Do's and Don'ts of CNC lathe overhaul?

[zoeper]

Hi All,
I am stripping down a old CNC lathe that seems to suffer from more aches and pains by the day.
Some of these include:
Oil system not working on all slides,
Cutting fluid leaking from turret backside,
Cutting oil only works when the last toolchange was in clockwise direction,
Cuts taper (about 2/100 of a mm in 100mmz travel
Hydraulic chuck only grips at the back of the jaws
Also some electrical problems (mainly with the spindle drive).

My intention is to clean everything up, replace seals and bearings where applicable (not going into the headstock or gearbox in this round )

Thus far i have removed all (almost) the covers and found some disintegrated seals at the back of the turret. The limit switches was also covered in old oil and chips.

The machine has got 3 limit switches per axis, 2 mechanical with rollers and one that looks like it works on magnetism. How does these work, and how can it possibly be acurate if there are chips in the vicinity? Presuming it is for homing??

Does anyone have any cautionary notes or advice before i start disassembling the turret? (hydraulics, oilers, electrics etc?)

Pieter

[StealthDumpKits]

What brand, model and style of lathe? Is it a slant bed or flat? What control?

[zoeper]

Lathe is a AJAX Victor TNS2 slant bed turning center with 10 station Hydraulic turret, 10" Kitagawa power chuck. Controll is Fanuc 10T and the machine was built in 1985. Tailstock has got hyraulic quill but the tailstock body is not automated. Spindle drive is fuji Frenic 5000V2 and spindle motor is 15KW ac
Shout if i missed anything
P

[David]

Originally Posted by zoeper The machine has got 3 limit switches per axis, 2 mechanical with rollers and one that looks like it works on magnetism. How does these work, and how can it possibly be acurate if there are chips in the vicinity? Presuming it is for homing?? Does anyone have any cautionary notes or advice before i start disassembling the turret? (hydraulics, oilers, electrics etc?) Pieter

Here is how the limit switches work for 99% of CNC machines equipped with incremental encoders:
You typically have one 'Reference' switch and one 'Basic position/home' switch. The sensor may be a secondary (redundant) 'home' switch and may be used for other functions, such as a bar feeder or automation. (Not all machines have a 'home' switch, some rely only on the reference switch.)
The reference switch only needs to be precise to within 1 revolution of the encoder, which is typically 1 revolution of the ballscrew.
The method of engaging the switch works one of two ways, either the axis drives off of the switch, then returns on it, or it drives on it first then off. Whichever way really makes no difference to referencing the axis, but does make a difference when trying to troubleshoot the switch.
Lets take the drive off/drive on type to explain.
You put the machine in reference mode, and hit the reference button for that axis, the axis drives off of the cam a set distance, (typically 2 mm), then reverses direction and drives back up on the cam, when it sees the switch change state upon going back onto the cam, it begins to look for the Zero pulse of the encoder. When it finds the zero pulse, it then moves whatever reference point shift distance is in the parameter and sets 'home'.
Note that some machines drive OVER the cam in only one direction (no reversing). The basics are still the same, they are looking for the reference switch to change states twice. So, as long as the reference cam (along with the switch error) is within one revolution of the encoder the reference point will always be the same.
Where you get into problems, is if the reference switch and the zero pulse of the encoder are too close together. When this happens, the error of your switch can cause the machine to reference one revolution of the encoder off, typically one rev of your ballscrew. Away from the chuck isn't too bad, but if it references too close...... BANG!. The solution to that problem is to move the cam half the distance of one encoder revolution to get it in the middle of the revolution.
The 'basic' or 'home' switch is used to physically ensure that the slide IS in the relative 'home' position. The setting of this switch can ensure that you do NOT reference up one revolution too close to the chuck.
To check this setting, you want to reference the machine, then increment the slide toward the chuck while watching the switch on the diagnostics screen, when you see the switch change states, stop incrementing and check the distance moved from home. This distance should be LESS than one revolution of the encoder. If you set it this way, and the slide references up too close to the chuck, the basic switch will not be made and the machine will not run, and thus, not crash. This will NOT, however, prevent the slide from referencing one revolution away from the chuck.

Having the machine make and control type would help.

[zoeper]

Printout of your explanation will definately go into my archive for future referance, and for when i convert my mill. Thanx David!

[Al_The_Man]

It sounds like most of your problems are Mechanical, if the 10T is still in fairly good shape, I would try and keep it, it was not a bad control, If it has DC servo's, remove them and clean them out and replace brushes or send them to a local Motor shop for cleaning.
For the limits, I would replace them with Honeywell hermetically sealed type.
The oil system would need a complete clean out as lube oil congeals when stood for a long time, replace the metering devices that are at the end of the lube points. The Original are probabally the same as Bijur style.
The Electrical problems with the spindle drive may be cured if you can fit a 3rd party drive, if it is analogue control, Is it DC or AC spindle?
If it has proximity switches (Unusual)! I would replace with the above type.
Al..

[zoeper]

Al,
Would you replace the motor brushes merely as a precaution, or would it actually improve the performance of the motors?

I do think that the spindle bearings would have to looked at at some stage as you can hear a very obvious tic tic tic sound from the headstock when you apply presure from the tailstock. This would have to wait for now, as i suspect we are looking at big bucks for a set of those???
Spindle is AC, but i think i have the problem traced to a faulty output transistor.(machine runs fine in reverse and only trips out(OC) when starting forward at high speed) The drive has got both analog as digital inputs and the guy from a motor drive company took one look at it and said No, their drives won't do! - guess i'll just have to fix it then...
PLR

[Al_The_Man]

Replacing brushes is cheap insurance! It may not improve performance but saves $k's on replacement motor, unless there is sufficient length left.
Is it permanent magnet AC spindle or Induction motor?
Al.

[zoeper]

How would you tell by looking at the motor if it is Permanent magnet or induction??

Did some more dismantling and found the cause of all my coolant problems.
The leak at the back of the turret was caused by a faulty o-ring.
The problem with coolant only working after a clockwise turret move was due to the disk that distribute the coolant to the working tool being loose on it's shaft. Clockwise movement would take up the slack and the holes would line up. As soon as you move anticlockwise, the disk would rotate +- 10 degrees on the shaft and the holes would not line up. Damage to the surface of this disk also caused the coolant to leak to the other tools.

There are 7 limit switches on the turret indexing. Why use 7 switches if you can 16 combinations with only 4??

Picture of X axis limit switches also included as discussed earlier in this thread.

PLR

[Al_The_Man]

It may indicate some clue on the motor name plate.
The only problem with setting the index up in hex or octal detection is that the 3 it or 4 bit code has to be registered at every position, rather than just one switch at each position.
I suspect the third switch is for home marker detection, as well as the two R.H. ones used for overtravel and initial home limit, rather than use an encoder marker that may not offer as good position repeatability on this particular machine, but It is very unusual way to do it.
Al.

[zoeper]

Al, there is no plate on this specific motor, unsess it is mounted in a really funny place.

I am not sure that i understand your explanation of the indexing arrangement for the turret. when you refer to octal and hex you mean 3 and 4 bit. On this machine they use 7 switches that would give you 128 possible combinations And it is only a 10 station turret.

As far as the third switch on the x axis is concerned, are you implying that they are not using the motor mounted encoder for home registering at all?

I have chacked the brushes on the two servo's for the x and z axis and both have about 10mm of compression left after the brush touches the commutator. How long would standard Fanuc brushes be when new and when the time comes to say good bye to them?

[Al_The_Man]

If you use 4 switches decoded for position, you require up to 4 switch operators at each position, to decode the 4 bits.
I have no idea what the L.H. switch is for, I am assuming it is for reference, unless there is a M coded position that look at the switch.
The Fanuc servo manual recommends a minimum of 10mm of carbon or 5mm for the 00m motor.
Al.