THCAD-10

More
01 Sep 2018 20:09 - 01 Sep 2018 20:12 #116957 by cruzinone
Replied by cruzinone on topic THCAD-10
My brother is KB8WMC and I am who he is helping. We do not know how to test thc.pid.volts. Does this have to do using the a 9v battery. Where do we hook up the 9v battery and what will we look for when we do this. We have looked the forum over and can not find details and instruction on doing this.
We have been using the hal meter to obtain data. Does my plasma machine pm45 50-1 voltage divider factor into the offsets or scaling or is that already factored in?
Can someone point us to instructions to test the thc.pid.volts? any help will be appreciated.
Last edit: 01 Sep 2018 20:12 by cruzinone.

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

More
01 Sep 2018 21:54 - 01 Sep 2018 21:54 #116961 by rodw
Replied by rodw on topic THCAD-10
You need to become familiar with Halscope
linuxcnc.org/docs/html/hal/tutorial.html#_halscope

Then you will be able to plot arc voltage and other parameters.
Last edit: 01 Sep 2018 21:54 by rodw.

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

More
01 Sep 2018 23:18 #116962 by PCW
Replied by PCW on topic THCAD-10
The other detail not mentioned about arc voltage scaling is the plasma supplies
divider ratio, unless the arc voltage scaling is right none of the THC setpoint numbers will makes sense either.

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

More
06 Sep 2018 17:29 #117156 by kb8wmc
Replied by kb8wmc on topic THCAD-10
I apologize for my ignorance in this situation...My lack of knowledge sticks out like a sore thumb...I did find some information on just how others have gone about obtaining a voltage reading through the THCAD-10...Rick G. suggests the following method:

"Can you disconnect the voltage input to the THCAD from your plasma power supply and apply a know voltage such as a battery to the THCAD and confirm that the thc voltage in hal is correct?"
Rick G

Rodw used a different method which appears from his pictures and written description to go through his plasma power supply and voltage divider with 96volts

forum.linuxcnc.org/show-your-stuff/32029...tter-build?start=190

Using Rick G's method we applied 9.45 volts input (from a battery) and the HAL meter indicated 3.46 volts...I don't know if I am correct but I am assuming this ratio between these values must be factored in so that the THCAD-10 input and output are equal and this is done by changing the scaling...

As we are using a PM45, it has a voltage divider of 50:1...How does the torch power supply fit in to the overall scheme of things, if in fact it does, and what method is used to obtain measurements?...We would prefer not to have to break into the plasma power box if at all possible, but will do so if necessary...Then how are the results used to make calculations?...
thanks in advance

Mike

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

More
06 Sep 2018 17:35 #117158 by PCW
Replied by PCW on topic THCAD-10
If you have a 50:1 divider, this means that the full scale reading should be close to 500V (and should read 450V if a 9V source is connected directly to the THCAD-10)
It should also read close to 0V if the THCAD analog input is disconnected
The following user(s) said Thank You: kb8wmc

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

More
06 Sep 2018 20:42 #117166 by Grotius
Replied by Grotius on topic THCAD-10
If you have a 50 to 1 voltage divider this piece of electronic will scale 50 to 1.

If your plasma inverter has a raw arc of 300 volts dc. The same piece of electronic will do 50 to 1 of 300 volt's.

If a mesa input card is 0 to 10 volt input. It's okey. It can be used for analog input.
If a mesa input plasma card is 0 to 500 volt dc input, they have to tell you, because, then you know your scaling.

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

More
06 Sep 2018 21:04 #117167 by rodw
Replied by rodw on topic THCAD-10

Rodw used a different method which appears from his pictures and written description to go through his plasma power supply and voltage divider with 96volts

forum.linuxcnc.org/show-your-stuff/32029...tter-build?start=190


The only reason for this approach was I was getting strange things. It seemed the documented divider ratio in my plasma differed from the published 16:1 ratio for that machine. So I was testing the internal plasma voltage divider circuit, not the THCAD-10.

Unexpectedly, I found by using a voltage near what we cut at was that the resistor network in the plasma itself was counted by the THCAD-10 as being part of scaling resistor network resulting in additional voltage division that I had not allowed for.

But I did this test after I had confirmed with low voltage as Peter suggests that the THCAD was working.

So do the battery test, then use Halscope, to capture a plot of the torch voltage when cutting and compare that with the published cut chart data to make sure you are in the ball park.

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

More
07 Sep 2018 22:40 #117209 by kb8wmc
Replied by kb8wmc on topic THCAD-10
UPDATE: IT WORKS!!...A Giant thanks to all that have contributed their information to our project...That information finally sank in...Although everyones input was instructing me properly, it was not sinking in...I was not understanding how the information fit in...I read and re-read the thc-pid manual, the Mesa thcad-10, and hypertherm manuals numbers of times and pored over all the postings in this forum, I just did not grasp the relationships of the info sufficient to be able to plug it in and be usable...
Yesterday as I read PCW's post something clicked and thought I was beginning to grasp what had to be done and place the proper numbers in the right boxes...
Today when I got to my brother's shop, it took about 1/2 hour starting with a battery and multimeter then applied the numbers based on what I thought I was understanding, and voila!, IT WORKS properly...
We made some test cuts...My brother then cut two parts he need for his work out of 20 ga. S/S...The THC follows the tool path adjusting elevation as necessary...We know we will have to do some tweaking on the correction velocity...But I now know what needs to be changed and why and the effects...I also realize that there is more to learn so I am sure I will be asking questions in the future...
Thanks to all for the help...I apologize if it seemed I was being a pest...

Mike

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

More
08 Sep 2018 15:59 #117223 by kb8wmc
Replied by kb8wmc on topic THCAD-10
In order that my brother and I not forget the processes we followed to get the THCAD-10 to function on his plasma table, in case we ever needed to do it in the future, I wrote an instruction manual for us...Although there is quite a lot of information that others have written and shared I have attempted to consolidate the methods which seem at present to be working for us...I am quite sure that some of the descriptions and nomenclature used in my writing may not coincide with actual descriptions that others would use, but for us I tried to keep it simple and understandable...
Mike

Torch Height Control

Description: The following describes the processes I used to set up a
functional THC.

Plasma Table: 50" x 50" with four stepper motors (2X; 1Y; 1Z)
Plasma Torch: HyperTherm PowerMax 45 with 50:1 voltage divider
Electronics: Mesa 5i25/7i76/THCAD-10 divide by 32 mode
Control: LinuxCNC ver 2.7; THC-PID component (by John thornton)

PROCESS: Using the equation provided in the THCAD-10 manual as a guide
for calibration of the THCAD-10 as follows:

V= VFS * (FOUT - FZERO) / (FFS - FZERO)

Both FZERO and FFS values are obtained from sticker on rear of THCAD-10 card.
where: FZERO = 118000 (frequency output at 0.0 volts input)
FFS = 924800 (frequency output at 10.0 volts maximum input)
FOUT = 806800 (difference by subtracting FZERO from FFS)
VFS = 10 (Maximum allowable volts through THCAD-10 card.

NOTE: FOUT is the total frequency range.

Substituting values in above equation provides:
V = 10 * (806800 - 118000) / (924800 - 118000)
V = 10 * (688800) / (806800)
V = 10 * 0.8537431
V = 8.5374318 (Estimation of THCAD-10 voltage at maximum range.)

Calculate an offset/scale which converts frequencies to voltages using
the following equation:

Scale = Divide By Mode * VFS / (FOUT - FZERO)
Scale = 32 * 10 / 688800
Scale = 0.0004645

Actual calibration was done using 9 volt battery and multimeter. The battery
provided 9.5 volts. Disconnecting the PM45 Plasma power supply cable
from the input side connector (right side) of the THCAD-10. The battery
provided the 9.5 volts input. Using the LinuxCNC HAL Meter function and
selecting the pin thc-pid.volts, the Hal Meter showed approximately 11+ volts.

This suggested that the existing Scale had to be reduced to be equivalent to
the input voltage of 9.5 volts...To provide an approximation, the Scale was
changed to 0.0004 and the Hal Meter was again read and was very close to the
actual input voltage of 9.5 volts.

Scale = 0.0004

To factor in the 50:1 voltage divider built in to the HyperTherm PM45, the new
scale (0.0004) was multiplied by 50 providing a new scale factor of 0.02. This
was tested by actual cutting and was found to perform well with the foregoing
approximate values.

Scale = 50 * 0.0004
Scale = 0.02

It should also be noted that further testing needs to be done with to obtain
more accurate measurements and values. Also, using the Correction Velocity
(thc-pid.p-gain) of 0.00002 as suggested by John Thornton seems to be a good
value to start with.

While we don't have the up/down indicator LED's working as yet, and I don't
yet have an understanding of what "Velocity Tolerance"; "Volts Tolerance";
and "Volts Limit" actually do and/or control, I am presently satisfied with
leaving the default settings of each until I understand more. The other
features seem to work properly and the cutting voltage setting is imported
from Gcode into the display screen "Volts Requested" and when cutting the
"Volt Meter" (actual volts) seems to function properly.
The following user(s) said Thank You: Grotius, Uthayne

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

More
27 Oct 2018 18:43 #119513 by Uthayne
Replied by Uthayne on topic THCAD-10
I've been following along in this thread and have been able to achieve decent functionality with a similar setup (THCAD-300 instead of -10). However, I think I'm running into corner lock issues. After looking into THC_PID.comp, I'm a little stumped by this line:
float min_velocity = requested_vel -(requested_vel*(1/velocity_tol));

where velocity_tol is a number between 0.1 and 1 (as per JT's shared config).
<spinbox>
                <width>"5"</width>
                <halpin>"vel-tol"</halpin>
                <min_>.01</min_>
                <max_>1</max_>
                <resolution>0.01</resolution>
                <initval>0.2</initval>
                <format>"1.2f"</format>
                <font>("Arial",10)</font>
            </spinbox>
Taking requested_vel as say, 1.0, min_velocity works out to be [1.0 - (1.0*(1/0.2))] = -4. This forces min_velocity to always be less than current_vel, and
if(current_vel > 0 && current_vel >= min_velocity){vel_status = 1;}
        else {vel_status = 0;}
always having vel_status = 1. Therefore, correction is always performed even while going into corners. Am I incorrect here?

I looked at the original THC.comp and the above line was a little different:
float min_velocity = requested_vel -(requested_vel*(velocity_tol*0.01));

I am guessing that the velocity_tol in this one assumed a number between 1 and 100?

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

Time to create page: 0.103 seconds
Powered by Kunena Forum