making efficient solutions...  Darn, i should have connected a couple of PC:s, pulse generators, and oscilloscopes to it to make it look complicated...  
/Morgan

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...