Seeking cut quality advice

[Owen]

Hi,

New user here. I have a Precision Plasma 4x8 Patriot table, CandCNC dragoncut bladerunner, hypertherm p65 run through mach3. Code is generated by sheetcam. I have had quite a few issues setting this table up, but the majority of them seem to be sorted out. I had a small air filter attached to my compressor, and was able to cut out this washer (1/4" thick, 1/2" diameter hole) with no issues other than bevel.

This was also cut with the ratio in my PWM module set incorrectly, so the torch height was quite high (voltage was off so I was basically just changing it through trial and error). I have since then fixed this. My z switch offset was also incorrect, so immediately upon fixing the ratio, the torch shield would be touching the material before the DTHC kicked in. This is now also fixed.

I then developed some slop in my table, so I put loctite on everything and made sure it was as tight as could be. There is still some slop on the z axis assembly that I can't seem to get rid of. It rocks in the y direction very slightly. Either way, I believe that slop was there since the beginning. I have attempted to tighten it multiple times with no luck. I have also added a black iron water trap at the compressor which removes quite a bit of moisture. Tried the blow gun on a mirror trick, and saw no water droplets.

I cut the same washer out the other day (with new electrode and nozzle) , and got very strange (but consistent) results. The first portion of the cut is serrated (lack of a better term), then it smooths out but develops a bevel halfway down the cut. The inside hole has no issues other than bevel and lead in lead out. The exterior cut also has an obvious bevel issue.






The only other change I have made is an update to my sheetcam software to support DCC. I have not dialed this in at all, and am not sure if that change could cause this. When the serration occurs the torch makes a pulsing noise.

If anyone could give me any insight I would greatly appreciate it. All cuts were made at 65 amps with a 65amp tip, due to the lack of lower amperage consumables (they have been ordered). I know that isn't the issue though.

[BTA Plasma]

If you don't have belt reduction you will not get a smoothness on any small arc. If you do have it try a 50% reduction is peed for the hole, arc lead in of .23, arc lead out of .18 and a overcut of .12. You can also overcut the outside.

[Owen]

Thanks for the quick reply BTA Plasma,

I do have belt reduction on my table. I am confused because I have gotten a very smooth cut, and no longer do. I have the typical small holes (60% feedrate under 1-1.5" diameter depending on material and other parameters). I was using an arc lead in of .25 (so it is in the center) but did not use a lead out or overcut. I will definitely try this today once I run a few cuts.

I am mainly looking for advice on how to eliminate the serrations at the beginning of the cut and how to address the odd bevel. I do appreciate the suggested values though!

[Owen]

Didn't realize that the odd bevel could be caused from the torch being uneven. I guess I thought the entire edge would be on an angle. I will attempt to address this by leveling and tightening the z axis once more. I might try dremelling out the bracket that holeds the z axis, so I can tighten it more to eliminate slop. Will report back after.

[Owen]

Okay, got a chance to get out to the shop and do some more cuts. I leveled my torch head which has gotten rid of the odd bevel. So that's taken care of. I did a few more test cuts with fine cut consumables (need to pick up more regular consumables) and am still having the jagged cut issue. When it occurs I can hear the plasma become inconsistent, as if it is pulsing.

Things that I have changed that could generate this issue:

I added a black iron water trap that may be reducing my flow significantly. Could lack of consistent air be causing the strange cuts on the washers posted above? It is doing a great job of removing water, but if its causing air flow problems I can easily find another solution to drying my air. When I first started using the water trap, I did have my hypertherm throw the "insufficient air pressure" code, so I bumped my compressor up. This was the first hint that the water trap is causing a loss.

I updated sheetcam through CandCNC's website. This gave me a great toolset but I've noticed that there is now an option for "DTHCIV response profile" where as I have DTHCII. I have been inputting a 0 here.

The table itself seems to be working great, there is no slop in the machine anymore, just a small amount of backlash between the rack and pinion gear on the Y and A. It didn't seem to effect cutting before, so I don't think this is the issue. I'm going to take down the water trap tonight and see if this changes anything.

[whiskeymike]

Consistent air flow was my first thought. Those serrations are pretty funky. Looking forward to seeing what you figure out.

[Owen]

Whiskeymike- I was hoping it would be the solution but unfortunately after removing my water trap, the table is still doing the same thing.

I did however notice that the z axis is vibrating up and down quite quickly when the serrations occur. It's very hard to see but I attempted to capture it on a video (sorry, forgot to flip to landscape. Portrait video fail).
https://www.youtube.com/watch?v=ZaZwyEcTmNc
This is cutting out a random test dxf I made that has all sorts of shapes, edges and curves. 16 gauge steel, fine cut @ 45 amps. Didn't change anything, right out of the toolset that CandCNC provided in their sheetcam update.

I realized I needed to go into my post and change the values from 1 to 0 for ethercut and DTHCIV, seeing as I don't have either of those. I'm going to uninstall sheetcam now, and reinstall without CandCNC's updates for DCC. When my machine was working I didn't have this update installed, although I would like to utilize it.

This is more of an entry log than a typical thread on here, so apologies for the thread bumping. I'm hoping I'm missing something stupid that will get pointed out.

[BTA Plasma]

IF its a DTHCII change your THC rate to 18

[illusivedf]

the THC is probably going up and down because its just getting a smooth cut, looks like some kind of backlash issue. I had a similar problem and it was because of some backlash/ slop in the gantry on my table. It had one x axis motor with a drive shaft going to the other side and a belt going down to the other gear rack. I added a second motor and it got rid of that problem. It also looks like it is right where the torch starts cutting the part maybe some kind of acceleration issue? My THC used to go up and down a lot too before I got rid of backlash/ slop issue. It stays pretty smooth now.

[Owen]

IF its a DTHCII change your THC rate to 18

I will try doing this, although I need to educate myself more thoroughly on what the THC settings do.

the THC is probably going up and down because its just getting a smooth cut, looks like some kind of backlash issue. I had a similar problem and it was because of some backlash/ slop in the gantry on my table. It had one x axis motor with a drive shaft going to the other side and a belt going down to the other gear rack. I added a second motor and it got rid of that problem. It also looks like it is right where the torch starts cutting the part maybe some kind of acceleration issue? My THC used to go up and down a lot too before I got rid of backlash/ slop issue. It stays pretty smooth now.

I thought it was slop originally too, so I tightened absolutely everything. I have a small amount of backlash between the rack and pinion on both my Y and A motors (A is slave on Y axis). Not sure how to get rid of this seeing as the rack tooth spacing is larger than the pinion gear spacing. Do you have absolutely no backlash at all? Maybe what I'm thinking is a small amount is actually a very large amount.

[acourtjester]

If you have a miss match between the spur gear and the rack that is not good. they are designed to interlock exactly for smooth transfer of power. It seems there are metric and standard gear pitches but most small gears in the US are 14.5 and 20 pressure angle.

gear.jpg (28.25 KiB) Viewed 4244 times

gear.jpg (28.25 KiB) Viewed 4244 times

[jimcolt]

Put your hand on the torch while it is cutting. Any vibrations will create striations (serrations) in the cut edge. If the x axis is oscilating rapidly...it will create striations. Shut the height control off and do a cut at a fixed height (.06" off the material)....again put your hand on the torch...you should feel less vibration and the cut will be smoother. The x axis usually needs to be detuned....either through drive gain settings or simply a slower speed. If there is a lot of backlash in any axis....striations will be the result.

Jim Colt Hypertherm

[Owen]

Thanks for the great replies. I have a lot of things to try today in the shop, will update the thread with results. Hopefully I can get the machine working as it was today.

[tcaudle]

Whiskeymike- I was hoping it would be the solution but unfortunately after removing my water trap, the table is still doing the same thing.

I did however notice that the z axis is vibrating up and down quite quickly when the serrations occur. It's very hard to see but I attempted to capture it on a video (sorry, forgot to flip to landscape. Portrait video fail).
https://www.youtube.com/watch?v=ZaZwyEcTmNc
This is cutting out a random test dxf I made that has all sorts of shapes, edges and curves. 16 gauge steel, fine cut @ 45 amps. Didn't change anything, right out of the toolset that CandCNC provided in their sheetcam update.

I realized I needed to go into my post and change the values from 1 to 0 for ethercut and DTHCIV, seeing as I don't have either of those. I'm going to uninstall sheetcam now, and reinstall without CandCNC's updates for DCC. When my machine was working I didn't have this update installed, although I would like to utilize it.

You will loose all of the automation features and have to use the manual settings for cut parameters from the stored settings in MACH

This is more of an entry log than a typical thread on here, so apologies for the thread bumping. I'm hoping I'm missing something stupid that will get pointed out.

You are confusing DCC with full TAP (hypertherm RS485 current control) and the DTHCIV sedttings versus the DTHCII. The DTHCII can and does use most of the automation settings like the Prest Volts, DTHC delay, etc. What is cannot use is the DTHCIV Response Profile because the DTHCII resposne is set by MACH . Having the parameter in there won't hurt anything but if it annoys you simply set the POST Option dthcTune = 0- and it will go away
You set any response numbers in MACH. THC RAte is basiclly the Gain with a higher number being more sensative. The Span Volts setting in the Cut Profile is a % that you set the "deadzone" (where it neither goes up or down) and acts as a dampening number. The lower the voltage you are cutrting at (like with finecut) you may find you need to adjust the Span Volts up slighlty and reduce the THC Rate by a few percentage points. If the Torch is oscillationg up and down rapidly (motor is acutally moving the Z) then lower the THC Rate in 5% increments. Since it only does that in cutting one side its very close and some small tweaks may be all that is needed.

No need to reload MCAH or SheetCAM.

[illusivedf]

My Dynatorch uses air cylinders to mesh the gear into the rack,I have heard some machines with springs have issues sometimes I'm not sure whats on yours?

[Owen]

IF its a DTHCII change your THC rate to 18

The THC rate was set at 40%, which is higher than I remember it ever being. Thank you for pointing this out, it is back up and cutting smoothly.

You are confusing DCC with full TAP (hypertherm RS485 current control) and the DTHCIV sedttings versus the DTHCII. The DTHCII can and does use most of the automation settings like the Prest Volts, DTHC delay, etc. What is cannot use is the DTHCIV Response Profile because the DTHCII resposne is set by MACH . Having the parameter in there won't hurt anything but if it annoys you simply set the POST Option dthcTune = 0- and it will go away
You set any response numbers in MACH. THC RAte is basiclly the Gain with a higher number being more sensative. The Span Volts setting in the Cut Profile is a % that you set the "deadzone" (where it neither goes up or down) and acts as a dampening number. The lower the voltage you are cutrting at (like with finecut) you may find you need to adjust the Span Volts up slighlty and reduce the THC Rate by a few percentage points. If the Torch is oscillationg up and down rapidly (motor is acutally moving the Z) then lower the THC Rate in 5% increments. Since it only does that in cutting one side its very close and some small tweaks may be all that is needed.

No need to reload MCAH or SheetCAM.

Fantastic advice, thank you. I am new to this, so I was basically going through the list of things that I changed since the table was last working correctly. I should have been more clear in which settings I changed in my post, as what I changed was setting dthcTune = 0, same with ethercut. I definitely do not want to lose all automation. Appreciate the needed lesson in THC parameters, now I can fine tune this thing!

[Owen]

Since I am stuck with the fine cut consumables for the next bit, I decided to try cutting some 18 gauge aluminum and 16 gauge stainless steel. Here is a test shape I made (looks like a strange whale) in aluminum cut at 30amps, 120ipm, volts @ 76.



This is a fairly small test piece (1/4" ratchet behind it for size). For a first time cutting aluminum I'm pretty happy with the results. The issues with the cut seem unrelated to the material, as I am getting the same divots/wandering lines with steel. In order to get sharper edges, I turned off CV on angles larger than 45 degrees. I also selected "CV distance tolerance" and input 20 units. This is probably what is causing those divots, as the machine is slowing down while the torch is remaining at the same amps. I need to read CandCNC's DCC with DTHCII manual as I have a feeling there is a dynamic way to control this, instead of using cv dist. tolerance.

My Dynatorch uses air cylinders to mesh the gear into the rack,I have heard some machines with springs have issues sometimes I'm not sure whats on yours?

My table just has brackets that hold the belt reduction components/motor and gear onto the rack. I was unaware that some tables use air cylinders/springs to hold the gears onto the rack.

[davek0974]

I've built spring tensioners into my table, it does need something with a little 'give' at that point I think.

I also turned off everything in Mach to do with CV stuff (except CV of course) and then use CV off at angles >89 deg. seems be helping on my table.

Those cuts look great BTW

[Owen]

Still no new consumables! I did however cut out a fairly long g-code to see how the table would hold up using fine cuts. Ran into a small issue with the metal warping slightly. During the touch and go the torch would press the metal down onto the slats, just to have it come back up when the torch raises. This made the torch come into contact with the metal during the delay on DTHC. Makes me feel like I'm really missing out on DTHCIV!

Anyways, normally I would just use something to hold the metal down during the cut, but seeing as this cut took around 5 minutes - it would be nice not to have to do that. The cuts still came out alright, but I had an issue happen twice where my Hypertherm threw the 0-21 fault code "Gas flow lost while cutting". Through a little research I found this thread which explained why this was happening.

http://www.pirate4x4.com/forum/torchmat ... issue.html

So, the contact between the material an issue. My consumables are still in good shape but have drag marks on them that could potentially lead to a gouge that would cut the lifespan a little short. I am using the shield that Jim Colt suggests to use at the end of the thread (to avoid contact between the nozzle and material) but still have contact marks on my nozzle.

Pictures of the parts! Friends are getting married and got a house together, figured they could use some art 14 gauge aluminum.


I did the same piece (but 6" smaller in diameter) in 16 gauge steel and it also turned out well. Steel is definitely easier to work with as there is basically no dross, and the edge is very smooth after a cut. With the aluminum the edge isn't smooth, its quite textured. Dross is fairly easy to remove though which is a plus!

I've built spring tensioners into my table, it does need something with a little 'give' at that point I think.

I also turned off everything in Mach to do with CV stuff (except CV of course) and then use CV off at angles >89 deg. seems be helping on my table.

Those cuts look great BTW

I'll have to look into retrofitting springs. Not sure if I need it quite yet but it does interest me. Glad turning CV stuff off isn't uncommon, it made a huge difference for me!

[davek0974]

Nice stuff there
Re the torch pressing the work down, I have improved mine a lot by fitting a spring acting against the torch so it reduces the effective weight of the whole head. Mine was pressing down with about 1.4kg of force before - enough to deflect thinner materials badly, now its down to about 500g and things are working much better.

It needs just enough weight to re-seat the head after the touch-off.

[Largemouthlou]

Tom (candcnc) also sells a feather touch unit that works very well.. Finally have all the new to me electronics figured out and feather touch installed and cutting well.

[Owen]

Tom (candcnc) also sells a feather touch unit that works very well.. Finally have all the new to me electronics figured out and feather touch installed and cutting well.

I am contemplating purchasing this kit now, would definitely solve all my issues with the touch and go. Thanks for mentioning that! The sensor attached to a clip on the retaining cap right? Going through the hypthertherm consumables catalog sparked my interest in that kit a few weeks ago.

Nice stuff there
Re the torch pressing the work down, I have improved mine a lot by fitting a spring acting against the torch so it reduces the effective weight of the whole head. Mine was pressing down with about 1.4kg of force before - enough to deflect thinner materials badly, now its down to about 500g and things are working much better.

It needs just enough weight to re-seat the head after the touch-off.

Thanks! That's also a fantastic suggestion, great alternative to the feather touch. I would like to stay away from springs for now. I don't want the z axis to be loaded at all during a cut. I'll probably purchase the feather touch once I have some money rolling in from this table.

[Owen]

Hey guys,

So today I tried cutting out some nameplates for a local company. Using Finecut 45amp consumables again and 14 gauge aluminum. Having some issues with getting nice cuts, needing a lot of clean up. Settings are: 24amps, 90ipm, 0.07 cut height, 0.1" lead in and lead out, 0.12 over cut. Using path rules small shapes, dropping the cut speed down to 60%. Did this a few times varying the speed/amps but usually get the same result. Forgot to set DTHC properly for the outer cut which is why there are some scratches from the torch shield scraping on the alum.

Any tips on getting smoother lines? I don't mind doing clean up with dross but I can't present these like this.
Before cleanup:



After a quick cleanup:

[davek0974]

Well, that wiggle on the tops of the letters/numbers is mechanical - either slop in an axis, loose gear, of wobbly z-axis. I had exactly the same issue on my old table which was the main reason i built a new one, i'm not suggesting you build a new one just that its exactly the same issue.

90ipm is 2200mm/min and my fault would appear at anything over 1500mm/min, i can still get it sometimes on the new machine but only over 4000mm/min. I would double check all mechanical features. To rule out THC do a cut with it turned off.

Aluminium does cut rough but that appears very rough, is the nozzle in perfect condition?

This is the back of 3mm/10g aluminium using standard consumables on a PM45, details as shown - there must be an issue somewhere as I would hope yours to be similar.

[Owen]

Well, that wiggle on the tops of the letters/numbers is mechanical - either slop in an axis, loose gear, of wobbly z-axis.

90ipm is 2200mm/min and my fault would appear at anything over 1500mm/min, i can still get it sometimes on the new machine but only over 4000mm/min. I would double check all mechanical features. To rule out THC do a cut with it turned off.

See, you would think that from the pictures - but there basically no slop in my table. Here is how the same piece (same size) looks when cut in 18 gauge stainless. I went crazy earlier somewhere earlier in this thread and took my table apart 4 times and made sure everything was as good as it could get.



There are a few imperfections in that piece, but those are due to not tuning the lead ins enough / not having enough space to have a proper lead in.

Ended up fine tuning acceleration a little bit for aluminum and also changed my CV units in mach3 down to 2. Seems to help. Got the aluminum to look half decent, but still need to go through with a dremel if I want it to be presentable.



In all fairness, these pieces are pretty small.... Should I be chasing after better cuts?