Volunteer for beta Firmware for TAZ 5 wanted

Development Software
1

I’m in the process of implementating an advance feature into Marlin. The first draft code is working ( :smiley: :sunglasses: ), but it needs more testing and it’s not bug free. It would be helpful to have one or two volunteers that can run a few tests and report to me. This persons have to know how to flash a FW, send manual gCodes (M-codes) to the machine and they should have a good tech. understanding of the printer to know what is going on in general.

What’s the advantage of “my” feature?
If you print a test cube at very low speed, say 10mm/s, you will see it has a perfect shape. There are no bleeding edges and the top surface is completely flat and perfect (assuming the esteps are calibrated).
If you do the thing again at high speed (say 80mm/s), it will not be a perfect square due to bleeding edges (excess of material) and the top surface is rough where it meets the perimeter lines. This happens because the Marlin firmware doesn’t take nozzle pressure into acount for it’s calculations.
With a working advance feature, you can have both: high speed and nearly perfect surface finish as in a low speed print. You might know something like that as JKN advance. I’m using also a linear aproach (k* delta velocity), but I have to rewrite the code from scratch due to license limitations.

I don’t want to publish the code in the current condition, and volunteers should not share the code with anybody! If testing and coding is completed, I will create a pull request to have this feature in the main Marlin branch.

If somebody wants to invest some of his/her free time, send me a PM with a e-mail adress. I will then provide the beta code and things that have to be tested.

2

Any volunteers yet?

3

I have several other irons in the fire right now but I’d be willing to devote some time to it.

4

I have the code ready to send, but I have no PMs with e-mail adresses yet. So - no, no volunteers now :wink:

Actual status:
.) Doesn’t work with FW retract. Use normal retract in your slicer instead.
.) I printed test cubes at 80mm/s with success, a test at 150mm/s failed. Sounds as the extruder looses steps…
.) Values around 0.007 seems to work best for TAZ 5.
.) You can send a new k factor with “M905 Sx.xxx” where x is your new value. This value is not stored during a power cycle, so reset it after each reboot of the printer!
.) Don’t use coast at end, extra restart length after retract or other fancy features in your slicer - the advance feature is a replacement for this work arounds!

If somebody want’s to test something, here is a list:
.) Play around, see what is working and what is not working.
.) k seems to be a fixed value, independent of layer height and print speed. Confirmation?
.) Is k constant for different filament materials? I can only test on PLA.
.) Is the range of k OK or do we need a multiplier to have a good value for every material?