My heated bed WAS working ok just took forever to reach temp (25 minutes). So when I overhauled my I3-2020 I flipped my heater over so grid side was up added a piece of cardboard between aluminum plate and PCB, then ran PID autotune with plugging the new values in my firmware. Well now I can not get the bed to 100 degrees by itself. If I use a heat gun to get it up to temp then the temp will not be stable for first few layers but then stays at my set point. I ended up removing the cardboard, re soldered the power wires, added high temp copper RTV around thermistor keeping it against glass build plate. Maybe when I removed wires I damaged the PCB?

Do these tune numbers even look correct?
KP 317.49
KI 23.77
KD 1885.6

This is on a mk2 PCB (the red pcb heater)

I don't think there's anything wrong. A PCB is a glass fiber epoxy which is a good heat insulator. When placing the thermistor on the component side the thermistor is thermally insulated from the track side of the pcb. so the temperature indicated by the thermistor is now different.
Another effect of flipping the pcb is that all the heat is going into the glass plate instead of heating up the thermistor directly and the air around it. The heat capacity of a sheet of glass is much higher compared to air, so it's harder to reach higher temperatures. But all that matters is the temperature above the glass plate, this is where the glass makes contact with the filament.

If you need higher temps and faster warm up times you'll need a 24V psu for the heated bed. I used a 12V mk2b with 24V and pwm with a duty cycle of 75% to limit the current and works very well. 110 degrees under 4 minutes, 140 degrees in 10 minutes. I measured with a thermistor under the heated bed, tracks side facing up. Actual glass side temps are 10-15 degrees lower.

My pid values are not suitable for comparison as my setup is very different.

Edited 2 time(s). Last edit at 08/26/2015 08:21AM by imqqmi.

I tried
kp 10.00
ki .023
kd 305.4
Could not get the bed over 85 degrees.
So frustrated I ordered a new PCB heater in hopes it was my fault and I damaged the bed trying to unsolder the wires when i flipped it.

I have seen people with this issue change the thermistor from table 6 to table 1 will try it.

Edited 2 time(s). Last edit at 08/26/2015 08:26PM by jasonsaffle13@gmail.com.

You'll need a temperature reference if you switch tables as the given temperature may be way off. If you own a dvm with thermo couple, that would be great. Place the thermo couple on the thermistor and check every 10 degrees. You can also adjust one of the thermistor tables to calibrate your particular thermistor. I did it with mine and checked with two thermo couples, the measurements were pretty close. Even with a matching table, a thermistor can be 10 degrees off. A thermocouple is much more reliable when using it with a thermo couple board. It's usually within 2 degrees C. I use one for the hotend. It can now run at higher temperatures than thermistors allow, which is great for printing nylon at 270 degrees.

12V isn't going to be great for the heated bed though, 24V is more efficient. Boosting a psu to 14V may just do it but it's still going to heat up slow and be subject to environment temperatures. You also want it to be stable, especially with ABS.

Quote
therippa
Quote
Lagomorph
Quote
therippa
Quote
Bert3D
So I upgraded to the latest marlin code posted earlier in this topic, and now by bed won't go much above 85 or 90 degrees, and takes much longer than before. Has anyone fixed this? What's the trick?

Also, is the throat on the extruder a 26mm or a 30mm?

Re-run the PID autotune for the bed.

I had the same problem until I changed the thermistor settings in Configuration.h to use table 1 instead of 6. At first I couldn't get a temperature above about 86C. I switched back to the original code and it was working again. Then I ran a diff on the new code versus the one from Folger Tech and saw a lot of changes to table 6 in the thermistor tables. (and whats with that note in the configuration about table 6 being not as accurate as table 1? Why even have it in there then?)

Also, does the PID autotune make any difference for the bed? It looks like the configuration is set to use bang-bang for the bed as #define PIDTEMPBED is commented out. On mine the LED for the bed is only on or off, it does not blink like the LED for the hotend.

Not sure if it makes a difference, but I also replaced the heat bed power wires with #14 gauge wires.

Interesting. I'll try changing to table 1 too, and if it works better I'll update the firmware settings on github.


SO HAPPY THIS WORKED!!! Though I will have an extra black mkb2 PCB heater now LOL!!! I changed the bed sensor from 6 to 1 and the bed heated up to to 105 in about 12 minutes!!!!!

The question is though, if what is displayed is the actual temperature or if it's off by 20 degrees. I guess you'll find out when you print abs and it still warps at 105 nay 85 degrees C.

Checked the glass on my bed with my laser thermo it would show from 2-5 degrees of my thermistor reading in repetier!!! Since this is on the glass I'll take it!! I guess I'll stick with profile 1