Advanced Search

I know I’ve tuned my other machines.
I usually bench test first. Then install and tune, then connect Linuxcnc and tune it. And the trouble start when I connect the analog control.
Even with analog Control and the servos have no load they have 50hz vibration. Which I can’t get rid of. With the spindle I don’t have the issue the wires could be routed differently and I can use other better shielded wire. For the drives I can’t do this easily. I twisted and try to shield the wires with tape. But it didn’t improve. And have no idea what else to do. I could start to add ferrites or try to reroute all other cables and chasing down the issues. But I figured going digital would be quicker

FF1 should not be over 1, at least i never ever had it over 1.
Spikes can be corrected with a bit of FF2.
As for f-error, that looks like a normal Z axis without counterbalance, so try adding just s bit of Bias.

Should be
sudo chmod 666 /dev/uinput

Yes, that seems the correct wiring.
-
Mesa 7i77 does +-10V, while your drives seem to be 5V PWM, so using a 7i77 would require limiting the output to +5V only, in hal.
Better use step/dir with 7i76.
You can use 6i25 and 7i76 for 5V PWM, does also require tuning and complicates things a bit compared to step/dir.
Edit:
The first line implies that using 7i77 with proper setting in hal can be used as PWM, and that is not correct, sorry for the confusion this might have made.
From PWM to analog can be done with the use of a resistor and capacitor, but not the other way around.

PFC smoothes the input current to power supply, removes the spikes caused by switching in the power supply. It also raises efficiency at lower loads.
In US anything under 100W can be without it, anything 100W should have it, although not enforced in any way, as far as i know.
Good source of info on youtube
@allthingsoneplace

It's under the "OLD" section, so it may no longer apply perfectly, but it has helped me.

Just a shot in the dark, have you tried the suggested
$ chmod 666 /dev/uinput

Thank you COFHAL, for share link to here topic
this is my old topic forum.linuxcnc.org/49-basic-configuratio...ey-how-to-map-number

I start with try from link : linuxcnc.org/docs/2.8/html/man/man1/sendkeys.1.html

all look good  except  ls -l /dev/uinput
In link show  : crw-rw----
But my Terminal show  : crw

---

      only  (no rw)

I try open linuxcnc from terminal 

cannot /dev/uinput. Suggest chmod 666 /dev/uinput

How to can access chmod 666 ?



FYI:
- I use 7i73 matrixkey 8x8  (good work)

- I already add loadusr -W sendkeys config=16t2
net scancodes hm2_7i73.0.0.keycode => sendkeys.0.keycodesetp sendkeys.0.scan-event-00 34 # Key G
setp sendkeys.0.scan-event-01 2 # Key 1
setp sendkeys.0.scan-event-02 3 # Key 2
setp sendkeys.0.scan-event-03 4 # Key 3
setp sendkeys.0.scan-event-04 50 # Key M
setp sendkeys.0.scan-event-05 05 # Key 4
setp sendkeys.0.scan-event-06 06 # Key 5
setp sendkeys.0.scan-event-07 07 # Key 6
setp sendkeys.0.scan-event-08 31 # Key S
setp sendkeys.0.scan-event-09 8 # Key 7
setp sendkeys.0.scan-event-10 9 # Key 8
setp sendkeys.0.scan-event-11 10 # Key 9
setp sendkeys.0.scan-event-12 20 # Key T
setp sendkeys.0.scan-event-13 11 # Key 0
setp sendkeys.0.scan-event-14 52 # Key Dot
setp sendkeys.0.scan-event-15 14 # Backspace
setp sendkeys.0.pin-event-00 29 # Left Ctrl
setp sendkeys.0.pin-event-01 57 # Space 


 

working on the bench is not the same as working with real load, Servo drive/controller should have an inertia tuning function, even cheap ass Chinese drives have that. usually it involves setting parameter or 10 on the drive and enabling, it them moves back and forth adjusting internal PIDs to get max accelerations without oscillations.

Mitsubishi mr-j4 servo drives require active laptop connection to do autotune
yasakawa/omron sigma 1-3 don't, didn't have a chance to play with later ones.

Can you please share a link with John Thorntons tuning guide?

Hello,I have Rhino motion controls 400W BLDC motors with hall sensor and supporting 750W drives that I'm planning to use for 3 axes of the CNC mill.

 

 

I'm confused with the PWM input mode and PULSE input mode. 7i76 step and direction signals are pulses right? 

 

TB2P1(GND)        ==>        COM
TB2P3(STEP+)     ==>        SV
TB2P5(DIR+)       ==>        F/R
TB2P6(+5VP)      ==>        5V
Is this wiring right after selecting the dip switches on the drive to enable pulse input? or should I do a PWM input with 7i77 instead of 7i76?

The smart serial spec can be found in the 7i73 manual starting on page 25.

I've attached the code that runs on the 7i90 in smart serial mode. The code is for one of the soft-cores that run on the fpga. The instructions aren't that hard to work out. It should give you an idea of how it works, shouldn't be that hard to port to another mcu.

I modified the code to use a colorlight rv901t as a sserial device to add extra MPG inputs.

It could also be done by bit-banging. There's example of using shift registers with Parallel Port on the wiki.
wiki.linuxcnc.org/cgi-bin/wiki.pl?Shift_Register_Port_Expander

Mesa also has smart serial IO cards
store.mesanet.com/index.php?route=produc...oduct&product_id=113

MDR is DIN rail and has power factor correction. LRS is the aluminum frame enclosed type with no PFC. LRS is pretty much the same as RS but it's physically smaller, usually a bit thinner.

I always use LRS or HDR series depending on where I have to stick it. PFC is more efficient but it doesn't matter that much unless your driving motors. If you have a 24v supply for steppers you'll probably want PFC, the 5v supply it's upto you but if it's a major cost difference I wouldn't bother. All of Meanwell's supplies are good, even the cheap ones.

I was shopping around on Aliexpress and found these 7" parallel LCDs:
7" LCD

These are good for being driven from an MCU, there's a pretty decent library on github with a bunch of this type of LCD. There's an SD card slot on it where I suppose images could be stored. Using an HDMI LCD would be a waste for something like this.

They also have similar LCDs with a built in ESP32 but there's no real GPIO, so it'd need some I2C IO expanders to actually read buttons. When I get around to it I'll whip up a PCB with an STM32 and an RS422 transceiver for the interface and something for the buttons, I'll probably have to experiment with that a bit. Either way I can see a DIYer being able to make something like this for well under $100

BTW tangentAudio, hope I'm not threadjacking too hard......I can take this somewhere else lol

Largest difference between the MDR and the RS or LRS is probably the arrangement of the input/output terminals.
(opposite sides vs same side) So wiring layout would be somewhat different. I don't have any of the LRS or RS so I
can't make much of a direct comparison, but looking at how similar the specs are, the price point on the LRS/RS is
a compelling point of consideration.