开云体育

--- In CAD_CAM_EDM_DRO@y..., Jon Elson <elson@p...> wrote:

AD2S90-EB PRODUCTION Evaluation Boards - COMMERCIAL $89.06
AD2S90AP PRODUCTION PLASTIC LEAD CHIP CARRIER 20 COMMERCIAL $29.16

snip

Making an Inductosyn seems a LOT simpler, as the geometry is pretty
straightforward. It has a planar square wave exiting winding which

forms the scale, and 2 planar square wave sense windings which have
the same wave length as the exciting winding, but they are placed
such that there is a 90 degree phase shift between the two windings.
These can be made by printed circuit technology. You just hook it
all to the 2S90 series of chips.

Great tutorial on resolvers @


The cost for commercial 10" inductosyn bars is about $250. Heads
another $250. The newall is definately cheaper unles you can make
heads and boards yourself. As Jon points out the bar is just a
square wave easily milled or etched. The commercial ones seems to
have a zillion lines per inch but give you .00001 resolution which I
don't need for a router.

I wonder how many lines per inch would you need for .0001 and if the
evaluation board would drive/read it? I'd be willing to give it a
shot if I could get by with a reasonable number of lines per inch and
the board has a chance of working.

Roger

vrsculptor wrote:

The cost for commercial 10" inductosyn bars is about $250. Heads

another $250. The newall is definately cheaper unles you can make heads and boards yourself. As Jon points out the bar is just a square wave easily milled or etched. The commercial ones seems to have a zillion lines per inch but give you .00001 resolution which I don't need for a router.

I wonder how many lines per inch would you need for .0001 and if the evaluation board would drive/read it? I'd be willing to give it a shot if I could get by with a reasonable number of lines per inch and the board has a chance of working.

I don't think they are very fine grain. Something like 10 waves / inch. The rest is done
by interpolation. A printed circuit board maker that can handle a 30" x 1" circuit board
could probably make a batch of these for a couple of hundred bucks. I might run a quote
to find out what my favorite board house can do.

Jon

AD2S90-EB PRODUCTION Evaluation Boards - COMMERCIAL $89.06
AD2S90AP PRODUCTION PLASTIC LEAD CHIP CARRIER 20 COMMERCIAL $29.16

-------
I expect to need a 2-axis resolver to quadratyre board, so I did a bit more
checking.

The evaluation board price above is quantity 1.

Unfortunately the above price for the chip is quantity 100 and is not
available for online purchase from Analog Devices WEB site.

Pioneer has ad2s90ap in stock with the quantity 1 price as $44.86. You
probably want an ad2s99 reference oscillator @$10.73 too, but out of stock.

At a guess a 2-axis board with just quadrature output could probably be hand
wired for around $125 with all new parts.

More thinking is in order; such as serial out to a computer.
--
Keith Bowers - Thomasville, NC

Keith Bowers wrote:

AD2S90-EB PRODUCTION Evaluation Boards - COMMERCIAL $89.06 AD2S90AP PRODUCTION PLASTIC LEAD CHIP CARRIER 20 COMMERCIAL $29.16

-------
I expect to need a 2-axis resolver to quadratyre board, so I did a bit more checking.

The evaluation board price above is quantity 1.

Unfortunately the above price for the chip is quantity 100 and is not available for online purchase from Analog Devices WEB site.
Pioneer has ad2s90ap in stock with the quantity 1 price as $44.86. You probably want an ad2s99 reference oscillator @$10.73 too, but out of stock.
At a guess a 2-axis board with just quadrature output could probably be hand wired for around $125 with all new parts.

More thinking is in order; such as serial out to a computer.

I have looked at these, and someone finally got smart! The cost for a 12 bit board back in 99 was very high. It took 2 8 bit convertor chips, a special chip to add their outputs together, and a high priced osc. The 8 bit chips were on the order of $300.00 each as I remember, and the adder was about the same.
Bill Higdon

On Wednesday 30 October 2002 22:53, Bill Higdon wrote:

Keith Bowers wrote:

AD2S90-EB PRODUCTION Evaluation Boards - COMMERCIAL $89.06
AD2S90AP PRODUCTION PLASTIC LEAD CHIP CARRIER 20 COMMERCIAL $29.16

-------
I expect to need a 2-axis resolver to quadrature board, so I did a bit
more checking.

The evaluation board price above is quantity 1.

Unfortunately the above price for the chip is quantity 100 and is not
available for online purchase from Analog Devices WEB site.

Pioneer has ad2s90ap in stock with the quantity 1 price as $44.86. You
probably want an ad2s99 reference oscillator @$10.73 too, but out of
stock.

At a guess a 2-axis board with just quadrature output could probably be
hand wired for around $125 with all new parts.

More thinking is in order; such as serial out to a computer.

I have looked at these, and someone finally got smart! The cost for a
12 bit board back in 99 was very high. It took 2 8 bit converter chips,
a special chip to add their outputs together, and a high priced osc. The
8 bit chips were on the order of $300.00 each as I remember, and the
adder was about the same.
Bill Higdon

Another thing I have been contemplating is the Maxim MAX157/159 10/20 bit A/Ds
along with something like a DS89C420 (33 MHz clock) or maybe the Basic-X
stuff to do the arithmetic, but I haven't come up with a simple way to read
two signals at the same time since the samples must be perfectly synchronized
to the peak of the reference signal. By the time you add a sample and hold
chip in front and then read each sample things start to get complicated.

A really interesting solution would be the one TI has in a TMS324F240 DSP
application note. The approach could be really interesting if it was
completed. They don't appear to publish complete code for what they do and do
not include quadrature output in the design. The code for one axis, including
reference signal generation, used less than 25% of the processor, so it would
appear that a 2-axis board might be hand wired for about $50. As usual TI
wants a bundle for the associated compiler, but there is a GNU version at the
usual price 8o). One would also have to come up with a way to burn the
program into a chip. I have never gotten into DSP and would really prefer to
make chips with machine tools.

The more I write, the better the AD2S90s look <g>.
--
Keith Bowers - Thomasville, NC

You might want to look at the Atmel line of microcontrollers. There are
several asm, c, and BASIC compilers around, some free. In particular,
the AT90S8535 has 8 A/D channels, as well as a serial port and other
nifty I/O. Plus they can be programmed in circuit. The Mega line has
even more features.

www.atmel.com

Regards,
JJ

Be Kind, Be Careful, Be Yourself

On Thursday 31 October 2002 16:25, JJ wrote:

You might want to look at the Atmel line of microcontrollers. There are
several asm, c, and BASIC compilers around, some free. In particular,
the AT90S8535 has 8 A/D channels, as well as a serial port and other
nifty I/O. Plus they can be programmed in circuit. The Mega line has
even more features.

www.atmel.com

Regards,
JJ

Be Kind, Be Careful, Be Yourself

-----Original Message-----
From: Keith Bowers [mailto:kabowers@...]
Sent: Thursday, October 31, 2002 11:21 AM
To: CAD_CAM_EDM_DRO@...
Subject: Re: [CAD_CAM_EDM_DRO] Re:Inductosyn DRO was balls

On Wednesday 30 October 2002 22:53, Bill Higdon wrote:

Keith Bowers wrote:

AD2S90-EB PRODUCTION Evaluation Boards - COMMERCIAL $89.06
AD2S90AP PRODUCTION PLASTIC LEAD CHIP CARRIER 20

COMMERCIAL $29.16

-------
I expect to need a 2-axis resolver to quadrature board, so

I did a bit

more checking.

The evaluation board price above is quantity 1.

Unfortunately the above price for the chip is quantity 100

and is not

available for online purchase from Analog Devices WEB site.

Pioneer has ad2s90ap in stock with the quantity 1 price as

$44.86. You

probably want an ad2s99 reference oscillator @$10.73 too,

but out of

stock.

At a guess a 2-axis board with just quadrature output

could probably be

hand wired for around $125 with all new parts.

More thinking is in order; such as serial out to a computer.

I have looked at these, and someone finally got smart! The

cost for a

12 bit board back in 99 was very high. It took 2 8 bit

converter chips,

a special chip to add their outputs together, and a high

priced osc. The

8 bit chips were on the order of $300.00 each as I remember, and the
adder was about the same.
Bill Higdon

Another thing I have been contemplating is the Maxim
MAX157/159 10/20 bit A/Ds
along with something like a DS89C420 (33 MHz clock) or maybe
the Basic-X
stuff to do the arithmetic, but I haven't come up with a
simple way to read
two signals at the same time since the samples must be
perfectly synchronized
to the peak of the reference signal. By the time you add a
sample and hold
chip in front and then read each sample things start to get
complicated.

A really interesting solution would be the one TI has in a
TMS324F240 DSP
application note. The approach could be really interesting if it was
completed. They don't appear to publish complete code for
what they do and do
not include quadrature output in the design. The code for one
axis, including
reference signal generation, used less than 25% of the
processor, so it would
appear that a 2-axis board might be hand wired for about $50.
As usual TI
wants a bundle for the associated compiler, but there is a
GNU version at the
usual price 8o). One would also have to come up with a way
to burn the
program into a chip. I have never gotten into DSP and would
really prefer to
make chips with machine tools.

The more I write, the better the AD2S90s look <g>.
--
Keith Bowers - Thomasville, NC

List Mom
List Owner



Your use of Yahoo! Groups is subject to

Thanks for the suggestion; It looks interesting. In this case we need to
measure two signals at exactly the same time which will require either two
separate A/Ds per axis or external sample and holds. I was looking at the
Dallas chip both because it appears to be fast and also because back when the
8046/8051 family hit the market I did some design with it. I didn't have
access to an assembler or development system, so had to make my own and hand
assemble the code. I may actually remember something about the code. It's
only been 25 years or so <g>
--
Keith Bowers - Thomasville, NC