Advanced Search

Some clarifications before this goes sideways:
-COM0/COM1/COM2 are referred to serial ports, be it RS232 o other, usually.
-LPT0/LPT1/LPT2 and sometimes only P0/P1/P2 are referred to parallel ports, usually.
What are you referring to with COMn ?
Serial ports are pretty much useless with LinuxCNC, except for some stuff.
Parallel ports, also known as IEEE 1284, are used for machine control still, are perfectly usable with LinuxCNC, but have limited pulse rate capabilities compared to Mesa board.
Reasonable pulse rate or encoder pulse rate for parallel port would be 10000-20000 PPS, sometimes can go to 50000 PPS with a bit of luck/tweaking the BIOS.
At 3000 RPM, that is 50 RPS, so for a 100PPR encoder ends up 5000 PPS, and for a 1000PPR encoder it is already way to fast for most PC setups, see above.
Now convert that to mm/m, as you gave us no info about drive/gear/reduction.
After all that, yes, you can use a parallel port, yes it will work with encoders (2 max normally per port), and yes, you can lower the speed if you have higher count encoders.
But first, fix the wiring and get rid of interference/bad grounding.

alangibson wrote:

My intention with the TVS was to limit input voltage to 250V.

OK, but you must also add resistors in front of it, or it will blow up the first instance the voltage goes above 250V.
-
Personally, i have a strict policy of doing the voltage divider using 3 resistors, always, saves a lot of headache in case of mishaps or wires shorts or any of plasma wires touching ground or machine.
Example, in case of Hakan's schematics, i would use a 2K7 in the middle of two 50K. The other side of both 50K would go to plasma arc voltage, and the measuring would be done on the 2K7.

hi, I am using a absolute multiturn spindle encoder , it is working but I need to simulate de Z index for G33 and G73
so I define the follow net

net spindle-index-en    and2.2.out => spindle.0.index-enable

but always say spindle-index-en and after some test I found spindle.0.index-enable it is already linked
So I delete may hal and leave just basic things like lcec  etc.. basic to make my ethercat network working 
but any way say the signal is already linked, I made allot names changes and this has the same result   
on ini file I have only a call to this hal file.  
after I comment the net line and run the linuxcnc
On hal show I look for the linked spindle.0.index-enable, and I can toggle  the input spindle.0.index-enable
on bash I look for the variable spindle.0.index-enable
at moment the linuxcnc probe basic stance running, there is not link result.
so mean that input is not linked jet, but after I close the linuxcnc instance and remove the comment the net on hal file I get the same error,
spindle.0.index-enable already liked 
I been using probe basic GUI for lathe. i found that you can simulate the spindle feedback or use the encoder feed back, is this option can declare this spindle input?


as additional comment I have the encoder count working and can calculate the right speed test as  tested whit tachometer and it is really precise , so I can control the spindle position net spindle-index-en    and2.2.out => spindle.0.index-enable


Any advice

Have you tried to use the example configuration?

github.com/scottalford75/Remora-RT1052-c.../remora-rt1052-basic

There is some latching logic in the hal file:

net user-enable-out <= iocontrol.0.user-enable-out => remora.enable
net user-request-enable <= iocontrol.0.user-request-enable => remora.reset
net remora-status <= remora.status => iocontrol.0.emc-enable-in

Thanks for the schematic. I wish I had found this a while ago.

I like how you use isolated and regulated voltage. I need to see if I can find a single component that does that.

> The torch is negative, ground is positive.

Ah right. I forgot that. Is that why U4 is inverted?

I took a stab at dissecting your schematic. This is what I got:

Voltage source
5V: Common 5VDC power source
U1: 1W Isolated DC-DC 9V out; Isolation 1.5k VDC/min
C3: Decoupling
U5: 5V LDO voltage regulator
C3: Decoupling
VCC_ISO: Isolated 5VDC

Input stage
R7,R9: Voltage divider, 38:1 ratio
D1: Limit input to max 5V
R6: Current limiting
U4: Voltage buffer (inverted ?)
R8: ?
R5: Current limiting?
C4: Decoupling?
C5: Decoupling?
U3: ADC with 12C output
R4: Pull-up resistor
R5: Pull-up resistor
C2: Decoupling?

Output stage
U2: I2C isolator; Isolation 2.5-kVrms
C1: Decoupling?
R1: Pull-up resistor
R2: Pull-up resistor

At a first glance, I think you have the wrong GND
connected to  the 7I96S spindle analog SPINDLE- (TB2 pin 22).
It should be the potentiometer GND not the digital input GND.

mitsubishi cnc never had step/dir controled drives...
meldas 3 had a parallel differential bus
meldas 5/6 is serial differential
meldas 7/8 is optical
i can interface them all..

My intention with the TVS was to limit input voltage to 250V. The ADC expects 0-5V range, and voltage division is 50:1, so anything above 250V is out of range.

In reality 250V is way high for a cutting voltage. I just came up with it by working backwards from the ADC voltage range.

Does that make sense?

Optoisolators are a good idea. I was planning on the TVS limiting the voltage, but I guess it's always possible that someone would hook this board up to a HF start torch.

X1 just has 5V power and rs-485 YIO bus...
 

What CPU do you have? How many cpus does it report?
what does echo $(nproc) say?
Its possible you are trying to isolate a nonexistent core

Hakan wrote:

One would like to have a speed output from the drive but there isn't one.

Should be possible to get rps or rpm in the frequency output
by setting proper scaling factor in frequency-scale. 
Have not tried it, though.

Should be possible

Hello guys,
im new in Linux CNC i wat to connect Linux CNC with my Spindel+YL620a Frequency Converter
via a 7i96s and failed. I dont understand what im doing wrong.....

I connectet the Components like the Picture.

Added this to the Hal
setp hm2_7i96s.0.pwmgen.00.output-type 1
setp hm2_7i96s.0.pwmgen.00.scale [SPINDLE_0]OUTPUT_SCALE
net spindle-vel-cmd-rpm => hm2_7i96s.0.pwmgen.00.value
net spindle-enable => hm2_7i96s.0.pwmgen.00.enable

Added this to the ini
MAX_OUTPUT = 24000.0
OUTPUT_SCALE = 24000
OUTPUT_MIN_LIMIT = 0
OUTPUT_MAX_LIMIT = 24000

Changed the Frequency Converter Settings like in the Picture.

Im grateful for any help or tips.

Thaks in advance
Tach0
 

Which series of drives do you have? If they are fibre-optic controlled, Dmitry has the cleanest, fastest solution to controlling the drives over the SSCNET III/H control fibres.

If they are copper controlled network drives, then Dmitry has a control for that as well, but you can also run the MESA cards for step/direction interfaces.

I thought I'd added my HAL file, guess not.

The M-code files still have remnants of my first attempt at switching the drive (internally) between velocity and position modes as in @Aciera's example.  That stuff is commented out; just ignore it if you aren't switching your drive.

If you ARE switching your drive internally, you'll probably want stuff like "M66, check for drive mode" in your M-code file.

Good job!

I've renamed my m-codes from the original example, and due to some jitter (servo dithering) when the motor/drive is in position mode I just leave the drive in velocity mode.  LCNC is switched between spindle/orient/positioning. 

My M-codes and main spindle HAL file are attached for your review; maybe they can help.

There is quite a bit commented out, but the gist of it is M254/M255 are used to switch between C-axis and spindle mode, using spindle-orient to realign the spindle before re-connecting the C-axis encoder.

M102 & M103 are 'backstage' M-codes used to adjust stepgen & c-axis PID settings (and disconnect index-enable) in a way you can't (I think) do from inside a normal .ngc g-code file.

Let me know if anything is really unclear.