New firmware 2.03RC2 available

gavatron3000

@dc42 the m291 isn't unlocking a tool change still when called. Or is this impractical to do within a tool change?

dc42

@ben_van said in New firmware 2.03RC2 available:

As for the M350 extruder microstepping, I will change to the following:
M350 X256 Y256 Z256 E256 I0

We don't recommend using native x256 microstepping. You will be limited to low speeds if you do, because generating the step pulses will use up too much processor time. It may be OK for a CNC machine, but not for a 3D printer.

ben_van

@dc42 said in New firmware 2.03RC2 available:

@ben_van said in New firmware 2.03RC2 available:

As for the M350 extruder microstepping, I will change to the following:
M350 X256 Y256 Z256 E256 I0

We don't recommend using native x256 microstepping. You will be limited to low speeds if you do, because generating the step pulses will use up too much processor time. It may be OK for a CNC machine, but not for a 3D printer.

I understand the computational expense of doing this.. going fast is the last thing i want Most of the stuff made with this machine is seldom larger than 1mm, at speeds less than 50mm/min, and we have never experienced speed issues being limited by the CPU. This doesn't solve the G0 problem though..
To give you an idea of why this matters, here is a sample pic of one of the things we can do with this machine + duet . It's too small to see properly with the naked eye.

jv43

@ben_van looks interesting, but what the heck is that?

ben_van

@jv43 said in New firmware 2.03RC2 available:

@ben_van looks interesting, but what the heck is that?

It is a generic 2D scaffold used for cell culture. Tissue engineering is not my field of expertise but for the people I am working with it is. This is just a test print to show we are working on a scale smaller than that of a stem cell, and about the size of a T-cell.
Cells are present while this structure is printed and are encapsulated inside the bio-compatible gel, where they then grow/multiply inside it, as it is like food for them. I like to think of it like bread crumbs, where we can coax the cells to grow and develop in a very controlled manner.
Whilst i can't dive into many specifics of how everything works until something is published, I can say that we have been getting some very good results
The benefit of using all open source software is once the groundwork is in place for this method, we can easily use readily available 3D printer slicers to make any micro 3D structure straight from a CAD model, and hopefully be able to provide this to people wanting to experiment without the $100k+ pricetag that comes with anything with the word 'bio' in it.

deckingman

@ben_van Fascinating stuff. Out of curiosity, are you using "standard" 1.75mm diameter filament? I've always thought that to extrude tiny amounts of filament, the way forward is to use something that has a much smaller cross sectional area to start with.

ben_van

@deckingman said in New firmware 2.03RC2 available:

@ben_van Fascinating stuff. Out of curiosity, are you using "standard" 1.75mm diameter filament? I've always thought that to extrude tiny amounts of filament, the way forward is to use something that has a much smaller cross sectional area to start with.

I didn't mention that this machine doesn't use filaments. It uses a tightly focused laser (we can easily change the dot size) to solidify tissue-like gels that are synthesised with certain proteins etc. It is somewhat similar to a traditional SLA machine but with differences in the layering methods.
Extruding filaments to that size would be cool though, and i definitely agree with you there. Other than there not being much of a demand for printing such small things, maybe filaments with such small diameters would just be too fragile..

dc42

@ben_van said in New firmware 2.03RC2 available:

Just did a few tests..
Everything works as expected for G0 moves if the feedrate set in M203 is higher than 60mm/min. If I set the max feedrate lower than 60mm/min in M203 for any axes, G0 moves will still request a speed of 60mm/min.

I found the problem. M203 enforced a minimum movement speed of 1mm/sec. Fixed in 2.03RC3.

I also found that DWC 2.0.0-RC5 rounds up the displayed top speed to the nearest 1mm/sec. So it displays 1mm/sec even when the actual speed is 0.5mm/sec.

dc42

@lb said in New firmware 2.03RC2 available:

Now somebody wants to print somewhere else on the table (on the same or new fabric). Instaed of just updating G55 in x&y for positioning the print (or doig an shift/offset in the the slicer), because of getting in a lazy fashion the printhead out of the way, the person just homes x e.g., THEN: Not only is the printhead moving in x only - as one would suspect and as it does usually with standard-homing after powering up - BUT (I think because of the delta between G54 and G55 in X&Y) it makes a diagonal move of the amount in X & Y AND if that offset was very big, this is of course a big diagonal move I do not understand AND it ignores only on that move totally any endstop? What am I doing wrong? Should I open a diffrent thread and post all homing and config files?

I have fixed a similar-sounding issue in 2.03RC3. Please try it when I release it.

gavatron3000

@dc42 did you see my post above regarding m291?

dc42

@gavatron3000 said in New firmware 2.03RC2 available:

@dc42 did you see my post above regarding m291?

There is a problem in supporting M291 commands that require acknowledgement or allow axis movement within tool change files. If the tool change command is sent from DWC or PanelDue, then RRF is executing that command for that channel until the tool change is complete. That means that it cannot execute an axis jog command or M292 acknowledge command that comes from the same channel as the tool change command. This is what causes the firmware to lock up in the tool change. If you send M292 from another device, it continues.

gavatron3000

ok no worries are we able to support it at the end of tpre and tfree instead?

dc42

@gavatron3000 said in New firmware 2.03RC2 available:

ok no worries are we able to support it at the end of tpre and tfree instead?

Not easily, because the tool change sequence isn't complete until tpost has completed.

dc42

I've just released firmware 2.03RC3 so I am locking this thread.