Software trouble with Stepping /mm

Hi

Ive got a problem with the software, ive tried to change the default stepping /mm setting for x/y/z in the software under the preference folder but whatever i type the software doesnt seem to set the hardware to use this value how can i fix this issue ?

Ive currently bypassed the problem with tampering in the firmware for the mainboard and setting the stepper / mm for X / Y but this is an anoying fix as it requires recompile / upload for everychange i make to my repstraps axis construction.

Also the software seems to complain about not finding some default values while starting up, but as i understand the configfile doesnt reside in the software folder so where is it ? (im running the java app under windows xp)

I'm pretty sure the step settings in the reprap-host software are remnants from older software/firmware combinations, and the correct place to adjust the settings is in the firmware. Digital calipers are the best tool to get the config right, but you should probably only have to do this once. Remember that the x and y values are probably the same unless you've got some weird machine that uses difference size sprockets.

Actually ive built a repstrap from scrap and tried to keep it to mcwireish design as possible, using nema 17 motors and v3 electronics from makerbot. Still working on the Z axis and a good way to attach it without using shitloads of wood to hold it stable in place... but some expermenting is still needed.

Ah, if you're using threaded rod to move the axes, you can just calculate the exact number of steps from the pitch of the rod.

The fomula is:
The threads per inch/mm gives you the number of turns to move one inch/mm.
The stepper motor has a degrees/step rating, which can also be stated as steps per revolution. This number is 200 for most steppers, or 400 if actually found some of the recommended 1.8 degree stepper motors. Simply multiply these two and you have your steps per inch/mm. Multiply or divide by 25.4 to get the other one and enter that into the appropriate firmware header file.

Threads per inch * steps per revolution = steps per inch.
steps per inch / 25.4 = steps per mm

---

--
I'm building it with Baling Wire

Im quite aware of the formula (its stated on the howto configure section) but the problem aint configuring the steppers.

The problem is that the software has a "stepping / mm" field for X / Y / Z but changing these values doesnt seem to change anything other than the software storing the values somewhere.... on the other hand if you change this value in the firmware itself it works almost instantly.

also the stepping formula is "almost" correct it has some errors as it always overshoots the value (atleast for me in my calibration effort)

Since the move to g-code the software sends distances in mm so the conversion has to be done by the firmware. So the software settings will have no effect when using g-code.

---

[www.hydraraptor.blogspot.com]

Yeah it would be good if some of the useless settings could be pruned from the config box, we don't need thermistor settings either. I don't think anyone is using SNAP any more, but at the very least the snap-only settings should be segregated. When you say the stepping formula is "almost" correct, it /is/ correct, but you must be using the wrong values, or there is backlash in that axis.

Well to be more precise about the formula you gotta compensate for design specs most devices have a margin of error in accuracy for instance resistors (if not specific low error margin ones) 5% is maximum off ratio. Hence the error in the formula (im talking about like <10 steps but still and the sample had fractions so shitloads of fractions/step sum up to some error in the end).

And yes it would be nice if all revised crap is removed from the software and the wiki / build page updated.

about the termistor part,i dont need the array in the firmware anymore ? as i got a standard 100k one at my local electronics store with peek of 250C (or close to it, not sure but software stops on 255C) and it messured ok on tests i guessed that the margin of error wasnt to large without changing the array table

I meant there are still some settings in the host software that deal with thermistors, even though that's all done in the firmware now, and the host software only deals with real temperatures. As for tweaking the tables, the actual temperature they report doesn't matter, as long as you can find a numeric value that gets the plastic to the correct consistency. Themistors are pretty lousy at being the same as each other, even from the same batch, but individually they're reliable and repeatable. Assuming the optimum real temperature of the nozzle gives readings that aren't wildly out, you're totally fine. Generally though, you want to change as many settings as possible in the firmware, otherwise you're trying to generate "wrong" gcode that transforms into what you want. I /think/ this post contains sense, but i've been drinking coffee for about 20 hours and I can't be certain.

alright so to sum up, settings in software may not work / legacy, settings in firmware bound to work but need tweaking ?

Yeah, as far as I know anyway. If you're not sure about whether a host setting still works, just set it to a stupidly large/small value and you'll see if its "active"