Fagor 8055 Error Codes: What the Number Tells You

An operator reads you a four digit number off a Fagor 8055 screen, and that number has already narrowed the fault down. Fagor groups the 8055 error list in blocks of a thousand, and the block names the source. The 2000 block is CNC hardware. The 4000 block came from a servo drive, not from the control. Anything below 2000 is program or data, and swapping boards will never clear it.

Below: how the 8055 and 8055i number their errors, the faults that turn up on machines from the 2000s, the battery behaviour that catches people fitting used modules, and what to order when the error does mean a dead board. The same reasoning on Siemens controls is in reading a SINUMERIK 840D alarm number.

The number block tells you who raised the error

The error manual is organised by source and the numbering follows it.

The manuals for the mill and lathe models are separate documents and the conversational errors differ. Use the one matching your control.

The hardware block, code by code

2000 is external emergency: PLC input I1 or the emergency mark has gone to zero, more often a circuit that lost power than a pressed mushroom head. Feedback errors run one per axis from 2001 to 2009 in the order X, Y, Z, U, V, W, A, B, C, so the number names the axis before you read the text. 2010 is the spindle.

Then supply and thermal. 2011 is internal overtemperature, either a cabinet that stopped breathing or a failing component on the axis board. 2012 means 24 V is missing at the axis board output supply. 2013 to 2015 say the same for I/O boards one to three, 2024 for the tracing board. Fagor's solution text for all of them mentions a blown fuse.

That fuse is the check that saves an afternoon. "There is no voltage at the I/O 2 board" reads like a dead board, and people order one. Meter the 24 V at the terminals, then the fuse. A board that is not being fed has not failed. On 2011, before condemning anything, put a meter on the cabinet fan supply and check the filters, because a blocked filter mat trips this error on a hot afternoon and clears overnight, which is exactly the pattern people misread as an intermittent board fault.

The memory errors name the board, which is rare and useful. 2017 is CNC RAM and sends you to the CPU board. 2019 is PLC RAM and sends you to the PLC board. 2016 is different in kind: the PLC program is stopped, missing or in watchdog, and that fix costs nothing.

A 4000 error is the drive talking

This is where money gets wasted. Much of the 4000 block is drive faults relayed through the control, and the message text carries a second number that belongs to the drive. Error 4056 reads "ERROR 100: Internal +5 V out of range". 4062 is "ERROR 106: Heatsink overheated". 4064 is "ERROR 108: Motor overheated". 4225 and 4226 report battery voltage under 2.5 V and 3.1 V in the drive system, nothing to do with the CNC.

So on a 4000 number the diagnosis continues at the drive: read the code after the word ERROR, then the LEDs on the module. Changing a CPU board because an axis drive overheated is an expensive way to learn the difference. 4000 itself, the Sercos ring error, is the exception, and that is usually fibre, connectors, or a drive that dropped off the ring.

Why a used 8055 arrives empty

The 8055 holds its machine parameters and tables in RAM backed by a 3.5 V lithium battery with an estimated life of three years. Fagor is specific about what follows: once the low battery warning appears, memory contents survive ten days at most with the CNC off. Ten days. Not ten months.

So if a machine flags low battery, the backup is this shift's job. And a used central unit that has sat on a shelf for a year arrives with empty RAM, which is normal rather than a fault with the part. You will be loading parameters into it.

The unnumbered start up messages belong here too. CHECKSUM ERROR GENERAL PARAMETERS Load CARD A? means that block has been lost and a copy exists on CARD A. Press ENTER and it restores. With no CARD A backup you are rebuilding parameters from the builder's documentation, and on a retrofitted machine that paperwork may be long gone. Keep a second copy off the machine as well, because a CARD A sitting inside the same unit that lost its memory is one fault away from being no backup at all.

When the error really does mean a module

The modular 8055 central unit comes in three module and six module versions, bolted together as CPU, AXES, then I/O. The 8055i puts electronics and monitor in one housing. Identify which you are standing in front of before quoting anything.

• CPU module. For 2017 and 2018, and a control that will not boot: the Fagor CPU 8055/AB-M.
• Axes module. For feedback errors that follow the module rather than the axis: the AXES 8055 module.
• Integrated control. When the unit is dead or not economic to repair: a tested Fagor 8055i.
• Monitor and keyboard. Backlight failure, dark screen over a live control, dead keys: the 8055i/C-M-MON-K monitor panel.
• Axis drive. For 4000 faults that end in hardware: the Fagor AXD 1.25-S0-0 drive module.
• Servo motor. When the motor or its feedback is the fault, such as overtemperature on one axis only: the FXM53.20A.E1.000 brushless servo motor.

A plain word on economics. Board repair on this generation carries a lead time, and the components inside a repaired board are as old as the ones that failed. A tested used module of the right type often costs less than the diagnosis. The exception is a machine whose parameters exist nowhere but in that unit's RAM. Get the data out first. Our stock is tested, held in the EU and carries a 30 day warranty: see the Fagor collection or the wider CNC controls range.

Quick answers

Where do I find the full Fagor 8055 error list?

In the error solution manual for your model, published by Fagor Automation. Mill and lathe versions are separate. Each entry gives four fields: detection, effect, cause, solution. Read the effect field, because it tells you whether the control opened the position loop and dropped the enables.

My error has no number. What is it?

Almost certainly a table data checksum message shown at start up, shaped like CHECKSUM ERROR ... Load CARD A?. The manual lists these without numbers, which is why searching for a code finds nothing. That block of parameters or tables has been lost from RAM, and pressing ENTER restores the copy on CARD A.

Does error 2001 mean the encoder has failed?

Not by itself. 2001 says the CNC is not receiving X axis feedback, which covers the encoder, the cable, both connectors and the axis board input. Where the cable runs in a drag chain, suspect the cable first. Move the axis by hand and watch the position display for a signal that drops at one point in travel.

Can I fit a used CPU module and keep my parameters?

They live in battery backed RAM in the unit you are removing, so they do not travel with the replacement. Back up to CARD A while the old control still powers up, even if it is throwing memory errors. With no backup anywhere, budget time for a full parameter reload. Our used industrial parts buying guide covers the checks before ordering.

The error returns seconds after I clear it. What does that tell me?

That the condition is still present rather than latched. A 2000 series error that comes straight back points at something physical sitting there now: an absent supply, an open fuse, a loose connector, a hot cabinet. Clearing it repeatedly delays the diagnosis, and on a feedback error it risks moving an axis on bad position data.

A diagnostic order that works

1. Write down the full number and the exact message text, including any second number after the word ERROR.
2. Use the thousands block to decide where you are going: program, hardware, PLC, drive or CAN.
3. If the number is below 2000, stop looking at hardware.
4. If it is a 4000, read the drive's own code and LEDs before touching the control.
5. On a missing voltage error, meter the 24 V at the board and check the fuse before ordering anything.
6. On a feedback error, check connector and cable through full travel, then the encoder, then the axis module.
7. Check whether monitoring parameters such as FBACKAL were switched off by somebody before you.
8. Back up parameters and tables to CARD A before changing any module, and again once the machine runs.
9. Quote the exact designation from the module label, for example 8055i/C-M-MON-K rather than "8055 screen".

Related reading on Samonde

• SINUMERIK 840D alarm codes: how to read the number
• Schneider Altivar fault codes: common faults and fixes
• Heidenhain encoder problems: error messages and fixes
• Used industrial parts buying guide

The 8055 was built when a control was expected to be readable by the person maintaining it, and that has aged well. The number gives you the subsystem, the text gives you the cause, and the drives keep their own counsel in the 4000 block. What the design did not anticipate was a three year battery guarding twenty year old parameter sets. Write those to CARD A, keep a copy off the machine, and the next error on that screen is a repair rather than a reconstruction.