Advanced Search

You only need a pierce delay for piercing not for edge starts.

John

I may not be doing things right. I was doing edge starts with a pierce movement and a slight delay. Maybe I shouldn't do that.

If your arc voltage is ok while piercing it should not dive. I guess you could "and" something to block the enable while piercing.

John

BigJohnT wrote:

Looking good! I like the scenery outside your door... The THC panel looks familiar...

John

Yeah, I like your panel layout. Thanks for all your help.

One thing I may need to do is turn off the THC during the pierce movement. I was cutting some 1/2" steel and the thc would cause the nozzle to dive and touch material until the torch stared moving horizontally. I think I could do it in classic ladder and have the THC turn on when Z axis is below a set point like 0.065". Do you know of a better way?

Tim

Looking good! I like the scenery outside your door... The THC panel looks familiar...

John

Got it working and it's working good. THC is almost a must have for long cuts. Made a small classic ladder program for the under, ok, over indicators.

link to vid.
www.dropbox.com/s/tgl8bx1dc1qu4vm/VIDEO0054.mp4

nozzle is old but still cutting ok with the thc.

Ok. Didn't know you could do that. Guess that's why there are 3 jumpers.

I would jumper the B pin as single ended and then ground it,
I would use differential on the 'A' pin (for noise immunity)

OK. That makes perfect sense. I didn't think about B being the signal for up/down. So in differential mode would you need to make B+ low(ground) and B- high(+5v)? Or is differential mode even necessary? It's supposed to have more noise immunity.

Thanks,

Tim

Don't know about the signal issues (probably need JT to help with that), but the negative scale issue is because the THC uses an encoder counter velocity signal to read frequency. The encoder counter is set in up/down mode (A input = count, B input = up/down) the default high state on the unconnected B input makes the counter count down rather than up, causing the velocity to be negative, You can fix this as you did, or by setting the encoder scale to a negative number (or setting the encoder 'B' pin to TTL mode and grounding the B pin)

Ok. Got my 5i25 installed in another computer. I was able to get John's 'plamsa-5i25' configuration to start up in linuxcnc after removing some lines in the ini and hal files. I have an adjustable power supply connected to the THCAD-10 voltage input to simulate an arc voltage. I have a couple questions.

1) For some reason I am getting a negative 'thc-encoder.vel'. I am able to correct this with negative values for 'thc-vel.scale' and 'thc-scale.offset'. Just wondering why it is negative value. All connections to the THCAD and 7i76 appear to be correct. I tried both single ended and differential connections to the incoder inputs on the 7i76 with same results.

my hal settings:

# Set up the Encoder for the THC
setp hm2_5i25.0.encoder.00.counter-mode 1
net thc-vel-in hm2_5i25.0.encoder.00.velocity => thc.encoder-vel
setp thc.scale-offset -119500
#setp thc.scale-offset 0
setp thc.vel-scale -0.00037866834
# this might need to be adjusted
setp thc.correction-vel 0.000075
# I added
setp thc.voltage-tol 1.0

question 2) I have to set 'thc.voltage-tol' manually in the hal file. The 'Voltage tolerance' setting in the pyvcp panel doesn't seem to be working or connected.

#net thc-volts-tol pyvcp.volts-tolerance thc.voltage-tol

If I remove the '#' Linuxcnc won't start up.

Also Arc 'Under' 'OK' and 'Over' lights don't seem to work or be connected in the postgui.hal file.

postgui.hal:

# THC connections
net thc-enable pyvcp.thc-enable thc.enable
net thc-voltmeter pyvcp.actual-volts thc.volts
net thc-voltset pyvcp.volts-requested thc.volts-requested
net thc-req-vel thc.requested-vel <= motion.requested-vel
net thc-cur-vel thc.current-vel <= motion.current-vel
net thc-vel-status pyvcp.vel-status thc.vel-status
net thc-vel-tol pyvcp.vel-tol thc.vel-tol

# connect the arc ok indicator
net start-motion-input => pyvcp.arc-ok

#net thc-volts-tol pyvcp.volts-tolerance thc.voltage-tol

# MDI commands
# rapid to home button
net remote-rapid-home halui.mdi-command-00 <= pyvcp.rapid-to-home

Appreciate any info.

Thanks guys. That is very helpful information. I will probably order the 5i25, 7i76, and THCAD and try to get it working in another computer.

It will work but the response time will be slower than when counted in hardware,
since a low pass filter is needed to get sufficient resolution.
At a 1 ms servo thread the THCADs ~1MHz count rate gives about
10 bit resolution or about 1/3 V resolution at 300V full scale

With a software counter, if you have a 50 KHz base thread you only get about 4-5 bits
of resolution at 1 ms update rate, so you need to use the low pass filter with about a
20 ms time constant to get to a useful amount of resolution.

I do know the THCAD card has dividers for the pulse to work with a parallel port but I don't know much about the C10 even though I used one on my plasma when I first started. At one time I used the parallel port directly I think with the THCAD...

John

Ok. Thanks for the info.

Right now I am just using the pc's parallel port with a cheap C10 breakout board from Keiling and everything is working very well. How well would it work just hooking up the THCAD card to one of the C10's inputs and using a software encoder (velocity pin) to get arc voltage? Would this cause problems at fast cut speeds because of slow response? Would it possibly work?