Repair / Retrofit of Milling/Machining Center TRUMP B8AC

[Quebecois_Sti]

Hello everyone,

I am helping my friend regarding his milling he acquired, details below. He knows the machine wasn't working
when he bought it but because of the condition and since it was Centroid, he was willing to take
the risk even if it involves doing a complete retrofit conversion. He ask for my help since I don't have issues
with English (but I am not perfect since this is my second language) and because I have some experience
with Centroid already. Just to put you in perspective, I have done a router conversion with Acorn
and I am in the process to finish a plasma table with Acorn as well. I have done a CO2 Laser controller conversion
recently as well. On my side I have a Milling/Machining center OKUMA Cadet Mate that I might do a conversion
later on but will try with the legacy controller first since it works at the moment.

Machine details
Brand: TRUMP
Model: B8AC
Year: 1999
ATC 20 tools, BT40
Travel: X/50", Y/22", Z/25"
Control: Centroid M-400, V7.08, serial 6672

Pictures, user manual, backup files, etc., here on OneDrive: https://1drv.ms/f/s!AuCfg0ciIehngitGMw0 ... 0?e=5TnfO8
I backup everything I could using the Utility Menu and floppy disk and I even take out the HDD to backup all files on it
My friend have pretty much all documentation with it, which includes: the TRUMP operating manual, the Centroid Intercon
and user manual, electrical blueprint, spindle drive manual, etc. In the OneDrive I didn't scan everything but the general needed
is there I think, if you might need something else let me know.

Here's the issues at the moment:
- Unable to move the machine at all
- Seems stuck in the Warm-up/booting sequence (Job name stated on display: WARM-U~1.NC)
- Control is turning OFF/ON automatically from time to time, happens randomly. Not only the display but the computer itself.
- The error shown on the display is "Spindle drive fault detected".

On the spindle drive in electrical cabinet, the ALARM LED is OFF

- Even with the electrical panel OFF, the error appear, with other errors (which is normal I assumed)

What I've done and checked so far:
- in the Centroid manual, it stated that the spindle drive fault solution could be: check inverter for fault,
reset spindle contactor OCR, then press & release E-STOP (see picture). We tried reseting the drive
(but doesn't seem in alarm) and my friend bought a new OCR contactor and I install it inside, swapping the older one.

- on the PLC DIAG, it seems the OUTPUT65 (spindle_fault_out) is ON and INPUT15 (spindle_ok) is OFF

- verified the connection between the RESET on the spindle drive and the board
where it's connected (there's continuity which is fine: the wire goes from A to B..)
- tried to force reset the drive by applying 24VDC to the RESET
- tried to exit the external mode (EXT) to manual mode (PU) on the spindle drive for testing
if we can make it move in FWD/REV (not able to change mode)

What my friend want is:
- Being able to move the axis and the spindle drive to ensure everything mechanical is fine
- Since the control is powering OFF/ON automatically randomly he understand it might be hard
to troubleshoot and can be dangerous, so he is willing to do a retrofit but don't know which package
he should get to be the most cost effective (is it possible to keep all motors and even the spindle drive ?)
- how to implement and make the ATC work on the retrofit ?
- Is there a way to check and troubleshoot the spindle drive, to know if it works ? I had the idea to
get an adjustable power supply and send 0-10V or 4-20mA to see if it could move.
- Any recommandations you could supply to us

Thanks alot for your support,
Best Regards from Eastern Canada

[cncsnw]

The current job name ("WARM-U~1.CNC") and "Machine Home Not Set" warning are not relevant at this time. You need to find out why the control thinks there is a spindle drive fault, because until that is cleared, no other operations (such as homing the machine, and loading a job you want to run) will be possible.

This control probably has a Koyo DL205 PLC in the gray cabinet. The spindle drive fault signal would be wired to a Koyo PLC input, and the Koyo PLC would forward that to the Centroid PLC software via a serial link (the OPTIC232 board in the gray cabinet, and a CPU7ADD board in the console, joined by three optical fibers).

The wiring schematics should tell you which input, on which Koyo PLC module, carries the spindle drive fault signal. You can then check the status LEDs on that Koyo input module to see whether the spindle drive fault signal is open, or closed. Unfortunately, it may be unclear whether the signal should be open when the drive is okay, or closed when the drive is okay. It was done both ways, in different controls.

Look on the Mitsubishi inverter terminals for an orange wire. Is the orange wire on a terminal labeled 'A', or on a terminal labeled 'B'? Alternately, if the orange wire is on 'C', then is the yellow wire on 'A', or on 'B'?

Look at the first input module (D2-16ND3) on the Koyo PLC unit. Can you identify the other end of the orange (or yellow) wire? Is it on terminal B3? If so, then slide the toggle switch for the indicator LEDs down to the 'B' position, and check LED #3. If the input circuit is closed, then the LED should be on. If the input circuit is open, then the LED should be off.

On a separate front, you can test and verify that PLC input information is getting from the Koyo to the Centroid. Watch the PLC Diagnostic screen while you trip and release any of the axis limit switches. If the limit inputs change between 0 (closed, okay) and 1 (open, switch tripped) then the PLC communication link is working fine. If you cannot get a limit switch input to show '1' when tripped, then there is a problem with the PLC communication link.


The spindle drive fault problem can likely be fixed, and when it is, the machine will probably run (in between console reboots).

The M400 console is probably going to be the fatal weak link, that pushes your friend toward a retrofit. The PC104 motherboards used in that style console are no longer readily available, and it sounds like this one is on its last legs (but first check the CPU cooling fan).

I would retrofit it with an Allin1DC control unit (plus a PLCADD1616 I/O expansion board, and plus a DC1 if it is a 4-axis machine) and a new M400 console. The existing servo motors, encoders, servo power supply, spindle drive and motor, E-stop contactor, and coolant contactor can be reused.

[Quebecois_Sti]

Thanks alot for the input.

Orange wire on spindle drive is on B.

On the DirectLogic205 Koyo PLC the terminal have continuity with #3 on the Input module. The little switch is on B.

The LED is ON but looks like the clip-on terminal thing is not perfect I can move it a bit and it looks like the status turn OFF/ON (bad connection behind probably, maybe the little spring pushing to make electrical connection isn't perfect).

The PLC with phone wire is going to a small board with 3x fiber cable, they all glowing. Does the telephone cable must be in Port #2 (D2-240 CPU) towards the Optic232 board ?

However I checked the PLC Diag and pressing the sensor +/- on Z and nothing change...

It looks like the comm between sensor and Centroid control aren't communicating with each other.

All the fiber cable inside electrical panel are glowing red and seems pushed correctly. When I unplugged some it trigger some component in panel (i hear it)

The fiber cables going to Centroid console some are glowing some aren't. However, when I unplugged some, it trigger some component in panel (i hear it)

Is there a way to verify fiber cables or buy fiber cable to hook it up from the electrical panel towards console ? How to troubleshoot communication between Centroid and panel, what to do ? Expensive to troubleshoot (need special equipment?) ? What are the options ?

Best Regards

[Quebecois_Sti]

It looks like the fiber modules TX/RX in the Centroid console and electrical panel seems to be Agilent/HP, Versatile Link series (HFBR-2524 and HFBR-1524)

From what I've read those fiber cable are called POF (plastic optical fiber). Not sure about the connector name itself, the POF seems to be related to their category or type ??

I heard the IT department at my work have some fiber cable diagnosis equipment (from FLUKE) I think but not sure if they will lend it to me for an evening because I heard it cost several thousands $$$...

I've seen some people using special flashlight hook-up to the fiber cable with pattern in it to see from one side to another (TX -> RX) if the pattern stays the same. It's not a high-end verification for sure but could check for leak of something in some way ??

[cncsnw]

Yes, the RJ12 (telephone-style) cable plugs into port 2.

If the Koyo inputs are working, then you should see the LEDs on the input module changing as you open and close a switch. For example, if you are tripping the Z plus and minus limit switches, then (with the slide switch in the 'A' position) you should see the LEDs for inputs A4 and A5 turning off and on (off when the switch is tripped/open, on when the switch is clear/closed).

If, while you do that, the inputs on the PLC Diagnostic screen (e.g. INP7 "Z_minus" and INP8 "Z_plus") are not changing, then there is indeed a problem with the communication link.

As a test, trip and hold a limit switch for at least ten seconds. If, after ten seconds, you see the input on the PLC Diagnostic screen change from 0 (closed/okay) to 1 (open/tripped), then the problem is with the INT fiber.

If the input on the PLC Diagnostic screen never changes even after ten seconds, then the problem is with the RXD and/or TXD fiber, or with the RJ12 cable, or with the OPTIC232 board or CPU7ADD board.

I am pretty sure that, if there were no communication at all with the OPTIC232 board, you would get the "External PLC Offline" message full-time (as you saw when you turned off the power to the control cabinet, and with it power to the Koyo PLC and the OPTIC232 board).

[Quebecois_Sti]

It looks like I found the problem!
The fiber cable radius was too steep.

I manage to bring the big black tube on the top of the rectangular tubing.

I was able to pull the fiber cable (especially the DATA one) and make some loose to bring them closer to the connector with larger radius curvature

I was able to do the Homing on all axis. The spindle and tool changer index using the console keys. The MPG wheel doesn't work but maybe the control itself is broken (even with the MPG button activated ON)

I've tried an empty tool change to see if it works (T01M06 in MDI) and it works good but now we wanted to try with a real tool in the spindle and it's asking for input (PLC waiting for input)

However, still having issues with the console computer turning OFF/ON... Weirdly we are able to JOG the axis in manual even when it's booting

See 3 videos here:
https://1drv.ms/f/s!AuCfg0ciIehnjQulbkb ... l?e=qxY1NK

[Quebecois_Sti]

cncsnw:
Do you have an idea what could be the issue regarding "Waiting for input #28 (M6)" ? or similar waiting ?

regarding the OFF/ON I am thinking it could be the thermal paste on the CPU that is worn out
and not being pressed enough on the heat sink so the fan isn't cooling it enough ? So the CPU went
overheat and reboot... I need to investigate and check that out.

Best Regards,

[cncsnw]

It is likely that the control is waiting for the carousel to be reported "out": that is, fully retracted out of the way.

You should probably be able to make the carousel advance and retract using M80 and M81 at the MDI prompt.

You can also test the inputs by disconnecting the air supply (and perhaps the air lines from the carousel cylinder) and pushing the carousel in and out by hand.

What do you see INP27 and INP28 doing on the PLC Diagnostic "Watch All" screen, when the carousel is fully advanced under the spindle, and when the carousel is fully retracted out of the way.

It is likely that INP27 should close (change to green, 0) only when the carousel is advanced in under the spindle; and that INP28 should close (change to green, 0) only when the carousel is fully retracted out of the way.

You may need to adjust the position of the magnetic reed switches on the air cylinder to get it to work properly.

On the Koyo PLC, it is likely that:
INP27 / Carousel-advanced = X22, input A2 on the second input module
INP28 / Carousel-retracted = X21, input A1 on the second input module