Torch Jumping Up And Down!!!

[Biker]

Having a problem with my cutting. The start pierce is fine then the torch start cutting but the torch starts jumping up and down. It wont loose arc but makes a bad cut. Some times it will settle down in a few seconds sometimes it wont. Also it will be going along cutting smooth and then all starts again. The machine has been doing very good. Cutting good clean plate. Is doing this with every tool sets I use now. 1/4 plate 65 Amp tip 3/4 plate 85 Amp tip. What is going on here.

[SeanP]

Mine isn't the ethercut, but I have had that happen on thicker material, I upped the span gap to 8/4 in cut profiles or up the thc rate to 30% from default 25%.

[Biker]

Sean,

Can you tell me where to find the arc rate. This problem is happen on 3/8 1/2 and 3/4 thick material. And this has just started.

[jimcolt]

moisture in the air system can cause the sewing machine issue. Another cause is that the THC is set too sensitive for thicker materials, you can slow down the Z axis speed and or decrease the sensitivity.....not sure how the sensitivity is decreased but I'm sure there is a setting for it.


Jim Colt Hypertherm

[Biker]

Thanks Jim

[SeanP]

Sean,

Can you tell me where to find the arc rate. This problem is happen on 3/8 1/2 and 3/4 thick material. And this has just started.

It's on the settings page in Mach/ thc rate.
The span gap is in the cut profiles button

[Biker]

I will look at this setting

[SeanP]

Sorry I got those number wrong,,,I have to keep checking these things myself!
Just looked at the pdf file and this is what it says right at the end

If you see oscillationin the Zaxis, and you are running with the default settings for THC RATE-(30% for stepperto 40% for a servo)and the SPAN VOLTS in Cut Profile is set to 4/4 (1 volt), then first lower the THC Rate by 5% and see if it stabilizes.If not try 10%. If still unstable change thespan volts to 8/4 (2 volts).Span volts is set per Cut Profile so you can change it on different types of cutting

Tuning The Plasma Program Parameters.pdf

(27.76 KiB) Downloaded 297 times

Tuning The Plasma Program Parameters.pdf

(27.76 KiB) Downloaded 297 times

[tcaudle]

none of that apples to the EtherCut. It does not use the internal MACH3 THC logic or the settings in THC RATE. It has its own inverse proportional gain and tracking parameters. (PID loop) It in fact, has its own separate velocity and acceleration when running in DTHC mode.

Oscillation on a THC shows instability in the loop.

First run cuts with the DTHC in MANUAL mode and OFF on flat material. Get a baseline voltage reading. If should be pretty steady. if it is bouncing around a lot on a straight cut it could be water in the air as Jim says or bad consumables.or even a poor Workclamp connection

You can "detune" the DTHC PID by:

running the Response Profile on SLOW and Stepper. (in MACH Cut Profile/DTHCIV Settings tab)
lowering the velocity on the DTHCIV motor tuning in the HUB ADMIN / DTHCIV Settings

Finally you can widen the Span voltage (aka Dead Zone) if you open the Resposne profile and select Custom and set the Voltage Tracking Tolerance to 2

I suspect that something else is causing the voltage to be unstable and the DTHC is reacting to it . The change in settings just make it less sensitive and maskshe underlying problem. especially if the settings ar the same as when it was working

[SeanP]

Sorry I just knew I should have waited for the boss to come along
I didn't realise the new system was so different, well explained!

[Biker]

Guys thanks,

Changed the two different things. Changed the Velocity Tolerance to 2 like Tom C suggested. Also changed the Velocity acceleration from 20 to 25 as someone else suggested. With these changes it is running with no jumping.

[Deezl Smoke]

I know this is an old thread, but the symptoms are just what I am experiencing right now with my Hypertherm 105, Linux Candcnc. I have a refrigerated air dryer. Cut several hundred feet today of 3/8 A36 mild. Ground the mill scale off at the ground clamp. Clean new plate. Saw the Z jumping, but not always, so changed out the electrode as it showed quite a bit of wear. Started the first part fine, then on random intervals, it jumps, as does the voltage, then quiets down, then jumps.
Is there a change in Candcnc that I can try?

[TJS]

There is a post on this on the CandCNC forum. Slow down the DTHC Response settings on thicker material. It will stop the bouncing.

[Deezl Smoke]

Thank you. I will go look for that thread. I need to find out how to do that.

[SeanP]

I used to have my thc response rate around 20% , now I'm at 6-8% maybe that's with having the Z going as fast as I can, not sure but it works damn well.
I've been running it nearly every day and all day quite often - candcnc bladerunner /parallel port

[Deezl Smoke]

This morning I found the setting menu for the DTHC. The responsiveness is on slow. I think my g-code over rides these settings when on automatic? I really dont want to start making changes until this job is complete as the parts are fine, just the Z axis will wear out etc.
Now, as I was stopping to relocate X and Y on the plate to resume cutting, (the customer supplied plate had some missing from an earlier job) I removed my ear protection while the torch was still in it's air purge, post cut. The air was pulsing. So I came back here to search pulsing air and found another thread on that topic. No codes at the HT head. Compressor settings good. Regulator setting good. So I turned off the air dryer for a cut and no difference, still pulsing at random times. Because it also pulses post cutting, and no codes show up, no faults in the messages either, I think I will try to put a new regulator on it.

[TJS]

Deezl,
I really do not use the DTHC on thicker material. over .250" I disable it in sheetcam (checkbox). It should not hop up and down. Do you have a pressure gauge at your plasma unit. If so what is it while cutting.

[Deezl Smoke]

The only gauge is on the regulator. I watched it while cutting and it does pulsate when the torch is jumping. It is set at 125psi. The HT internal does the reduction as I understand things? But the gauge on the compressor does no pulsing.
I pretty much have to use THC as the plate I end up using the most is usually customer supplied and may have some warpage from the previous job I would have done for them. Also, my table slats get a bit of build up on them and the plate wont set true to the gantry anymore.
For example, I just a 19"x123" block of .375 A36. On the Y return the plate had already begun to warp upward, and without torch height control it would have hit the plate. Right? Or is there another setting I dont know about?

[Deezl Smoke]

I have my new regulator, but before I have the opportunity to install it, I had to cut some .75" mild. I chose 65amp. The air flow is higher as is the pressure. The torch still made a few short hops, but for the most part it was stable and normal. Also no post cut pulsing. So I still think it is the regulator and will update when I get it swapped out.

[TJS]

Put a guage at the plasma unit inlet.

[Deezl Smoke]

March 25 2023, I installed the new dryer, filter and regulator. It worked great at first cutting 1/4 milt with a 65 amp nozzle, then started the pulsing again. No codes. Gauge at regulator pulses with the torch.
So now I'm not sure where to go.

[Deezl Smoke]

March 30 2023. Further cutting required a replacement electrode. For the most part, the pulsing seems to be in check now. Though I still can not figure out why, on occasion, the pulsing occurs post cut after torch has raised for rapid. It still will pulse from time to time during a cut, and it does seem to be worse after the system has cut some and "warmed up".

[adbuch]

I know this is an old thread, but the symptoms are just what I am experiencing right now with my Hypertherm 105, Linux Candcnc. I have a refrigerated air dryer. Cut several hundred feet today of 3/8 A36 mild. Ground the mill scale off at the ground clamp. Clean new plate. Saw the Z jumping, but not always, so changed out the electrode as it showed quite a bit of wear. Started the first part fine, then on random intervals, it jumps, as does the voltage, then quiets down, then jumps.
Is there a change in Candcnc that I can try?

Make sure you don't have your work lead cable ("ground cable") coiled up under the table, as this will act like a large inductor and could cause the "sewing machine" effect where the torch appears to bounce up and down as it is cutting. I had this happen to me early on with my Hypertherm Powermax 85. The solution is to shorten the work lead - I shortened mine from around 25 ft. down to about 3 ft. - just long enough to reach the table.

David

[Deezl Smoke]

I know this is an old thread, but the symptoms are just what I am experiencing right now with my Hypertherm 105, Linux Candcnc. I have a refrigerated air dryer. Cut several hundred feet today of 3/8 A36 mild. Ground the mill scale off at the ground clamp. Clean new plate. Saw the Z jumping, but not always, so changed out the electrode as it showed quite a bit of wear. Started the first part fine, then on random intervals, it jumps, as does the voltage, then quiets down, then jumps.
Is there a change in Candcnc that I can try?

Make sure you don't have your work lead cable ("ground cable") coiled up under the table, as this will act like a large inductor and could cause the "sewing machine" effect where the torch appears to bounce up and down as it is cutting. I had this happen to me early on with my Hypertherm Powermax 85. The solution is to shorten the work lead - I shortened mine from around 25 ft. down to about 3 ft. - just long enough to reach the table.

David

Oh holy crap!!! You know what? I did not think about this, but the issue did in fact start after I moved the table. I moved the table to a new location within the shop for easier loading. Once I did that, I was able to go back to the 25' torch lead that came with the plasma. In it's old location, I had to buy a 50' lead. All that required was replacing the tool set from Hypertherm, but the original 25' torch lead was new and in box yet. Anyway, I drove a new rf ground rod 10' into the soil and hooked it up. When I moved the table, I also made all new cross slats. But the ground cable is unmanaged, and on a completely unrelated note, I just watched a video about sending magnets with a AA battery through coiled copper wire.
I will check on this today and see what I can find. Thank you for pointing this out. It would have been the last thing I would ever thought about.

[adbuch]

Jim Colt put me onto this 6 years ago. He called it the "sewing machine" effect.
David