Another plasma component...

More
02 May 2019 22:18 #132570 by BigJohnT

Someone needs to do a PR on that, I can't quite figure out how to start one, I know how to merge one lol.

JT


I was kinda hoping if Phill implemented it in his repository, it would find its way into master branch eventually!


Andy mentioned on the IRC that doing that is a huge increase in the NML memory, so I'm assuming that is the reason it has not be done. He also mentioned that he thought Rene was working on the tool table. And that a limit is silly because changing tools breaks blending anyway so motion only needs one tool... which would mean an unlimited amount of tools in the tool table. I wish my C and C++ was as good as my Python.

JT

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

More
02 May 2019 23:14 - 02 May 2019 23:15 #132575 by phillc54
John,
Thank you, what are the following?
Pause at end of cut=0.5
Wiggle length=9.999979999999999
DTHC delay -sec=1
NO DTHC - 0=1

Cheers, Phill.
Last edit: 02 May 2019 23:15 by phillc54.

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

More
02 May 2019 23:36 - 02 May 2019 23:42 #132577 by phillc54

Dee436 wrote:
I feel as though I have a path wrong or some files missing. I have tried a new clone and copy of the files to my Plasmac directory but it was still the same. I also used to have a test window pop up that I no longer see.


This section is needed in the ini file for the test window:
# this section only required if using the sim testing panel
[APPLICATIONS]
DELAY                   = 3
APP                     = ./test/plasmac_test.py

You seem to missing:
# sets the type of IHS skip, if not specified the default is 0
# 0 = if start of cut within skip-ihs-distance of end of last cut and THC is enabled
# 1 = if start of cut within skip-ihs-distance of start of last cut
SKIP_IHS_TYPE           = 0
from the [PLASMAC] section of the ini file.

This is the ini file for the sim, make sure that all commented settings are in the correct place in your ini file.

Cheers, Phill.
Last edit: 02 May 2019 23:42 by phillc54.

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

More
02 May 2019 23:53 #132578 by islander261
Phill

Some of these items are left over from from the CandCNC Mach legacy. They are needed with their hardware but may not be needed now.

Thank you, what are the following?
Pause at end of cut=0.5
Wiggle length=9.999979999999999
DTHC delay -sec=1
NO DTHC - 0=1


Pause at end of cut == just what it says, G4P?, this is to wait for the CandCNC external THC to finish doing it's thing (serial comms).
Wiggle Length == the distance for the torch to wiggle back and forth when doing a wiggle pierce to limit dross blow back. Mostly used on thicker material.
DTHC delay == the time after the arc ok signal before the THC started servoing to the voltage set point. This was required by many external hardware THC solutions because of their poor transient handling.
NO DTHC 0 == THC Enable == 1, THC Disable == 0

From what I have seen of Plasmac so far we may need wiggle length for the job shop guys that cut thick plate. The THC Enable is a requirement for me!

John

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

More
03 May 2019 00:09 #132580 by phillc54
John,
So the pause at end of cut is not the same as torch off delay.
Wiggle length I haven't thought about as I recall reading that HT (or at least Jim Colt) recommend to not use it.
THC delay may be needed depending on feedback for auto volts mode.
THC enable/disable I can add to the cut parameters rather than a global setting.(is that the way to go)

Cheers, Phill.

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

More
03 May 2019 00:24 #132581 by rodw
Replied by rodw on topic Another plasma component...

Someone needs to do a PR on that, I can't quite figure out how to start one, I know how to merge one lol.

JT


I was kinda hoping if Phill implemented it in his repository, it would find its way into master branch eventually!


Andy mentioned on the IRC that doing that is a huge increase in the NML memory, so I'm assuming that is the reason it has not be done. He also mentioned that he thought Rene was working on the tool table. And that a limit is silly because changing tools breaks blending anyway so motion only needs one tool... which would mean an unlimited amount of tools in the tool table. I wish my C and C++ was as good as my Python.

JT


Is memory a limiting factor these days?I tried to buy 2gb of memory for my PC when I bought it but the minimum size they sold was 4 gb and that was a couple of years ago. Most plasma systems are going to be running PREMPT_RT so they can use a Mesa ethernet card so I'm sure they are not running a 32 bit OS anymore.

It would really be nice to define "huge increase". Thats only going to happen if NML dynamically allocates memory. based on the changed buffer size. Does it do that?

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

More
03 May 2019 01:50 #132589 by islander261
Phill

My reverse engineering of the CommandCNC code and PP shows the pause at end of cut being a G4P? after the M5 before the distance moved is read from the external THC.

The wiggle pierce is so out of touch with what I do that I have never given it any thought YMMV.

I think that Auto Volts is for thick sheet guys. I doubt that I will use it much but others will find it handy. Looking at Halscope traces of the cuts I've recorded Plasmac grabs the torch and drives it to the commanded arc voltage quickly and accurately. So I don't think I need it right now YMMV.

I don't really care where the THC Enable is stored as long as I can change it from cut to cut by changing the tool or the material when generating Gcode. The way my system works right now if I enable the THC from the GUI it over rides every thing, if I clear it in the GUI then the Gcode drives. Not perfect but works for me.

Sorry, I didn't get to cut today. Three days ago I filed for Social Security (old age pension) and Medicare (health coverage from working years payroll taxes). The next day my main old client called after 4 years of no work with a proposal for a nice new design/prototype contract! so now I have to figure out how to undo all of this because of income limitations while still having affordable health care. Sometimes you just can't win.

John
The following user(s) said Thank You: phillc54

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

More
03 May 2019 02:07 #132591 by rodw
Replied by rodw on topic Another plasma component...

. Three days ago I filed for Social Security (old age pension) and Medicare (health coverage from working years payroll taxes). The next day my main old client called after 4 years of no work with a proposal for a nice new design/prototype contract! so now I have to figure out how to undo all of this because of income limitations while still having affordable health care. Sometimes you just can't win.

John


John sounds like you need a company (and a smart accountant). Dunno how it works over there but my wife and I both draw income up to the tax free threshold of $20k plus a bit more.. During all of the formative years we ended up with some carry forward losses so the company does not have a tax liability on its profits. Between that and repayment of directors loans, we won't have a taxable income for a few years.... But our private health car insurance cost is horrific!

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

More
03 May 2019 02:52 #132592 by islander261
Rod

I am actually pretty good at manipulating how we generate income or profits and the tax we have to pay on them. When you don't have outside investors or need maximum liability protection here there are better forms of organizing entities to avoid double taxation and tax free savings limitations. The problem I have is that the limits while drawing SSI are total income from all sources before adjustments not just earned income.

Phill

Why is the plasmac.skip_ihs_type pin a float type? At least for now it only has two states and can be bit (boolean) type. My control switch is causing LinuxCNC to crash because of this. I guess I will need to insert a type conversion component in my hal.

John

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

More
03 May 2019 03:43 #132594 by phillc54

Why is the plasmac.skip_ihs_type pin a float type? At least for now it only has two states and can be bit (boolean) type. My control switch is causing LinuxCNC to crash because of this. I guess I will need to insert a type conversion component in my hal.

No particular reason other than my stupidity.
Originally it was to be an S32 as I had four different states.
I will fix that. You would prefer 0 for skip if near last successful probe and 1 for skip if near end of last, is that correct?

Cheers, Phill.

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

Moderators: snowgoer540
Time to create page: 0.307 seconds
Powered by Kunena Forum