EEPROM Settings and mins and maxes?

[Circus621]

Hey guys,
I'm searching the forums and internet for clarification on the EEPROM item definitions and mins and maxes and I'm not really finding anything but conflicting information. Is there a breakdown on what the settings do and what the mins and maxes are? I know a stock head has a max temp around 240-255C. Im working with my extruder PID settings and cant seem to get a solid understanding of what they do. Im not needing formulas or anything but my understanding so far is the pid max (theres 2 of them for some reason) increases power to the resistors. One would be power the other temp (1-255)? The pid gain increases the power when there is a deviation from the set point and the I and D can be left alone for the most part? If there is a breakdown somewhere I would love to look at the setting descriptions and try to wrap my head around it.

[Polygonhell]

From the source code it looks like PIDDriveMax is the maximum allowable PWM setting i.e. it clamps the output of the PID calculation.
PIDMax scales the output of the PID calculation i.e. it's some sort of normalization.

Both are 0-255 255 is always on for the PWM pin.

[Eric]

Look through Configuration.h for the comments about various settings.

And https://www.repetier.com/documentation/ ... tallation/ should be helpful to you too.

[Xenocrates]

So, Power is how much potential energy will be driven into the heater. It's a term that goes into a somewhat complex equation, with I and D. Now, a higher P only goes so far, as PID max sets a cap on the scaling of the equation. PID Drive max sets a limit on the duty cycle of the heaters (At least, that is my understanding). I is the interval, and determines response time. So if you have very large sags and overshoots, and they go on for a while, you need to adjust I. D is Damping, and will slowly settle the rest out, so you'd change D if you had rapid oscillations. This allows you to eventually get to a very stable temperature (For example, I currently get oscillations of ~.3C at a set temperature of 235, and they are rather slow, so I'm less than concerned by it)

[Polygonhell]

PID stands for Proportional Integral Derivative, not Power Interval Damping.

Everything is based on the difference between the target temperature and the actual (the error),
P basically is multiplied by the difference to give a direct relationship,
I is based on the integral, in this case that's the product of the error term and the time the error has been outstanding, without it as you approached the target you would never reach it as the P term would tend to 0, or you'd overshoot and oscillate wildly.
D is based on the rate of change of the Error, and can be used to damp out oscillation you can't tune out any other way. Generally for heat control loops it should be small the bulk of the tuning based around P and I.

[Circus621]

So, Power is how much potential energy will be driven into the heater. It's a term that goes into a somewhat complex equation, with I and D. Now, a higher P only goes so far, as PID max sets a cap on the scaling of the equation. PID Drive max sets a limit on the duty cycle of the heaters (At least, that is my understanding). I is the interval, and determines response time. So if you have very large sags and overshoots, and they go on for a while, you need to adjust I. D is Damping, and will slowly settle the rest out, so you'd change D if you had rapid oscillations. This allows you to eventually get to a very stable temperature (For example, I currently get oscillations of ~.3C at a set temperature of 235, and they are rather slow, so I'm less than concerned by it)

Are results like this normal after an Autotune?

[Xenocrates]

Are results like this normal after an Autotune?

Nope! That's not autotuned at all, as the Duet lacks an autotune, and that's what I'm using. As far as autotune goes, I usually found ±2.3C swings. Just persistence in tuning, and a lack of drafts.

Polygonhell, I knew I had the term names wrong (That's what I get for not looking it up), but it's also the way a lot of the industrial stuff describes it, the same way they say "Triangle math" instead of Trigonometry.

[Polygonhell]

Auto tune I find to be about as useful as a random number generator.
It's usually an OK starting point, but I always end up manually tweaking the results, you can get <1C variation with some messing about.

[Circus621]

Look through Configuration.h for the comments about various settings.

And https://www.repetier.com/documentation/ ... tallation/ should be helpful to you too.

This is what I was hoping to get. Thanks!!