Set the current for the plasma generator from the Plasmac

phillc54
Offline
Moderator

21 Aug 2020 23:50 - 22 Aug 2020 00:15 #178969 by phillc54

Replied by phillc54 on topic Solved: Set the current for the plasma generator from the Plasmac

hermann1976 wrote:
I found out why I couldn't edit any entries (neither nor feed rate, ampere, ect.)
In the settings it is possible to activate or deactivate the keyboard operation in the hardware tab (gmoccapy).
If I activate it then I should be able to use the keyboard, right?
But when activated I cannot set any keyboard entries. If I deactivate it then I can edit the entries. Is that twisted?

It seems that if you enable keyboard shortcut in Gmoccapy then it takes control of all keystrokes and if if doesn't use them as a shortcut it doesn't pass them through.

Now linuxcnc no longer starts. If I comment out the entry "net run_current plasmac_run.cut-amps-f => hm2_7i76e.0.7i76.0.0.spinout" then linuxcnc works again, but the current value is no longer output (before it was working).

Was this actually working before. I thought that the postgui hal file was run before the gladevcp files.

Edit: This seems to work in Axis but not Gmoccapy. As an experiment I changed the name of postgui.hal to postgui.tcl and also changed it in the ini file and it now works for both Axis and Gmoccapy.

Last edit: 22 Aug 2020 00:15 by phillc54.

Please Log in or Create an account to join the conversation.

taloot
Offline
Elite Member

22 Aug 2020 08:03 - 22 Aug 2020 08:17 #178982 by taloot

Replied by taloot on topic Set the current for the plasma generator from the Plasmac

PCW wrote:

taloot wrote:

rodw wrote:
PCW, thats a cool idea. I wonder how easy it would be to replace the current knob on a thermal dynamics with a digital pot controlled with the spindle?
good idea but u will gain nothing

also i dont think 0-10v will be stable and amp stability important to control the arc density

Actually with the Telwin you gain something quite significant, the ability to change the current in real time

Why do you think the current control would be unstable? The 7I76E analog output stability depends
chiefly on the 10V supply stability (Whether or not you could replace the Hypertherms POT with a voltage
source remains unknown)

coz i see this signal in laser fluctuate a bit sometime 9.5 9.6 9.7 and its set to 10v but it might be due wire resistance and power supply ac in voltage
unlike set the power using serial
2nd
u cant control it dynamically. because the arc voltage will be changed. so THC will have problems
anyway i make qtpyvcb with function for dynamic control..based on speed..with virtual gas console.. it might help.. for plasma

Last edit: 22 Aug 2020 08:17 by taloot.

The following user(s) said Thank You: hermann1976

Please Log in or Create an account to join the conversation.

rodw
Offline
Platinum Member

22 Aug 2020 08:45 - 22 Aug 2020 08:45 #178984 by rodw

Replied by rodw on topic Set the current for the plasma generator from the Plasmac

taloot wrote:
u cant control it dynamically. because the arc voltage will be changed. so THC will have problems
anyway i make qtpyvcb with function for dynamic control..based on speed..with virtual gas console.. it might help.. for plasma

But when the the cutting speed falls as you approach a corner, the THC is disabled by the velocity anti-dive (or Cornerlock). But you do raise an interesting point. Plasma is a constant current device so voltage is varied to keep the current at the desired level. It might be possible to manage the current to keep voltage the same during velocity slow downs. This should result in maintaining the intensity of the arc (is that called power factor?) This should prevent the widening of the kerf during velocity slow downs...

Last edit: 22 Aug 2020 08:45 by rodw.

Please Log in or Create an account to join the conversation.

PCW
Offline
Moderator

22 Aug 2020 14:44 - 22 Aug 2020 17:27 #179000 by PCW

Replied by PCW on topic Set the current for the plasma generator from the Plasmac

If the voltage fluctuates with a 7I76E output that's because the external reference supply is not well regulated, that's easily fixed. The 7I76E analog output stability should be better than 0.1%

Lots of high end systems vary the current dynamically. A first order model of voltage vs height versus current would allow THC scale compensation at different currents with nothing more than MX+B done by LinuxCNCs scale component (the coefficients could be determined with a few trial cuts at different currents and checking the THC set Z height)

Last edit: 22 Aug 2020 17:27 by PCW.

Please Log in or Create an account to join the conversation.

hermann1976
Topic Author
Offline
Senior Member

22 Aug 2020 18:30 #179021 by hermann1976

Replied by hermann1976 on topic Set the current for the plasma generator from the Plasmac

taloot wrote: Edit: This seems to work in Axis but not Gmoccapy. As an experiment I changed the name of postgui.hal to postgui.tcl and also changed it in the ini file and it now works for both Axis and Gmoccapy.

Thank you, this is working.
Is this a temorary solution or could it be for fix?

Please Log in or Create an account to join the conversation.

rodw
Offline
Platinum Member

22 Aug 2020 23:10 #179044 by rodw

Replied by rodw on topic Set the current for the plasma generator from the Plasmac

PCW wrote:
Lots of high end systems vary the current dynamically. A first order model of voltage vs height versus current would allow THC scale compensation at different currents with nothing more than MX+B done by LinuxCNCs scale component (the coefficients could be determined with a few trial cuts at different currents and checking the THC set Z height)

Interesting Peter. From the graph in the Plasma primer, we know for one machine at one particular current we know with 99.4% confidence that

Volts  = 7.53 * Height(mm) + 89.2

If this experiment was repeated at different currents (and halstreamer had a switch to set CSV output), it would be very easy to collect statistically significant results. Shouldn't the results of a change in current vary in accordance with Ohms's Law?

But I think what we need to do is to create a model that predicts the required current to keep voltage constant where cut velocity varies. If we could do this, we should maintain a constant kerf width. (I think)

Please Log in or Create an account to join the conversation.

phillc54
Offline
Moderator

22 Aug 2020 23:33 #179047 by phillc54

Replied by phillc54 on topic Set the current for the plasma generator from the Plasmac

hermann1976 wrote:

taloot wrote: Edit: This seems to work in Axis but not Gmoccapy. As an experiment I changed the name of postgui.hal to postgui.tcl and also changed it in the ini file and it now works for both Axis and Gmoccapy.

Thank you, this is working.
Is this a temorary solution or could it be for fix?

I am reasonably confident it could be permanent as it seems to work consistently and it only seems to be a Gmoccapy issue. I did some work a while ago on moving to a tcl Hal file and using a conditional test to make sure the relevant components were loaded before doing any hal commands. Something like this would wait up to three seconds:

set start [clock seconds]
set timeout 3
while {1} {
    if {[hal list pin plasmac.state-out] != {} &&
        [hal list pin plasmac_run.cut-amps-f] != {} &&
        [hal list pin plasmac_config.arc-fail-delay-s] != {}} {
        break
    } elseif {[clock seconds] >= ($start + $timeout)} {
        puts "\n*** TIMEOUT WAITING FOR PLASMAC HAL COMPONENTS TO LOAD ****\n"
        break
    }
}

But it looks like it may not be requried.

Please Log in or Create an account to join the conversation.

taloot
Offline
Elite Member

23 Aug 2020 14:47 #179104 by taloot

Replied by taloot on topic Set the current for the plasma generator from the Plasmac

rodw wrote:

taloot wrote:
u cant control it dynamically. because the arc voltage will be changed. so THC will have problems
anyway i make qtpyvcb with function for dynamic control..based on speed..with virtual gas console.. it might help.. for plasma

But when the the cutting speed falls as you approach a corner, the THC is disabled by the velocity anti-dive (or Cornerlock). But you do raise an interesting point. Plasma is a constant current device so voltage is varied to keep the current at the desired level. It might be possible to manage the current to keep voltage the same during velocity slow downs. This should result in maintaining the intensity of the arc (is that called power factor?) This should prevent the widening of the kerf during velocity slow downs...

yes very good idea.. so u compensate speed vs amps.. to keep the volt the same..during acceleration and deceleration

plasma is good with straight lines,, but when it come to turns and hard edges.. the problems appear
thats why light gentry plasma always better.
also the uniformity of the cut based on speed and arc relationship.. the sweet spot windows is so small.. the plasma arc eating metal so fast

The following user(s) said Thank You: rodw

Please Log in or Create an account to join the conversation.

Moderators: snowgoer540

Time to create page: 0.161 seconds