DMC104V linear Spindle Cooling System Fault

The control display gives the alarm message:

Error description 60 [PLC]
Cause of error:
Backfeed signal from spindle cooling missing

Effect: NC stop, feed interlocking
Corrective action:
Check spindle cooling, contact Technical Service, if required

The machine adopts the Heidenhain iTNC 530 system. When the I6(X42.7) end of its I/O has a low level, the alarm will be given, indicating that the cooling of the spindle and X axis is abnormal.
The specific fault is related to the cooling system. If the heat exchanger is configured (C1-09A21, the location is shown in Fig. 1), it can be repaired according to Fig. 2 and Fig. 3. The pressure switches C1-09S51 and C1-09S52 of the heat exchanger are respectively connected to terminals 2.2 and 20 of X13 of the machine tool electrical cabinet.

Fig. 1 Location of heat exchanger in DMC104V linear

Fig. 2 Schematic diagram of heat exchanger

Fig. 3 Alarm signal access point of heat exchanger

If the external Pfannenberg EB60WT FL cooling device (Fig. 4) is configured, you can refer to fig. 5 and the following instructions for maintenance.

Fig. 4 Pfannenberg EB60WT FL cooling device with cover plate removed

Fig. 5 Cooling system configured with external Pfannenberg EB60WT FL

At this time, the alarm signal connection is shown in Fig. 6, that is, the contacts of EB60WT FL alarm relay C1-09A21 are connected between terminals 2.2 and 20 of X13.

Fig. 6 Alarm signal connection when configuring Pfannenberg EB60WT FL


Repair of Machining Centers with Vickers A2100 Control

A (cromatic) 2100 CNC control (Fig.1) is a product of Vickers company in the United States in the late 1990s, which is a so-called PC based CNC system. Its hardware is mainly composed of the left main board (usually Intel 486 general-purpose computer motherboard with a 500M hard disk, which is responsible for the system software and hardware operation, called Workstation Board), the right main board (also a general computer motherboard, for real-time processing, called Realtime Board), Bridge Board (which is the interface of the boards and cards, the main control chip is an Intel MCS 51 series single-chip microcomputer), Servo Board (the interface between the system and the servo module of each axis, which is also a 51 series single chip computer application system), I/O Board (programmable logic application circuit, I/O interface of machine tool), etc. The software is an application program loaded automatically on Windows NT platform when startup.



(1) Appearance of control                                     (2) System boards

Fig. 1: Vickers A2100 CNC system


Given based on PC, the repair will be more convenient and the cost can be reduced. For example, the hard disk can be replaced by a general hard disk and hung to the desktop for backup or check. However, the slot connection between the boards and cards is easy to cause poor contact. During maintenance, the pull-out should be reduced and the electrical cabinet should be kept clean. If necessary, the electrical cabinet can be separated from the bed to reduce vibration, and the electrical cabinet can be equipped with air conditioning to maintain the stable and suitable temperature.

The following is a brief description of the repair methods and examples, of course only for those A2100 machine tools.



1 System troubleshooting

This category includes system board, hard disk, display, serial communication, etc. Many faults are caused by poor contact of connectors. It is necessary to judge calmly during maintenance, and do not dismantle and replace at will. It should be noted that the system hard disk should be backed up as soon as possible, otherwise it will be difficult to recover in case of hardware and software failure. The backup method is the same as that of general computer hard disk, but great care should be taken. The setting and initialization of hard disk partition, the selection of operating system and backup software are very important. The wrong backup method will make the original normal hard disk inaccessible.


Ex. 1 A SABRE1250 screen did not display, inspection showed that the relevant connectors were normal. Checked to find that FU5 was burnt out. After replacement, powered on, the fault was still there, and FU5 was burnt out again. According to the circuit diagram, the fan of Y-axis servo module was found to be damaged and the machine tool returned to normal after replacement.


Ex. 2 A SABRE1000 always stopped at the following text line when starting up:

AMIBIOS (C) 1999, American Megatrends, Inc.

BIOS Version: 1.00.05.AZ0

016384 KB

Please Wait…

Keyboard……Not Detected

After eliminating the possible bad contact of the hard disk, replaced the disk with a backup, and the fault disappeared. Another example also stayed on the page, but there was a more line of text: Drive not ready error. Here “drive” means hard disk drive, not floppy disk drive even servo drive module. If this is well understood here, it is easy to solve the problem. Pulled out and pushed in the hard disk connectors again, the boot became normal.


Ex. 3 When a SABRE1000 started to boot windows (in the startup screen, it is normally Vickers logo with light blue and green background), the display was out of order. In the previous text screen, the date and time in CMOS were hinted not set, so it was suspected that the CMOS parameters of the left motherboard were missing. Replaced the button cell battery (had been low) and cleared corrupted CMOS data. Connected an external keyboard (cable lengthened) to set the CMOS parameters following a backup. Restarted the control and the machine tool returned to normal.


Ex. 4 The 8004d red alarm appeared in a SABRE1000, which means that the workstation message service program cannot receive the response signal from the real-time processing module. Pressed the “Reset system” key, the system became normal after restart. However, when the magazine returned to the reference point, the alarm appeared again. Turned off the machine and separated the right main board and plugged it in again and the fault disappeared.


Ex. 5 when a SABRE1000 booted, 8004e red alarm appeared, which means that the power supply of control is abnormal, so as to make the real-time processor reset and restart. Cleaned the connectors to make the contact reliable and tried to reset, switch on and off for many times. But the fault persisted. In order to facilitate repair, replaced the power supply module directly and the machine restored to normal immediately.

The power supply module can be replaced by ordinary computer power supply. If similar products cannot be found, just buy a new one on the market and modify the connectors slightly.


Ex. 6 The alarm page appeared after a SABRE1000 was turned on:

Name                                Part Number                             #                         Result

Bridge Board [WS]              3-542-1260A+                         1                         Failed

Bridge Board [RT]                                                                0                         Unknown

Conv. Servo                                                                        0                         Unknown

High Density I/O                                                                 0                         Unknown

Pendant/Panel [MSL]                                                           0                         Unknown

Realtime Processor                                                              0                         Unknown

Keypad/Touch Screen         3-542-1223A.B                         1                         Passed

Workstation Processor        3-424-2123A                            1                         Passed

Obviously, the system self-diagnosis found that the bridge board was abnormal. Because the servo, I/O, etc. at the back were connected with the bridge board, so they could not be identified (unknown). In order to determine the fault, replaced the bridge board with that of a VMC1000C which worked normally, and the alarm page changed to:

Name Part                         Number                                    #                                Result

Bridge Board [RT]                                                                0                                Unknown

Conv. Servo                                                                        0                                Unknown

High Density I/O                                                                 0                                Unknown

Pendant/Panel [MSL]                                                           0                                Unknown

Realtime Processor                                                              0                                Unknown

Bridge Board [WS]              Unknown C                               1                                Passed

Keypad/Touchscreen          3-542-1223A.B                         1                                Passed

Workstation Processor        3-424-2123A                            1                                Passed

“Failed” disappeared in WS side of bridge board, and “unknown C” was caused by mismatch. But after a while, that “failed” appeared in the check again. There seemed poor contact between the left main board and the bridge board. And the bridge board of Sabre1000 could work normally when installed on VMC1000C. This further confirmed the judgement.

The original boards of the machine tool were restored and the connector between the left main board and bridge board was carefully cleaned, and the machine could now run normally.


Ex. 7 A SABRE1000 could not transfer program in DNC system, so pulled out DNC system plug from 25-hole RS232 socket on side plate of machine tool, connected a laptop computer with it directly, and tried with the communication software such as MTTTY and TERMINAL (empirical parameters: 19200/8/1). But still could not receive and send data.

Before turning to the left main board 9-pin COM1 socket and try again, it was found that one of the pins had been bent, and everything was normal after correction.


Ex. 8 The serial communication circuit of SABRE1250 motherboard was burnt out, but other functions were normal. Tried to transfer program through floppy drive, however, the floppy drive was always inaccessible at any security level (Operator/Setup/Service/Worldclass). Entered CMOS setup through an external keyboard and found the floppy drive had been disabled. Corrected it, OK.


Ex. 9 The touch screen of a SABRE1000 was in confusion. When clicked on a place with finger, the cursor arrow appeared at a symmetrical point about the center of the screen. It was normal after correction this way: hold down the CTRL and ALT keys simultaneously, and click anywhere on the screen with a finger, and a target icon will appear on the upper left corner of the screen. Click it, then it will appear in the lower right corner, and click again to finish. Another example was that the central part of the screen had no response, and so many operations could not function. Replaced it with spare to make it return to normal.



2 Repair of servo fault

A2100 system gives a more detailed diagnosis and treatment of servo fault. It is difficult to solve the problem only by replacing the module.

In addition, readers interested in chip level maintenance of servo module can refer to the following post:

Three examples of Kollmorgen VFS5 servo repair

Ex. 10 A VMC1000C could not be powered on, and the following alarm appeared:

Alarm 40-239

Machine Not Ready

Cause: The machine has not reported a ready condition within the time limit specified by the Machine Ready configuration item.

Remedy: Check MTB machine ready conditions.

While checking the electrical cabinet, it was found that QF5 relay (I/O 24V power supply) tripped, and the alarm was eliminated after it was turned on.

Another time, the same phenomenon occurred, but QF5 did not trip. From I/O state on display knew CR_AXES_RDY (I00:00/32) signal missing. After measuring the voltage of relay pins 1604, 1704 and 506 of servo modules to ground, it was found that the relay pin of X-axis module was poorly soldered and it pulled down the voltage of CR_ AXES_ RDY. The machine was normal after re-soldering.


Ex. 11 When the X-axis of a SABRE1000 moved rapidly, there was a lot of noise in the working cabin, sometimes it disappeared and sometimes it occurred, even when it was not moving.

There is a shock absorber composed of 4 pieces of iron cake about Φ 60×10 connected to the tail of X-axis lead screw. Some looseness was found in the inspection. Dismantled the shock absorber, cleaned it, installed it again and adjusted it properly to eliminate the noise.

If it still can’t be eliminated, can adjust the STABILITY potentiometer on the X-axis module, or replace the DV5 servo board of the system, or try to exchange the X-axis and Y-axis servo modules.


Ex. 12 A SABRE1000 sometimes lost power accidentally. In serious cases, It did even when not moving or machining. The following alarm would appear on the screen:

Alarm 39-41

“Axis ready” signal missing

Cause: servo drive ready signal missing

Input CR_AXES_RDY (I00:00/32) is low level


Kollmorgen BDS4 drive


Could not measure CR_AXES_RDY related voltages due to random failure, so observed the machine tool for a long time while it was powered on but not machining. Later, we saw that with a sound like “Kong”, the CONTROL VOLTS indicator of Z-axis module went out, and then the contactor M20 controlling the 3-phase AC power supply of servo was disconnected. After a few seconds, the UNDERVOLTS light of each servo module was on, and then the CONTROL VOLTS indicator of Z-axis module remained off.

Seems that the interruption of the signal CR_AXES_RDY was caused by the poor contact (C3 socket and plug) of 110V AC power supply of Z-axis module or the module itself. Exchanged the C3 plugs of Z-axis and A-axis modules to change the contact state of the connectors, and the machine became normal.


Ex. 13 A SABRE1000 had the following alarm:

Alarm 39-81

Spindle not responding

Cause: A2100 system has given the spindle enable signal M1SPDL_ENBL, but did not receive the ready signal M1SPDL_READY fed back from KOLLMORGEN spindle driver.

This is still the case after the system sends the spindle reset signal.

Remedy: Check whether the “Active” indicator of spindle drive is on, and confirm that the “Fault” indicator is not on.

Check whether the spindle drive is overloaded (OVL1, OVL2).

Severity 30: Deceleration stop, data reset

It was found that the “Active” light is on, but the “Fault” light is not on, which indicates that the spindle module is normal, but the signal is not sent to the system when it is ready. Took down the plug connecting the C8 socket of the spindle module, and the plug connecting the other end to the Output Board. Plugged them back after changing the position, and the machine was normal.


Ex. 14 A SABRE1000 had the following alarm:

Alarm 39-29

The spindle does not reach the specified speed

Cause: The spindle does not reach the programmed speed within the specified time, which may be due to:

(1) The measured speed exceeds ±10% of the programmed speed

(2) No CR_SPDL_SPD signal from spindle servo module

Remedy: Observe whether the spindle is rotating, if not, check the fuse of the spindle circuit and whether the contact is poor. Observe the LED indicators of the servo module. Check whether there is poor contact from driver to CR_SPDL_SPD signal terminal of A2100 I/O card. In MDI mode, input M3S1000, use the servo display page to monitor the actual spindle speed, and compare with the programmed speed. Then observe with M4S1000. It may be necessary to adjust the spindle offset and balance and check the spindle drive configuration data.

Severity 40

Checked I/O status and found CR_SPDL_SPD signal was on.

When the programmed speed was 1000, the measured speed was less than 900, Reduced the open-loop gain, the speed was increased to the required range, and the alarm was eliminated. (on Servo Setup page, in Gain column, check Open Loop Settings option to adjust the value in the Rpm/Volt box.)


Ex. 15 A SABRE1000 had the following alarm:

Alarm 39-31

Spindle stop error

Cause: The spindle did not stop within the allowed time.

Input signals CR_ZERO_SPD and CR_SPDL_SPD should be all high.

Remedy: Check the connection of spindle driver. Check for spindle offset. Check the “Spindle and rigid tapping” option and program M3S0 in MDI mode to check the spindle offset.

Use A2100 system to balance the spindle. It may be necessary to use Motion Link software to check the drive parameters, which requires special equipment. Check the regeneration resistance of the spindle. Turn off all power before checking, because there is high voltage inside.

Close the “Spindle Setup” page before recommissioning.

Severity 40

Checked I/O status and found CR_ZERO_SPD was low level.

Under Configuration/Machine Application/Run-off Service, checked the “Spindle and Rigid Tapping Setup” option, and run M3S0 in MDI mode. Observed the spindle offset, and found that the spindle rotated at a speed of 10rpm. Appropriately increasing the value of Balance Offset Open (from +5000 to +5001 in this example) eliminated the spindle offset. However, when executing M3S1000, found that the speed increased a lot. At this time, adjusted the Open Loop Gain (in this case, to +1050 rpm/v) to make the spindle speed normal.


Ex. 16 An alarm came seconds after the power button of a SABRE1000 was pushed down:

Alarm 39-6

Axis drives not enabled

See cause (1)


(1) Drive ready signal CR_AXES_RDY(I00:00/32) not received.


Severity 10: Machine powered off

Sometimes the power could be on, but after returning to the reference, moving the axes or rotating the spindle for a period of time, the alarm suddenly appeared. And sometimes other alarms successively appeared, such as “Spindle does not reach the specified speed”, “Spindle lubrication alarm”, “Spindle stop error”, “Spindle does not respond”, etc.

From the phenomenon the loss of the signal CR_AXES_RDY seems to be caused by the fault of the spindle module, which makes the three-phase AC power contactor M20 supplied to the servo power module open, so that the relays of X, Y and Z modules are disconnected. It may also be caused by poor contact.

First, checked the spindle module, and exchanged it with that on another SABRE1000. But both could work right.

No attempt was made to remove the fault one by one according to the above alarms. Checked all connectors, wiring, CR relays, etc., and no problem was found.

After a long time of exploration, it was found that after the spindle rotated for a period of time, the voltage of servo DC bus decreased gradually, especially when the speed was high (such as 2000rpm), and then the M20 contactor was broken. Thus, replaced the servo power module, and the machine tool became normal after replacement.


Ex. 17 The A-axis of a SABRE1000 had fault, it was necessary to wait for spare part. The A-axis can be easily sealed with the following method: press the setting key on the manual handle to activate the machine setting mode, and then select “REM AXIS” and “REM” to select the A-axis. At this time, the coordinate of the A-axis disappears from the screen. Select “PRES” to restore. Press “==>” and “ESC” to switch in the menu. Refer to the machine tool operation manual for the shielding method of each axis.



3 Repair of spindle lubrication fault

The main spindle is usually lubricated by VIP, VIP2 and other lubrication controller systems of Dropsa company of Italy, and other lubrication devices or methods are also used. But in case of failure, it is through I/O port line I00:00/33 (PS_LUBE_ALARM) to inform the CNC system. Therefore, when the spindle lubrication alarm appears on the operation screen, as long as the spindle lubrication device is checked, the problem can be solved quickly.


Ex. 18 The following alarm appeared on a SABRE1000 screen for many times:

Alarm 39-82

Spindle lubrication alarm

Cause: Dropsa spindle lubrication device alarm

Input PS_LUBE_ALARM (I00:00/33) becomes low level

Remedy: If it is VIP lubrication controller, check the status of LED indicators of the controller

Pump On  yellow = pumping oil

Heathy  flashing green indicates that the system is operating normally

Pump Alarm  red = alarm status. Check whether the oil pressure is too low or the filter is blocked when pumping oil.

Low Level  red = low lubricating oil level

Air Low  red = air pressure below lower limit

Air High  red = air pressure above upper limit

The nominal pressure is 2 bar. Press the red reset button to restart the oil pumping process.

If it is Dropsa advanced lubrication timer (VIP2), check the alarm diagnostic code on the display panel:

A-01 No parameters set

A-02 Oil level too low

A-03 Oil pipeline blocked, resulting in oil pressure cannot be released

A-04 Pumping is abnormal, oil pressure cannot be established

A-05 Air pressure too high

A-06 Air pressure too low

A-07 Abnormal triggering, unable to start pumping cycle

A-08 Servo motor cannot adjust air pressure

Refer to the manufacturer’s documentation for more information.

The alarm information clearly shows the maintenance method. In this case, VIP lubrication controller was used. The Pump Alarm light was on and could not be eliminated by pressing the red reset button for many times. After cleaning the oil tank, pump and filter, the alarm was gone.


Ex. 19 When the spindle of another SABRE1000 rotated, No. 39-82 (see Ex. 18) alarm appeared on the screen. This machine used VIP2 lubrication controller, and the alarm number was A-07.

Looking at the VIP2 technical document, we know that A-07 alarm will be generated in the step (SEP) mode. Changed parameter 1 from “3” to “1”, or to single line injection mode 1 (PS1), the alarm disappeared.



4 Precision adjustment

When the accuracy of the machine tool is poor or the servo shaft bearing is replaced, the machine tool needs to be detected and adjusted by corresponding instruments, such as Renishaw ballbar, laser dual frequency interferometer, etc. The detection and adjustment method of A2100 system machine tool is basically the same as that of other machine tools. For details, please refer to the documents of corresponding detection instruments. After detection, the modification of backlash and pitch compensation parameters is completed in the System Configuration/Error Compensations/Bi-directional Axis Comp page.



5 Repair of other faults

This involves tool changing, electric cabinet circuit, etc. The last example shows that repair sometimes goes through a lot of repetition, but as long as the principle and thinking are clear, and the diagnosis and maintenance resources provided by A2100 system are well used, the most difficult problems can be solved.


Ex. 20 A SABRE1000 gave an alarm, which means that the system needs 10# tool, but the tool magazine turned to 9# position, and the actual observation shows that the current tool position was 12#. According to the operator’s reflection and some observation, found that the disorder was irregular and sometimes normal.

The machine tool adopts a hanging disc type tool magazine. The rotation of the tool magazine is driven by an ordinary single-phase motor and counted by a proximity switch. It was suspected that the proximity switch was faulty. During the inspection, it was found that the tool change of the machine tool had become normal. It is estimated that the proximity switch was dirty or loose.


Ex. 21 A SABRE1000 gave an alarm:

Alarm 39-28

Draw bar not clamped

Cause: the spindle draw bar is not clamped.

Remedy: check the air pressure supplied for the draw bar, and test the action of it with the button Draw Bar Unclamp on the manual pendent.

Confirm that the door of the working cabin is open, and select the Tool Setup function on the pendent. The correct solenoid/proximity switch signal is as follows:

SL_TL_CLMP  high when the draw bar clamped, and low when loose

SL_TL_UNCL  low when the draw bar clamped, and high when loose

PR_TL_CLMP  high when the draw bar clamped, and low when loose

PR_TL_UNCL  low when the draw bar clamped, and high when loose

Check the compressed air first, during which It was found that the proximity switch PR_TL_UNCL did not function. Moved it up about 0.5mm, and the machine became normal.


Ex. 22 A VMC1000C had the following alarm:

Alarm 44-43

Unable to communicate with unused channel

Cause: feedback signal of unused servo channel cannot be detected. This is usually because the encoder in the channel is not configured or installed.

Remedy: install loopback plugs in all unused servo channels.

During the repair, it was found that the servo board and I/O board were wrongly inserted, and the alarm disappeared after the exchange.


Ex.23 The QF16 jumped off when a VMC1000C moved Z axis. According to the circuit diagram (omitted), QF16 is the 110VAC power supply protection switch for Z axis brake. 110VAC is applied to both ends of the brake coil after full bridge rectification, and the protection diodes V1 and V2 are connected in series with the coil in parallel.

It was found that a solder joint of the brake coil had worn through the isolating sleeve and touched the metal end cover inside the motor, causing a short circuit. Re-welded, improved the isolation and replaced the damaged V2 with a 24V voltage stabilizing tube BZX85C24 (1N4749).

It is worth noting that before dismantling the Z axis motor, the spindle must be supported by wood blocks or other things, otherwise the spindle WILL slide fast to the workbench after dismantling the motor!

First, disconnect the Z axis motor from the lead screw for debugging. Note that the motor will ROTATE or JUMP when adjusting the position of the revolver. Hold it carefully by hand! Adjust it bit by bit and fix it after adjustment. (No. 44-45 alarm will appear during adjustment, meaning Z axis servo failure, which will be automatically eliminated after adjustment.)

Installed the adjusted motor back to the Z axis. The Z axis could now move normally, but when it returned to the reference point, the following alarm appeared:

Alarm 44-6

Z axis grid alignment error

Cause: Z axis grid alignment has been completed, but no mark pulse is found.

Remedy: check encoder, wiring and mark pulse settings. The mark pulse setting is in the axis data of the servo configuration. Index Invert A and Index Invert B should be set correctly, otherwise the mark pulse is not easy to find.

On the Configuration/Machine Application/Z AXIS page, adjusted the value of Move To Align Zone (generally between -4 and 4). Finally, when it was adjusted to -3, the alarm was eliminated.

After properly adjusting the Z axis limit, on the Configuration/Machine Application/AUTO/MAN TL CNTL Z AXIS page, modified the value of TL Store 1 Auto TL Chg Pos item (this example was finally changed to +564.3000mm), executed T1M6, T2M6 and M30 commands under MDI to adjust the tool change point.

Now the machine run normally.


Ex. 24 After a SABRE1000 spindle dropped down and crashed on the machining part, all three driving belts were broken, and the spindle orientation needed to be adjusted after replacing the belts.

It needs to be adjusted separately in automatic mode and MDI mode.

Through adjusting the belts and repeated tests, the spindle orientation adjustment under the automatic mode was performed. When adjusting, put the tool magazine disc close to the spindle. On the Mchine Application/Runoff-Service page, Tool Changer Setup, check Spindle TC Posn, and slowly lower the spindle with the pendent to observe the orientation angle value.

The VFS5 spindle module can also be adjusted by serial communication with an external laptop:

1 Power on, run MotionLink, baud rate automatically set, after successful serial connection the = = > prompt appears

2 B↙  (single arrow appears)

3 LOCK■OFF↙  (open parameter, ■ denotes space, the same below)

4 2100↙

5 P■P1↙  (see the actual position of the current spindle)

6 Turn the spindle to the correct orientation position by hand and record the value (P1)

7 P1■ (the value of P1)↙

8 RUN↙

9 LOCK■ON↙  (close parameter)

The orientation adjustment in MDI mode can be completed by changing the value of the Spindle Orient Position item on the Configure/Machine Application/Spindle Range1 page.


Ex. 25 When a SABRE1000 spindle returned the tool to magazine, an alarm appeared:

Alarm 39-76

Spindle orientation error during tool change

Cause: the spindle cannot be oriented to the tool change position

The spindle driver did not set the input signal CR_OR_Comp. The software has made five attempts to orient the spindle driver correctly.

Remedy: correct equipment status:

Input CR_OR_COMP set

If it is a high torque spindle, then

PR_C_TPA needs to be set when M_ACT_DRUM = 1

PR_C_TPA needs to be reset when M_ACT_DRUM = 2

Output CR_SPDL_ORN or CR_SPDL_ORN2 set (not at the same time)

Note: CR_SPDL_ORN2 is only applicable to the spindle orientation of the second tool magazine (option) of ARROW series machine tools.

Check the spindle drive setting and adjust the offset in the system.

In case of high torque or 125000 RPM spindle, check the connection cables between PR_C_TPA switch to A2100 and spindle driver.

If the spindle driver is KOLLMORGEN VFS5, check, with MotionLink, feedback from PR_C_TPA to the spindle driver and view the value of the HOME signal. When the input is reset, it should be level 0, and when the input is set, it should be level 1.

First checked CR_OR_COMP signal, was reset state. Then observed if the relay corresponding to CR_OR_COMP signal acted when trying to change the tool, did not. Used a multimeter to measure the potential voltage at the O3 end of the spindle module (pin 43 of the C8 plug), during which it was found that CR14 acted. Therefore, after cleaning the C8 connector, the machine tool worked normally.


Ex. 26 A SABRE1000 had a red alarm, which means that the real-time processing has timed out. After disassembling the right main board, it started normally. However, when executing the program, it should grasp 1# tool, but it became 5#. The same phenomenon persisted for three times.

Due to the good repeatability of the fault, the factors such as the proximity switch of the tool magazine disc and the brake of the magazine motor can be eliminated.

The tool table was as follows:

Pressed the Data Reset key on the pendent, the “0” in the above table became “5”, and changed it to “1”, and the machine returned to normal.


Ex. 27 A SABRE1250 could not be powered on, and there was no alarm. Looking at the circuit diagram, the power on circuit is completed by the normal operation of ESR1 (K1, K2, K3 contacts are all closed) of the emergency stop relay (combination), and the abnormal operation of any point in the branches composed of the emergency stop switchs (main console, pendent, chip conveyor, etc.), the limit switch of each axis (X, Y, Z, etc.) and the door switches (left, right, etc.) will destroy the normal operation of ESR1. It makes the machine power unable to be maintained, which means that the machine power cannot be on.

It was judged that ESR1 was not working normally by the indicator light on ESR1. The potential of each point of the branch was measured point by point with a multimeter (the reference point is the T21 end of ESR1). It was quickly found that the switch of the left door was not closed. The accumulated aluminum chips were stuck in the crack of the door slightly and deformed it. After cleaning, the machine tool returned to normal.


Ex. 28 A VMC1000C had the following alarm:

Alarm 42-11

APC Start Fail

Cause: This is an internal error. The Programmable Controller failed to start because of the following error:

Hardware Configuration Table Error.

Returned by: cr_build_hwct.

Cause: A board with an unrecognized board ID is plugged into a rack.

Remedy: Turn the power off and on again. If the problem persists, try reinstalling the system. Call your service representative.

Replaced the hard disk with the backup, and the fault still existed.

Pulled out and pushed in the right main board, and the alarm changed to:

Alarm 40-33

Output/Fuse Failure in APC Rack 0

Cause: A fault condition was detected in one or more of the output devices in APC Rack 0.

Remedy: Check for current overload, excessive temperature or the external power source. If possible, correct the problem and restart the control. If the condition persists call your service representative.

Exchanged the right main board with that of a SABRE1000 to test the machine. The result showed that the machine was normal, but the SABRE1000 gave the same alarm. Later replaced the right main board with a new one, the machine returned to normal.


Ex. 29 A SABRE1000 had a serious fault. After testing and repairing all six servo modules (X, Y, Z, A, spindle and power supply), installed them again, checked all fuses and possible short circuit, and then tested the machine.

After the self-test of the system was completed, 40-33 alarm appeared (see the above example for the meaning). Found two copper foils were burnt out after checking the output board. One of them is the connection from +24V to the I/O board. Recovered them by wires. But the machine tool could not be powered on and there was no alarm.

Replaced the I/O board, 39-6 alarm came (see Ex. 16, Cause 1), checked the circuitry from +24V to the pin 32 of the plug J1 (CR_AXES_RDY), normal. After carefully observing the startup process, it was found that after pressing the startup key, the Drive Ready indicator of Y or Z axis module was on, while the X axis was not, and an alarm appeared after a few seconds.

The X and A modules were then sent for inspection and maintenance, and then installed back into the machine tool. The Y, Z and A axes now could be moved, but the X still could not. Each axis could move normally after X module being sent for repair again.

When the magazine returned to zero, it was interrupted by a proximity switch alarm. Four proximity switches related to tool change were replaced, and the magazine return finished.

After replacing the four coil protection diodes and sending the spindle module for repair again, the machine tool basically returned to normal. After the spindle orientation adjustment, the machine tool was fully available.



The machine tools of Vickers A2100 system have been used for more than 20 years. If the failure rate is too high, the system can be replaced by others, such as FANUC 0iF, Siemens 828D, etc.

Repair of Control Power Unit CPS-150F of System YASNAC XRC

The control power supply unit (as shown in the figure) provides +5V and +24V DC power supply for I/O unit, power on unit and programming box through CN01, CN02 and CN03 sockets. The remote switch control function is provided through its CN04 socket. It through CN06 provides +5V power supply, status signals for system control board and receives control signals. Rated input voltage: 200/220VAC, allowable fluctuation range: -15%-+10% (170-242VAC), frequency: 50/60Hz±2Hz (48-62Hz). The output voltage is +5VDC/10A, +24VDC/4.0A. There are two green LED indicators, SOURCE and POWER ON, respectively indicating that the AC power input is normal and the DC power normal. Three red LEDs +5V, +24V and OHT respectively indicate +5V over-current or over-voltage alarm, +24V over-current alarm and internal power output silicon rectifier overheating alarm. In addition, when the temperature in the module reaches 65℃, pulse signal is sent to the system control board.

The circuit is mainly divided into three parts: auxiliary power supply, main power supply and fault detection circuit.

Auxiliary power supply

AC200/220V is input from CN05 socket. After rectification and filtering, it outputs about 310VDC voltage from full bridge RC2 to supply power to NMOS FETs M12 and Q11. Another AC branch is through transformer T4, diode full bridge, capacitor C55, etc. to generate about 5V reference voltage on resistor R69, which is sent to the fault detection circuit.

The self-excited oscillation circuit is composed of M12, 3-4, 1-2 coils of transformer T3, D25, C53, R48 and C50. Through 6-7 coil of T3, D23, D28 and filter capacitors, about 23VDC power supply voltage VS is obtained, and the SOURCE indicator is turned on. The power supply voltage gets VP power supply through thermal switch OHD3-100B, which supplies power to PWM control chip M51995AFP and comparator chip 393. Once the components RC3 and RC1 of +5V and +24V output circuit overheat, VP will be cut off. In addition, the +5V VCC power supply is obtained from the VS through a 3-terminal voltage regulator M1 to supply the fault detection circuit.

Main power supply

The +5V/10A and +24V/4.0A are respectively obtained from the secondary side of transformer T1 of main oscillation circuit through silicon diodes RC3, RC1 and capacitors. The frequency of oscillation pulse generated by PWM control chip M51995AFP is determined by the resistors and capacitors connected to its pins 11-13, which is about 189kHz. The feedback signal is provided by the voltage dividing circuit composed of R45, RV1 and other components, connected to the DET end, and through the resistor-capacitor circuit connected to the F/B end, and the voltage at DET is about 2.5V. The output Vout is applied to the grids of Q12 and Q11 through transformer T2, and coupled to the main power output side through (3,4-5,6) coils of transformer T1. At the same time, from the (1-2) coil of T1 the induced pulse signal is rectified and filtered and applied to the thyristor TH1 to make the resistor R18 nearly short-circuit.

The current inductors CT2 and CT1 detect the current changes in the output circuits of +5V and +24V main power supply respectively, and input them to the CLM+ end of M51995AFP through diodes and resistors, etc., so as to realize the so-called “pulse by pulse current limiting”. In addition, the +24V over-current signal is taken out from the resistor R68 and sent to the fault detection circuit.

Fault detection circuit

The +5V main power over-current detection circuit is composed of voltage comparators M3A, M3C and NAND gates M6B, M5B, etc. In case of over-current, M3A outputs high level, M3C and M5B output low level, and the +5V alarm light is on.

The +5V overvoltage detection circuit is composed of 6V voltage regulator 621, triodes L6 and M6, etc. When overvoltage occurs, 621 turns on, L6 and M6 turn on, the reverse input end of voltage comparator 393B turns to low level, the On/Off end of M51995AFP turns to high level, the chip stops oscillation, +5V turns to 0, and the +5V alarm light is also turned on.

The +24V overcurrent detection circuit is mainly composed of voltage comparators M3B, M3D and NAND gates M6C, M5C, etc. When the +24V main power supply is overcurrent, the signal taken out from the R68 makes the output of M3B become high level, and then causes the output of M5C to be low level, and the +24V alarm light is turned on.

The overheat detection circuit of main power supply is composed of thermal switch OHD3-100B, buffers M4F, M8C and NAND gates M6D, M5D, etc. When overheat occurs, the thermal switch will be disconnected and VP will be disconnected too. This will make the output of M5D low, and the overheat indicator OHT on.

Through a series of gate circuits such as M9D, M9C and M9B, resistors and capacitors, alarm signals make the reverse input of voltage comparator 393A into a high level, so that the On/Off end of M51995AFP becomes high and the chip stops.

The circuit composed of M9A, M6A, M5A and M7E, etc. is used to detect remote control signal, system control signal, etc. When the power supply is overhauled separately, the pin 1 of CN04 and the 14c and 14b pins of CN06 should be short circuited to signal ground to make the output of M7E high level and ensure the normal operation of each detection circuit.

The system control board receives various status signals of the power module through CN06, including the temperature switch signal in the module sensed by the thermal switch THS2.

Repair cases

Ex. 1 The SOURCE and +5V indicator lights were on when the power was turned on separately (the same below).

This shows that the auxiliary power supply VS was normal, but the +5V main power supply had overvoltage or overcurrent alarm. Whether it was overvoltage or overcurrent, the +5V power supply disappeared, and previous voltage value of each point could not be measured anymore. It seems difficult to judge which situation made the alarm light on. First checked whether the working conditions of M51995AFP were met. Measured the supply voltage of pin 20 or VCC end, that is VP, 46V. The recommended working voltage of the chip is 12-36V. It can be seen that although the SOURCE light was on, the VS was still too high. Checked the components of the self-excited oscillation circuit with a multimeter and found that the capacitor C53 had changed from 10μF to 0.135μF. Replaced it with a 47μF capacitor on hand, the VP dropped to 23V, and the whole power supply returned to normal.

Ex. 2 After power on, SOURCE and POWER ON lighted up normally, but after a period of time, POWER ON went out and +5V alarm light lighted up. The power on time of each test was different, which was irregular from a few seconds to a few hours.

From the phenomenon, it should be caused by the poor contact of the main power circuit. Careful inspection of the main circuit components showed no obvious poor contact or copper foil lead fracture. The power supply VP of M51995AFP was also normal and stable. When the oscilloscope was used to observe the waveform of the relevant pins, the tiny claw of the lead-out measuring clip picked up a pin. It turned out to be that the pin of M51995AFP was poor soldered. But because it was a SMD and small, it appeared OK while measured with a multimeter. Soldered and checked these pins again, the poor contact disappeared.

Ex. 3 After power on, only SOURCE light was on.

POWER ON is not on, indicating that the main power supply is not normal. However, if the main power supply is abnormal, +5V alarm light should be on, unless the output of buffer M7E is low level and the output of NAND gate M5B is locked to high level. The output of M7E was measured, 1.2V, which was really low level. The voltage (potential) of the pins 1, 2 and 3 of the front NAND gate M5A were 4.9V, 4.9V and 4V respectively, which obviously did not conform to the logic relationship of NAND gate. After replacing the four NAND gate chip M5, still only one light was on, and the voltage of pins 1, 2 and 3 became 0V, 3.3-4.8V (unstable) and 4.9V respectively. Continued to check the voltage of each point from back to front while considering the logical relationship. It was found finally that the short wire between 14b and 14c of CN06 socket (convenient for offline maintenance) have been disconnected or touched other solder joint. After solving the problem, the power supply became normal.


[1] YASNAC XRC Instructions for European Standard, Manual No. RE-CTO-A205, Yaskawa

[2] Datasheet: M51995AP/AFP Switching Regulator Control, REJ03D0835-0300, Rev.3.00, Jun 06, 2008, Renesas

Using Renishaw ballbar to check and improve the accuracy of CNC machine tools

New CNC machine tools, such as milling centers and lathes, must be tested and adjusted according to the agreement after installation and commissioning, which is a necessary procedure for qualified acceptance. Regular or irregular testing has been made, if the accuracy loss is serious, the machine needs to be overhauled, transformed or scrapped. This has become one of the contents of many enterprises’ quality assurance system. In addition, the accuracy of machine tools should be checked and improved after serious fault to ensure machining quality.
It is usually difficult to detect, adjust, restore or even improve the accuracy of machine tools, which requires a lot of experience. But Renishaw’s ballbar instrument makes it easy.
QC10 dynamic ballbar system is the latest product of Renishaw precision calibrator for CNC machine tools. It is a precision instrument to evaluate the dynamic contour precision of machine tools. This technology has been quickly adopted by international machine tool inspection standards, such as ISO230 and ANSI B5.54.
The instrument consists of a precision rod with a precision ball at one end and a dynamic positioning precision ball seat at the other end. The rod itself is a linear displacement sensor. It is precisely installed between the worktable and the spindle of the machining center or between the tool rest and the spindle of the lathe. Then the machine tool performs circular interpolation motion, and the deviation error between the interpolation trajectory and the best fit circle is measured by the sensor. The instrument also has optional accessories for radius calibration to calibrate the length of the rod, so as to ensure the high accuracy of the instrument.

Fig. 1 Installation of ballbar instrument on milling and lathe machine

This fast and simple detection can provide the data of various errors of machine tool, analyze them automatically, and give various error optimization tables. It includes:
Overall contour non-circularity error
Servo gain mismatch error
Periodic error of ballscrew
Creeping error
Machine tool vibration
Perpendicularity Error
Grating ruler/encoder error
Clearance and other commutation errors
Straightness error
Reversal spikes.
We have applied this type of ballbar instrument to the inspection and maintenance of more than 20 imported and domestic machining centers and lathes with different controls and different precision, and achieved some results. Here are two examples.
Ex. 1: The X-axis of a German FP5CC/T 5-axis machining center sometimes made abnormal noise. The XY plane was detected by the ballbar instrument, and the results are shown in Fig. 2.

Fig. 2 Test results of ballbar instrument when X-axis of FP5CCT 5-axis machining center had abnormal sound

It is clear from the test results that the circularity is poor, which is 121.2μm. According to the experience, when this kind of high-precision full closed-loop machine is normally used, its accuracy should be in the order of micron. The main reason is that the reversal spikes of the X-axis are larger, and the positive and negative spikes are 98.0μm and 103.1μm, respectively, the error of which accounts for 47% of the total roundness error.
The analysis of the reasons for the reversal spike error of the testing software is as follows:
• An inadequate amount of torque has been applied by the axis drive motor at the axis reversal point causing it to stick momentarily at the reversal point, as the frictional forces change direction.
• The servo response time of the machine is inadequate on backlash compensation. This means that the machine is unable to compensate for the backlash in time; causing the axis to stop while the slack caused by the backlash is being taken up.
• Servo response at the crossover point is poor, causing a short delay between the axis stopping movement in one direction and starting movement in the other.
Recommended measures:
• If your machine has a controller with a spike removal capability use this facility to limit the effects of reversal spikes while using the machine.
• Attempt to identify a machine feedrate on the machine at which the effects of reversal spikes are minimized by performing a number of tests at various machine feedrates. Use the best feedrate identified for finishing cuts during circular interpolation.
Because the reversal spikes were too large and accompanied by X-axis abnormal sound, first checked whether the X-axis mechanical motion system was normal. It was found that the X-axis screw nut was partially damaged and loose, and could not be disassembled or adjusted. The machine was imported from Germany, and had been purchased for nearly ten years, so it was difficult to replace the screw nut pair. Inserted copper sheets with suitable shape and thickness into the gap between the screw nut pair while checking the effect with the ballbar instrument. The data was read, and the copper sheet thickness was corrected. After two or three repetitions the circularity had been reduced to 18.3μm, the reversal spikes X was positive 9.3μm, negative 13.4μm, and the abnormal sound disappeared.
Ex. 2 A domestic TH5632D vertical machining center was reflected by the operator for not milling round shapes well. The XY plane was detected by the ballbar instrument, and the result was shown in Fig. 3. The circularity is 91.3μm, but according to experience, FANUC 0 series semi-closed loop machine tools like this kind usually have a circularity between ten to thirty microns when leaving the factory. The main reason for the poor roundness is unequal backlash in Y-axis, which is positive 48.7μm and negative 9μm, which accounts for 33% of the total roundness error.

Fig. 3 Testing results of ballbar instrument when circle milled by TH5632D vertical machining center was not round enough

The cause of unequal backlash is as follows:
• An excessive amount of windup in the ballscrew is causing a backlash effect. This depends on the position of the axis relative to the driven end of the ballscrew, which causes the unequal backlash type plot. Machines with backlash compensation may be adjusted to ‘average’ this difference, resulting in the opposing steps. The windup may be caused by a stiff or worn ballscrew, nut or guideway. This type of backlash is common in tests involving vertical movement, due to counterbalancing effects.
Recommended measures:
• Remove any backlash compensation which is applied to the machine. This will allow the extent of the problem to be assessed.
• Check the ballscrew and guideways of the machine for signs of wear. If necessary maintain or renew these components.
• If unequal backlash plots are the result of tests involving vertical machine movement, then the counterbalance is likely to be the source of the problem. Adjust the counterbalance system of the machine as required.
The Y-axis ballscrew was inspected, and it was found that the ballscrew nut was seriously corroded due to metal chips and coolant entering the protective sleeve. After replacement, the test results showed that the circularity was 28.8μm, the Y-axis unequal backlash was still the main factor, the positive and negative direction were -21.9μm and -22.6μm respectively, accounting for 34% of the total roundness error.
Adjusting the parameter of Y-axis backlash made Y-axis change from unequal backlash to negative backlash (Backlash-Negative), both positive and negative backlash errors were -10.1μm, which became a secondary factor. At this time, the circularity was lowered to 22.5μm, which met the machining requirements. Of course, the X-axis could be adjusted to further improve the roundness index.
QC10 ballbar instrument system is precise and compact, easy to carry and operate and fast to test. It is really a powerful tool for inspection, adjustment and maintenance of CNC machine tools.

An example of repair of the fault automatic restart

There appeared an abnormal situation of frequent automatic control restart in a SABRE500 machining center of Cincinnati company of the United States, the intervals became shorter, from months to minutes, even continuous. Due to its irregularity, repair was very difficult. Checked from NC, servo to I/O, still could not find the cause of the fault, the machine was in a state of ‘paralysis’.
The machine, with a GE-Fanuc 18i-M control has been used for about ten years. The circuit diagram of the system and power supply is shown in Fig.1 and Fig.2. (for convenience, kept the original style).


Fig.1 GE Fanuc 18i control connection diagram


Fig.2 Connection diagram of SABRE500 power supply (partly)


The power supply of the control is provided by +24V via CP1A. Firstly, checked the wiring and connector to eliminate the possibility of poor contact.
It was known that the 24V power supply module had been replaced for many times, and has a capacity with large margin. In order to further confirm, two of them were taken out and a piece of resistance wire (1kW) was used as load. The static load capacity was measured to be above 5~6A. However, the fuse FU9 of the machine tool for all the load of +24V power supply is only 3A, so it was considered that the power supply can basically meet the requirements.
To locate the fault, disconnected the load one by one to observe the system response. After repeated tests, it was found that the system would not restart when all plugs of VDU except CP1A (or CA55, CK2 on MDI, JD36A, JD1A, COP10A, JA41 and CP1B) were pulled out, but would as long as CP1B was plugged in, even when the remaining plugs (CE56, CE57, JA3, JD1A, JD1B) on 48/32 I/O board were pulled out. Therefore, I/O board was judged to be poor.
Exchanged the I/O boards with an ARROW500 machine (with the same model, A20B-2002-0520/02A), the failure was still the same, but the ARROW500 was still working normally, indicating that the board was normal.
So far, it was considered that the 24V power load of the system has been in a critical state. Although it does not exceed the load capacity range of the power supply (FU9 seemed to have not been fused and replaced), it sometimes causes system restart.
Since the 24V power supply has no obvious heavy load and feeds quite a few branches, decided not to check one by one any more, but to add additional power supply module to the control. For convenience of the connection made changes to the circuit shown in Fig.3. The +24V power supply of the control was disconnected from TB4-46, and TB4-46 was then connected to the 0V output terminal of the original 24V power supply through the new relay coil CRp, and its normally open contact controls the new power supply to the control. Pay attention to the joint of the 0V terminals of the two power supply modules, i.e., “common ground”, to ensure the correct closing of each circuit.
After renovation, the machine tool worked stably.


Fig.3 Connection diagram of power supply after renovation (partly)

Starting from Principle, Quickly Restore DMU125P 5-Axis Machining Center

A DMU125P 5-axis universal milling machine was purchased from Germany in 2004 at a cost of $430,000 . According to the operators, the machine has frequent faults in recent years, and its working efficiency is very low. It needs to be started many times or even half a day each time, and it will repeatedly give an alarm every few hours, you can do nothing but to restart; both the X-axis and Z-axis at G0 move slowly with abnormal noise; after machining, the tool on the spindle cannot be returned to the magazine, you solve it only by modifying the next program; there are obvious tool marks on the workpiece, etc.. It has been repaired many times before, but the problems still exist. It has become a headache for operators and workshop leaders.

In view of a lot of handover and repair of different departments, operators and servicemen, paid special attention to the principle rather than the experience this time.

First of all, tried to make the power-on process normal. It was found that the indicator light is off after the start of machine power. It is easy to see from the schematic diagram that as long as the machine power is normal, the indicator light should be on. It was found through inspection that the bulb is burnt out and normal after replacement.

The A7-00A31 safety relay must work before the machine can be powered on normally (Fig. 1). However, it was observed that during the unsuccessful power on process, its two channel indicator lights were not lit. Therefore, checked its related circuits first, namely itself, the A7-00S31 safety switch, the emergency stop buttons of the control console and manual control box, as well as the connecting wire, etc.

Fig. 1 DMU125P emergency stop circuit (partial)


Since there was no spare part, made an exchange of A7-00A31 and F8-23A31 (tool loading, the model is the same, or PNOZ X3 24VDC 774 318). The machine went well on the start, but after that E220 alarm (“No mag. sw. > running time checkb.”, meaning, according to the German document Error List, that no counting pulse is in the magazine and the tool change is fault,) occurred, which indicates that A7-00A31 has problem. Through the console interface manually closed the small door of the tool magazine, made the spindle returned from the magazine, turned the revival table to the normal position. Shut down. Replaced back the two safety relays, restarted the machine, and the power-on became abnormal again (sometimes good or bad), but the magazine was normal. When the spare part arrived, replaced the old one, the machine power-on was fully restored to normal.

Now observe the frequent alarm during the machining process. The alarm is E02 (Lubrication pressure switch high), which means that the lubrication pressure switch is closed and the system receives the high level signal.

In the machine, the lubrication of each axis guide rail and lead screw, A, C axes and spindle vertical horizontal conversion rotating head bearings are all through the central lubrication device (VOGEL MKF1-KW2-20036+428), sent to 18 lubrication points through distributors.

When the lubricating device pumps to 30 bar, preset by pressure switch, the one-way valve of the distributor will open, and the grease will flow to each lubricating point. At this time, the system will let the lubrication device continues to pump oil for 5 seconds and then stops. After 5 minutes, check whether the pressure switch (I_AL_SCH_Rueckmeldung) is open, if still closed, which indicates that the system pipeline is blocked, the system gives E02 alarm.

The failure of central lubricating device, improper adjustment of pressure switch or blockage of pipeline will make the lubricating system abnormal. Too much pressure will cause nozzle crack and the alarm E01 (meaning the pressure switch is off, the pressure cannot be maintained) appear.

Without a good understanding of the above process, it is not easy to adjust the seemingly simple lubrication system. This is the case with this machine, the pressure switch had been adjusted to an inappropriate position, most of the pipelines had been blocked, the lubricating grease had caked and deteriorated. The machine gave an alarm and because it could not be cleared the machine was forced to shut down and restart. The process repeated every few hours. Since the moving parts were not effectively lubricated, abnormal noise occurred. Guess (by checking the machine parameters) that the speed of X and Z axes were restricted by the operator just because they caused much noise. It can be imagined that if it is not handled in time, the accuracy of the machine tool will be seriously lost and even the moving parts are damaged.

Thoroughly cleaned the pipeline including lubricating device, distributors, joints, etc. Refer to the lubricating device (VOGEL MKF1-KW2-10006) of an adjacent machine DMU80P, the main technical parameters are similar) adjusted the position of the pressure switch, and manually pressed down the lubricating pump contactor B1-01K71 (in the electrical cabinet, because the lubricating device has no manual oiling function) fill the pipeline with grease, and the lubrication alarm appeared no more. Then adjusted the parameters No.3103 and No.3303 of this machine (respectively the maximum PTP speed of the first axis and the third axis) made the X-axis and Z-axis G0 speed return to normal.

As for the problem that the spindle cannot return the tool, in order to observe the phenomenon, executed T0M6 in MDI mode and return the tool on the spindle back to the magazine, and everything after that became normal, it seems that there is no more chance of chaos. This phenomenon also happens from time to time: simple problem is made complicated.

The tool mark of the workpiece is caused by the loose of the main shaft pull rod and disappears after tightening.



[1] Deckel Maho Pfronten GmbH.DMU125P Circuit Diagram [EB/CD].

[2] Pilz GmbH & Co..PNOZ X3:Emergency Stop Relays, Safety Gate Monitors[EB/OL].

[3] 马力.安全继电器PNOZ X3 维修方法[J/OL].

[4] Deckel Maho Pfronten GmbH.DMU125P Users Manual [EB/CD].

Hard Disk Fault


A DMU80P 5-axis universal machining center, manufactured by the German company DMG in the late 1990s, recently had a hard disk failure. When it was started up, it prompted “Failure Fixed Disk 0” and “Operating system not found”. Then the cursor blinked, unable to complete the start-up process. After eliminating the possibility of poor contact, took out the hard disk and connected it with a desktop computer through an adapter card. The hard disk could not be accessed by many methods. When the shell was removed, found that the disc had been seriously scratched, and the magnetic head could not move out of the large area of scratches like rut, so it could only be regarded as scrap.

After contacting with the service office, in addition to the purchase and service fees of tens of thousands of Yuan of hard disk, it would take 3-4 months to arrive. The impact on production could be imagined, so we decided to try to repair it by ourselves.

The hard disk is a Hitachi DK224A-14, with a 2.5-inch parallel port and a capacity of 1.4G. Because it was not easy to buy, we used a 60G of an old laptop and connected it to the desktop through a LGH-IDE-K interface card (available online, RMB Yuan 10 to 20).

Through observation and analysis of manuals, floppy disks and the start-up process of a DMU80T 4-axis machining center nearby, which was built in about the same period, made some attempts and finally restored the machine to operation in the following ways:

Checked ten floppy disks of this machine, nine of which are system software, including DOS commands, Heidenhain Millplus numerical control application and some tools, and the other is machine specific data, such as IPLC, machine parameter, etc.

Run the INSTALL program on the first disk in the pure DOS environment on a desktop computer. It extracted the contents of nine floppy disks on the hard disk with prompts. Then it automatically run a CNCINST program, which prompts the serial port setting, transfer configuration and the host waiting for the remote PC to be ready. It could be seen that the installation is to install the system on the PC, then configure the application according to the specific machine tool, and then transfer it to the machine hard disk through the serial or network port. The hard disk of the machine tool must be able to start normally on the machine tool, have corresponding settings and be able to automatically or manually run the communication software. After the failure of proper attempt to this method, we decided to install and configure the hard disk manually, because the “installation” of INSTALL is actually decompression, and CNCINST is transfer.

Under the CNC_VERS\8321_331.00B\ROOT path after decompression, checked that the number of bytes of the COMMAND.COM file is 54619, and from this confirmed that its DOS version is 6.21. When ” HEIDENHAIN NUMERIC Service Menu” appeared in DMU80T startup, pressed the “S” and “7” keys, and run Fdisk at the DOS prompt to check the partition situation. Based on this, using DOS6.21 created a 50M primary partition C: and a 2G extended partition D: on the 60G hard disk, then formatted and installed DOS6.21 to C:.

The INSTALL program creates the CNC_VERS directory under the root of the hard disk, which contains 8301_331.00B, 8321_331.00B and UTIL subdirectories and a text file. The text file records the compressed files and attributes in each floppy disk. Under UTIL, there are system configuration, transfer and other tool softwares. 8301_331.00B is applicable to (DAX/VAR) analog and digital servo system, while DMU80P adopts (SDC/DDC) digital servo system and applies 8321_331.00B (for convenience, the following always refers to subdirectories or files under this directory if there is no special note).

Copied all files under the ROOT subdirectory except DOS system files COMMAND.COM, IO.SYS and MSDOS.SYS to the root. The contents of DOS subdirectory can be ignored because the hard disk was installed with complete DOS.

The UTIL subdirectory (8321_331.00B instead of CNC_VERS) contains various hardware self-test, communication and other tool softwares. For system startup, the entire subdirectory was copied to the root.

The SYSTEM subdirectory contains system files and copied them to the root.

The NUM subdirectory contains files such as Interactive Part Programming (IPP). Referring to the file configuration of DMU80T machine tool, created an OEM directory in C:\, which includes three subdirectories: CFG, LANGUAGE and MSG. Created LANGUAGE and MACROS subdirectories under the SYSTEM directory. Copied three .331 files under NUM\IPPM to OEM, and all .331 files under NUM\IPPM\LANGUAGE to OEM\LANGUAGE. The .331 files under NUM\UIMS were copied to SYSTEM, and all .331 files under NUM\UIMS\LANGUAGE were copied to SYSTEM\LANGUAGE.

By the way, many files have similar names, but the character(s) at the end or extensions are different, such as A830M*.331, A830T*.331 and U8300*.331. Different parts represent languages, such as D German, DK Danish, E Spanish, F French, GB English, I Italian, NL Dutch, P Portuguese, S Swedish, SF Finnish. Only the corresponding files need to be copied.

CM subdirectory contains the default machine parameter file, CNCINST and INSTDATA subdirectories contain files about installation, tool, technology, etc. OEM subdirectory is empty. They were left not considered.

The following describes the processing method of the files in the tenth floppy disk:

Copy the machine parameter .CM and temperature compensation .CT files to the SYSTEM subdirectory. If the machine parameters are lost, can run the STARTUP CM command to load at the DOS prompt. The .CT file is loaded in the menu of Millplus.

The DSK_HINW subdirectory contains the version information of IPLC, which is stored in the database files DMG_INFO.DBF and DMG_INFO.DBT in German. If you run the DMG_HINW.EXE program (self extracting through DMG_INFO.EXE) and enter the version number range as prompted, can generate corresponding text file. Ignore this information and copy the I213*.* files under the IPLC subdirectory to C:\OEM\LANGUAGE.

Run the 8321_331.00B\SYSTEM\EASYOPER.EXE program, and copy the macro programs self-extracted and those under the MACRO subdirectory in the 10th floppy disk together into the C:\SYSTEM\MACROS subdirectory.

The meaning of German Meldung and English message is similar. Copy the text files in the subdirectory of MELDUNG to C:\OEM\MSG.

The TECHNO subdirectory contains .MG (machining type table), .MA (material type table), .TE (technology table) and .TT (tool type table) files. Usually not used, ignore them.

The UK subdirectory contains the user softkey table file. Can copy it to C:\, and load it under the machine menu Control/Installation/User softkeys. Can also view and edit it in a text editor under the menu.

Through the process above, the machine was restored to work normally. However, if you use some unusual functions, such as online interaction programming, may cause problems. At this point, it can be solved by properly arranging the file location according to the error prompt. In fact, if make further analysis, as using decompiler tools, can understand its function and proper location, so that the configuration is complete and concise, but it takes time. In addition, if conditions permit, the installation through serial or network port should be well tried.

Alarm 120201 (Communication interrupted) & 120202 (Waiting for connection to NC)


Alarm 120201 (Communication interrupted) came randomly and frequently. Sometimes it turned to be 120202 (Waiting for connection to NC).


The information on the 840D diagnostic manual is as follows:
120201 Communication interrupted
The operator panel is connected with the NC and PLC via a serial bus. This alarm occurs if the communication to these components is faulty. Together with this alarm, all display values connected with NC/PLC become invalid. Such malfunctions are normal during startup of the controls (e.g. after reset).
Alarm display.
The alarm disappears automatically as soon as the error situation is terminated. If this alarm is pending permanently, this may have very different error causes (e.g. cable break, no startup of NC/PLC, incorrect address/baud rate configuring of one of the bus nodes etc.).
120202 Waiting for connection to NC
The operator panel is connected to the NC and PLC via a serial bus. This alarm occurs when the MMC is started up for the first time if the NC/PLC power-up is not yet complete, or if an error occurs during communication between these components. All display values associated with the NC/PLC are rendered invalid by this alarm. Such malfunctions are normal during startup of the controls (e.g. after reset).
Alarm display.
The alarm disappears automatically as soon as the error situation is terminated. If this alarm is pending permanently, this may have very different error causes (e.g. cable break, no startup of NC/PLC, incorrect address/baud rate configuring of one of the bus nodes etc.).

It can be seen that the failure of the machine is probably caused by the poor condition of the bus line between MMC and NC (purple) or connectors. After removing, re-inserting and fixing the plugs the alarm disappeared.
If it does not work later, can try cleaning the connectors or replacing the line.

Alarm 20000 (Channel 1 axis WZM reference cam not reached)


Alarm 20000 (Channel 1 axis WZM reference cam not reached)


The alarm can be generally understood as the tool magazine failed to return to reference point (for details see SINUMERIK 840D/840Di/810D/FM-NC Description of Functions Basic Machine,FB1,Reference Point Approach). It’s easy to find problems with a roughly check to the reference point device. In this case, the broken connection was found to the proximity switch S32.3. And the alarm disappeared after welding.

Poor tool change


Poor automatic tool change, with loud noise.


With method “grabbing empty tool” can observe relative position of tool changer and taper of spindle. Found the position of spindle was slightly lower. By adjusting the MD34090 parameter of Y axis (REFP_MOVE_DIST_CORR, in the menu Start-up/Machine data/Axis specific) can raise its position. In this case adjusted it from -96.403 to -93. After resetting NCK the tool change was basically normal.