Fanuc Repair Gurus - Brain teaser of the week... 21iT tale of woe..

[RayG]

Seeing as how the problems appear to have started with the encoder battery replacement, it seems logical to see if there is anything stored in the encoder bbram, that would affect comms, from what I can read the only thing stored is
axis position...

I came across this http://support.ge-ip.com/support/re...n_US/Fanuc Series 0i Model A CNC Battery.pdf
So, If the encoder batteries were changed while the machine was powered off, you might have to do a Manual Reference Position Return.

Ray

 

[Milacron]

Don, tried to sent PM but your box is full.

It's always full... try procyon at hargray.com

 

[Milacron]

Seeing as how the problems appear to have started with the encoder battery replacement, it seems logical to see if there is anything stored in the encoder bbram, that would affect comms, from what I can read the only thing stored is
axis position...

I came across this http://support.ge-ip.com/support/re...n_US/Fanuc Series 0i Model A CNC Battery.pdf
So, If the encoder batteries were changed while the machine was powered off, you might have to do a Manual Reference Position Return.

The encoder batteries were changed because voltage was low enough to loose reference return...so it didn't matter a whit if I changed them with power on or off as reference return was lost already. But that is alarm 300. Alarm 301 is the bugaboo here... 300 is easy to clear once 301 is banished. Alarm 300 has nothing to do with Alarm 301.

Having said all that, it is interesting that the 301 alarm did start about the same time as I changed a battery...but thinking back on it now, I don't think it was exactly the same time. Frankly I had just got the machine in and was having fun going thru the screens and playing with the handwheels and now damned if i can remember if the 301 popped up before or after I changed a battery. (I only changed the Z battery at first as the X still had 5.7 volts and no 300 alarm) Bottom line is changing the encoder battery almost certainly had nothing to do with the 301 alarm.

 

[jCandlish]

Frankly I had just got the machine in and was having fun going thru the screens and playing with the handwheels and now damned if i can remember if the 301 popped up before or after I changed a battery.

Its always the worst not knowing why.

 

[Milacron]

Its always the worst not knowing why.

Looking back on it now I think the exact moment the axis stopped working was after I moved the X toward me, heard a beep that warned me I had reached what I know now was the soft limit* (the machine has no physical limit switches), I reversed direction about one turn and the axis and encoders eased to function...both of them. I'm not 100 percent sure when I reversed direction of X that it actually moved but I think it did for about one turn.

This gave credence to the exceeding soft limits theory expressed early in this thread. But I concluded days ago that soft limits have nothing to do with this problem.

----------
*If you are wondering how the machine knew the soft limit position remember I said at first the X battery had 5.7 volts and apparently still retained it's zero home position memory...the 300 alarm on X came later. The Z axis had lost home position from the get go... battery was only 1.5 volts (supposed to be 6 volts)

 

[jCandlish]

(the machine has no physical limit switches)

Just WOW! That is scary. Really? No hard limits???

The hateful thing about not knowing how something became broken is the nagging feeling its going at any moment to happen again.

Without hard limits maybe you drove into the region of buggeredness caused by the previous owner and tripped something up?

Thats the thing about not knowing the causes of things. Sure you could get it all back together and working, but its like you've reassembling a complicated assembly only to find that you've got leftover parts. T'aint right.

 

[Milacron]

Without hard limits maybe you drove into the region of buggeredness caused by the previous owner and tripped something up?

I thought about that, and had some notion that perhaps I actually went beyond the soft limit and maxed out the screw, which then overloaded something. But I was turning it real slow and two Fanuc techs I mentioned that to said "nah, that's not even remotely a crash..nothing to do with it". Plus I've overloaded Fanuc servos on other machines before much worse than that, with no ill results at all.

If the servo card does turn out to be the culprit I have a theory as to why that could have gone bad. There is no filter on the lower intake hole of the electrical box. And right at that hole is the Z servo drive. Although the cabinet insides were very clean in general, the lower half of the Z drive was filthy with black dirt. I took the Z servo drive cover off and found that one IC had some of it's legs so gunked up there was some continuity between the legs there shouldn't have been. This was one of those "computer in a chip" deals with a gazillion legs very close together. It took a pointed Exacto knife to clean out the spaces between the legs. I had some hope doing that would have fixed the whole deal but it made no difference whatsoever. But I can't help but wonder if the problem might have gradually caused the servo card to go nutty.

Having said that, it's not impossible there is nothing wrong with the lathe and never was.... my Phase Perfect going into excess RF mode could still account for the entire problem. Highly unlikely, but not impossible. I have a filter on order to solve that issue just in case.

 

[jCandlish]

my Phase Perfect going into excess RF mode could still account for the entire problem.

At the exact moment the lathe alarmed out (overtravel?) and ever since then?

 

[Admin5]

At the exact moment the lathe alarmed out (overtravel?) and ever since then?

The operative phrases from that post are "I think..." and "highly unlikely"...wish I'd been paying better attention but was having too much fun at the time...

Milacron

 

[Admin5]

Sheesh. Time for the glass-fibre bristle brush, ethanol rinse (AND NOT denatured with meths..), hot air-gun on low/no heat.

And if the maker has no filter? JF add one.

Yeah I did that too but it still took the Exacto knife to get it all. Re filter..my place is clean enough I don't really need one...but I will anyway.

Milacron

 

[WILLEO6709]

Really? No hard limits???.

I have name brand HMC's running absolute encoders, no home dogs, no hard limit switches. Don't knock it until you have tried it, these are the most up-time friendly machines I own... all switches get sticky and trouble prone with age... IMHO.

 

[RayG]

Frisket knives can cut when one wishes otherwise. Grab a set of stainless Dental Hygienists' tools next time you see 'em. Mine were made in Pakistan. Not the best for Dental work, perhaps, but cheap enough for something used once in a great while around PCB's.

Bill

Scratching around fine pitch ic's with an exacto knife ( or dental picks ) is not a good idea... Spray can of PCB cleaner will remove the bulk of the grime, but also pushes little bits into corners where you don't want them. After you get the majority off with the spray cleaner, dunk the whole board in an ultrasonic cleaner with isopropanol.

If it's a new ( or repaired) board and you want to deflux, I use a water based product "Chem Tools PCB Wash" and then thorough rinse in hot water, and into a drying cabinet to bake..

Ray

 

[Admin5]

Scratching around fine pitch ic's with an exacto knife ( or dental picks ) is not a good idea..

I was aware it wasn't the greatest idea but electronic solvent plus air blast (thru a 6mm ID hose) didn't get it all so very carefully removed the remaining residue with a dull Exacto blade being attentive to not scratching the board itself. I took a photo of it afterwards with the iPhone 5, which takes such nice closeups it serves as a sort of microscope when you enlarge the photo on the phone's retinia display. The results looked like a complete success.

Remember you are talking to someone that once desoldered 400 legs to remove 20 static rom DIP IC's packed as tight as they come, without damaging an IC or trace. Both of you would have told me that wasn't a good idea as well.

Milacron

 

[RayG]

Both of you would have told me that wasn't a good idea as well.
Milacron

That's wasn't a good idea...

Ray

 

[Admin5]

That's wasn't a good idea...

LOL... without knowing the circumstances that led to it, and the $$$ replacement cost of the board, true enough...

Milacron

 

[metaltech]

Just WOW! That is scary. Really? No hard limits???

We should make the distinction of hard limit switches (physical switches) that signal the control to decel the axis to 0 and display an overtravel alarm versus emergency hard limits that are wired in the E-stop string, such that hitting one causes the servo drives (and whatever else) to shut off, stopping the axis before mechanical lock-up or jamming. I would think Milacron's lathe has emergency switches, which stops the axis if the soft-limit doesn't. It used to be common for machines to have soft-limits, hard-limits, and emergency-limits, but for many years, the hard-limits have been deleted on most machines. I've never seen a machine without emergency-limits, though they might exist. It would be putting way too much faith in the electronic soft-limit, and likely leading to a premature death of the ballscrew or support bearings, etc.

Related, but uncommon, are position switches. These are a software switch that turns on when an axis is within a certain range of travel, defined by parameters. You set a plus position param and a minus position param so when the axis enters that range, the switch signal is output by the CNC to the PMC. A 240" travel VMC I work on uses them to "fire" air cylinders that support the very long X ballscrew.

 

[RayG]

S'pose that might depend on one's experience and skill level? And what one considers 'fine pitch'?

Bill

Yes, I was only being half serious, I admit that, I have on occasion poked between ic pins with sharp objects in a usually vain attempt to remove solder bridges. I do hand solder 100 pin TQFP 0.2mm chips on a semi-regular basis, doing prototypes, I use an Elmo video presenter a bit, but close up I like hand held magnifiers, and good lighting to see properly. Good flux and a bit of practice. It's not that hard, just time consuming.

In this particular case the servo interface card had already stopped talking to the encoders.. so it was probably already screwed.

Ray

 

[jCandlish]

I have name brand HMC's running absolute encoders, no home dogs, no hard limit switches. Don't knock it until you have tried it, these are the most up-time friendly machines I own... all switches get sticky and trouble prone with age... IMHO.

It just seems like false economy to me. Whats a pair of switches and inputs cost the MTB anyway, as a percent of the total cost of the machine? And Hall Effect switches are good for like a gazillion cycles.

But I suppose if you make the servos small enough with respect to the other kinematic elements there isn't much to break.

We should make the distinction of hard limit switches (physical switches) that signal the control to decel the axis to 0 and display an overtravel alarm versus emergency hard limits that are wired in the E-stop string, such that hitting one causes the servo drives (and whatever else) to shut off, stopping the axis before mechanical lock-up or jamming. I would think Milacron's lathe has emergency switches, which stops the axis if the soft-limit doesn't. It used to be common for machines to have soft-limits, hard-limits, and emergency-limits, but for many years, the hard-limits have been deleted on most machines. I've never seen a machine without emergency-limits, though they might exist. It would be putting way too much faith in the electronic soft-limit, and likely leading to a premature death of the ballscrew or support bearings, etc.

Looking back on it now I think the exact moment the axis stopped working was after I moved the X toward me, heard a beep that warned me I had reached what I know now was the soft limit* (the machine has no physical limit switches)...

.

 

[Heavey Metal]

Sounds like machine is parked on the dead stop limit.

What is the procedure to move off limit?(could be manually rotating the axis)

Is the ball screw mechanically locked from over travel?

 

[machtool]

It just seems like false economy to me. Whats a pair of switches and inputs cost the MTB anyway,

Thats akin to saying your axis needs limit switchs to hit any other arbitury position. Whilst I dont mind hard / e-stop switch at the extreme end's of travel. Simple fact for at least the last decade and a half, Absolute encoders with battery back up, regard the over travel positions, just the same as any other number. I.e, why could you turn a diameter of say 250.00mm, and your soft limit is -305.00mm. Why would you trust one, but not the other?