What's new
What's new

Fadal 4020 Wall Finish Problem

Juz

Plastic
Joined
Dec 8, 2019
Hello All.

Have had some great info come from this forum over the years and I think everyone appreciates the help that is still available here.

One of our 3-axis machines is an old 4020 Fadal which is something that is becoming harder and harder to receive support for.

Anyhow, we have run into a problem that we solved in the past with a Y-axis drive, but we are in quite deep into it this time.

This is the situation. Slowly towards the end of 2021 the machine started producing less-than-stellar finish on Y-axis direction walls (this of course includes radius'd corners and angles, etc). We can produce an evident difference feeding purely in Y vs purely in X. The direction in Y does not matter, inwards and outwards we produce the same finish. The finish looks like chatter but it is not, I would describe it as very fine "striation" lines.

So our trouble shooting led us through the following progression:

- Did a general check over the machine, spindle had next to no runout: .0001" but Y-axis had quite bad play so its where we started.

- Replacement of Y-axis thrust bearings as we had over .0014" in movement on the ball-screw.

- Upon replacement of the Y-axis thrust bearings we found the bearing block and motor mount had a fine fracture in it and was flexing when tightening into the machine. So we replace the entire motor mount / bearing block assembly on both the X-axis and Y-axis.

- We then cleaned the brushes and tachometers on both the X-axis and Y-axis motors (DC motor machine).

- We then swapped the X and Y motors with still no change.

- We tried swapping the controller cards for the X and the Y with no change. We then did the same with the amp cards, and still no change.

- We then replaced the resolvers in both the X and Y axis motors, still no change.

- I personally went thru the entire power supply and verified +/-5V and +/-10V and all was good. Verified voltage on the resolvers were ~1.75V at the bullet connectors and on the pins of the controller cards. I found that the Shielding lead on the cable for Y-axis had real high resistance (this turned out to be a real corroded connection on the wire to shielding in the Y-axis motor housing. I repaired this myself and got the resistance back down to nominal) All of this still proved to have no change.

- Looking thru the electronics we went as far as observing the voltages in the cards and resolvers while the machine is feeding constant rates in both the X and Y axis directions

- At this point we brought our local service guys in who checked the machine out, adjusted the gibs and did a bar-ball test. They said our gibs and our ways are still good and they had it all adjusted back to factory specs. They said the Y-axis was really bad for play still and found a ton of thrust issue with the ball-screw nut. On top of adjusting the gibs they also did a quick tune of the amp cards.

- So now we started tossing parts at the machine. We bought a brand new Y-axis ball-screw assembly, this had zero effect on our problem (although now our Y-axis is whisper quiet haha).

- Then we got a brand new DC servo which we tried on both the Y-axis and X-axis with again zero change.

- Being very stumped, yesterday and today we started inspecting the Head. The belts and tensioners all seem to be fine (this is a Hi/Lo spindle machine). We actually have a brand new tensioner assembly on this machine, with brand new belts from just a few months ago. We played around with the spindle motor tensioner wire and gave everything a good once over but have no leads here in the slightest.

So now we are somewhat stuck. Mechanically the machine seems fine according to us and the service guys. This problem exists with even brand new end mills and this problem is consistent with any tool holder. The strange thing is we still get good finish in steel (Good as in, decent for the old beast that this thing is). Not going to pretend the X-axis is superb finish but it is miles better than Y and we would like to get this machine back to the state where it produces a part with walls that don't look like they were cut with a dull hss mill in an old manual mill haha

Y-axis Problem 1.jpg

X-axis Lines 1.jpg

Y-axis Problem 3.jpg

X-axis Lines 2.jpg
 

Attachments

  • Y-axis Problem 2.jpg
    Y-axis Problem 2.jpg
    86 KB · Views: 43

wrrocket

Plastic
Joined
Jul 14, 2018
Kinda a dumb question, but is the fretting present if you offset the work to different locations on the Y or X-axis?

One possible item would be on the cable harnesses going to the resolvers or the servo motors, considering that is about the only thing from what you said that is still common on the electrical side after all the parts you swapped. If you can, I would do an inspection of the cable harnesses as there might be frayed wires, damaged connectors or cracked insulation that is interfering with the motor feedback. I would however, expect cabling issues to be more intermittent than what you are seeing, but I have seen some weird things before. Some cracked insulation with packed in chips/coolant can potentially do odd things.
 

rklopp

Diamond
Joined
Feb 27, 2001
Location
Redwood City, CA USA
If the striations are on the face parallel to Y, that suggests the table or spindle is moving slightly in X as Y feeds along. How is the X thrust control, encoder, tuning, etc? How are the Z gibs for head side-to-side (X) looseness?
 

ferretlegger

Aluminum
Joined
May 15, 2009
Location
San Jose, CA
rklopp may have it right. When I saw the striations on the wall when feeding y direction I thought that such striations could be caused by the tool moving in the X direction in a periodic manner, OR the wall is moving in X in a periodic manner. Is it possible the x position servo has a low level oscillation? If you feed faster does the "wavelength" of the striations lengthen? if you feed faster does the striation wavelength shorten?

If you know the feed rate (inches per minute) and you can estimate the wavelength of the striations (inches), if you divide the velocity of the Y axis by the "wavelength" of the striations you get a frequency (1/minutes). If you were to look at the X axis encoder signal with high resolution, or possibly the X axis motor drive you may find a tiny ripple frequency. This may be caused by the servo tuning. In digital servos, there is something called a "limit cycle oscillation" which is highly technical but may stir some neurons in readers of the forum or the service techs.

In any case, something that has always helped me in dealing with obscure problems is Sherlock Holmes's dictum “When you have eliminated all which is impossible, then whatever remains, however improbable, must be the truth.”

Best wishes,
Michael
 

ViktorS

Aluminum
Joined
Apr 26, 2021
We have the same problem on a cinnci arrow with glass scales.

Same kind of lines on Y axis feed.
The a2100 control have a built-in "ocsilloscope" to monitor position and servo command in real time with a graph, and it turnes out that the X axis is oscillating, not Y.

I have not yet concluded why that is on our machine, but I would suspect that it is the X axis on your machine too.
 

Juz

Plastic
Joined
Dec 8, 2019
If the striations are on the face parallel to Y, that suggests the table or spindle is moving slightly in X as Y feeds along. How is the X thrust control, encoder, tuning, etc? How are the Z gibs for head side-to-side (X) looseness?

This is a suspicion we have arrived at as well. We are checking and adjust the gibs on the head right now. Right off the hop I could get the head to move the indicator .001" quite easily along a Y face (so moving the X-direction on the head).


rklopp may have it right. When I saw the striations on the wall when feeding y direction I thought that such striations could be caused by the tool moving in the X direction in a periodic manner, OR the wall is moving in X in a periodic manner. Is it possible the x position servo has a low level oscillation? If you feed faster does the "wavelength" of the striations lengthen? if you feed faster does the striation wavelength shorten?

If you know the feed rate (inches per minute) and you can estimate the wavelength of the striations (inches), if you divide the velocity of the Y axis by the "wavelength" of the striations you get a frequency (1/minutes). If you were to look at the X axis encoder signal with high resolution, or possibly the X axis motor drive you may find a tiny ripple frequency. This may be caused by the servo tuning. In digital servos, there is something called a "limit cycle oscillation" which is highly technical but may stir some neurons in readers of the forum or the service techs.

In any case, something that has always helped me in dealing with obscure problems is Sherlock Holmes's dictum “When you have eliminated all which is impossible, then whatever remains, however improbable, must be the truth.”

Best wishes,
Michael

We thought that X-axis may be oscillating but hopefully our replacing the motor, the resolver, swapping the controller cards and the amp cards would have eliminated this.

As yourself and rklopp stated it could be in the Z gibs which we are going through now. It is quite a logical place to check when thinking about this now!

Thanks guys, will report back soon.
 

Juz

Plastic
Joined
Dec 8, 2019
Kinda a dumb question, but is the fretting present if you offset the work to different locations on the Y or X-axis?

One possible item would be on the cable harnesses going to the resolvers or the servo motors, considering that is about the only thing from what you said that is still common on the electrical side after all the parts you swapped. If you can, I would do an inspection of the cable harnesses as there might be frayed wires, damaged connectors or cracked insulation that is interfering with the motor feedback. I would however, expect cabling issues to be more intermittent than what you are seeing, but I have seen some weird things before. Some cracked insulation with packed in chips/coolant can potentially do odd things.

You may be on to something here as well. In the Fadal machines the X-axis motor cable moves with the table so it sees a fair bit of wear-and-tear. We are going through the Z-axis gibs and then may inspect this.

We should put an indicator on the X-axis ballscrew and feed Y-axis and observe to isolate if X-axis is jumping during operation (the output to the motor shows not as I observed this from the amp card as the machine is feeding Y, but you never know)!!
 

Juz

Plastic
Joined
Dec 8, 2019
We have the same problem on a cinnci arrow with glass scales.

Same kind of lines on Y axis feed.
The a2100 control have a built-in "ocsilloscope" to monitor position and servo command in real time with a graph, and it turnes out that the X axis is oscillating, not Y.

I have not yet concluded why that is on our machine, but I would suspect that it is the X axis on your machine too.

Seems like the consensus is we really need to look more towards the X-axis here! Appreciate the input!
 

Juz

Plastic
Joined
Dec 8, 2019
Has the spindle been inspected other than a run out test?

We checked the run-out as well we used the draw bar pressure tester to make sure tools are being held in adequately. I believe we had 1400lbs in that regards. Is there something we may be missing here you reckon?

Thank you!
 








 
Top