Z Crawling?

[MecTechMike]

Dave,

By any chance, is your VFD housed with the Acorn? VFD's are known to produce a lot of EMI noise. Putting a VFD into it own housing may be a solution.

Mike

[Dave_C]

Dave,

By any chance, is your VFD housed with the Acorn? VFD's are known to produce a lot of EMI noise. Putting a VFD into it own housing may be a solution.

Mike

Mike, yes this one is as are all the other CNC conversion I've ever done! Never had an issue before and this time only with the Z axis and not the X and Y.

If that is the issue then apparently the Acorn is VERY sensitive to noise as I had never been able to get the Quadrature errors to go away either.

Oh yeah, I can take the entire control cabinet apart and rebuild it from scratch, spend some more money, but if I go that far the Acorn will go in the TRASH!

I should have bought a product that is not so susceptible to noise I suppose.

Dave C.

[Reedfirst]

Just thinking out loud and kind of randomly........

If I understand this correctly, the z axis servo is moving on it's own and the dro on the acorn isn't.

I would imagine that the clearpath servo would fault if there was a position error between the encoder and the commanded position of the drive......

Can you disconnect the step and direction inputs to the z axis clearpath servo (after homing the machine) and run your cnc program ( with x and y axis clear of any obstructions) and see if the z axis moves on its own?

Perhaps that way you can isolate if the clearpath is seeing something on the step signal side of things that it is causing it to think it's seeing a step pulse.

Do you have the step and direction cables going to that servo controller running parallel with some other high voltage wires in your cabinet?

Can you try a shielded twisted pair with the shield connected to ground on one end from the acorn to the clearpath?

Are there any ground loops in you cabinet/ and or outlet wiring that you may be aware of?

Like I said, just thinking out load.....Maybe it'll help maybe it won't.......

[Dave_C]

Reedfirst posted:

Just thinking out loud and kind of randomly........

If I understand this correctly, the z axis servo is moving on it's own and the dro on the acorn isn't.

I would imagine that the clearpath servo would fault if there was a position error between the encoder and the commanded position of the drive......

Good thinking out loud and all things that I have thought of.

What has me stumped is that the Acorn DRO thinks it is at "0" where it was homed but after the part has ran for a bit it is not at the same location. I'm with you that the clearpath servo would error out if it were not at the right location but I think the reality is that clearpath is not using "absolute" positioning but instead is using a logic of "you said go 1,000 steps and I did, if not, I'll let you know." So if it is getting stray signals it just thinks the control told it to move and it does.

So yes, this all could be a clearpath issue on the Z axis. My Z servo is a NEMA 23 while the X and Y are NEMA 34's so there is a difference. I have no issues with X and Y creeping and the wiring is all clearpath wire and all ran in the same bundle with one another and not close to any high power runs.

I'm waiting on a drive belt for the spindle motor and it is coming today. I'm changing the gear ration from 2/1 to something like 1.571/1 so that will let me lower the drive HZ a bit.

More testing tonight!

Dave C.

[Dave_C]

I'm waiting on a drive belt for the spindle motor and it is coming today. I'm changing the gear ration from 2/1 to something like 1.571/1 so that will let me lower the drive HZ a bit.

The new belt came in this afternoon and I started testing again. I had mentioned that after retuning the servo I thought the problem went away. Well, it's back!

As best I can tell it is realted to the speed of the drive. Anything below abuot 100Hz and the Z axis stays put as it should. Anything above that and it starts to crawl!

So I ordered some EMI filters, 20 amp ones, and we'll see if that helps. I'll put one on the incoming power and also one on the 240 volts to the Acorn power supply.

It will take a week to get them!

With the new pulley and V belt setup I can hit 2800 RPM on the spindle if I want so I like that and I still have plenty of power to do rigid taping at 640 RPM. Or I can stay below 2000 and the Z axis doesn't do the crawl.

Dave C.

[Dave_C]

If I tell you I have the problem solved would you believe me? I'm not sure that I even believe it myself as what I did last night should not have fixed the issue in my opinion.

I had ordered 10 5mm ferrite clip on beads that are supposed to eliminate EMI and I'm suppose they have their purpose. I had tried these on the Encoder to get ride of quadrature errors and they did not help.

But since I had them, I thought what the heck, I'll try something. So I put one on each incoming 240V power lead (2) One on the ground wire going to the motor (after the drive), one on each power lead to the 24VDC Acorn power supply and mounted them as close to the PS as possible, one on each power lead to the 75VDC toroid power supply and then I powered it all up.

I ran the spindle full tilt boogie, up, down, sideways and no Z crawl. Hum, looks good so I think I'll go design a flood coolant drain cover and run the part. It has 37 holes all with 1/8" holes so the spindle is running at 2250 RPM and then an 1/8" end mill to cut out the cirlce.


And wouldn't you know it, it ran just perfect!


So is it fixed? We'll see! I attached a pic of the part and I held it up by my flood coolant plumbing to show off the arc I made to mount the nozzle and flow control valve.

Dave C.

[Reedfirst]

Man that's awesome news!

Sounds like it's sorted out.

And nice part's too!

[Dave_C]

My wife remined me at lunch "didn't you make some parts earlier before you added the encoder?" And all of a sudden I started to put two and two together! (But I came up with five, LOL)

Yes, I made the servo cover/tool holder plate that was quite a long run so I took a straight edge and measured the large pocket to see if it was level. It is dead on, no Z variance! I checked a couple of other parts, all good, no Z issues.

So I went to my parts log and looked at the program numbers of the good parts, all were made before I added the spindle encoder! Since I added the encoder, none of the parts have been accurate on the Z axis until today.

So a question for Marty and the Centroid techs:

Why would this have worked so well before the encoder was added and then start having Z crawl issues afterwards? Noise? Yes, I agree but why did it only show up after the encoder was added? And why can't I get the quadrature errors to stop?

Could this possibly be an issue with the (my) Acorn board itself?

Dave C.

[Dave_C]

#1 rule in trouble shooting a problem: "If it worked at one time but it does not work now, or works unreliably, then what changed?"

In trying to answer that question I came up with the following:

1. I had made 4 parts that had no issues!
2. I looked at the date of the part runs and they were all prior to me turning on my Central AC unit.
3. I added an encoder (about the same time as when I turned on the A/C for the year)
4. Every part I ran after turning on the A/C and adding the encoder had issues.
5. There are only three circuits fed from the sub-panel, one is the A/C unit, the second is my Mach 3 lathe and the 3rd is my new Mill.
6. The lathe has been running parts for 4 years without any issues!

I had posted several post saying "I think I fixed it" only to find the next afternoon or the next day the problem was back! (So why)

Answer: I was going down to the basement shop in the early morning hours before the A/C unit was needing to run so everything was appearing to be fixed. Then when I came back later in the day, the A/C unit was on and I was having Z crawl issues!


So to fix the issue, I bought a 20 amp EMI filter and installed it this morning. I'll run some test later toady but if I don't post anything else, the problem will be considered fixed.

Here is what the EMI filter looks like:

EMI Filter

Installed in control cabinet

Dave C.

[Dave_C]

Just so I don't leave anyone in suspense:

I re-ran the same facing program that I had ran before. This was the program that revealed the "Z Crawling" issue.

Just to be clear, the A/C was on, the program was ran at the same spindle speed and so on. The program ran perfect!

Not wanting to leave anything undone, I tried to re-enable the quadrature error detection. As soon as I did I got quadrature errors before the spindle even gets to 10 RPM.

No idea what that is about but I am leaning toward the Acorn board itself as I've eliminated every variable possible.

Even if it is not defective, it must be WAY too sensitive to something and there is no way to determine what it is at this point. It's amazing that my Mach 3 Lathe runs an Ethernet smooth stepper, an Automation Direct G2 drive, it is in the same type metal cabinet, same proximity of drive and control board, same power supply/panel and yet I've never had an issue.

No one else has this issue and yet many of the other builds have drives and components in closer proximity than mine.

Dave C.