KNozzle - Hot on startup

[CNC_Machines]

Greetings,

Just plugged in my KNozzle to test it out. As soon as I run "KNozzControlTemptsMCode.c" the watchdog signal kicks on, and the heater turns on. Checking the console, it says the nozzle and bed setpoints are 0C, and I dont have any thermistor hooked up. Does this seem right? The documentation is slightly confusing about where the power goes. Is my image here correct?

Thanks,

Scott

[TomKerekes]

Hi Scott,

The setpoints would be the setpoints you set. Did you configure an MCode to use Var 30? What values did you set with the MCode P Q parameters?

The ADC reading should be 0 with out any resistance connected (open circuit). See here.

I don't see any image.

[CNC_Machines]

Hi Tom,

I tried to add a picture a number of times last week, but it wouldnt attach. It shows in the bottom, but when I click the "preview" button the attachment is removed. I tried a JPG, TIF, and PNG file format.

No I didnt use an M-Code, the P and Q parameters are both Zero. That was why I was surprised that the coil started to heat.

Thanks,

Scott

[CNC_Machines]

Looks like the attachment worked !!

[TomKerekes]

Hi Scott,

The default KNozzControlTemptsMCode.c code assumes a Thermister 3rd order function of:

Code: Select all

// Convert Temperature in C to equivalent ADC Counts
// based on 3rd order data fit to this measured data
//
// Temp C, ADC
// 16.0	199
// 35.0	400
// 49.1	600
// 54.7	700
// 60.7	800
// 66.9	900
// 73.6	1000
// 78.4	1100
// 85.2	1200
// 93.8	1300
//107.5 1482
//115.5	1570
//125.0	1688
//132.0	1722

// function to convert Temp C to ADC counts
float TempToADC(float T)
{
    return  ((-0.000837 * T + 0.170202) * T + 4.565045) * T + 78.452228;  // 3rd order polynomial
}

Here is a plot of the Function. This polynomial was intended to be used over the range shown here

A setpoint of 0C corresponds to an ADC reading of 41 counts. Without any Thermistor (infinite resistance) an ADC reading of 0 counts would be "cold" and therefore activate the heater.

It would probably make sense to add a test for a nonsensical ADC reading (ie < 100 counts) and disable the heater.

[CNC_Machines]

Thanks Tom, I agree.. I would be afraid that in the event of a wire failure we would roast the print head.

[CNC_Machines]

Tom,

The 3D printer I am making needs to be able to handle temperatures close to 300 (475 F) degrees C. So I will need to use a different function to convert temp to ADC. Here is the one I am trying to use(part number 104GT-2) sold on Amazon, rated to 300 degrees C.

https://www.mouser.com/datasheet/2/362/ ... 202937.pdf

My plan was to plot the profile, and make an equation for the function. Here it is:

This is my first time trying to profile a thermistor and I am not certain I did it right. If I did, my concern is how flat the curve is at the upper range of temperature. At a fraction of a KOhm the temperature jumps a lot. Do you think this would make for stable control?

If you think it will work, my next step is to convert KOhms into ADC counts, and put the equation into the function?

Thanks,

Scott

[TomKerekes]

Hi Scott,

Amit tried to use the same sensor some time ago and had poor resolution at higher temperatures. There was a discussion on cnczone here.

In the Thread there is an Excel sheet to calculate KNozz ADC counts vs Temperature the result is copied below.

The sensor resistance becomes too low when used with the 10K ohm resistor built into KNozz. A better solution is to pick a sensor with a higher resistance such as the 105GT2. Or add a 1K resistor to the KNozz Terminals to reduce it's resistance.

[CNC_Machines]

Thanks again Tom, always very helpful. I have been reading through data sheets for thermistors. I cant help but wonder if this is really stretching the upper limit of their usable range. Do you think I might be better off using a thermocouple instead? Problem is I would need to get a thermocouple amplifier and get it to communicate with the KFlop. Any thoughts on this?

Scott

[CNC_Machines]

Tom,

I am now looking into an "RDT". Looks like this may be the way to go. Looks like it has a more linear profile over a larger temperature range. Platinum wire inside. Looks like this should plug directly into the Knozzle?

They also appear to be pretty cheap, still trying to understand the datasheet on how to characterize the profile for one of these.

100 Ohms RTD Platinum ±0.3% ±3850ppm/°C -50°C ~ 400°C 2-SIP

Here is a quote from the website - "The DIN standard specifies a base resistance of 100 ohms at 0°C, and a temperature coefficient of .00385 Ohm/Ohm/°C."

If I understand this correctly the temperature is linear from 0 degrees (100 Ohms) to my target temperature 230 degrees (188.55 Ohms). Will that be a good resistance range for the Knozzle, or will I need to change the resistance?


https://www.digikey.com/product-detail/ ... ND/6677753

https://www.te.com/usa-en/industries/se ... oMQAvD_BwE

Thanks again!

Scott