could someone make me a temptable please?

Hi folks,
normally the CreateTempTable.py script works fine to do this, but now for some reason
it keeps failing with an error about "raising a negative number to a fraction"

The cmd line I'm using is,
CreateTempTable.py --t1=60 --adc1=767 --t2=98 --adc2=406 > blah.txt

It always seems to have worked in the past. Stumped what is going on.
Really appreciate this if someone can do it. Calibrating the heatbed is the last thing
to do before I start trying to print.

For t1=60 adc1=379 and t2=112 adc2=314, standard 25^c 100K NTC


Many thanks,
Andrew

Edited 2 time(s). Last edit at 01/09/2016 03:03AM by friarfish.

Probably the place to start is take the command line that is working and change one parameter at a time until you understand which one is causing the failure. But do you really need to create a custom table? Aren't you using standard thermistors?

Hi Roxy,
the problem is if I use the values in the example, MakeTempTable and CreateTempTable work.
But if I try to generate a table with t0=60 and t1=110 then I get the error. It is obviously some
thing strange in the code, even though it seems elementary.

I was hoping to generate a 2nd table specifically for the heatbed.

Yes, my system can be described as different. I'm using umm, 104's but they are returning
distorted values. Hence the adjusted temptables.

Andrew

I don't know Python, so I can't really debug that code. But I do have a couple of comments. First, isn't 60 C a very high starting point for the first temperature in the table? And the second comment is you don't actually have to use the generation script. You can edit the Thermistortables.h file to have the values the way you want them.

Can you explain why you need to generate a new table if you are using a standard thermistor?

The MakeTempTable script by hydraraptor has the t0 etc NTC values already plugged in, so you only have to provide 2 temp:adc pairs to generate a new table.
But with the starting value of 25℃:100K being inaccurate and main table being created using t1=185(PLA) and t2=235(ABS), any temp readings in the bottom part of table at out of whack by around 10℃ or so. I was going manually generate a new r0 value at t0=25℃ or as close as I can get dependent on room temperature.


Andrew

So if you know the correct values, can't you take an existing table that is close and just manually shift the values to get it right for your setup?

Unfortunately not.
As the temptable is rough or smooth, depending on the number of points in the table,
approximation of a curve, manually entering say the t2 pair of 235:50 would mess this up.

If you were lucky with a table with three points, ie p1-3, and the temp pair you wanted
to enter lay between p1 and p3, you probably could just whack it in.

But and I think this would be the big but, no longer having a smooth progression along
the approximated curve would likely really mess with the PID tuning for Hotend and Heat bed.
Maybe not a case of finding a mushroom cloud where your printer was, but probably
not a fun time.

It could very well work, but I don't plan to do it.


Andrew

OK. Well, I don't know Python so I can't help with what ever is blowing up the script. But if you post this issue over at: [github.com] I think you will get an answer. That is where the code is evolving and things like this get dealt with on a routine basis. If you do get a good answer and a fix, please bring it back to this thread.

Given up on this.
Thanks for trying.

Andrew

Edited 1 time(s). Last edit at 01/19/2016 05:31AM by friarfish.

No guarantee this will be of any help or pertain to this issue. But there was a patch to the createTemperatureLookupMarlin.py script a couple months ago that hasn't been published yet. Check out the version here: [github.com]

Thanks Thinkyhead.
It has me stumped what is going on.