Still haven't been able to find stand alone brass barrels, anyone have a source?

In USA:

http://www.makergear.com/products/operators-pack

Confusing page but you can order just the brass barrel.

AUS:

https://www.bilbycnc.com.au/DisplayCat. … bCatID=101

A resistor that works as a replacement is AC03000006808JAC00 (6.8 Ohm, 3 Watt).

Edit: Here's a picture of the original and the replacement where you can see the part number: http://i.imgur.com/0URHujUh.jpg

Everyone should do a M303 S190 when they get their printer and enter the PID values into the EEPROM storage.

My friend gave me the nick name tesla when I built a tesla coil in my parents garage at 15 years old and convinced my parents it was safe.

No one knows the part number, Solidoodle seems to consider it to be IP. Best bet is to find the ID of the hole in the heater block, and match a 6 ohm resistor with an appropriate size OD to fit in the block. You want a slight interference fit, and preheat the block before installing the new power resistor. Not sure if SD does this at the factory, which could contribute to them going bad - high thermal resistance resulting in resistor overtemp.

Keep in mind that Lawsy fork may not be what they're flashing currently.

Side note - do you work at Tesla?

Anyone know the part number of the power resistor used in the hot end.  I would like to just fix it myself if I can. 

I am looking at the code right now.

I also see that PID loop only calculates if the error is +/-10.  If the error is +/-10 then it is either full on or full off.  This could be why I had seen Bang-Bang behavior.  So, not a full PID system. PID only when your in that +/-10C window.  Outside of that range it never calculates P, I or D(or if you swing outside of that it freezes the PID terms). 

I do see the limit for the I term in there, but now I see it is not full PID I get what I was seeing.  Still curious why I saw no power being put out when I the temp just dropped below setpoint(P should have been calling for power). 

Also since the resistor failed I may have been fighting that in the tune.

looks like I have more PID tuning to do because mine is not maintaining that level of accuracy.  Do you heat your extruder up to temp for a few minutes before autotuning?

I did, yes.

How close you can hold depends on a number of factors, like the thermal resistance between the resistor and the heater block, the thermal resistance between the thermistor and heaterblock, how well insulated the thermistor is, etc. You should be able to hold +/-2C 95% of the time, though.

From temperature.cpp:

    temp_iState_min[e] = 0.0;

FF control is in my opinion an unnecessary complication for this plant. Properly tuned PID with a normally functioning heater and thermistor holds temps within a degree or two for everyone I know of with a solidoodle.

By the way, watching the graphs isn't a reliable way of figuring out each term's contribution. You should be able to turn on debugging, which will echo actual temp, each term's contribution, and desired heater power to the log window.

When the temp goes above set-point you can tell the I is winding up in the negative direction by watching it just as it falls past the set-point(doing this with D at zero).  In a normal thermal PID system you clamp the "I" at a min of 0 so it cannot windup backwards.

On a normal PI thermal system AS SOON as you drop below set-point P should be calling for power, but it does not.  It waits till the "I" unwinds and then slams the system with 100% power.  This ends up being a 10 C drop, until it calls for power.  If I was clamped at 0 like  it should, then even if you drop 1 deg below set-point it should be calling for at lease a small percentage of power.

ADDED:
Autotune are for static systems.  These printers are both static and dynamic systems.  It really needs a FeedForward+PID system.

Since we know how much plastic we are putting though we can use that to call power ahead of time before the system sees a drop it temp.