Strange Temperature problem ... reposted from firmware folder...

I haven't recieved any responses in the firmware forum, and am desperate for some ideas...
====
I'm running an older Botmill (Mendel Prusa), with Gen6 w/Marlin, MK-IVB hotend, 100k thermistor and I'm having a devil of a time controlling temp. I've been able to print, but was always baffled by having to use 220 deg C just to print PLA. Anything below that and the extruder would jam and skip. Once I got past that temp, it would print, but never to a quality you'd like. I was searching the net and it made we wonder what temp I was actually getting at the hotend vs the pronterface reading. I got a temp probe and when pronterface says 235 deg C, the hotend is ~135 deg C - BIG DIFFERENCE. At room temp both pronterface and the temp sensor read the same ~24 deg C, which is accurate. I used the thermistor supplied with the hotend, and have read repeately that I'm to use thermistor table 1, which I always have. The marlin config for temp looks like this;

#define HEATER_0_USES_THERMISTOR
#define THERMISTORHEATER_0 1
#define HEATER_0_MINTEMP 0
#define HEATER_0_MAXTEMP 275
#define BED_CHECK_INTERVAL 5000
#define TEMP_RESIDENCY_TIME 10
#define TEMP_HYSTERESIS 3
#define TEMP_WINDOW 1
#define TEMP_SENSOR_AD595_OFFSET 0.000000
#define TEMP_SENSOR_AD595_GAIN 1.000000
#define CooldownNoWait true
#define HEATING_EARLY_FINISH_DEG_OFFSET 1

I checked wiring - seems ok. I used the PID settings that came wth the marlin setup for this machine;

#define PIDTEMP
#define PID_MAX 255
#define PID_INTEGRAL_DRIVE_MAX 255
#define K1 0.95
#define PID_dT 0.128
#define PID_PID
#define DEFAULT_Kp 63.0
#define DEFAULT_Ki (2.25*PID_dT)
#define DEFAULT_Kd (440/PID_dT)
#define PID_ADD_EXTRUSION_RATE

But I've have read I need to adjust those. Could this be the problem? I entered M303 in pronterface (as suggested in another post), and while it sent it, nothing happened after that - no response (was I supposed to get a response?). Then I entered M301 and got an instant response that simply matched the original settings.

I've had the machine for 9 months, but still feel like a nube on it. Any thoughts on this will be greatly appreciated....
T.
========

---

in beautiful southern, Maryland...

I never thought a thermister could read high, low I can understand.

I think I have the same problem as you, I just haven't bought a laser thermometer yet.

Seller says filament should extrude and look good at 170-180 but I have to get it up to 220+ to get it to extrude properly. Problem is my hot end wont stay above 220 when printing, when it hits fill it drops to 200-210. I wonder what temp mine is actually getting when it reads 220 or 230.

I need to get a laser thermometer and see what the actual temp of my hot end is.

I'm using a printrboard Rev D running Marlin.

Edited 1 time(s). Last edit at 03/27/2013 10:55PM by hendo420.

more info: Instead of focusing only on the software portion of the potential problem, I looked closer at the thermistor. I had some other 100k thermistors I used on the heatedbed (Vishay-NTCLE203E3104), so I taped one to the side of the hotend, near its thermistor and connected it to a DVM. Then I heated the hotend to ~150, unhooked the hotend thermistor from the Gen6 board and put another DVM on it. I watched the resistance readings of both meters as the hotend cooled. Since I had the mfg's temp table for the taped-on thermistor, I could coorelate the readings, or so I thought.

Readings;
(Both DVMs were set at the 200K range setting.)
Hotend Thermistor Taped-on Thermistor Temp from taped-on therm spec sheet
.6 1.9 ~122
.8 2.2 ~119
.9 2.4 ~116
1 2.7 ~113
1.1 2.9 ~110
1.5 3.9 ~100
2.0 5.1 ~92
5.0 10.6 ~71
10 18.3 ~58
20 31.7 ~44
50 62.3 ~32
65 76.1 ~23
102 103 ~18

Since I have no idea what thermistor was supplied with the hotend, and it doesn't seem to match anything in the thermistor tables - I'm lost. So I figured the best would be to replace the thermistor with something known, so I tried to buy what I thought was the correct thermistor (104GT-2), but what I received was a (104GTA-2). There is no mfg temp table available anywhere that I could find. I've tried to use the "createTemperatureLookup.py", but I'm not up on python, so I'm not sure how to enter the command, nor where is puts the output file. For that matter, I'm confused on the entries in the "thermistortables.h" file. Is there someplace that defines what these entries represent?
ANY THOUGHTS ARE GREATLY APPRECIATED....
Thanks....

---

in beautiful southern, Maryland...

Curious,

I'm new to this whole 3D Printer scene.

Why do people use thermistors and not RTD's?

RTD's temperature to resistance is at least linear.

So a couple of thoughts. Most likely it is the wrong
temp table for that specific thermistor.

The temp tables in Marlin are based on a 4.7K resistor being in-line with the thermistor. I believe the Gen6 has
this (I have printrboard). But there is at least one printer where the resistor is 10K. The comments in the Marlin
code note this in the temp table selection.

The following web site will generate a temp table for you:
[nathan7.eu]

I just pasted the new one into the correct .h file and gave it it's own number.
The nice thing about that web site is you can generate a much more dense temp table (100 entries).
However, note - in Marlin (or at least the version I have) the adc readings have a multiplier
in them.

Also, check your actual temp table code. I recall reading and seeing a bug in one of the Marlin versions
where the ADC or Temp value was incorrect. The ADC value should be in sequence.

Thanks everyone for the help. I got it fixed. Was a mixture of correcting the table and measuring the actual temps with a thermocouple. And as nophead said, here and elsewhere, the key is measuring from the inside of the hotend, not the outside. Big difference. I also got the PID values corrected, so now I have rock solid temps.....

---

in beautiful southern, Maryland...