PID question

Any topics that don't fall into above categories. Please note that your thread may be moved to other discussion boards.
firesped
Posts: 594
Joined: Mon Mar 21, 2016 9:23 pm
Location: Tucson, AZ

PID question

Postby firesped » Thu Mar 02, 2017 11:11 pm

I'm trying to figure out what my Trinus is telling me on the M105 report line.

It is showing this.

Send: N451 M105*39
Recv: [VALUE] T:200.0/200.0 B:26.5/0.0 S:0 F:'' P:-1 I:0 D:100 E:100

other people are assuming the P, I, and D are showing the PID information but these numbers don't seem to be acceptable numbers for the PID.

My question is what would a Kp:-1 Ki:0 and Kd:100 actually do?
RL name: Michael Nolen
printers:
raise3D N2 kickstarter Early Bird
Trinus Deluxe (running smoothieware on Azteeg X5 GT board)
Monoprice Maker Select v2

Jetguy
Posts: 1238
Joined: Tue Mar 22, 2016 1:40 am

Re: PID question

Postby Jetguy » Fri Mar 03, 2017 3:00 am

So PID values and what they mean:
P is proportional. This means say your set point is 200C and the reported temp is 190C making a 10C difference. The percentage the heater PWM is on is proportional to the temp difference from target. Most PID loops don't even operate until within 10C anyway. To understand that, if the temp difference is 10C or greater- then the heater is 100% on. As you cross the 10C threshold, you begin proportional tuning. So as you near say 2C away from target temp, the heater is proportionally less on and as actual and target match= heater OFF. Now seeing a negative 1 value, I would think you have a 12V heater being pushed at 24V in a dangerous situation. The heater would actually draw 4X more power (double Voltage times double Amperage=4X Power) and thus heat insanely fast. So that's why proportional would be a less than one value and in this case negative- so something you might want to measure that resistance of that heater.
I is integral. The way I explain this value is that this is time related. The longer the reported temp is lower than target temp, the higher this value keeps cranking up the heater power. So say you are within 2C of target temp, but instead of creeping up to the target temp, you just kind of hang there (a common complaint of untuned systems AKA heater sag). The idea is that this value again, says I've been waiting far too long and so proportional is not getting it and I'll crank the heater on full blast to get us to target temp. Again, this is a TIME based factor, where as proportional is just based on the difference between target and actual temp
D is derivative, but I've also seen it called Dampening. The idea here is this is also a time based values, but also proportional because what it's measuring is the rate or speed that we are approaching the target value. So say your heater is really overpowered and heating up really fast. Proportional is going to turn the heater down as it narrows the difference, but it's not fast enough and we would overshoot and end up hotter than we wanted. So this sees that last little bit of difference and the rate (change in temp over time) and turns down the heater output to slow the rate to not overshoot.

Honestly, we typically see proportional "P" as a lower value, say 1 to fractions of 1 (but can be other values), and "I" can be single or double digits, and yes, quite often D is sayi double digits and even 100 in some cases. So the values don't seem grossly off, although a negative 1 value for P find curious but maybe that's just that system. Or as i said, that could mean a overpowered heater. Just something to check.

Jetguy
Posts: 1238
Joined: Tue Mar 22, 2016 1:40 am

Re: PID question

Postby Jetguy » Fri Mar 03, 2017 3:04 am

Again, the main difference is P and I are values that turn the heater on.
D is the term that tries to turn down the heater to prevent overshoot.

Because the various terms measure different aspects of heating (time and temperature, and then rate) they work together to smoothly control the temp rather than bang bang ON OFF systems.
Example, a cheap thermostat, would just say that anything less than set temp is 100% on, and then every time you would overshoot, then wait for it to cool back down and bang on full blast again.

firesped
Posts: 594
Joined: Mon Mar 21, 2016 9:23 pm
Location: Tucson, AZ

Re: PID question

Postby firesped » Fri Mar 03, 2017 8:23 am

Thank you for the answer on this Jetguy. I think the Trinus community is looking at replacing the boards and are looking at going with smoothieboards, which will cover all the applications for the device, it already has a laser engraver and a 3d printer and there is a third option they are working on, which I think is a CnC router. The unit came with a 120 watt external PSU which seems to be ok for printing PLA but not enough for anything requiring higher temps.

Many users are upgrading their PSU to bigger ones but my concern is if PID is doing as you say and is dangerous that it may just make it even more so.

I'm thinking of picking up the Azteeg X5 GT for the trinus but I may also look into it for upgrading the N2, if I can get it to talk to the touchscreen. I'm not sure how much firmware adjustments I can do.
RL name: Michael Nolen
printers:
raise3D N2 kickstarter Early Bird
Trinus Deluxe (running smoothieware on Azteeg X5 GT board)
Monoprice Maker Select v2


Return to “Others”

Who is online

Users browsing this forum: No registered users and 1 guest