Overcut support on CandCNC Ethercut

[Diesel]

Awesome - thanks for chiming in, Jim!

Cheers,

Matt

[beefy]

So at this point in time can anyone confirm if this issue has been fixed with the Ethercut system. I went to the thread I started on the Candcnc forum and no mention there of a fix http://www.candcnc.net/supportforum/vie ... ?f=2&t=746

The last words about this matter were:

admin13o5 » Wed Jul 08, 2015 10:12 am
This is a problem that is not going to have an easy answer. We got the M10M11 to work but it still comes with a somewhat variable delay that makes the exact spot it turns off vary as well. We have not released that fix as we do not consider it reliable.

[whiskeymike]

So at this point in time can anyone confirm if this issue has been fixed with the Ethercut system. I went to the thread I started on the Candcnc forum and no mention there of a fix http://www.candcnc.net/supportforum/vie ... ?f=2&t=746

The last words about this matter were:

admin13o5 » Wed Jul 08, 2015 10:12 am
This is a problem that is not going to have an easy answer. We got the M10M11 to work but it still comes with a somewhat variable delay that makes the exact spot it turns off vary as well. We have not released that fix as we do not consider it reliable.

Beefy, I had posted the question on the CandCNC forum and Tom replied that it was in the latest firmware update with some information on how to use it. It was above my pay grade, so I haven't implemented it. I'm hoping someone else tries it out and get's it working. You might want to call CandCNC and get more details if you are interested.

Here's the link - http://www.candcnc.net/supportforum/vie ... ?f=8&t=968

On a different note, I added a lead out and made it .3 in/out and the divot is gone on the few cuts I made this weekend. There's a tiny mark where the lead out crossed, but it wasn't a divot like the others. So I'm going to continue working with that before I update firmware.

[beefy]

Beefy, I had posted the question on the CandCNC forum and Tom replied that it was in the latest firmware update with some information on how to use it. It was above my pay grade, so I haven't implemented it. I'm hoping someone else tries it out and get's it working. You might want to call CandCNC and get more details if you are interested.

Here's the link - http://www.candcnc.net/supportforum/vie ... ?f=8&t=968

On a different note, I added a lead out and made it .3 in/out and the divot is gone on the few cuts I made this weekend. There's a tiny mark where the lead out crossed, but it wasn't a divot like the others. So I'm going to continue working with that before I update firmware.

Cheers Mike,

checked out that thread. It appears it's still a bit experimental (the word "TRY" was emphasised). I'm just keeping tabs on the system for when I might be ready to buy. The Linux system tempts me but crikey I've found Linux a bitch to learn after being so used to Windoze.

Might PM some of the lads that already have the ethercut system and see if they've gave it a go.

Keith.

[exapprentice]

Hi everyone
I have now found some time to play with the “CandcncPlasmaRev11M1” post which allows the use of the M10P1 command (quick torch off).
I have had some success with being able to stop the divot at the end of an open line and it seems at the moment repeatable?

I have been testing the M10P1 command on 0.9mm / 20 gauge steel, cutting a 95mm (3 ¾”) open lines, just thought the thinner materials show the divots up the best, I am using 30amp consumables and the 30amp setting on my PM45.

Using the previous 11D post to make sure I was using the best settings from a known starting point, then loading the “CandcncPlasmaRev11M1 post” and using those settings. After some trial and error I have found a rule that uses M10P1 command 2mm / 0.080” before end works well.
Initial setting:Feed @ 8100mm/min, Pierce height @ 1.6mm, Cut height @ 1.5mm, Pierce delay @ 0, Plunge rate @ 5000mm/min, Pause @ end 0, Preset volts 105 (more on volts later???)

My first cuts were made with the DTHC off and all seemed good, I did start to get some problems though when I turned the DTHC on, for some reason (I don’t understand) the torch would raise when the DTHC kicked in which made me think the voltage was off (I did check the torch height at the beginning @ 105 volts and it was steady @ 1.5mm gap) so for no other reason because I don’t know any better, I ended up reducing the volts to 88 and got some of the best cuts I have had using the 0.9mm material???(can someone explain that to me)

I have noticed a definite delay at the pierce with this setup, maybe it’s the addition of more code in the sequence?? Namely the “M11P1” after M03.
In conclusion so far I can say that the M10P1 (quick torch off) does appear to work at stopping the divots at the end of open lines with the tests so far and it seems repeatable

I am now stumped on how to stop the divot at the pierce point, I have tried a pierce height of 1.5mm and a cut height of 1.5mm but that gave me issues with not triggering the S20 DTHC on command, I think a pierce height of 1.6mm works as there is a plunge distance (yes small but apparently enough)

What next, Different material thickness, and hopefully someone can tell me the best way of stopping the divot at the pierce, ideally I think the noticeable slight delay at the pierce before the machine moves is the culprit and needs eliminating but how??
Feel free to comment and or add to this
Best regards to all
exapprentice

ps if some one can tell me how to add pictures to posts I have some of of the cuts for all to see!

[Diesel]

Very good news on your success in eliminating the divots at the lead-out with the new firmware/post processor!

Here are some step-by-step instructions on adding photos.
Note: Keep the file size small by cropping or whatever means necessary.

Click 'Attachments':

Attachments_1.jpg (24.42 KiB) Viewed 3987 times

Attachments_1.jpg (24.42 KiB) Viewed 3987 times

Click 'Add files':

Add files using Windows:

Attachments_3.jpg (51.27 KiB) Viewed 3987 times

Attachments_3.jpg (51.27 KiB) Viewed 3987 times

If the files are below the size limit, you'll see the status turn green:

Clicking "Place inline" embeds the photos into the post where your cursor is in the text box.
This allows you to annotate the photos instead of just piling them in at the end of the post.

Cheers,

Matt

[islander261]

Hello

I have a request that anyone who has this working post the g-code for your test piece. The reason is I don't have a CandCNC controller or use their post and want to see how this shows up in the files. Then I can edit it and test with my controller (won't work as planned if docs are correct).

John

[Diesel]

I have noticed a definite delay at the pierce with this setup, maybe it’s the addition of more code in the sequence?? Namely the “M11P1” after M03.
In conclusion so far I can say that the M10P1 (quick torch off) does appear to work at stopping the divots at the end of open lines with the tests so far and it seems repeatable

Food for thought:
We know for sure we have delays in the system causing issues with cut quality due to torch on/off timing.

I wonder if we have delays in the system with regard to the signal that gets sent from the power source (Hypertherm) when the arc is OK/established? The motion electronics are patiently waiting for this before it starts motion...and any delay would make the pierce divot bigger.

Cheers,

Matt

[beefy]

Food for thought:
We know for sure we have delays in the system causing issues with cut quality due to torch on/off timing.

I wonder if we have delays in the system with regard to the signal that gets sent from the power source (Hypertherm) when the arc is OK/established? The motion electronics are patiently waiting for this before it starts motion...and any delay would make the pierce divot bigger.

Cheers,

Matt

Good thinking Matt.

I've sort of had some direct experience with this in the real world. I designed a circuit board for marking hole centres with nothing more than the shortest blip of the torch. It effectively gives me a tiny round crater in the top of the metal. Mach3 just could not give me a short enough torch on duration using the known workaround methods. Another issue was post cut airflow being on for 10 seconds after each torch "blip". So the circuit I designed sends the torch on signal and WAITS for the Arc OK to come. Once it gets the Arc OK it either terminates the Torch On signal immediately or after a user settable time delay (milliseconds). It's all controlled with a little computer chip so any time delays in that department are negligable.

Anyway this electronic circuit is at the mercy of the plasma cutters reaction time to the Torch On signal being removed, i.e. how long from the time the Torch On signal is removed to the flame actually extinguishing. All I can say is my Hypertherm 1250 seems to have a very good reaction time and shuts off very quick, giving me the little "centre punch" marks I want https://www.youtube.com/watch?v=yg-s-EAqMfI

Now, regarding divots at the end of a cut, and using the fast M10 command. Even if a plasma cutters reaction time is a bit slow, as long as that delay is CONSISTENT and is always the same then you can allow for that in the software, i.e. switch the torch off a greater distance before the end of the travel path. BUT if that delay varies then you are pretty much screwed and won't get repeatable results.

Keith.

[Diesel]

Hi Keith:

I've read about your electronic circuit here on the forum for plate marking - that's a nice workaround for sure!
Have you had any issues with consumable life by cutting the post flow?

I agree - if things are consistent, we can just turn the torch off earlier/later in the cut to get the plasma arc to extinguish on time to avoid divots.
Makes perfect sense on paper.

A few folks are experimenting with this, but I have to admit that I'm not yet comfortable hacking around with post processors...so I'm waiting patiently for a while.

Cheers,

Matt

[beefy]

Hi Matt,

the postcut airflow is to cool the torch but with such a short duration "crack" repeated every few seconds the tip (sorry Jim, I should say nozzle ) the nozzle won't even get warm. You're talking only a few milliseconds for the torch on duration.

Jim (Colt) told me that what wears electrodes the most is the initial torch firing so I modified the circuit again and it now uses current sensing instead of waiting for the ARC OK signal.

I have no idea at what current level the ARC OK signal is given, PLUS I'd bet it varies across machines / manufacturers. If your main cut current is set at 105 amps, you certainly don't need anywhere near that to give a "centre punch" mark. At torch switch on, the plasma current ramps up to the set cut current. For hole marking I have my plasma current detection set at about 15 amps then the circuit starts timing from when that 15 amp level is reached. So it doesn't matter what my main CUT current level is at, 25 amps or 105 amps, the hole marking is done at low current regardless. I can now get "centre punch" marks that are so small I'm trying to find them LOL.

I'm HOPING that using the low current hole marking method will dramatically reduce the electrode wear compared to using the old ARC OK sensing. And as for electrode nozzle heating I think you'd have to do a lot of holes in rapid succession to even get things moderately warm. The rapid succession hole marking in the video was more to show how post cut airflow doesn't get in the road, but I can't see too many times when you'd want to mark holes THAT fast.

There's also another benefit for the nozzle life itself. Pierces are the most damaging to nozzles. If you could look in slow motion at a pierce, it starts with the sparks flying out sideways. As the crater gets deeper, the sparks start doing a U-turn and tend more to fly back upwards towards the nozzle. With a low marking current and a very short torch on duration the crater doesn't get deep enough to make the sparks do a U-turn. I actually mark at about 1mm above the surface now to increase the accuracy of the "centre punch" mark placement.

In the Youtube video I got over 700 holes marked from one electrode so I was pretty happy. Let's think, the video was cutting 16 slats with 24 holes in each one so 384 hole marks. Then I had the corresponding slat toppers after that so 768 "centre punch" marks from one electrode. The downside was the drilling and deburring of those hole, I didn't enjoy that

Keith.

[exapprentice]

Hi everyone
hopefully below are the pictures I mentioned in an earlier post?
with help from Matt, thank you

DSC00219.JPG (148.68 KiB) Viewed 3935 times

DSC00219.JPG (148.68 KiB) Viewed 3935 times

DSC00222.JPG (134.53 KiB) Viewed 3935 times

DSC00222.JPG (134.53 KiB) Viewed 3935 times

The pictures show a short open line cut the end with the "DIVOT" is the pierce point???
the other end without the divot was using the M10P1 (quick torch off) and so far so good it does seem to be repeatable
I have cut some parts using the CandcncPlasmaRev11M1 post and the M10P1 code and all went well

Just need to find a way of reducing / stopping the "DIVOT" at the pierce point, any ideas, how do you reduce/stop the start "DIVOT"

Best Reagards to All
exapprentice

[BTA Plasma]

Try the ramp function

[exapprentice]

Hi BTA Plasma

Thanks for the input
I thought the ramp option was the same as a lead in in sheetcam, dont you need a lead in for the ramp option to ramp along eg a lead in of 6mm 1/4" will let the lead in taper over that length? How do you specify the details of the ramp option on a open line at the start

I am probably totally wrong, still learning on that very steep curve though

best regards to all
exapprentice

[whiskeymike]

Hi everyone

Just need to find a way of reducing / stopping the "DIVOT" at the pierce point, any ideas, how do you reduce/stop the start "DIVOT"

exapprentice

Exa, how long is your pierce delay? It may be set too long and it's lingering after the pierce occurs.

[exapprentice]

Hi Whiskey mike

Thanks for the suggestion
I already have the pierce delay set at 0, but I do agree with you it is lingering in the same spot after the pierce
I don't think we are talking even a single second but obviously enough to create the divot.
I am sure that this delay would'nt be such a problem on thicker materials, but as above I thought I would do the testing on thin (0.9mm/20 gauge)
As the thinner the material gets the more the divots show up.

I am more than happy with my results of using the M10P1 command and the latest CandCNC post at the end of an open line using the 0.9mm material
its just trying to squeeze those last few milli-seconds between the M03 and the machine actually moving (I think)
If only you could have a - 0.1 set in the pierce delay, it would be solved
lets hope someone out there has the answer and done it already

Best regards to all

Pete

[exapprentice]

Hi everyone

Just thought I would confirm the Kerf width of the cuts in my pictures above is approx 0.7mm / 0.027", just to get some sense of scale.

Best regards to all

exapprentice

[tcaudle]

Try the ramp function

Ramping causes the DTHC ON trigger point to not happen and causes other issues. If you have the varaible current option on your Hypertherm (RS485) you can define a "soft pierce" that lowers the starting current but it jumps up to full current as soon as you reach cut height. There is also an option for a "Peck Pierce" that goes do to a predefined height and short cycles the torch. If you do have the variable current option then you can combine that with a peck pierce to also lower the current in the peck cycle

[tcaudle]

Hi Whiskey mike

Thanks for the suggestion
I already have the pierce delay set at 0, but I do agree with you it is lingering in the same spot after the pierce
I don't think we are talking even a single second but obviously enough to create the divot.
I am sure that this delay would'nt be such a problem on thicker materials, but as above I thought I would do the testing on thin (0.9mm/20 gauge)
As the thinner the material gets the more the divots show up.

What you are asking for is instant motion because on thin material it only takes a fraction of a second for the divot to form beacuse the pierce is virtually instant. Because the system waits for ARC OK there are a couple of options:

1. The easiest would be to jumper the ARC OK inputs at the PWM module and just leave it on all of the time. Only downside is if for any reason the torch fails to fire or loses arc , motion will just keep truckin'. That should remove ANY delay from ARC OK.

2. Get the DCP-01 and set the current trip point to 10A BUT even that will have some delay between it sensing and the siganl being sent to the DTHC which may still show up as a divot on the start.

Since the majority of cutting is closed shapes the best method is to use a lead-in and let the devot drop out with the scrap. Lead ins were pretty much invented because of divots at the start.

[motoguy]

Try the ramp function

Ramping causes the DTHC ON trigger point to not happen and causes other issues. If you have the varaible current option on your Hypertherm (RS485) you can define a "soft pierce" that lowers the starting current but it jumps up to full current as soon as you reach cut height. There is also an option for a "Peck Pierce" that goes do to a predefined height and short cycles the torch. If you do have the variable current option then you can combine that with a peck pierce to also lower the current in the peck cycle

My PM85 has the RS485 port. Does your description above require the HyT option, or is that something different?

[exapprentice]

Hi everyone

following on from Toms suggestions above, I did "Jumper" across the "Arc OK" on the CandCNC PWM module and the divot at the pierce is noticeably smaller almost to the point of nearly gone, I have now cut more than a few open lines in the 0.9mm sheet of material and I am pleased so far with the quality of the cuts on open lines

I have now replaced the "jumper with a switch wired across the Arc OK terminals of the PWM, this way I can decide when and on what thickness materials I want to use it on

as I have already mentioned in an earlier post I don't think the divot issue is as pronounced on thicker materials and shouldn't be an issue if you are using lead in and lead outs on closed shapes.

I would recommend this tweak especially on open lines and the thinner materials, thanks again to Tom for chipping in and making his suggestions.
I will carry on testing and post back again

If anyone does get around to trying this out, please add your results to this topic for all to share

Best regards to all and thank you
exapprentice

[tcaudle]

My PM85 has the RS485 port. Does your description above require the HyT option, or is that something different?

No, Just the HYT-Connect RS485 SIM option so you can control Cut Current from the G-Code . Then you need an advanced POST that has the "soft Peck" built in . Its basically a "dril " tool that finds the centers of holes on your drawing and the peck parameters are defined in the tool settings.

[exapprentice]

Hi every one

So a little update, I have been cutting 1.2mm & 0.9mm (18 & 20 swg) CR4 material about 10 sheets of each since my last post.
I am really happy and pleased to say it all seems to stack up.
very minimal divots at both the start and ends, and I mean minimal, after hitting the parts with the wire wheel you can hardly see anything
so all in all a very good learning experience, thank you to all who contributed to this subject and hopefully it may have helped some others

[acourtjester]

Pleased with another tidbit of info from Plamsa Spider I put a small toggle switch across the "arc OK" .
Ct some 16 Ga. today and used no offset and it is very hard to see where the start was.


following on from Toms suggestions above, I did "Jumper" across the "Arc OK" on the CandCNC PWM module and the divot at the pierce is noticeably smaller almost to the point of nearly gone

[exapprentice]

Hi Tom

Glad to get some feed back and it works for other people as well

I use it all the time now on open lines on thinner materials