Hypersensing questions

More
16 Apr 2020 02:22 - 16 Apr 2020 02:25 #164103 by snowgoer540
So I have my hypersensing circuit wired up, and I have a few questions that I did try to self answer via search, but haven't been too successful.

1. For the 390K resistor, I assume 1/4 watt is fine?
2. In the .pdf that is in the hypersensing thread the wiring example shows the voltage divider jumper on F/1, but the .hal and the .comp both show F/32. Which is correct?
3. In one post, it is suggested to point the .ini file to the hal file Rod attached. In another, it's suggested to copy and paste it to the hal_connections file. In yet another, it's suggested to put it in the <config>.hal. Which is correct? (I am guessing the hal_connections).
4. 2.9 which I am running, comes with the ohmic.comp. But I haven't seen anywhere that says I need to modify this file to match my THCAD board's frequency calibrations. Do I need to modify this file, or is the numbers in there just an example of what the .comp file is doing, and it pulls this information from the .hal? Obviously I need to change these in the .hal, I just wasn't sure what else was going on/needed changed.
5. If someone had 2.8, where would they place the ohmic.comp?
6. In the ohmic.comp (again not sure if this is an example that is commented out or I need to change it) the encoder is pointed to the A encoder, and I believe it should be pointed to the B encoder, since most will use A for THCAD. (hm2_7i76e.0.encoder.00.velocity) Plus the PDF says to hook to encoder B.
7. In the .hal file, it uses "hm2_7i76e.0.encoder.02.velocity" which I believe would be the IDX pins, and not Encoder B... wouldn't Encoder B be "hm2_7i76e.0.encoder.01.velocity"?
8. One last point of potential confusion... the .comp data (probably example) and the .hal input THCAD data don't match (different frequencies).

I find that a lot of the stuff on the forums assumes a certain level of Linux/Linuxcnc proficiency, and new users dont necessarily have that, and may be afraid to ask, or easily frustrated to figure it out. So I figured the clearer this all can be, the more everyone benefits.

I think that's all for now :laugh: thanks for the help! Really looking forward to trying this out.
Last edit: 16 Apr 2020 02:25 by snowgoer540.

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

More
16 Apr 2020 04:12 #164110 by Uthayne
Replied by Uthayne on topic Hypersensing questions
I'll do my best to answer your questions as I have just been going through this myself. I'm on mobile, so apologies in advance.

1. Yes, a 1/4w should be fine
2. F/32 is the correct setting
3. I put my settings in its own hypersensing.hal file. Wherever you put it, it needs to be loaded after plasmac is loaded as it unlinks pins. If you place it in its own Hal and put that file at the end of your Hal file section in your ini, it will be fine. I do have my encoder setup in my _connections.hal file. Makes no difference though.
4. Generally, you don't edit comp files. Once they are installed, you have to recompile them and reinstall to push any updates. The way you interact with them is through the Hal files via setp and net commands. The comp does have default values, but they are overwritten with the setp lines. Notice the comp is loaded first, and then setp changes different pin values.
5. If you want to edit or add components, they are found in /src/hal/components/. However, you have to run halcompile -- install <component-name>.comp in order for it to be used and called from your Hal files.
6. The formatting makes it difficult to read, but that is all commented out as example config data. Don't worry about it
7. Correct, encoder.01 is/was the B channel. You'll need to flash your board (different firmware exists for 7i76e vs 7i76) that will turn your B and IDX pins into two more A channels. However you still identify the second encoder spot as encoder.01 if you are wired into what was previously the B channel. Don't forget to enable a second encoder in the top of your machine.hal file!
8. Yes, just an example. Fill the Hal file in with your numbers on the sticker. Use the F/1 values even though the card is set in F/32 mode.

Learning linuxcnc is like learning a second language. It's challenging and has a steep learning curve, but the forum has many helpful people. Don't be afraid to ask! It's how you learn

Good luck with your setup, I've learned quite a bit about this topic just this week
The following user(s) said Thank You: Clive S

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

More
16 Apr 2020 05:59 #164115 by rodw
Replied by rodw on topic Hypersensing questions
In my view, the documentation takes precedence over the forum (which probably discussed this before it was published to master branch. See linuxcnc.org/docs/devel/html/man/man9/ohmic.9.html
I do agree that we are using encoder.00 here but it does not matter which encoders are used for arc voltage and ohmic sensing.

Please note for my config, I blew up encoder.0 quite a while ago and PCW created a new bit file for me that is similar to what is now available for hypersensing. So my choice of encoders works for me but you may have something different.

I don't think ohmic.comp is distributed with 2.8 yet. What you could do is grab the component from master branch
github.com/LinuxCNC/linuxcnc/blob/master...omponents/ohmic.comp

download it to a folder, open a command line window and move to that folder and then do
halcompile --install ohmic.comp

But if you are using master branch, it should be installed already.

If you have not compiled from source, I think you need to preface that command with sudo.

Note that this comp uses my approach to calibrating THCAD's. That is there are pins for every parameter and including the THCAD /f divider so you just need to change those settings to what you choose to use.

When I wrote my hal file for this, I Put it in a totally separate file that is referenced in the .ini file. I used unlinkp so it could be added to a working config without breaking anything plasmac was using.

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

More
16 Apr 2020 14:38 #164144 by snowgoer540
Replied by snowgoer540 on topic Hypersensing questions
Good deal, thank you both, this is very helpful!

I will be doing the rest of this in the next few days, I may have other questions.

Actually, I do have one. Somewhere I read that Plasmac can be configured to fall back on the float switch if the ohmic doesnt trigger for some reason. Is that done by default, or do I need to do something to my config to get this to happen?
The following user(s) said Thank You: rodw

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

More
16 Apr 2020 14:40 #164148 by tommylight
Replied by tommylight on topic Hypersensing questions
You just need to have them both set up and working properly.
Floating switch means floating head with two sets of rails, one very short.

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

More
16 Apr 2020 19:31 #164193 by rodw
Replied by rodw on topic Hypersensing questions

SomewPlasmac here I read that Plasmac can be configured to fall back on the float switch if the ohmic doesnt trigger for some reason. Is that done by default, or do I need to do something to my config to get this to happen?


As Tommy said this is done for you by default. With pending changes to the Plasmac probing algorithm to speed up probing and to reduce the number of probing errors, the float switch may become a requirement so that there is room for the torch to slow to a stop when the ohmic probe is triggered without breaking anything (Or a float switch is required to implement the new probing algorithm).

Please report back when you are up and running so the feedback can be incorporated into the component. Once we get this proven with some feedback, we should be able to get it deployed to V 2.8 as well.
The following user(s) said Thank You: tommylight

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

More
17 Apr 2020 02:36 #164224 by snowgoer540
Replied by snowgoer540 on topic Hypersensing questions

SomewPlasmac here I read that Plasmac can be configured to fall back on the float switch if the ohmic doesnt trigger for some reason. Is that done by default, or do I need to do something to my config to get this to happen?


As Tommy said this is done for you by default. With pending changes to the Plasmac probing algorithm to speed up probing and to reduce the number of probing errors, the float switch may become a requirement so that there is room for the torch to slow to a stop when the ohmic probe is triggered without breaking anything (Or a float switch is required to implement the new probing algorithm).

Please report back when you are up and running so the feedback can be incorporated into the component. Once we get this proven with some feedback, we should be able to get it deployed to V 2.8 as well.


Good to know. My float switch throw is very short, probably .030-.040... hopefully that doesn’t become an issue.

I enjoy testing things and finding bugs so if there is anything specific you want me to focus on, please let me know.

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

More
17 Apr 2020 03:04 #164226 by rodw
Replied by rodw on topic Hypersensing questions
That throw should be fine. The torch needs that movement to decelerate to a stop at the ini file acceleration setting when the ohmic is triggered so you can control that with your probing speed. Faster = more travel.

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

More
17 Apr 2020 03:53 #164231 by phillc54
Replied by phillc54 on topic Hypersensing questions

you can control that with your probing speed. Faster = more travel.

Not quite right, Probe Speed only comes into play after Probe Height is reached. It uses Setup Speed from the config panel to get to Probe Height.

From the yet to be completed docs:
The following formula can be used to calculate your machines approximate overrun.
0.5 × a × (v ÷ a)2
a = acceleration in units/sec2
v = velocity in units/sec
So with Z axis acceleration of 2000mm/s2 and Z axis velocity of 33.33mm/s we would have an overrun of 0.278mm.
This is roughly 78in/s2 and 1.3in/s for an overrun of 0.01"

acceleration is obtained from MAXIMUM_ACCELERATION from the joint associated with your Z axis in the .ini file
velocity is Setup Speed in the Config Panel divided by 60
The following user(s) said Thank You: tommylight

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

More
17 Apr 2020 18:49 - 17 Apr 2020 18:50 #164295 by snowgoer540
Replied by snowgoer540 on topic Hypersensing questions
In addition to changing the "num_encoders=" to 2 in the top of the main hal file, another critical piece of information is that you need the other setp values that are similar to the THC:

setp hm2_5i25.0.encoder.01.scale -1
setp hm2_5i25.0.encoder.01.counter-mode 1

I am honestly not sure if you need the filter or not (setp hm2_5i25.0.encoder.01.filter 1)? I've always been a little fuzzy as to what this did?

I think I read somewhere the filter was ONLY for the 7i96... but I could be wrong. Note that I DO NOT have an Ethernet setup, I am using a 6i25 with a 7i76 with Peter's custom Plasma firmware...
Last edit: 17 Apr 2020 18:50 by snowgoer540.

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

Moderators: phillc54
Time to create page: 0.166 seconds
Powered by Kunena Forum