Advanced Search

What are the joint and axis MIN_LIMIT values in your .ini file for Z?

When  HOME = 0,  MAX_LIMIT  should be greater than 0, to avoid get stoped by softlimit error

Hi, which module have you been trying to use? The software Encoder module or the QDC module? Both support the index pulse detection and encoder count latching so LinuxCNC can acquire the index position count.

There are a few issues. Should we break them into new threads?

1. The discussed issue about being on or over a limit. The resolution/variance comments.
2. Axis UI will allow an AXIS to be moved back within limits, but Probe Basic does not
3. The docs state: "The [JOINT_N]MIN_LIMIT/MAX_LIMIT settings are enforced while jogging in joint mode prior to homing. After homing, [AXIS_L]MIN_LIMIT/MAX_LIMIT coordinate limits are used as constraints for axis (coordinate letter) jogging" but that is not the case. This is different for Probe Basic and Axis.
3a. Using the Axis UI, the JOINT MAX_LIMIT and MIN_LIMIT are ignored before homed, meaning no limits are enforced.
3b. Probe Basic seems to enforce the AXIS_L limit, whether homed or not

  The velocity range only seems to be around 6500 hz from 0-10mm (3/8")

That does seem quite small, maybe too much stray capacitance on the sensor?

The BCL-AMPs ~2.5 MHz /2 frequency of ~1.25 MHz only gives you about 1250
counts per 1 ms servo period which will end up adding quite a bit of noise.
 This may be improved if you set the hires-timestamp of the encoder true but
a better solution may to use the periodm's frequency output as it does frequency
estimation by 1/period_average:

forum.linuxcnc.org/plasma-laser/51568-in...e-sensor-to-linuxcnc

Is there a home switch on the Z axis?

Rember that you can touch-off an axis.

So G0 Z0 ≠ "Go to actual, physical 0 coordinate"

Sometimes LinuxCNC is weird.

Look at the stepconf default config and try to make your config similar if you can.

Worked for me!

7I97 or 7I96?
 

Hola a todos soy muy novato en esto pero lo intento.
Monte un  plasma cnc con mesa 7i97s que consequi hacer que funcione bien hasta que le puse el thcad2, le hice solo 3 pruebas de antorcha y los motores ya no se mueven, save alguien si puede ser que se haya quemado la tarjeta 7i96. Se conecta a linux puedo encender antorcha pero no se como verificar cual puede ser el problema, gracias.

The ELS project I listed above is my design, yes. I've talked about it before here on this forum but interest seems low. Most people go for Clough42's much simpler+cheaper approach: github.com/clough42/electronic-leadscrew

My ELS system is LinuxCNC based with a web interface frontend. I use a 7i96 board which has only one encoder, hence this rp2040 based project.

My system also has X-axis steppers, not just a Z-axis stepper. One my goals is to be able to do NPT taper threads easily since I do this often and using dies is a PITA on stainless.

no checksum for SPI, it's only for short distance. only the UDP and RS485/RS422 connections have checksum

I was checking your GITHUB repos - great content. I am actually working on ELS myself right now. Fitting mechanical parts first to the Atlas lathe - need to add spindle encoder and stepper tot the lead-screw.

What ELS do you use ? your own design? LinuxCNC based ?

Yes, I am actively using it. I just finished hooking it up in a closed loop system so it can do canned cycles on my lathe. So far the results look good, it works great!

I have my own ELS system for which I use it for: github.com/tinic/el-linuxcnc-electron

My hal file is here, though I doubt this will be meaningful to you since my system is rather complicated: github.com/tinic/el-linuxcnc-electron/bl...p/elle-hal/lathe.hal

what do you think of my configuration:

# --- Encodeur configuration ---
setp [HMOT](CARD0).encoder.00.counter-mode 0
setp [HMOT](CARD0).encoder.00.filter 0
setp [HMOT](CARD0).encoder.00.scale [JOINT_3]ENCODER_B_SCALE
#setp [HMOT](CARD0).encoder.00.index-enable true
setp [HMOT](CARD0).encoder.00.index-mask 1
setp [HMOT](CARD0).encoder.00.index-invert 0
setp [HMOT](CARD0).encoder.00.index-mask-invert 0

# --- Feedback position/velocity via encodeur ---
net b-pos-fb <= [HMOT](CARD0).encoder.00.position
net b-pos-fb => pid.b.feedback
net b-pos-fb => joint.3.motor-pos-fb
net b-vel-fb <= [HMOT](CARD0).encoder.00.velocity

# --- Index pour homing ---
net b-index-enable joint.3.index-enable <=> [HMOT](CARD0).encoder.00.index-enable
net phaseZ [HMOT](CARD0).encoder.00.input-index

# --- Raw counts (optionnel debug) ---
net b-pos-rawcounts <= [HMOT](CARD0).encoder.00.rawcounts

# --- Switches (home et fins de course) ---
net b-home-sw => joint.3.home-sw-in
net b-neg-limit => joint.3.neg-lim-sw-in
net b-pos-limit => joint.3.pos-lim-sw-in

So, has there been any progress on getting this sensor to market? 

I'm trying to get a laser machine built, but, this is one of the toughest issues I'm facing.  I have a MESA 7i76EU that I have been trying to interface a BCL-AMP output to.  I have the BCL-AMP connected to a Raytools BT210S002A cutting head sitting on the bench with a chunk of steel in front of it and grounded to the case of the Raytools cutting head.  Between the BCL-AMP and the 7i76EU, I have a TI SN74HCS74DR dual Schmidt Trigger Flip Flop that I'm using to divide the BCL-AMP output frequency by 2 and "clean it up" at the same time.  However, it's just not very stable.  The velocity measured by the 7i76EU jumps around about 600-700 hz.  The velocity range only seems to be around 6500 hz from 0-10mm (3/8") sensing distance, so, my positional sensitivity sucks at around 0.1 inches.  Just for giggles, I ran the signal through 4 Flip Flops to see if the 'jumping around' would decrease.  It didn't.  Which leads me to believe the instability is introduced somewhere in my circuitry or wiring.  I even tried connecting Q and Q not outputs into the MESA card as differential signals and twisted the wires.  No help.

Anyhow, This CV2 height sensing device sounds very promising.