Hi,
Looking for some help on the following:
I have a persistant 400 servo overload alarm on my Fanuc OM. This seems to occur after the machine has been switched on for a few hours. If I switch off for about an hour it goes away (I have to wait this long for it to 'reset').
Is there a pot setting that could raise its tollerance, or a thermal overload reset like on DC motor gear. Any advice would be great and please assume nothing as I do not know my way around the machine yet! Thanks
Darryl
Does the servo motor show high temperature? if so it may be getting hot for a reason and the thermal cutout is tripping.
It can be that there is mechanical resistance or a problem of an axis having a problem holding position, 400 is the X axis, so check this first for over temp.
Do you have the manuals?
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I have 3 Fanuc manuals: Maintenance, operating and descriptions. Nice to know which drive it is - thanks. I would not say the x motor feels anything other than slightly warm when touching the casing. The comment about resistance.....I have had a faulty lube pump and have just replaced the motor. Could a sticky or dry slideway cause the overload alarm? Though, I have to say I have been applying oil with a brush in the interim and when I jog the x axis it appears to be nothing but very, very smooth. One other thing I have noticed is that the overload alarm doesn't trip when the motors are actually moving - it seems to always go when the machine is dormant, after it has been switched on for about 2 1/2 hours (if I wait about an hour the alarm drops without any intervention). Not sure if it is a related problem but once, and only once, I have had the 401 alarm (VRDY off). All the servo motors seem to buzz quite a lot but I am told this is normal, although one engineer remarked it was excessive (but used a m/c 15 years younger than mine). I have to say that x is the quietest and Z the noisiest.
Please get your drives checked. This alarm happens if yout drive modules become weak
which drive you are using and what is the alarm on drive
harsh
An Fanuc engineer has now resolved this problem. By posting this 'fix' I hope other machinists will benefit from it - and also hope that those that advise on this site will advance their knowledge too. This particular problem is a major machine problem but was utterly simple to fix, and despite warnings from the forum about failing and defective drives or servos, it was nothing of the sort.
Just to recap, after the CNC machine had been switched on for an hour and a half, the alarm tripped ....... every time, regular as clockwork. Once tripped, the machine was inoperative for anything between 24 to 48 hours, where upon it mysteriously reset without operator intervention (note, I tried everything to force a reset after the alarm tripped, but to no avail).
The Fanuc engineer informed me that alarm 400 was the Z drive (not X, as I was advised on this forum). Over the phone, I was advised to trace the Z drive cable from the system board (I have it labelled) back to the respective drive (on my machine, X and Y go to one drive board, and Z has its own independent drive board). On both drive boards, there are a small number of potentiometers or 'pots', which increase the 'gain' on the drive. The 'pots' on the Z drive board were labelled R1, R2, and R3, which were all set to about 2 O'clock. I was told to change R1 and R2 to about 12 O'clock and switch on. The new settings did not work at first but I noticed the Z motor did not buzz as much as it had done recently. A little more trial with the settings revealed that a pot setting on R1 and R2 of 11 O'clock was correct and I have not had the 400 alarm trip since. I suggest anyone that gets this alarm tries a different 'pot' setting. I found that if the drive noise (buzz) increased I was going the wrong way, and turned the pot in the opposite direction.
Here is the manual about alarm 400.Originally Posted by darrylh
The axis can be confirmed by looking in 720~727 xyz...etc
That is why I asked if you had the manual, which I could have sent if you had needed it.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I report what the engineer advised. Some additional info: I was asked to check diagnostic 802, which read a non-zero value (diag 800 and 801, which are X and Y drive diags respectively, both read 0 or 1, which is what they sghould be). The fact that diag 802 was non-zero meant that 'gain' on the drive was required by adjustment of the potentiometers.
I have a Tecnics cnc wt Fanuc controls O-mate? I'm getting 400 servo alarm right at start up from what I can see this machine has about 500 hrs on it an I need to get it running.
See chart
Fanuc 0T 0M Troubleshooting Procedures Alarm List
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
any steady values in 800-802 are a sticky slide most of the time, will cook a motor...increasing gain will help it pull into position, but I'd be sure the slides are wet with oil, or 'fixing' the gain might just hide a problem that can tear up the machine. excess error with dc motors can mean one of the 4 brushes is not making contact too...a loose brush screw or worn brush...I report what the engineer advised. Some additional info: I was asked to check diagnostic 802, which read a non-zero value (diag 800 and 801, which are X and Y drive diags respectively, both read 0 or 1, which is what they sghould be). The fact that diag 802 was non-zero meant that 'gain' on the drive was required by adjustment of the potentiometers.
here is how I set the analog drives- wont say its right or wrong, but been fighting finish issues off and on for going on 30 years, this has worked for me:
1) turn off auto drift comp (parameter 1.0 I think in a zero control, might be
1.7, know its in the first parameter- check the book)
2) zero out drift comps (param 544-546 or thereabouts?)
3) jog each axis slowly +, stop. note value in 801~803
4) jog each axis slowly -, stop. note value again.
ideally values will be flickering 0 +/-1, if so, go to next axis (if any servo is 'buzzy' might need to decrease gain on that one, then match others to it.
*note- buzzy dc can be (often) just a slightly worn oldham coupling on the servo...carefully pull the encoder off (dont rotate it or its center shaft or you'll mess home/grid shifts up), mark the outside for orientation, soon as its pulled straight off, mark the center so you dont turn it 180. the little brass oldham coupling must have NO shake...Ive just dimpled a couple spots with a center punch/put a dab of heavy grease on it, usually will keep it tight for a couple years...best permanent fix, I spottec blind holes in diagonally opposite sides (both keys) and put a little piece of model airplane sized silicone fuel line in the hole- sticking out just a tiny bit...lube/reassemble, the silicone tube acts like a spring, never had one of those wear loose again. even simply dabbing grease on the loose key will tighten it up for a few months.
if unequal/opposite signs, adjust center pot(offset adjust) to make equal/opposite
if equal/opposite sign, increase gain pot slightly (left or top pot), retry...should pull in to 0-1 both directions if slides not really dry/screw/bearings ok... repeat for other axes. note, gain pot 99% of the time should run near straight up- if you start seeing it turned by more than a click either way, something might be screwed up.
[*note-
on DC servos, if theyve ever been apart, many rebuilders dont time the magnets...if the motors are off the machine, you can just hook it up to a (well charged) cordless drill battery or other solid dc source(only needs a few amps to run, but maybe 10 to accelerate), and tach the shaft, reverse polarity retach. if mismatch is seen, just loosen the nuts under the pulse coder cap, note the holes are slotted...twist the brush endbell as needed to match the rpm. if the rpm is the same on a solid dc source, the motor will build the same torque in both directions. I used to note voltage/rpm too for future reference. if you ever tach a similar motor and it runs >10% faster, its magnets are failing...weaker field= higher rpm/less torque]
5) program 2 (all three if your machine will do 3 axes simultaneous) axes to move same distance in G1 mode... due to 'fading' linearity on analog stuff this old, I usually feed at the finishing feedrate typically used.
Watch 800-803, they NEED to match within 1% ( OR 50% within 1% on some machines due to DMD/CMR/servo size/screw pitch ) adjust 'tach' pots to match these values. tach is either bottom or right of the three pots. if the 'tach' feedback gets too far out of whack, whenever you interpolate a radius, one axis will 'outrun' the other by a tad, then slow for it to catch up, kinda turning a precise radius into a polygon like surface...its more cosmetic than dimensional, but ballbar plots show a little more mismatch too if its out- finishing aluminum with a diamond or small radius tool it can make a huge cosmetic difference.
on the auto drift comp, I prefer to leave it off...a sticky slide can throw comps in there at faster feedrates than finish feeds, and cause finish to no longer be optimized...if you elect to turn it back on, at least check those 544? params on occasion to make sure the controls not trying to comp for a mechanical issue... if its left off, you will need to check them every so often and touch up adjustments- usually a couple years or big temperature swings will affect them a little, but not bad...
also, while on analog servo issues, regarding a 400 alarm- normally 40_ means X, but not 400 or 401...
400 is a thermostat in the servo or the one in the transformer. if the motors not hot to the touch, could be a bad thermostat- normally closed hard contact on pins 8/9 of the honda 20 pin plug for that axis(or R/S pins at the motor end on most motors- all the incremental encoders with a 20-29S sized
amphenol plug that ive ever seen). if you want to check the transformer thermostat, on standard fanuc supplied transformers its transformer terminal 52/53 i think...the last 2 terminals, usually the only blue wires there too). to check transformer input to the control, look at your servo drives- one or more will have five wires on CN1 white plug...the 2 extra wires on the right are from the transformer(s) thermostat, should read shorted.
also note, if 5 wires on CN1, then the 'toh' jumper on the drive should be open.
if you get a 400 again, might shut off power and check those thermostats.
oops- shoulda mentioned bigger machines could possibly have 'regen units' on the drives- if so they also have thermostats in series with the transformer wires...
thats my 2 cents for the night
Last edited by tc429; 10-27-2011 at 10:06 PM.
Posting Permissions
LinkBack URL
About LinkBacks
