Spurious engagement of Plasmac's Anti-dive

rodw
Offline
Platinum Member

02 Jul 2019 06:59 #138380 by rodw

Replied by rodw on topic Spurious engagement of Plasmac's Anti-dive

thefabricator03 wrote:
So what is the main problem you guys are facing in regards to getting it to work?

I am gearing up to do a bunch of cutting as my machine has been down for 4 months with my CandCNC dramas and building the new controller and I have been outsourcing all of my cutting,

I could try a few things and get some data if needed, I just need to know what you guys are looking for.

Its really only the start of the journey. You can't work on this until you have a working table and you have to use plots in halscope to see what is going on. An alternative is to use the sampler component and analyising data in a spreadsheet but its time consuming to get the data set up for analysis. So we are really only able to look at it now and Phill needs to finish his table.

So step 1 is to reliably sense a kerf cross (where we are now) and step 2 is to re-enable the THC once the danger is over.

When you can cut again, let me know and if we have not got something working, happy to enlist your support.

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

PCW
Away
Moderator

02 Jul 2019 13:57 #138395 by PCW

Replied by PCW on topic Spurious engagement of Plasmac's Anti-dive

rodw wrote:
I've added a white line somewhere I think the dv/dt threshold should be. But as I said, we need more data!

And we have not even thought about re-enabling. I kinda think that should be when the arc voltage falls back to being near the desired cutting volts.

Shouldn't the threshold be above the DV/DT baseline, so it detects the leading edge of the kerf? (sudden voltage rise)

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

tommylight
Away
Moderator

02 Jul 2019 16:58 #138414 by tommylight

Replied by tommylight on topic Spurious engagement of Plasmac's Anti-dive

I have to agree with PCW.
For any situation in cutting, the voltage will rise first followed by THC trying to correct for that rise.
I think we have discussed this before, but it should lock the THC whenever the DV over DT goes high, not after the cross and while the voltage is settling.

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

PCW
Away
Moderator

02 Jul 2019 17:22 - 02 Jul 2019 17:28 #138421 by PCW

Replied by PCW on topic Spurious engagement of Plasmac's Anti-dive

Right, also if the THC is locked, the subsequent negative peak at the kerf trailing edge should be almost the same magnitude as the positive peak (its most likely larger because of the unwanted THC generated dive into the kerf)

I was thinking that this negative peak could be used to re-enable the THC (with appropriate gating/timeouts etc etc)

Last edit: 02 Jul 2019 17:28 by PCW.

The following user(s) said Thank You: tommylight

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

rodw
Offline
Platinum Member

02 Jul 2019 19:53 #138430 by rodw

Replied by rodw on topic Spurious engagement of Plasmac's Anti-dive

PCW wrote:
Shouldn't the threshold be above the DV/DT baseline, so it detects the leading edge of the kerf? (sudden voltage rise)

It is!
Look closely and Zoom right in. It is above the base line but it can't be too far away
The leading edge of the kerf is about 3 divisions in from the left edge. Thats the event we wish to capture.
The divisions are 50 ms apart.
Look at the arc voltage plot below dv/dt. It rises 25 volts in 100 ms. We have to catch it in a few volts, not after 25.

The dv/dt is currently calculated as a deviation from the moving average which for this plot is based on 10 readings. Maybe it should be averaged 100 readings to make the average less twitchy. Currently the component allows this to be changed by a pin up to a maximum of 1000 readings.

The output pin to disable thc is called void-hold and currently, the code to turn the THC back on is as follows:

          if(movingdown && void_hold && (torch_volts < first_dvdt_volts))
                void_hold = 0;                  // arc volts has fallen below where trigger point

So if the arc volts falls below the voltage at which it was enabled, THC is resumed.

Anyway, keep the ideas flowing

For mkardasi I'm a bit worried about the variation in the signal. I wonder if you should repeat the test with 100 readings and 1000 readings for the moving average. In theory, that should smooth out the average more.

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

PCW
Away
Moderator

02 Jul 2019 20:38 #138434 by PCW

Replied by PCW on topic Spurious engagement of Plasmac's Anti-dive

Ahh you are moving much slower than I expected.

So the slow initilal rise is possibly the arc pulling away from the leading edge of the kerf, and the very fast rise I though was the kerf is perhaps losing the arc altogether in the middle of the kerf

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

rodw
Offline
Platinum Member

02 Jul 2019 20:53 - 02 Jul 2019 20:54 #138436 by rodw

Replied by rodw on topic Spurious engagement of Plasmac's Anti-dive

PCW wrote:
Ahh you are moving much slower than I expected.

So the slow initilal rise is possibly the arc pulling away from the leading edge of the kerf, and the very fast rise I though was the kerf is perhaps losing the arc altogether in the middle of the kerf

Yes, I think that is the case. If you consider the generally accepted rule that the voltages changes 10 volts per mm and I measured 7.53 volts/mm, then 25 volts represents between 2.5mm-3.3mm of height change the controller sees in 100 ms. That is much more than the recommended cutting heights of most plasma machines so it won't be pretty when it makes the height adjustment!

A I said earlier the DV/DT graph is hard to understand because its a rate of change so flat lines don't mean nothing is changing!

Last edit: 02 Jul 2019 20:54 by rodw.

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

rodw
Offline
Platinum Member

02 Jul 2019 21:16 #138439 by rodw

Replied by rodw on topic Spurious engagement of Plasmac's Anti-dive

Actually when I think about it, if we can accurately pick the edge of the material with this algorithm and save its position, we could calculate the length of the arc using Pythagoras therom. eg. the Hypotenuse of a triangle defined by the distance travelled and the height of the torch tip.

This would be totally useless from the controllers perspective but it might be helpful in increasing our understanding.

Thanks for the idea Peter!

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

tommylight
Away
Moderator

02 Jul 2019 22:55 #138450 by tommylight

Replied by tommylight on topic Spurious engagement of Plasmac's Anti-dive

Since I do not have a machine to test yet ( put your jokes here , thank you ), can any of you do some test cuts of 10 to 20 cm on different thickness sheets and plot the THC voltage only, with THC on and THC off.
I wonder how much does voltage change usually with no averaging, just plain 1000 reading per second.
Thinking that if there are no fluctuations bigger than 5 to 6V with the THC on, setting the THC to lock if the voltage changes by over 8 or 10V should do the trick nicely. Setting that inside of 2 readings might not work properly, but inside of 5 to 10 reading might work.
Averaging 1000 samples is not an option as that is 1 second and when cutting at 6m/minute that would happen after 100mm of cut. At 100 samples that is still 1cm and at 10 samples 1mm .
Are we reading that chart from left to right or am I missing something?
Oh well, had a mentally tough day so brain is fried properly

. Went to see a house today, same size as this one, placed very badly on the plot as there is no way of getting in the back yard except walking, has a parking garage under but none of our cars can go in since the slope is to big and both our cars have a long wheel base, could be used as a workshop by the ceiling is very low at 30cm over my head so my younger brother would most probably be scraping his head on it as he is 205 cm tall, and it is about 1 km nearer to city centre.
There' s more but again I am flying off topic.

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

mkardasi
Offline
Senior Member

02 Jul 2019 23:34 - 02 Jul 2019 23:36 #138460 by mkardasi

Replied by mkardasi on topic Spurious engagement of Plasmac's Anti-dive

Gentlemen, I'd like to split off discussion of dv/dt into this thread. My scatter brain is having a hard time keeping up with two discussions in one thread. Phil posted how I can increase the adjustment range of his implemented algorithm so I would like to get this thread back on that track and continue the dv/dt discussion in the thread linked above.

I have replied to your comments in the new thread. Thank you for your cooperation

Last edit: 02 Jul 2019 23:36 by mkardasi.

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

Moderators: snowgoer540

Time to create page: 0.117 seconds