Deviation observed on 3+2 axis CNC milling setup.

[j.wilson]

Hi Tom,

We have a 3+2 axis CNC milling setup. It uses a XYZAC configuration where X and Y moves the table, Z is the spindle height and A axis rotates along X axis and C is the table bed and rotates along Z axis. I am noticing a strange phenomenon.

When are milling a cube with a bore in center of the top face.
Below is the process followed:

1. Top face machined at A0 C0
2. Bore on the top face at A0 C0
3. Side faces are machined in A-90 condition and C0, C90, C180, C270 respectively.

When we measure the pitch of the bore with respect to the side faces we notice an error. The face that is cut in A-90 C0, the bore pitch with respect to that face is correct. However the face that was cut in A-90 C180, the pitch is off be 0.1mm.

What do you think might be the reason for this? Could you help point me in a direction to begin diagnosing this? Your help is appreciated.

[TomKerekes]

Hi Wilson,

What do you mean by “pitch”? Pitch is usually a angle or slope not a distance. How big is the cube?

What is the situation with the other two faces?

Do you understand what backlash is and how to test for it?

[j.wilson]

Hi Tom!

Thank you for replying

Regarding your questions,

I've attached an image with this reply, I hope it gives you a bit more insight.

https://postimg.cc/7GVCgCgY

I've marked A and B as distances (24mm) that should be same and C and D (24mm) that should be same. Currently they are not.

A is 23.9 and B is 24.1
C is 23.9 and D is 24.1

The sun of both interestingly seems to add up to the required measurement 44mm though.

Regarding backlash: backlash has been disabled.

P.s. I apologise for my delayed response, I hadn't seen your reply earlier, does this forum have an option to send an email notification if a thread I've posted on is replied to by any chance? If not that is perfectly fine, I was just curious.

[TomKerekes]

Hi Wilson,

I've attached an image with this reply, I hope it gives you a bit more insight.

Its still not clear what faces are machined in what order. Please put more effort into your explanations.

I would describe that as an offset not a pitch. The bore is offset from the faces. If you were to make the bore at X=0.1 and Y=0.1 instead of x=0 y=0 is the final result correct?

In that case I think that would indicate the X and Y Origin is not at the point of rotation for the A and C Axes. See here as I attempt to show how a bore Y offset becomes a Z offset after an A rotation. The top diagram attempts to show no Y offset that results in no Z offset for machining the face. The bottom diagram shows an exaggerated Y offset that results in a Z offset for machining the face.

The sun of both interestingly seems to add up to the required measurement 44mm though.

I assume you meant to say "sum" and "48mm"

Regarding backlash: backlash has been disabled.

I was referring to backlash in the hardware.

I apologise for my delayed response, I hadn't seen your reply earlier, does this forum have an option to send an email notification if a thread I've posted on is replied to by any chance? If not that is perfectly fine, I was just curious.

Yes you will normally be notified/emailed. It seems you are changing user names. Anyway there is no rush from our end. You can respond whenever you wish.

[j.wilson]

Hi Tom,

I apologise, yes it was 48mm and not 44 mm, in the rush to respond quickly I made a few errors in my response. You are correct about this being an offset issue and not pitch too, I got the terms mixed up.

I shared your response with my team as well and they are trying to diagnose the problem given your insight, I'll get back to you as soon as I head back from them too.

While they do, there was another bit of guidance I needed.

Currently the configuration is an open loop, we use kmotion with Konnect to help with I/o expansion, I would like to move it to closed loop.

My understanding is that I would need to hook up the kflop along with the kanalog. Is it possible to have 6 axis in a closed loop as well as konnect to handle all the i/o needs? If yes, how would I go about this?

P.s. I actually just signed up to this forum recently, I didn't try changing my user name, or atleast I want aware of it.

Also I found the "notify me when a reply is posted" setting
I do greatly appreciate your help. But since you are giving your time, the least I could do is keep you updated with promp responses.

[TomKerekes]

Hi Wilson,

Currently the configuration is an open loop, we use kmotion with Konnect to help with I/o expansion, I would like to move it to closed loop.

What kind of drives do you have? Can they accept analog Velocity commands? Do you have encoders? Do the Drives have encoder outputs?

[j.wilson]

Good day Tom,

What kind of drives do you have?

- We use LEADSHINE ELPD750Z drives.

Can they accept analog Velocity commands?

- Yes

Do you have encoders?

- Yes

Do the Drives have encoder outputs?

- Yes

[TomKerekes]

Hi Wilson,

I think you were already operating closed loop at the Drive but not back to the Controller (KFLOP).

To operate closed loop back to the controller I would add Kanalog, switch drives to Analog Velocity mode, connect the DACs to the drives and the Encoder outputs to Kanalog's differential encoder inputs.

[j.wilson]

Hi Tom,

I'm sure you are wondering what is the progress on the feedback that you have given.

Here is an update, the team is still working on checking and fixing axis the X and Y Origin issue.

Meanwhile, I have tried following your suggestion on how to operate closed look back to the controller however I'm facing a few issues.

1. The drive that I am experimenting on has two analog input pins one for +(positive) and one for -(negative) as well as a ground.
I notice that there are 8 outputs ranging from 0 to 7, and 8 ground terminals as well. Connecting the ground is straight forward but how would I go about wiring the positive and negative this to the JP11 DAC outputs?

2. Regarding the wiring for the encoder feedback, with the information i was able to gather while researching on the Kanalog board, it seems that there is a provision to receive only 4 sets (A channel, B channel and Z channel) of input. Ie. inputs from 4 drives.
"4 encoder's A/B signals are connected to JP1 and any Z index inputs are connected to JP2"

Is it possible to add a secondary Kanalog board or a daughter board or a slave of some sort to increase the capabilities to support 8 drives?
Or is there any other solution you could suggest?

Also how would I go about setting this up in the Kmotion software? Do you have any resources I could refer to?

Thanks for taking the time to go through my post, I appreciate your support.

P.s. On an entirely different note, I have another query. Does Dynomotion currently support a feature similar to AICC, HPCC or Smoothening?

[TomKerekes]

Hi Wilson,

1. The drive that I am experimenting on has two analog input pins one for +(positive) and one for -(negative) as well as a ground.
I notice that there are 8 outputs ranging from 0 to 7, and 8 ground terminals as well. Connecting the ground is straight forward but how would I go about wiring the positive and negative this to the JP11 DAC outputs?

Connect the + to the DAC output and the - to GND

2. Regarding the wiring for the encoder feedback, with the information i was able to gather while researching on the Kanalog board, it seems that there is a provision to receive only 4 sets (A channel, B channel and Z channel) of input. Ie. inputs from 4 drives.
"4 encoder's A/B signals are connected to JP1 and any Z index inputs are connected to JP2"

Well no. Kanalog JP1 and JP2 supports 8 channels of A/B encoder inputs. Any unused A or B inputs can be used for Z inputs. Z inputs aren't required unless you need them. Z inputs are handled in software and be connected to any input. See also here.

Is it possible to add a secondary Kanalog board or a daughter board or a slave of some sort to increase the capabilities to support 8 drives?
Or is there any other solution you could suggest?

No. See above.