damn that thing

[MIYANO vs FANUC]

Hello everyone

I haven't been in the zone for a while because my CNC has been working perfectly until a couple weeks ago. I've been hunting around on the internet trying to solve this problem(s) but I figured it was time to ask the experts.

These are the alarms I’m getting, both are spindle related.

AL-24
AL-27

Now I got this AL-24 only a couple random times today but there doesn’t seem to be a rhyme or reason but here is the alarm…

Alarm Code 24
The power to the CNC is turned off. (This symptom does not represent an
error.) Serial communication data transferred between the CNC and
spindle amplifier module contains an error. Troubleshooting when this
alarm is issued
(a)Noise occurring between the CNC and spindle amplifier module
(connected via an electric cable) caused an error in communication
data. Check the condition for maximum wiring length. Referring to "Connection," in "FANUC SERVO AMPLIFIER i series Descriptions
(B-65282EN)," check the condition of electric cable connection.
(b)Noise exercises an influence because a communication cable is
bundled with the power lead. If a communication cable is bundled
with the power lead for the motor, separate them from each other.
(c)A cable is faulty. Replace the cable. If an optical I/O link adapter
is used, the optical link adapter or optical cable may be faulty.
(d)The SPM is faulty. Replace the SPM or SPM control printed
circuit board.
(e)The CNC is faulty. Replace the board or module related to the
serial spindle.


Here is the main offender…

Alarm Code 27
The signal of the position coder is disconnected.
(1)If this alarm is issued when the motor is deactivated
(a)The setting of a parameter is incorrect. Refer to "FANUC
AC SPINDLE MOTOR i series Parameter Manual (B-
65280EN)," and check the parameter for sensor setting.
(b)The cable is disconnected. If the connection of the
feedback cable is correct, replace the cable.
(c)The SPM is faulty. Replace the SPM or SPM control
printed circuit board.
(2)If this alarm is issued when the cable is moved
(a)The connector has a bad contact, or the cable is
disconnected. The conductor may be broken. Replace the
cable. If coolant has penetrated into the connector, clean
the connector.
(3)If this alarm is issued when the motor rotates
(a)The shielding of the cable between the sensor and the
SPM is faulty. Referring to, "Connection," in "FANUC
SERVO AMPLIFIER i series Descriptions (B-65282EN),"
check the shielding of the cable. (b)The signal cable is bundled with the servo motor power
lead. If the cable between the sensor and the SPM is
bundled with the servo motor power lead, separate them
from each other.


Some things to note: When I turn the CNC on and after it has booted up, if I grab the spindle by hand and turn it I will get the AL-27 on the spindle display. On the control panel it comes up as a 408-alarm. This alarm cannot be reset unless the power is shut off on the control and the main switch on the back of the machine that shuts of the spindle.

NOW! If I let it boot up and then MDI a M3 S(whatever) the spindle will turn on and spin. But if I MDI a M5 or hit reset I’ll get a 409 alarm (AL-27) along with a 1000-alarm. WTF is a 1000 alarm? Also make note that I don’t get the 408 alarm if I first turn on the spindle with some M commands. If I spin it by hand after the M commands I will only get the a 409 alarm (AL-27) along with a 1000-alarm. WTF is a 1000 alarm? From what I’ve read it’s a machine builder alarm. So this 1000-alarm is Miyano’s biz?

Now both the 409 alarm (AL-27) along with a 1000-alarm can be cleared by pushing the reset softkey twice…. Then I can turn the spindle and get both alarms again and clear them again. I can also MDI a M3 with whatever speed I want as long as it’s above maybe 10 RPM. I can MDI an M3 S10 or S15 and turn the speed down with the override but it’ll tag the 409-AL-27.

I should also point out that the spindle seems to “ripple” at low speeds. Let say I MDI a M3 S10 and grab the spindle. When I grab it and try to stop it I can feel it pulse at it spins. If I try to stop it the control gives the 409-AL-27 alarm.

Does this sound like a serial cable issue or the encoder or something else?

As always, many thanks in advance

[tc429]

saw similar situation once- it was a broken lead to the spindle motor- they can kinda run single phase if they start, but cant regen to a stop. BUT that wouldnt explain the al-27 if spun by hand (maybe if theres a broken lead touching ground though?)
unhook your motor leads from the drive and check between with a ohmmeter- they should read nearly shorted if intact... hopefully thats all it is- stringing new cables can be a pain, but a lot cheaper than a amp.
might open the motor junction box and look for a terminal burned off- they get pretty crappy inside sometimes. usually if its a broken lead, it will be near the mounting ends of one of the cable tracks- seems the wires always slide a little more there than elsewhere.

a kinda common failure in the older spindles is the pulse generator- sucks as the only thing that goes wrong is the magnet gets weak...most the ones we had get weak had trouble starting or running at low speeds, when they go out the motor will squeal at low rotation speed then trip with a al-12 same as if it had a short...sometimes you can get a few more months out of them by closing the gap down to just a few thou, long as it dont rub on the rotor...

[MIYANO vs FANUC]

Thanks for the reply!

I did pull the top off the spindle motor (red top) and removed the fan so I could look at the encoder. It was clean and I did set the gap @ .006" Is that close enough?

Everything looked fine in the junction box but I'll check the resistance. It always starts and has good torque.

Just thought od something... The load meter pegs 100% when stopping the spindle but it doesn't use the regen brake, it coasts down to around 1000rpm then brakes to a stop with an alarm 27.

I still wonder what the 1000 alarm is all about. Could I find it in the ladder?

I probably should check the cable between the SPM and the motor too.

Any more ideas on what I should do are welcome!

[tc429]

what control are you running? weve got mostly 0-C/D, but with the newer drives, guessing you might have a 18 or 21 or 0-i...

if you do have a zero-c or d control(think it CAN be used in this format in newer controls, but theres easier ways) :

as for the 1000 alarm, look thru the ladder for a 'disp' block instruction with a whole bunch of numbers listed below it...if no numbers on the crt, try your paper copy of the ladder, some controls wont show on the screen, but list in the printout...the block with have several parameters,will kinda give example/tell what they mean:
160
10
D300
would mean it stores 160 pieces of data, and splits 10 per message, message calls start at d300.0 (16 means the calls end at d301.7) note the 'message number' is 4 digits, counted as one piece of data, but text is pairs of digits- each 4 digit chunk is counted as 2 pieces of data.

ill throw some bullship numpers like the table shows:
000 1001 6566 6768
003 6970 7172 7374
006 7576 7778 7980
009 8182 1000 8384
012 8586 ...............

in that case, alarm 1000 would be called by d300.1 (as d300.0 calls alarm 1001) any text is just a conversion, 65=a thru 90=z, 32= space. the above alarms would read 1001 abcdefghijklmnopqr, and 1000 stuv....was a really stupid way to enter text. many foreign machines(like yours) have no text, just the number...still called the same way

newer controls like 15 (think 16/18/21/0i) offer 'dispb' instruction, its easier- just a block with a number of available messages...the message data in these is stored in a message table as normal text, starting at address a0.0

one common thing with foreign equipment- actually US stuff too- is programmers like to get fancy and hide their logic using a bunch of compb, moveb instructions to move byte/bit parts 16 pcs at a time into the msg call address- programmers trying to justify their own existence in my opinion, as the typical maintenance guy might need help looking thru the extra crap to find what hes looking for...hopefully miyano didnt, but a lot of foreign machines do...our moriseikis are some of the worst.
we had some big makinos that actually had a separate pascal controller(mpc-2 they called it) just to hide the ladder from the user...we chopped that crap out, wrote a 30 page ladder/cassette that ran the machine just fine, but maintenance friendly. we had some mv-jr moris with a 298(!) page ladder I rewrote in under 30 pages with text- operators couldnt tell the difference...why so many oems insist on overcomplicating simple machines is beyond me...guess they like to sell service work.

sorry for writing a book here...anyways guessing you will find a serial drive address calling the 1000, and book will probably show 'al27' as the cause...but never know

[MIYANO vs FANUC]

It's fixed!

I pulled the spindle amplifier out and washed all the boards down. The colling fan on tbe IGBT was siezed up from years of coolant mist. Seriously gross but the fan now works. I checked the resistance of the serial cable with an Ohm meter, all was good. Replacing the pulse encoder didn't fix it so I went back through the books and read that my problem was most likely the precharge resistor on the board with the led display & button pad was the culprit. This is board a16b-220-0440/07b.

I had a spare so I replaced it but it didn't work, no brake. I chatted with a Fanuc guru and he tried to explain how to set the parameters for the spare board but I got confused and decided to just swap the eprom chip from the old to the new board.

TADA! It works.

Now I just need to probe around on the old board and compare resistor values and hopefully repair it.

Thanks for the help guys!