Searching \ for '[OT] Digtally interfaced Analog Math circuits, are' 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/method/math.htm?key=math
Search entire site for: 'Digtally interfaced Analog Math circuits, are'.

Exact match. Not showing close matches.
PICList Thread
'[OT] Digtally interfaced Analog Math circuits, are'
1999\06\13@160044 by Graeme Smith

flavicon
face
Having a limitation with clerical tasks, I ran into a problem with analog
electronics in college... there was just too damned much math in it, ...
I burned out. (other personal factors didn't help... divorce, kids being
taken away etc....)

Anyway, problem is, I burned out before I got to understanding active
analog circuits very well. Now I have a function I want to play with,
that might be more efficient as an analog circuit....

I suspect I need at least 3 opamps... two unity gain, with both positive
and negative bias and one variable gain... of course I am not sure how to
rig them as multipliers.... etc.


It involves floating point multiplication so its probably too expensive
timewise for a PIC.

So... Assuming I could regulate factors such as powerline voltage, and
temperature fine enough to approximate accuracy....

An analog circuit MIGHT be more to the point.

I would want to supply some sort of 8 bit DAC inputs, and read an ADC
output.

I haven't tried this yet with a PIC, so what I want to know, is....

How long does it take to:

1. Supply an 8 bit DAC value ( I would need two) (PIC?)
2. Stabilize a relatively simple analog circuit to a particular value
after input. ( a variable gain multiplier ) Actual formula is proprietary
figuring out how best to generate the gain, is probably beyond me right now but
I can learn how if it is worth it.
3. read the analog value out using an ADC (PIC?)


in comparison how long would it take to:
lookup a 16 bit Float,and an 8 bit number using an 8 bit value as index
(Its an orthoganol function so I only need to store 128 values for each)
768 bytes total? seems steep for a PIC!

Subtract the lookup from the second value
do a 16 bit floating multiply to
an 8 bit number that was the result of the previous subtraction
add a range adjustment check it fits within the range using another lookup
or two from the same table as the value subtracted came from
getting the result, an 8 bit integer


I marked this OT, because I suspect that I am not going to be able to use
a PIC for this. (too slow for one thing) However, if it works out that I
can build a parrallel processor that does these things in a pipeline like
fashion, maybe the actual PIC timing would become minimized as a factor.


The application I envision for it is an accelerator function, for a large
data array calculation similar to a neural network If I can offload some
of the math to a peripheral accelerator circuit, it might speed the
base application, or allow me to use this function in smaller circuits.

I am on welfare... so COST is DEFINITELY a factor.... which is why I will
probably end up building this circuit myself... some year....

                               GREY

GRAEME SMITH                         email: spam_OUTgrysmithTakeThisOuTspamfreenet.edmonton.ab.ca
YMCA Edmonton

Address has changed with little warning!
(I moved across the hall! :) )

Email will remain constant... at least for now.

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...