Welcome! Log In Create A New Profile

Advanced

New approach to pulse generation

Posted by DragonFire 
New approach to pulse generation
October 11, 2016 07:41AM
It strikes me that current Reprap controllers ALL have the same flaw... they use the processor to generate pulse steps for stepper motors.

Why not offload the work to dedicated frequency controllers, freeing up processor time and permitting faster movements from 8 bit electronics?
Re: New approach to pulse generation
October 11, 2016 11:22AM
Because 32-bit processors are so cheap that it is not worth using 8-bit processors any more.

Following their takeover of Atmel, Microchip has almost doubled the price of the attiny25 chip that I use in the IR sensors I manufacture. I can buy a much more capable 32-bit processor for around half the price of the attiny now. So it looks like I will soon be shipping IR sensors with 32-bit processors in them, even though the attiny is entirely adequate for that application.

The atmega2560 commonly used in 8-bit controllers has been around double the price of 32-bit processors with sufficient I/O pins for a long time. It's now £7.99 if you buy 1500 of them. By comparison, the atsam3s4ca is now £3.37 in quantities of 1000 and features useful things like built-in USB port, high speed SD card interface, and much more RAM.

I believe the Smoothieboard 2 design includes a plan to offload step generation from the main processor to either a separate ARM Cortex M0 core or an FPGA. But this seems to me unnecessary for ordinary 3D printing applications.

Edited 2 time(s). Last edit at 10/11/2016 11:44AM by dc42.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
VDX
Re: New approach to pulse generation
October 11, 2016 03:22PM
... another point is the needed "4-dimensional" synchronisation of the XYZ- and E-steppers to get the correct amount of material out of the nozzle with differently accelerated/deccelerated moving speeds.

How will you achieve this with "dedicated frequency controllers"?


Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]
Re: New approach to pulse generation
October 11, 2016 06:25PM
Klipper is a new firmware which will calculate most stuff on the host side.


Triffid Hunter's Calibration Guide --> X <-- Drill for new Monitor Most important Gcode.
Re: New approach to pulse generation
October 11, 2016 06:57PM
You can pick a mega r3 up for under a fiver, sad to say I ordered 2 more....but I really want a cleaner all in one, just to keep the clutter to a minimum...32bit will be a bonus, DC42 can you put a direct link in your sig' to the board.
Re: New approach to pulse generation
October 11, 2016 07:01PM
Trinamic Stepper controllers are also a solution. They do not need pulses any more. Each change on the Step line causes a step. So Low-> High causes the first Step High-> Low causes the next step to happen. They can also do Step interpolation (They make 16 steps when ordered to do one).

I Agree with @dc42 that 32bits are the way to go right now. So for a new electronic go looking for something with a 32bit CPU. If that is not good enough for you then put all step generation (to keep the synchronization between X, Y, Z and E) into an FPGA.

If you want to stay on you 8bit hardware then offloading some of the work to a host is the only way I know that works.

Klippy(Klipper) and Minnow(Pacemaker) are Firmwares that do that for you.

Disclaimer: I write Pacemaker (The Host needed by Minnow).
Sorry, only registered users may post in this forum.

Click here to login