Hypersensing questions
28 Jun 2020 14:02 #172991
by Aciera
Replied by Aciera on topic Hypersensing questions
Ha! I'm glad somebody gets something out of that. When I brought my last car in for the mandatory checkup the guy made a comment about my car being an older model. It was four (4) years old.
Sorry, totally off topic.
Sorry, totally off topic.
Please Log in or Create an account to join the conversation.
26 Jul 2020 02:06 - 26 Jul 2020 02:06 #175839
by rodw
Replied by rodw on topic Hypersensing questions
Attached is a revised component.ohmic3.comp.. You will need to edit your halfile from ohmic2 to ohmic3.
I had some issues where this was a bit forgetful and forgot to turn the probe off when retracting so this does a bit more checking. Seems OK now after cutting a full sheet.
I had some issues where this was a bit forgetful and forgot to turn the probe off when retracting so this does a bit more checking. Seems OK now after cutting a full sheet.
Last edit: 26 Jul 2020 02:06 by rodw.
The following user(s) said Thank You: snowgoer540
Please Log in or Create an account to join the conversation.
- snowgoer540
- Topic Author
- Offline
- Moderator
Less
More
- Posts: 2386
- Thank you received: 779
28 Jul 2020 23:31 #176302
by snowgoer540
Installed this tonight. Will report back if anything goes amiss.
Replied by snowgoer540 on topic Hypersensing questions
Attached is a revised component.ohmic3.comp.. You will need to edit your halfile from ohmic2 to ohmic3.
I had some issues where this was a bit forgetful and forgot to turn the probe off when retracting so this does a bit more checking. Seems OK now after cutting a full sheet.
Installed this tonight. Will report back if anything goes amiss.
Please Log in or Create an account to join the conversation.
29 Jul 2020 08:39 #176341
by rodw
Replied by rodw on topic Hypersensing questions
I am in terror now the official tester has shifted his focus to my code!
I could not really get back into the headspace where I was when I wrote ohmic2 so I simplified the conditions abit. One thing I did do was turn the probe off if the voltage fell below the low limit which Ohmic2 was ignoring. I might still need to find a better pin to turn the relay on with.
I could not really get back into the headspace where I was when I wrote ohmic2 so I simplified the conditions abit. One thing I did do was turn the probe off if the voltage fell below the low limit which Ohmic2 was ignoring. I might still need to find a better pin to turn the relay on with.
Please Log in or Create an account to join the conversation.
- snowgoer540
- Topic Author
- Offline
- Moderator
Less
More
- Posts: 2386
- Thank you received: 779
29 Jul 2020 10:39 #176347
by snowgoer540
I find if I don't take copious notes between high-thought-provoking tasks like that, I require too much time to get back to where I was mentally.
I was going to compare the codes to see what might have changed, but honestly I didn't understand what was going on in the ohmic2 comp lol.
Replied by snowgoer540 on topic Hypersensing questions
I am in terror now the official tester has shifted his focus to my code!
I could not really get back into the headspace where I was when I wrote ohmic2 so I simplified the conditions abit. One thing I did do was turn the probe off if the voltage fell below the low limit which Ohmic2 was ignoring. I might still need to find a better pin to turn the relay on with.
I find if I don't take copious notes between high-thought-provoking tasks like that, I require too much time to get back to where I was mentally.
I was going to compare the codes to see what might have changed, but honestly I didn't understand what was going on in the ohmic2 comp lol.
The following user(s) said Thank You: rodw
Please Log in or Create an account to join the conversation.
29 Jul 2020 11:18 #176353
by rodw
Replied by rodw on topic Hypersensing questions
Yeh, some days the brain does not want to do the hard stuff. It probably won't ever make sense to you becasue a prerequisite is to have spent the many hours studying halscope plots like I have
First off there is a moving average calculator. Its high level advanced code. so just ignore it and accept that it calculates the average of the last num_readings eg (5,10,100 or whatever you set the pin at)
So I know from said hours of study that the THCAD is sensitive enough to see the arc voltage ramp up from 0 volts to 24 volts over say 0.04mm of movement as the torch contacts the plate.
When the torch is wet, there is very little change in the arc voltage as it breaks contact from the material but we want to build some hysteresis into the probe on and off signal to prevent false triggering. This means we can't really use a separate lower threshold and be confident the voltage will fall below it.
So what I did is use the moving average and just guessed that with say a 22 volt high voltage threshold, the first time its crossed, it might be say 23 or 24 volts, not 22.01 volts. At that point, the average volts should be quite high becasue the buffered readings that are averaged would be increasing. So once we start moving away from the plate, it should take a few servo cycles before the average falls below the high threshold. So what we've done by using the average is build in some hysteresis between the peak voltage reading and the high threshold.
Ohmic3 takes this one step further and will also turn the probe off if the voltage falls below the low threshold as a failsafe backstop. Without this backstop (say on a dry table), I had a couple of instances where the torch probed backwards forever without the probe turning off. So now its reliable in wet or dry conditions ( I hope!)
So that is as clear as I can make my twisted thought process behind this component. Its really quite cool with a 50c resistor and this code we can do something that is not really possible. That is to implement reliable ohmic probing on a water table (touch wood!!!)
First off there is a moving average calculator. Its high level advanced code. so just ignore it and accept that it calculates the average of the last num_readings eg (5,10,100 or whatever you set the pin at)
So I know from said hours of study that the THCAD is sensitive enough to see the arc voltage ramp up from 0 volts to 24 volts over say 0.04mm of movement as the torch contacts the plate.
When the torch is wet, there is very little change in the arc voltage as it breaks contact from the material but we want to build some hysteresis into the probe on and off signal to prevent false triggering. This means we can't really use a separate lower threshold and be confident the voltage will fall below it.
So what I did is use the moving average and just guessed that with say a 22 volt high voltage threshold, the first time its crossed, it might be say 23 or 24 volts, not 22.01 volts. At that point, the average volts should be quite high becasue the buffered readings that are averaged would be increasing. So once we start moving away from the plate, it should take a few servo cycles before the average falls below the high threshold. So what we've done by using the average is build in some hysteresis between the peak voltage reading and the high threshold.
Ohmic3 takes this one step further and will also turn the probe off if the voltage falls below the low threshold as a failsafe backstop. Without this backstop (say on a dry table), I had a couple of instances where the torch probed backwards forever without the probe turning off. So now its reliable in wet or dry conditions ( I hope!)
So that is as clear as I can make my twisted thought process behind this component. Its really quite cool with a 50c resistor and this code we can do something that is not really possible. That is to implement reliable ohmic probing on a water table (touch wood!!!)
The following user(s) said Thank You: snowgoer540
Please Log in or Create an account to join the conversation.
29 Jul 2020 11:22 #176354
by rodw
Replied by rodw on topic Hypersensing questions
So I should mention to the Chief test pilot that I don't think the number of readings averaged should be too big becasue we really want to only average the voltage for that period the voltage is increasing as it makes contact.
Please Log in or Create an account to join the conversation.
29 Jul 2020 11:30 - 29 Jul 2020 11:32 #176357
by rodw
Replied by rodw on topic Hypersensing questions
A load of crap! deleted sorry.
Last edit: 29 Jul 2020 11:32 by rodw.
Please Log in or Create an account to join the conversation.
- snowgoer540
- Topic Author
- Offline
- Moderator
Less
More
- Posts: 2386
- Thank you received: 779
29 Jul 2020 11:36 #176361
by snowgoer540
Replied by snowgoer540 on topic Hypersensing questions
Thank you for the explanation, I think I mostly understand what you're trying to do. So far my experience has been that it works. However if it decides to not work, I'll be sure to let you know
I did notice sometimes it pauses while probing. I think that is when the material was rusty and it got to the float switch instead. I'll have to do some testing around that, see if it's really an issue or a perceived issue.
I did notice sometimes it pauses while probing. I think that is when the material was rusty and it got to the float switch instead. I'll have to do some testing around that, see if it's really an issue or a perceived issue.
Please Log in or Create an account to join the conversation.
29 Jul 2020 11:44 #176364
by rodw
Replied by rodw on topic Hypersensing questions
I think the pause happens when the probe has to back away for longer before it breaks contact. I observed this sometimes pre hypersensing. This can happen if the material is warped or near an edge hanging over a slat so the material bends down before the probe triggers. You also see this if you probe into a puddle of water as the contact breaks once the torch clears the meniscus resulting in being far too high!
My feeling was it is quicker probing with hypersensing.
I have not upgraded to Phil's fast probing. I suspect that might make this more prevalent by smashing into the material.
My feeling was it is quicker probing with hypersensing.
I have not upgraded to Phil's fast probing. I suspect that might make this more prevalent by smashing into the material.
Please Log in or Create an account to join the conversation.
Moderators: snowgoer540
Time to create page: 0.125 seconds