Servo draws too much current near edges of machine.

[machina]

I have a 1999 Bridgeport VMC800 with Fanuc 18i control. I've been getting an over-current alarm 436 - Soft Thermal occasionally on the X-axis and I've figured that it only seems to happen when I have left the table sitting near the travel limits of the X-axis for a minute or so. I'm not sure how long this has been a problem as the machine has only been used occasionally for the last couple years and I don't remember if I was getting this alarm before.

I turned on the current monitor page and can watch the X-axis servo current climb over 100% when I jog to either the far right or the far left of X-axis. When at rest in the middle of the table the motor is typically drawing 40 - 50% current. As I jog towards either end the current slowly climbs and will go above 100% for the last 4 inches of travel.

One interesting thing is that if I switch directions and turn the jog-wheel 0.0001" in the opposite direction the current will drop to ~5% and stay there. Reverse back 0.0001" in the original direction and current returns to 100%+. It doesn't matter which direction I and traveling, as long as I am near the extents of travel my current remains over 100%.

My initial thought was that the ballscrew nut is not exactly colinear with the ballscrew. The closer I get to the middle of travel the ballscrew can bend slightly and it doesn't load too much. The closer to the edge of travel the ballscrew is much stiffer and resists more. I'm not sure this is true however because the ballscrew mounts are pinned in place after being factory aligned and don't seem able to be shifted. Also, I loosened the ball nut from it's bracket under the table and jog it to the end of travel and the problem persists. There is plenty of play in the screw holes and the bracket for the nut to shift in any direction it wants to but no reduction in current occurs even if I remove all four mounting screws.

If I jog the ball nut to the extent of travel when not attached to the table current stays ~5% always.

I loosened up the gibbs significantly to see if it had any effect and it did decrease the current draws at the extents a bit but still goes above 100% near the ends. Machine is level.

Any ideas what could be wrong?

Thanks,

Aaron

 

[ViktorS]

There is not too many reasons for this to happen. I just type it all out:

1)Table binding
Remove the screws to the ballnut (so you can push the table without the ballnut attached) and push the table CAREFULLY (it might be able to be pushed of the rails).
Does it bind? If not, this is not the problem. If it is, probably uneven worn rails.

2)Ballnut binding.
Remove the screws of the ballnut and push the table out of the way to one direction. Rotate the ballnut manually along the screw towards the other direction. Can you go all the way
or does it bind? If not, that is not your problem. If it is, it's because of a worn screw.

3)Axis-Ballscrew alignment binding
If none of the above this is the most likely one.
Ballscrew axis is
a) not parallel to table axis or
b) parallel but the bearing blocks for the screw are not on the same line as the ballnut.
Either of the above would have the screw bend and get progressively more binding closer to the bearing blocks.
To test this I suppose one could let go of the screws to the ballnut or bearing blocks and see if it still binds. If not this and none of the above...

4) Way covers?
Do they collapse as they should or to they try to push back? Might be swarf caught between them. Detach them from table and collapse them by hand. Then run the axis and see if it helps.


The servo is not the issue in any way here. The reason the load drops when you use the MPG to move the table back a bit is because the set position is closer to the real position,
and the position control loop lowers the servo command output. In other words it's closer to where it should be and therefore does not fight as much.

I personally would suspect no. 2 as the culprit if you have already tested no. 1.

EDIT: I see that you have gibs and you tried loosening them. The gibs should be somewhat tight, there are several threads about gibs adjustment and how to check it.
But as the loose gibs did not produce any large change in the problem I would say that they are not related to this. The small change you see is because of the correctly adjusted gibs adding some load, as they should.

 

[guythatbrews]

I vote for chip packing at the end of travel. I have a machine that behaves this way due to chip packing under y axis slide and must be cleaned periodically. Just a mass of solid crap under there.

 

[dandrummerman21]

You took the nut off the table? So can you push the table by itself in the middle?

Sounds like a box way so might be difficult to get it started but you should be able to move it with a 2x4 as a prybar with relative ease. I wonder if you'd feel it that way.

Could be lack of oil?

How do the ways look?

Seems to me even the 50% axis load you see in the middle of travel is a tad high. I dont know your machine and servos but I would have guessed under 20%, especially for the x axis (which isn't nearly as heavy as the y)

 

[ViktorS]

Oh crap I didn't see the line about 40-50% load normal. Yeah thats way to high (for a flat axis atleast).
You might have some sort of lubrication issue.
The small amount of lube (or wear) might be collecting in the middle and the dry unworn ways at the end will probably make it go that much heavier.

 

[gustafson]

I had a pair of the Heidenhain versions [1000s] and they had no problems ever that way. I am going with the guys saying lube. Check the lube pump, maybe hot wire it to make sure oil comes out. Again from sitting it might just be goo

 

[Ox]

This machine doesn't have linear rails?


---------

Think Snow Eh!
Ox

 

[dandrummerman21]

This machine doesn't have linear rails?


---------

Think Snow Eh!
Ox

I found a tired one at HGR from 1996 that appears to have box ways for Z.

Used Bridgeport Bridgeport VMC800 CNC VMC | HGR Industrial Surplus

 

[Ox]

I had a cpl of early/mid 80's Cinci's that had linear rails in X and Y, but had box in Z, so "box in Z" doesn't really = Box in the others.


------------------

Think Snow Eh!
Ox

 

[gustafson]

This machine doesn't have linear rails?


---------

Think Snow Eh!
Ox

Mine are probably already Kias by now, so I cannot go check but I think you are thinking of the other series of the same vintage, torque cut .....had to look it up

 

[dandrummerman21]

Hmmm... The 720 was box XYZ.
The newer (1998 era) VMC600 was box Z and linear (pissy little things about an inch apart) XY.
This being 1999 VMC800 looks a bigger version of the 600, so is probably linear XY.

If so, I wonder what the gib he loosened up was?

OP, can you clarify? And hey, pictures of what you've got going on would help.

 

[Ox]

I'm not thinking of anything.
I have the next newer generation mills, but they are [I believe] on a whole different level playing field, so ...

But I'm thinking that this is an entry level machine, thus the rails would be cheaper to build?
No?


End game being - he mentioned loosening gibs, and the issue stayed the same, so apparently this thing has box, and I am just really surprised by that.
And if he's been that deep, I haft'a think that chip packing is NOT the cause, although that was my first thought.

And if he loosened the gibs, and it's the same - ???



------------------------

Think Snow Eh!
Ox

 

[machina]

LoL @ Ox saying he's not thinking


So, my ass-umption looks to be very wrong.
Here's a '99 800 digital and look at them there thighs!
it looks as well built as the 720 which was a good machine.

Yes, this machine is box ways, no linear rails. Chips are cleaned out so it's not packing. The table is very difficult to move by hand when disconnected from the ball screw. I can put my back on the wall and push with both feet against it without moving it when it's at the end of travel. I'm thinking this is a lube issue combines with table wear that makes it worse when traveling near the ends.

I removed the turcite covered bars that capture the table from the underside (what you would remove if you wanted to lift the table out of the machine. The servo power required to move the table dropped by more than 50%. Stays well under 100% at the ends of travel. It looks to be in decent shape as far as I can tell.

A tech I spoke with told me he has seen turcite swell up over the years on some machines and caused binding. Ever heard of such a thing? I could re-mount these bars with some 0.001" or 0.002" shims to relieve some of the pressure as I work out whether the potential lube problem. Bad idea?

The metering vavles are all in manifolds on the saddle so they are easy to replace. They seem to be obsolete metric models from Dropsa so I may have to replace the entire manifolds. I want to see which ones may be clogged. I'm planning on removing the 4mm output lines from the valves and running the oil pump to see which are not flowing but I can't figure how to activate the pump. Any ideas? I can't find a parameter that controls pump timing. It comes on for about 2-3 seconds when I power on the machine and then has an unknown interval before pumping again. I also see a timers page that could control it but no details on which timer to choose.

Aaron

 

[machina]

The machine has box ways. There aren't any chips packed in, I thought of checking that first

I'm pretty sure its a lubrication issue being lessened in the middle of travel by wear. The ball screw is fine and there's no binding there. I removed the turcite covered pressure plates from under the table that keep the table captured. The current draw fell by more than 50%, staying well below 100% even at the ends of travel. Stays at about 30% now for most of the table travel. The Turcite looks fine as far as I can tell. I'm thinking of putting these pressure plates back on but with some 0.001" or 0.002" shims to relieve some of the pressure off the ways (and servos) until I can sort out the lube issue. Bad idea?

A tech I spoke to told me he had seen turcite swell up after years of use on some older machines and cause binding. Anyone else ever heard of this

I'd like to find out if and which metering valves may be clogged. I want to run the lube pump with the 4mm output tubes disconnected from the metering valves and see which ones are not outputting oil. I cannot figure out how to trigger the pump though, any ideas? It comes on at power-up for 2-3 seconds then waits some unknown interval ( I read that its about 30 minutes for another Bridgeport VMC user). I can't find a parameter that controls this. There is also a TIMERS page in the PMC section that might control it but I don't know which to try.

Aaron

 

[machina]

I'm trying to upload some photos. Apologies if this doesn't work.

Aaron

 

[Ox]

The machine has box ways. There aren't any chips packed in, I thought of checking that first

I'm pretty sure its a lubrication issue being lessened in the middle of travel by wear. The ball screw is fine and there's no binding there. I removed the turcite covered pressure plates from under the table that keep the table captured. The current draw fell by more than 50%, staying well below 100% even at the ends of travel. Stays at about 30% now for most of the table travel. The Turcite looks fine as far as I can tell. I'm thinking of putting these pressure plates back on but with some 0.001" or 0.002" shims to relieve some of the pressure off the ways (and servos) until I can sort out the lube issue. Bad idea?

A tech I spoke to told me he had seen turcite swell up after years of use on some older machines and cause binding. Anyone else ever heard of this

I'd like to find out if and which metering valves may be clogged. I want to run the lube pump with the 4mm output tubes disconnected from the metering valves and see which ones are not outputting oil. I cannot figure out how to trigger the pump though, any ideas? It comes on at power-up for 2-3 seconds then waits some unknown interval ( I read that its about 30 minutes for another Bridgeport VMC user). I can't find a parameter that controls this. There is also a TIMERS page in the PMC section that might control it but I don't know which to try.

Aaron

I just recently had to doo some oil line work, and IDK what cheat the OEM has to make that cycle, but like you said - at power-up it cycles every time, so I just kept going in and out of E-stop several times to get it to clear the ends of the lines.

I had mine all apart while I was in there.
Not sure if that was a good idea or not? LOL!
There's some tiny little parts in there!
Miraculously I didn't lose any tho.
(this time)

BTW - my lines were down low and had gotten knocked off by a build-up of chips in the conveyor.
Mine weren't plugged. (AFAIK)
I just had to dock the end off the tube and put it back together...


-----------------------

Think Snow Eh!
Ox

 

[CarbideBob]

Why okay in center and poopey near the outside?

 

[machina]

My theory is that the ways are worn more in the middle and therefore looser. Reasonable?

 

[Ox]

I don't think that it matters how long the pump timer is on.
A Bijure unit will only fill a certain void in the metering stem (?) and once that is full - it's full.
Then it seemed that after the pressure cycled off, that the new volume of oil was shifted to the other side of the piston, and then was pushed out by the next pressure cycle.

So a shorter time between cycles would push more oil, but a longer pressure cycle should not change volume.
At least on Bijure.

If not, then the line with the most resistance (plugged?) wouldn't get any,and the most free flowing line would be flooded.

At least that is how it seemed to work when I was working on mine the other day, and also how it makes sense for the system.


I could be wrong, but I don't think so.


---------------------

Think Snow Eh!
Ox

 

[gustafson]

I don't think that it matters how long the pump timer is on.
A Bijure unit will only fill a certain void in the metering stem (?) and once that is full - it's full.
Then it seemed that after the pressure cycled off, that the new volume of oil was shifted to the other side of the piston, and then was pushed out by the next pressure cycle.

So a shorter time between cycles would push more oil, but a longer pressure cycle should not change volume.
At least on Bijure.

If not, then the line with the most resistance (plugged?) wouldn't get any,and the most free flowing line would be flooded.

At least that is how it seemed to work when I was working on mine the other day, and also how it makes sense for the system.


I could be wrong, but I don't think so.


---------------------

Think Snow Eh!
Ox

My VMC1000's had a rotary pump, rather than a displacement pump.

My older machines had a bijur with clockworks and a cam that only ran when the spindle ran, creating a pump once every x minutes of spindle time