[bonesbr549]

Ok I posted a thread back i Aug about after getting my 48x96 pro up and going (vcon) that I was getting ghost lines. Most replied that it looked like a tramming issue. I worked on it and got it close and had to get a sad project out the door so lived with it.

http://www.cnczone.com/forums/cnc-ro...-mach-3-a.html

I figured i had ordered the linear rail upgrade I would wait and see if that fixed somethings. I completed my upgrade and using the edge tramming tool, I got the spindle dead on left to right and .001 front-to-back. I ran my first play piece which was a simple set of letters and a pocket path boarder (vcarvepro 8.5) and got similar results to the old system almost identical even with all the change out. The bit is an Amana 1/4 x 1/16 ball nose bit.

Attachment 336754
Attachment 336756

Being frustrated I posted the thread yesterday asking if the .001 out could possibly still be the issue. Then comes the interesting part. I watched the video about fusion and changing the tolerance to .001. I thought can't hurt lets try. So I changed the tolerance from .0004 to .001. I recalculated and just redid the pocket path and the results improved dramatically?

Attachment 336758

Is this normal? Feedback would be appreciated. Any other settings or am I off base and this is a coincidence?

[ger21]

As I said in your other thread, tramming can't cause that. It's either lost steps, or a mechanical issue.

So I changed the tolerance from .0004 to .001. I recalculated and just redid the pocket path and the results improved dramatically?

Not an equal comparison. My belief is that it's losing steps plunging or retracting into the letters. When there are no letters, you don't have the lost step issue.

Try reducing the Z axis acceleration by 50% and run the same part.

And make sure that the Z axis is not binding anywhere along it's travel.

[bonesbr549]

As I said in your other thread, tramming can't cause that. It's either lost steps, or a mechanical issue.



Not an equal comparison. My belief is that it's losing steps plunging or retracting into the letters. When there are no letters, you don't have the lost step issue.

Try reducing the Z axis acceleration by 50% and run the same part.

And make sure that the Z axis is not binding anywhere along it's travel.

Thanks i'll look into that but not knowing the acceleration location it will take some research. . As to the z axis binding, IT's a completely diff axis with the upgrade.

[ger21]

Config > Motor tuning.

[CRP]

Thanks i'll look into that but not knowing the acceleration location it will take some research. . As to the z axis binding, IT's a completely diff axis with the upgrade.

In parallel with Gerry's suggestion - Sounds like Cory gave you some things to try. Let us know how you get on!

-Nathan

[bonesbr549]

In parallel with Gerry's suggestion - Sounds like Cory gave you some things to try. Let us know how you get on!

-Nathan

Yes, I have a lot to try, and Cory had some ideas so I'll let all know how it goes.

[bonesbr549]

Ok so I reran the original merged toolpath last night with letters and pocket and the lines are back. So at least apples to apples now and the .001 change did not affect. So going down the possible mechanical lost steps path. So going to reach out to Cory again. Another question is what about the quality of the PC pushing the signals to the drivers. Its an old Laptop I had with a celeron processor and virus software and other apps installed. wondering if a better PC without virus software or any thing else loaded might be playing a piece here. Also could it be the motor? I do have a spare motor for a 4th axis I've not hooked up yet could swap that out as well.



Anyway will keep plugging. Thanks Ger21 and all for the assistance.

[beone]

Are the lines equal depth? Up or down? Random? Do the dishes all occur on the same side of the path? It still looks like a tram issue to me. Oh, is the field being cut diagonal or straight? Could the z head be tilting when pushed or pulled by the lead?
Dave

[bonesbr549]

Are the lines equal depth? Up or down? Random? Do the dishes all occur on the same side of the path? It still looks like a tram issue to me. Oh, is the field being cut diagonal or straight? Could the z head be tilting when pushed or pulled by the lead?
Dave

Tramming pretty dead on. Here's a pic - of the path

[DaveCNCnewb]

Ok so I reran the original merged toolpath last night with letters and pocket and the lines are back. So at least apples to apples now and the .001 change did not affect. So going down the possible mechanical lost steps path. So going to reach out to Cory again. Another question is what about the quality of the PC pushing the signals to the drivers. Its an old Laptop I had with a celeron processor and virus software and other apps installed. wondering if a better PC without virus software or any thing else loaded might be playing a piece here. Also could it be the motor? I do have a spare motor for a 4th axis I've not hooked up yet could swap that out as well.



Anyway will keep plugging. Thanks Ger21 and all for the assistance.

My original laptop (old Pentium 2gb) died. It was having issues losing connection with the smoothstepper. My new laptop (i5 8gb) had similar problems until I disabled wireless and was only running Mach3. So I suspect the old laptop would have worked if I had taken similar steps.

Sent from my SM-G930V using Tapatalk

[ger21]

Are you using a motion controller? or a parallel port?

Did you change the acceleration yet?

[bonesbr549]

Unfortunately I got a day job so gonna try it, how much should I start with as a cut 25% 50%?

I'm using Ethernet smooth stepper

[ger21]

50%. What's it currently set at?

Yeah, those day jobs suck....

[ger21]

Another thing to try.
Are you running this in the same place on the machine every time? I'd try running it in a different spot, and even rotating the g-code 90° and 180° to see if it's repeatable in all cases.

[bonesbr549]

Well problem is finally solved!

First thanks to all the replies. Gerry, I tried the reduction and it did and unfortunately it did not resolve but thanks for all the suggestions. When I first posted the thread I had also sent a note to The folks at CRP and Cory reached out and worked with me as well.

Cory started me out by having me open the controller (off) and check that the z axis controller was seated properly and the ess card was seated good and the wiring was not loose. I did that and reran a test. No change

Next was removing the dust cover and placing a small piece of tape on the z axis coupler and whatch it and mach to see if mabee there is a flutter. I noticed that the lines appeared and the z axis srew did not move during each pass and it would appear on the very first pass.

Next was removing the DC since mine is a big unit that is for all my shop, to see if some noise was causing an issue. Nope no good

I then wondered if the problem was related to the Laptop I had for mach3 as it was an old celeron processor and had lots of crap on it to begin with. So I took another latop i had that had an i5 and wiped it cleat and put a new copy of win7 and nothing else not even virus software to make it as clean as possible.

I also called Amana and talked to their Tech support and described the problem and as the suggested IPM rate was 35-45, he suggested I cut my feed way down to the max and keep going down. (i had been increasing it). He also suggested I lower my stepover down to 3% or half of what it was.

I reran the test, and at 40ipm the lines were there and I manually lowered it on mach to 30 and still there, and finally to 25 and they disappeared. I let it finish to make sure it was gone and it was, Ran another one and still gone.

It was a slow process as I have a day job so it was one test an evening was tough. I have to say I'm getting good at installing and setting up Mach & ESS now. So at least I'm ready to start some real things now.

Again thanks to Cory they were there all the way.

[ger21]

You didn't solve the problem. You should have no issues at 200ipm.
I think you have something loose.

[nlancaster]

I run Vcarve toolpaths on my CNCRouterParts machine at 160-200IPM depending on DOC.

[LeeWay]

I agree. The slower speed eliminated the source of the issue in the material. It didn't solve the issue. That would take 8 times longer than normal to run a part and when it is 3D, that can be a really long time.

[bonesbr549]

Great feedback. After reading the last responses I went back to G-wizzard which I use and as a novice rely on.

Even thought the feedback from the bit manufacturer indicates it should run 35-45 IPM. Gwizzard indicates with the details of that bit (.25 shank, 5.6 taper, and .0625 tip, and 4 flutes) at the most rough setting (fastest) it could do 400ipm with a plunge rate of 100. The most conservative with highest quality 143ipm and a plunge of 3 (rouneded).

So, I did the test again and started at 200, and here is the results. So, now not sure what to think. Very interesting.