Tweet
ADC
....
....
-
-
@JC1 - "...FPGAs can outperform DSPs, says study...."
see https://www.geotech1.com/thuntings/s...highlight=FPGA
post 9Comment
-
Not ADC,Originally posted by TheWizard View Post
Sample must very very fast for low error....Comment
-
I currently use the PicMicro development board from Matrix (click here), I have both Rev 2 & Rev 3. I have been intending to buy an EasyPic4 board for some time, maybe soon.Originally posted by Qiaozhi View Post
- CarlComment
-
The study compares some pretty high-end chips ($200-plus) with pretty high-end power (1-watt-plus) that aren't the least bit suitable for a battery-operated detector. Both DSP and FPGA start out at sub-$10, and you should always buy only what you need, especially keeping an eye on power. It's very possible that a lowly Microchip DSPic is plenty good enough, depending on what you want to do.Originally posted by JC1 View Post
In a current PI design, I am using a PIC chip with a 10-bit ADC (good enough), throttled back to 1MHz (good enough), and pulling only 300uA. My battery will last a very long time.
- CarlComment
-
??????
Originally posted by Carl-NC View Post
So what type of Digital Signal Processing are you
doing to the signal with this setup?
Carl.Comment
-
DSP not suitable
Now if this type of processing IS requireOriginally posted by Carl-NC View Post
do to any REAL digital signal processing
then you are right.
DSP IS NOT SUITABLE FOR METAL DETECTORS.
Still Garrett pulled if off for real.
Now if the discussion is about what kinda of
processing a little 1 MHz pic is going to do,
don't kid me or yourself.
Tell me what you are doing.
I ain't buying the story.
About any real deal.Comment
-
Exponential Moving Average,
Integrator //y[n] = a * y[n-1] + x[n]
Create Audio Tone.....Comment
-
DSP Part....Comment
-
Running average
Hi Wizard,
Yea this is the typical stuff done
by all the detector manufacturers
with their microprocessors which
is NOT Digital Signal Processing
which some my wish to act like it is.
1. Get some Engineering books that say
Digital Signal Processing on the side of
the book and look inside. That IS DSP.
2. If you don't believe that, then understand
that the little slow toy microprocessors
ARE NOT going to yield "state of the art"
design. They NOT going to do all the sharp
filters and noise reductions techniques or any
thing but run an average {easier done with
leaky integrator} and light some lights and display.
Now I have NO doubt NO real DSP will be done
by the Collective design team. We will be lucky
if everybody/anybody even understands what this
really is.
Not someone's made up pointless la la land definition.Comment
-
Hi JC1
What for you is "Digital Signal Processing"
FFT, DCT.... ?
I not need FFT like Garret for Pulse Induction.....
"A digital signal processor (DSP) is a specialized microprocessor designed specifically for digital signal processing, generally in real-time computing."
http://en.wikipedia.org/wiki/Digital_signal_processor
http://en.wikipedia.org/wiki/Digital_signal_processing
http://en.wikipedia.org/wiki/Filter_...igital_filtersComment
-
Are you happy with the features on the PicMicro board, and do you have any comments concerning the MikroElektronica offering?Originally posted by Carl-NC View PostComment
-
DSP Classroom
Originally posted by TheWizard View Post
Hi Wizard,
Just what it says:
a specialized microprocessor designed specifically
for digital signal processing
NOT a general purpose slow toy designed specifically
for lighting lights.
Wizard don't take me wrong, remember this is still
left over from the Design Dream Team stuff,
where Sean Goddard who at least does understand
what is meant by any Serious DSP was proposing a
DSP programmer and a laptop computer. That didn't
seem to be a problem, though 1 Watt now does, even
though Minelab PIs use more that 1 WATT. sure you
have to carry a GEL Cell battery around, but heh
you know what????? Guys WILL carry that battery
around if the detector works well.
But the real deal is I didn't buy into some kinda of
DSP wet dream. And I said something about it.
And I'm still waiting from the beginningg o0f this
story, for anyone to tell me WHAT they in their dreams
are going implement in DSP. And then I am going
to tell them just how hard or easy that really is to do.
Now I am not in any way putting down what you
have done. I think that is great !!!! because there
are many who actually do very little. And a moving or
running average is very typical with a general purpose
microprocessor to do. But it is not BRAND NEW STATE OF
THE ART, which is about the only specification that the
new great sell and get rich detector has at the moment.
It was not to be constrained in design, but already the
processing must not be capable, but the first design
requirement is that it be cheap and low power.
So now it will be easy to show that the new great state of
art detector will be nothing.Comment
-
JC1, you very underrate us...
Ok I understand you, Now the MCUs is very fast and DSP core is not need,
but Integrator like y[n] = a * y[n-1] + x[n] is the basic of Digital Signal Processing....... never mind....
In my old PI design I use not analog integrator, How I make this...?
How I make noise reduction, detection, tone creation....
I make many test and source is optimized to normal C algorithm(I have free MCU time for this and use not DSP core, but if I need....
Yes, Power Supply... fast MCU - current is 200 mA, I think and about this...
MineLab - GP3500 - 2,400kg, 12Ahr - 10h work...
JC1, I make this research more 10 years... Now technology is ready for this...
If Minelab, Garret can not do this - some other CAN or will try.....Comment
-
...
Garrett did do this. With a REAL DSP CHIP,Originally posted by TheWizard View Post
Analog Devices DSP part number is in Patent.
Think it is 40 MHz part. But to be doing DSP
you are going to have to do more than integrate
the ANALOG OUTPUT. That is my point.
Plus to have new "state of the art" DSP
detector you will have to do everything
in software, the front end will only be an
amplifier. Nothing more.
And I'm glad Carl's new low power detector
only uses 300 uA of current. Which seems
like the Xmit power is a bit on the low side for
this to be a very exciting detector.
But no this is going to turn in a big ego thing
real quick, I know what level of effort is required.
But we will see many years from now if anything
comes from this endevor.Comment
Comment