Forum Replies Created
-
AuthorPosts
-
cmcgrath5035
ModeratorI don’t speak for Synthetos, but see no indication that support for gShield hardware will disappear anytime soon.
You are correct, lots of global activity on CNC and particularly 3D printers . Particular up-tic in NEMA23 machines and Multi (>3axis) machines.
gShield drivers might not be a good long term choice for driving NEMA23’s depending on your torque needs.
Shields supporting the 8825 stepper driver likely a better choice.Many folks have tinyG2 on DUE running with 3 motor drivers on gShield.
There is a 4 driver board, “GAUPS” available at least in Europe. Search on GAUPS at Shapeoko Forum.
You can also search on RAMPS boards (an early 3D printer).
The Ultimaker II boards (support for 5 stepper carriers) is an evolution of the RAMPS project.
There are others out there as well.As a general observation, the DUE can support most UNO shields, give or take potential interface issues between 5V logic and 3.3V logic.
The DUE pinout configuration is same as Arduino MEGA, again with potential issues with I/O voltages.
The Ultimaker board, as an example, connects via the MEGA footprint, but appears it should work fine with tinyG2 if driver carriers compatible with 3.3V logic are used (such as Pololu 8825 carriers)Good luck with your project.
ModeratorFollowing your numbered comments/questions:
1. Jog “a lot”. As in many clicks, or long distance?2. PWM Frequency requires a tinyG reset to take effect, so you cannot change PWM Frequency from active session, because the default parameters reload with reset.
Solution is to recompile with the PWM parameters you want on bootup.
You did not say if you are building Master or Edge, so you will have to fill in the blanks here:
A. navigate in the github area on your machine to the ../TinyG2/settings/ directory
B. Copy the settings_name h file you specify in the build process to a new file. For example, copy settings_default.h to settings_qtfp.h.
C. Open settings_qtfb.h and edit the PWM parameters to your liking. To do this you will need to add some #defines to the basic file as they do not exist in the base settings_default.h file. Have a look at my exampleAround line 218 you will see a block I inserted.
Pick your values. Also, set #define P1_PWM_PHASE_OFF 0.000
by default it is 0.1, which is why it still pulses after M05
D. Recompile tinyG2, specifying your personalized settings file, settings_qtfp.h3.What are your settings for m4? If M4 is attached to Axis, $m4tr should be $m4tr=360 and G0 A1 should result in 1/360 revolution. Are the all the $m4 parameters consistent with a rotational axis? That will depend on what settings file you chose at compile.
4. AFAIK, tinyG2 only supports 2,4,8,16 or 32 microsteps per step. It is my understanding (not actual experience) that steps per revolution (200, 400, etc) are a motor property but microsteps per step are a property of the DRIVER device. Try ms=16 or32 to see if you get sane results.
5. B and C position – you are blazing new territory here. Your redefinition of the status report looks correct. It could well be Chilipeppr issue, I don’t believe the CP widgets are instrumented for B and C parameters.
A. Are the B and C axes enabled?
B. Are motors assigned to B and C?
C. Are the motor parameters assigned to B and C consistent?
D. If yes, try a simple move from the Serial Port Console, e.g. G0 B1 then read back the position from the Serial Port Console with $posb to see if it movedFYI, I am not sure what you mean by “tinyG2 suspend”. TinyG2 hangs (requires reset)? Are you sure it is tinyG2 and not CP?
ModeratorLipo – ah so, a battery!
FW release 440.14, made some refinements to the Power Management implementation, this is mow a good reference:
Good Luck with your build
ModeratorWhat build number of tinyG2 are you flashing?
You built with Platform=”gShield” and default settings?If tinyG2 and Chilipepper connect correctly, you should not need the Logic Analyzer, but of course it should work.
Any jogging commands or G code entered into the Serial Port Console command window should result in movement in the 3D display and XYZ readout in the AXES widget.A quick way to verify that you are connected is to issue a ‘?’ command (position report reques)’ or a $ command (short parameter list) in the Serial Port console and see an appropriate response.
Are you sure you have m1 mapped to the X Avis?
If you built with Platform=gShield, then DUE pin 2 is m1step and DUE pin , which is what you are monitoring.5 is m1dirModeratorYes, the tinyGV9 prototypes have a SAM 3x8C 100 pin version of the SAM3x8E on the DUE.
Otherwise, the structure of the tinyGV9 is very much like V7 and V8 – 4 8825 stepper drivers, I/O for Limit switches, PWM, etc. All 3.3V logic.
It is unclear what schedule might be for release of a tinyGV9.Should I assume ‘Lipo’ is Linear Power supply?.
I can only say your results are not like mine.
I normally run my V7, driving NEMA17 steppers, from 24V supply (6Amp, I believe).
I interpret your description to imply that the uC function is still operational, just no motor movement? That is very strange; only mechanism I am aware of would be the automatic thermal shutdown in the stepper devices.ModeratorGood progress – yes, running higher voltage is better(cooler).
I have no good idea what you ultimately might need to automate your plasma machine, but offer this as a thought experiment if you are adventuresome:
The next generation of tinyG is tinyG2, porting the tinyG logic to a faster uC (ARM based)platform. Currently tinyG2 runs on an Arduino DUE and on a tinyGV9 prototype, which is functionally very similar to tinyGV8.
I am currently playing with a custom build of G2 that will support a DUE + a Ultimaker shield (here is a typical clone)This shield supports up to 5 steppers, has a bunch of built-in IO including 3 hi-power FETs for driving heaters (focus is 3D printer)
A good place to start would be the G2 Wiki at
All this is still WIP, but something to consider if you need a lot of control.
ModeratorSounds like an interesting project.
We have not had much plasma traffic to date, sort of a high cost of entry unless you score a find as you did.TinyG is 3.3V logic. It is straightforward to interface external 3.3V drivers if necessary.
Many folks have CNC systems using NEMA23s. If you drive at maximum current, be mindful of driver heating – do you have plans for a cooling fan of some sort.FYI – Edge FW 440.14 is not really bleeding that much, rathaer mature tweaks on the tinyG system. 440.14 will be Master soon as well.
Resources:
WikiGo to the Forum level and use the Search Box
ModeratorHmmm, nothing intentional. Perhaps you caught the system in a maintenance period(?).
ModeratorYou might want to follow this:
ModeratorThe most current tinyG FW is build 440.14,
tgFX has not been fully compatible with tinyG builds since 425 builds, so depending on youw tinyG build I am surprised tgFX worked at all.You should upgrade tinyG to the latest 440.14.
tgFX is no longer supported (or being developed), Chilipeppr is suggested.
CoolTerm is also an option if you prefer basic console interface.If you want some custom features for your custom machine, CP is an easily forked and tweaked interface.
-
This reply was modified 9 years, 10 months ago by
cmcgrath5035.
ModeratorYou might want to post this as question about CP macros over at
More CP types hang out there
ModeratorTroubleshooting wiring issues is tough in these.
Knowing what wire was detaching, can you explain tinyG restarting?
I am sort of guessing it was caused the first time the wire separates while under step power, likely a voltage spike due to the winding to driver opening(windings are rather large inductors).ModeratorIf you wish to not perform the remainder of the move after the feedhold send a queue flush character (‘%’) after the feedhold character. Motion will start with the next Gcode command entered. A cycle start is not necessary after a queue flush.
I looked and looked and could not find the % reference – where did you see it?
ModeratorI believe % command is same as $QF, Wiki sez
-
https://github.com/synthetos/TinyG/wiki/TinyG-Configuration-for-Firmware-Version-0.97#qf—queue-flush
all commands are flushed from planning buffers
ModeratorIt is confusing!
Enjoy!
-
This reply was modified 9 years, 10 months ago by
-
AuthorPosts
