Manual PID tweaking

[Jimustanguitar]

Hey guys! I'm printing some ABS lately, which I haven't done in a while (at least one firmware upgrade ago), and I'm having trouble getting my hot-end to keep up to temp. I'm running an E3D V6 with a 40w heater cartridge, and I'm printing at about 240, but only able to get it up to between 220 and 230 when it's running. I'm printing at 60mm/s and .3mm layer heights, so the volume of plastic that I'm pushing is taxing the heater circuit.

I've done the PID autotune, and I know that my temp is accurate because I'm using a thermocouple, but it seems like I'm still missing something. What PID values can I tweak manually to get a little more heat out of my nozzle?

I also remember Oly saying something about the newest firmware having one of the settings wrong, but I don't recall what it was. Perhaps it was the I drive max?

[DLS3141]

You could try manually upping KI, which is the PID term that the controller uses examine the difference between the temperature history and the setpoint, the longer the system runs under temp, the greater this gain becomes. Increasing KI tends to eliminate steady state error in a system.

However, if your system is already powering the heater constantly (100% duty cycle), your heater may simply be underpowered and unable to keep up with the heat being removed by the mass flowing through the hot end. I've had similar issues on industrial systems and had to either add a second heater or a larger heater.

[JFettig]

PID autotuning is pretty worthless, keep the P value, bring the I and D values down around .5 and dial it in from there. I utilized repetier server(host might also work) to dial it in live while I printed and got it to hold a relatively constant power % on my nozzles.

What you're looking for - P gets you up temp and slows the heating, I controls the slow wave oscillation, D is rapid. You shouldn't need much I or D. P takes care of most of it. You only want to use I for the slight corrections and D will make it a little more rapid. Every time you retract and move, the temp will shoot up since the filament stops moving, then you start extruding again and the temp undershoots, thats where you need the D number tuned in.

[Jimustanguitar]

I'd assume that 255 is the max value for the drive max if it's at 205 right now?

I'll look up the other values and figure out what's firmware and what's EEPROM this evening. I'll also measure the resistance on my cartridge. I've got more than one, and I don't remember if this one is OEM E3D or not.

[Jimustanguitar]

Well... I discovered a few things. For one, I had a broken wire going to my YellowJacket board, so there was a higher resistance on the heat circuit as a result. Repairing the wire helped. I also had my drive max at 205 instead of 255, so I've upped that. I didn't see the KI value in my EEPROM, so I'll save that note for my next firmware update.

The other thing that I discovered is that my newest design for the fan ducts blows a lot of air right at the heater block, so I designed a quick ring shaped deflector to solve this problem.

After it's done printing, I'll get on with the high temp / high speed print that I was attempting the other day and I'll see what happens.

[Jimustanguitar]

Keeping air off of the heat-block helps a ton.

[img]http://i.imgur.com/31HUWg1.png[/img]

[bvandiepenbos]

Keeping air off of the heat-block helps a ton.

[img]http://i.imgur.com/31HUWg1.png[/img]

Nice! Jim, I have been wanting to do something like that but with a metal cone shaped heat shield.