Controlling temperature of reprap extruder hot-end

More
23 Sep 2011 11:40 #13383 by fma
Mmm, such log-amplifing circuits don't look easy to build! I think a software-based linearizer is much easier...

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 11:41 #13384 by Rick G
Try looking at Analog Devices AD594 or AD595

Also their Application Note AN-369 on the circuit.

Rick G

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 11:43 #13385 by Rick G

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 11:46 #13387 by fma
This chip is for thermocouples, not for thermistors...

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 12:15 #13388 by andypugh
fma wrote:

BTW, is there a way to linearize an signal?...
If not, I guess I will have to write this component?

I keep toying with the idea of writing a 1D / 2D interpolator for HAL.
(I have written a few in Excel for my job, I write engine maps)
Perhaps I will look into it in more detail this weekend.

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 14:34 #13395 by Rick G
Yep, thermocouples small, rugged, inexpensive. You can build the circuit to be very accurate in the temp range you want to work.

But there maybe a software solution also.

Rick G

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 19:13 #13406 by fma
@andypugh,

that would be nice!

@Rick,

A friend will give me a platiniul sensor, wich is even less expensive, and linear enough for my needs. I will just have to drive the NE55.

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 19:41 #13408 by cncbasher
a MAX6675 is a much better thermocouple solution ( single chip & 2 cap ) , and theirs plenty of circuits and code available

Please Log in or Create an account to join the conversation.

More
23 Sep 2011 22:04 #13414 by fma
But I would like to avoid using an external ยต-controller, and instead use EMC.

Please Log in or Create an account to join the conversation.

More
24 Sep 2011 18:47 #13426 by andypugh
I had all sorts of ideas, but in the end ended up with something very simple indeed.
component polynomial "Applies an up-to-7 order polynomial transformation";
pin in float in "input value";
pin out float out "output value";
pin in float coeff-#[8] "polynomial coefficients";

function _;

license "GPL";
author "Andy Pugh";
include <rtapi_math.h>;

;;

FUNCTION(_) {
    int i;
    out = 0;
    for (i = 0 ; i < 8 ; i++){
        out += coeff(i) * pow(in, i);
    }
}

Should actually do all that you need.
The actual values you need to use would probably be most easily obtained using Excel or Open Office, using a curve-fit on a graph of calibration data.
As you will have an (unknown to me) transfer function between the value in HAL and the resistance of the NTC I can't suggest any numbers at the moment, but if you have any calibration data I can show you what the HAL snippet would need to look like.

Please Log in or Create an account to join the conversation.

Time to create page: 0.233 seconds
Powered by Kunena Forum