Calculator For Temp Mods

[roller]

If I were to do a temp mod I would use a trimpot so I could dial in the temp I want but it's annoying to read a resistance off the trimpot and calculate the temperature I want so I made this spreadsheet. It presumes you have a switch on your mod and can switch between normal temp and using your trimpot to reduce the temp. So how do you use it:

- Set your switch to normal operation and read the starting temp of your Up (effectively the extruder at room temp reading)
- Put in the desired extrusion temp
- Flick your switch and dial your trimpot until you reach the temp the spreadsheet tells you. Remember there is a slight delay as the temp only gets updated every second or so (never paid attention to exactly how long it takes to update)
- Switch back to normal and check your start temp hasn't change (give it a few moments to update) and if it has re-enter in the spreadsheet and dial in again.

The target temp is accurate to +/- 0.2 degrees so don't get too obsessive about getting that decimal place too close. I recommend doing it before you start heating anything otherwise you will find you're chasing a moving target with the start temp and it gets a little annoying.

Before anyone asks I would recommend something like a 100ohm multiturn trimpot to make it nice and easy to adjust.

Here's the spreadsheet:
https://dl.dropboxusercontent.com/u/652 ... stment.xls

For those that just want to know what the resistance should be at a given temp the formula (accurate to +/- 0.02 ohms equivalent to +/- 0.05degree) is:
R=T/10*(3.982-0.006*T/10)+100 where R is in ohms and T in degrees Celcius

I am sure I could simplify that formula but its 3AM and I can barely keep my eyes open so no algebra critiques thanks :p My formula was generated from the sampling the RTD table so I know it works between 0 and 300 degrees +/- 0.05 degrees as I said - way more accurate than the rest of this method allows.

[ktronik]

nice work!! that will come in handy...

I was thinking along the same lines as you....and came up with the same problem, with a dial, how do we know what temp we are at...

so i used a 12 position rotary switch, and added 5deg jumps to each position...so now its easy to tell what temp I am at by looking at the dial. this filled my temp feedback needs.

hope this helps

K

[JuliaDee]

For those that just want to know what the resistance should be at a given temp the formula (accurate to +/- 0.02 ohms equivalent to +/- 0.05degree) is:
R=T/10*(3.982-0.006*T/10)+100 where R is in ohms and T in degrees Celcius

Roller's formula above (nice work and thanks, Roller) gives the resistance of the thermocouple at a given temperature. To know what series resistance is required to fool the Up into extruding at a lower temperature, you first need to plug the Up's software temp setpoint (260ºC for ABS) into the formula, which gives 199.5 Ohms. The combined resistance of the thermocouple+resistor needs to be this value at your desired temp. Then enter your target temperature in the formula to get the thermocouple's resistance at that temp and subtract that from the first. For example, at 230 degrees the thermocouple will be 188.4 Ohms. The Up's setpoint is 199.5 Ohms, so you need to add 199.5-188.4 = 11.1 Ohms in series. The Octave switch has an 11-Ohm resistor.

To raise the extrusion temperature you need to put a resistor in parallel with the thermocouple instead of in series with it. The formula for the parallel resistor's value is

Rp = 1/(1/Rtn - 1/Rtt)

where Rtn is the thermocouple's resistance at the Up's setpoint (199.5 for ABS setting) and Rtt is the thermocouple's resistance at the target temp. For example if we want 280ºC, we find via Roller's formula that the thermocouple will be 206.8 Ohms at that temp. Plugging the numbers into the formula above, we find we need 5,638 Ohms in parallel with the thermocouple.

Julia

EDIT: I now see that Roller's spreadsheet actually has the intermediate values of the thermocouple's resistance at 260ºC and the required series resistance in hidden columns C and D, but they might not have been obvious to all.

[roller]

Oh yeah, sorry I should have said - I did bodge this up at 3AM - I hid the columns to make my own easy tool for dialing in a pot but all the formulas are there for converting a temp to a resistance (Cell C1) and total resistance to a temp (B7). I did comment in the hidden fields so that should help anyone with a different intended purpose.

[outsider]

What kind of temperature range is useful? Would I ever want/need to go below 200ish deg?
I'm thinking I just get a trimpot that is 15 or 20ohms (so that even at max rotation I only get 200ish deg), and make a housing for it that has markings on it with temperature numbers.

[roller]

What kind of temperature range is useful? Would I ever want/need to go below 200ish deg?
I'm thinking I just get a trimpot that is 15 or 20ohms (so that even at max rotation I only get 200ish deg), and make a housing for it that has markings on it with temperature numbers.

20 gives you a great range because if you want lower than 200 you can use the PLA setting which gives you a normal temp of 190C. 25ohms would be ideal to give you a full range.

Anyway, to specifically answer your question a range of 205-260C (which 20ohms will give you) is ideal for just about all the currently available ABS. in fact 215-260 (from a 15ohm max) would serve you fine in 95% of circumstance. Mostly you want a range of 220-240C for aftermarket ABS.

[JuliaDee]

What kind of temperature range is useful? Would I ever want/need to go below 200ish deg?
I'm thinking I just get a trimpot that is 15 or 20ohms (so that even at max rotation I only get 200ish deg), and make a housing for it that has markings on it with temperature numbers.

20 gives you a great range because if you want lower than 200 you can use the PLA setting which gives you a normal temp of 190C. 25ohms would be ideal to give you a full range.

Anyway, to specifically answer your question a range of 205-260C (which 20ohms will give you) is ideal for just about all the currently available ABS. in fact 215-260 (from a 15ohm max) would serve you fine in 95% of circumstance. Mostly you want a range of 220-240C for aftermarket ABS.

You might conceivably want to go higher than 260 for polycarbonate, as discussed on another thread here not long ago. Assuming, that is, that the Up is capable of heating the extruder to the desired temp.

But yes, for ABS, 215-260 oughta do it. Make sure your potentiometer has a low end resistance, or include a switch to bypass the pot completely - a couple of Ohms would limit you to 255ºC. Or do as ktronic auggests and use a multi-position switch with fixed resistors; that's probably the most stable, easy, repeatable, and reliable method, albeit at the cost of some fineness in settability.

[roller]

If you fit a switch then you can use my spreadsheet and calculation method in the top post to set the right resistance without having to pull out the multimeter.

Julia, I am thinking about a second calculator on the sheet and a three way switch to allow a trimpot in parallel to allow setting temps over 260C though this makes me a little nervous about roasting the hotend so I probably wont actually do it myself.

[mr6k]

Roller suggests using the PLA setting to extend the temperature range downwards.
A question, does anyone know or has observed whether the extrude feedrate or any other stepper rate parameters are changed, in the background, when switching to PLA from ABS ?

cheers

Peter

[roller]

It's very hard to know for sure without an official answer - in reprap there does need to be a shift of feedrate to compensate for the fact PLA is softer and gets more bite in the extruder resulting in a slightly different level of feed. I suspect there is no difference except for temperature though because the Up doesn't exert sprung pressure on the filament as it hits the drive pinion so bite should be fairly consistent.

Best test would be to set it to PLA and use a resistor in parallel to trick the temp back up to 260C and see if normal ABS print comes out the same.

[JuliaDee]

Here are seven standard 1% metal-film resistor values which per Roller's formula will yield extruder temps at 10-degree intervals (plus 0 Ohms, i.e. a short, for full Up temp). Note that these temps assume, as does Roller's formula, that the Up's standard temp is 260 deg C.

Screen shot 2013-08-11 at 11.32.10 PM.png (21.29 KiB) Viewed 20103 times

You could also use standard 5% values, which at 27, 22, 18, 15, 11, 7.5, and 3.6 Ohms, wouldn't be very far off and which are available at Radio Shack.

All of these can be purchased from Mouser, Digi-Key, or other electronics distributors. I'm going to build a little board with an 8-position rotary switch and will post details when it's done; it will be a lot cheaper than the Art Taylor controller. Be sure to use a switch with low contact resistance; generally the cheap ones meant for "DIP switch" type applications are not suitable since they can have a contact resistance of 2 Ohms or more. This is the one I've ordered:

http://www.mouser.com/Search/ProductDet ... 1712F-18NS

julia

[roller]

I now prefer an 8 dip switch layout which simply adds 3.7R (ohms) per step giving pretty close to a 10C drop per switch. This has three added advantages:

- Obviously it's much easier to change the temp in the fly than a pot.
- It gives nice even 10C steps which I can easily read to see my current setting
- It lets you walk the temp up or down (need to be gradual else an exception will be triggered by the firmware if the temp is too far from expected) unlike the Octave single switch method which has an open period in the circuit while you switch.

I just cascade the 3.7ohm (actually a common 2.2+1.5) resistors in series with the switch providing a tap at various points. With one tap before the resistors to switch to full temp and all switches open gives me the lowest temp. You can have multiple switches set to on and whichever is the highest temp is the one that will be set. It's nice and simple.

This setup gives 180-260 (or 190-270 - who can be dead sure) on the ABS setting. If someone wants a circuit diag because this doesn't make sufficient sense (too be honest I wasn't trying to write a how to in this post - just a follow up) then buzz a reply and I'll try to get one up though it may just be a napkin sketch.

[JuliaDee]

I know prefer an 8 dip switch layout which simply adds 3.7R (ohms) per step giving pretty close to a 10C drop per switch. This has three added advantages:

- Obviously it's much easier to change the temp in the fly than a pot.

But not necessarily easier than a multi-position rotary switch.

- It gives nice even 10C steps which I can easily read to see my current setting

So can a rotary switch.

- It lets you walk the temp up or down (need to be gradual else an exception will be triggered by the firmware if the temp is too far from expected) unlike the Octave single switch method which has an open period in the circuit while you switch.

Good point, but a break-before-make ("shorting") rotary switch should take care of that. The Octave switch is wired stupidly: it's set up as an SPDT that switches the printer's thermocouple input between the 11-Ohm resistor and the thermocouple, so that it's open in between positions. It should be wired as an SPST that just shorts out the resistor, then there'd be no open condition.

Again, if you're going the DIP-switch or cheapo-rotary switch route, make sure the contact resistance is low or you'll never get the full nominal temp when the resistors are shorted out. CTS's are spec'd at 25-50 milliohms which should be fine, but I've seen some that are several Ohms, not good.

julia

[roller]

I would have preferred rotary but nothing decent is available locally - I am waiting on another option. An SPST slide would usually still have a momentary gap (i.e. no state) though you would have to be unlucky to get it to upset the sensor read but it was possible. Have an SPDT is just dumb ... but both options are moot because switching 11 Ohms while printing has a tendency to trigger an overheat or underheat condition and the printer stops and then the print fails.

[JuliaDee]

I would have preferred rotary but nothing decent is available locally - I am waiting on another option. An SPST slide would usually still have a momentary gap (i.e. no state) though you would have to be unlucky to get it to upset the sensor read but it was possible. Have an SPDT is just dumb ... but both options are moot because switching 11 Ohms while printing has a tendency to trigger an overheat or underheat condition and the printer stops and then the print fails.

My point about the Octave switch wiring is that if the switch is across the resistor, then even if the switch has a momentary open condition, the circuit will never go open because the resistor is always connected. So you'll either have

a) Thermocouple only (switch closed and shorting out the R) or
b) Thermocouple + resistor (switch open, either deliberately or due to switching "gap").

The way Octave wires it there are three conditions:

a) Thermocouple only (switch in position 1, connected to thermocouple only) or
b) Thermocouple + resistor (switch in position 2, connected to thermocouple+resistor) or
c) Completely open circuit, due to contact gap in switch action

You may be right about the printer not liking the resistor being switched in and out on the fly, but I'm sure that it doesn't like the circuit going completely open while printing.

[roller]

Oh, yar yar yar. Sorry, 4 hours sleep ... being dumb. The other issue is in use the jump is a bit much to not upset the printer and kill your job.

[LonV]

I don't know about where you guys are, but Radio Shacks in Arizona have gone down hill. They had such a small range of resistors that I couldn't really build what I wanted to build...and ordering online is so weird for these components. On eBay I can order 5000000000 small resistors all for a nickle, but it will take 4-6 months from Shanghai.

I ended up ordering from Futurelec because I've bought from them many years ago, but even then it's still a week or two for delivery.

[JuliaDee]

Mouser has everything you could possibly want, no minimums, they're in Texas, and they ship the same day.

[LonV]

Cool, thanks...I'll check them out.

[roller]

The reason I chose 3.7 ohms is because it's almost exactly right for an average 10C drop over 8 steps and because it's easily made form two of the more common resistors ie 1.5 and 2.2 ohms. I wasn't particularly set on 10C drops per step ... it just worked out conveniently.