Working documentation

This page is generated from the project's working notes. It records a real retrofit in progress, corrections included. Anything marked 🔴 or ⚠️ is unresolved or carries a caveat - read those before acting.

Every fuse and breaker, in plain English¶

9 September 2026. What each protective device on TuChop is, what it protects, and why some of them do not do what you would assume.

The single most useful thing to know¶

🔴 Not all fuses protect against overload. Two different fuse types are fitted here and they behave completely differently:

Type	Blows on	Does NOT protect against	Why you would use it
`gL` / `gG` — general purpose	Overload and short circuit	—	Protecting cables. The everyday fuse
🔴 `aM` — motor rated	Short circuit only	🔴 Overload	Feeding a motor or transformer, where a big inrush current at switch-on would blow an ordinary fuse

⭐ An aM fuse will happily sit there while a cable slowly cooks. That is not a fault, it is the design: aM is only ever fitted where something else handles overload — a thermal relay, a motor breaker, or the device's own protection.

On this machine every aM fuse feeds either a transformer or a motor circuit that has its own overload protection downstream. ✅ The arrangement is correct; it just is not obvious.

The three kinds of device fitted¶

1. Motor breakers — `GV2-ME`, the ones with a dial¶

Telemecanique GV2 units. These do both jobs in one box:


Magnetic trip	Instant, for short circuits
Thermal trip	Slow, adjustable, for overload
The dial	⭐ Why they have a current setting — you turn it to the motor's nameplate current

✅ Resettable. No fuse to find at 9 pm.

⚠️ The "range" and the "setting" are different numbers. A GV2-ME10 covers 4-6.3 A; what matters is where the dial is actually pointed.

2. Fuses — `gL` and `aM` as above¶

3. The incoming protection¶


`-QM1`	Main isolator, Telemecanique `V4`, 80 A, with a door lock and extension shaft — you cannot open the cabinet without turning the power off
`IN = 40 A`	The supply's own overcurrent protection
🔴 `ID = 0.1 A`	An earth-leakage device set at 100 mA. If 100 mA finds its way to earth instead of returning down the neutral, it trips

The whole list, and what each one is really protecting¶

Motors and drives — all on motor breakers¶

Device	Breaker	Range	Set at	Protecting
Drills motor `-M101`	`QF101` `GV2-ME10`	4-6.3 A	4 A	The 1.5 kW motor that spins all ten drills
Vacuum pump `-M601`	`QF601` `GV2-ME21`	17-23 A	18.5 A	The 7.8 kW pump
X drive `-UX`	`QFX` `GV2-ME10`	4-6.3 A	4.6 A
Y drive `-UY`	`QFY` `GV2-ME07`	1.6-2.5 A	1.6 A
Z drive `-UZ`	`QFZ` `GV2-ME07`	1.6-2.5 A	1.6 A
Carousel drive `-U560`	`QF560` `GV2-ME10`	4-6.3 A	5 A	The Panasonic, on 220 V
Spindle inverter	`QFINV` `GV2-ME21`	17-23 A	20 A	The VS-606V7's supply
🔴 Brake resistor	`QFRI` `GV2-ME14`	6-10 A	6.5 A	The 30 Ω 1300 W resistor that stops the spindle

⭐ X is set nearly three times higher than Y and Z because it is a bigger drive — SGDH-15DE against SGDH-10DE. The gantry is the heavy axis.

🔴 QFRI does more than protect the resistor — see the section below, because it is the cleverest bit of wiring on the machine.

Transformer primaries — `aM`, because transformers have huge inrush¶

Fuse	Rating	Feeding	Sheet
`-F1`, `-F2`, `-F3`	6 A `aM`	`-T1`, the 3 kVA 400→220 V transformer for the carousel	4.8
`-F7`, `-F8`	2 A `aM`	`-T2` primary, 700 VA	3.1
`-F11`, `-F12`	4 A `aM`	`-T3` primary, 1500 VA	3.5 / 3.7

⭐ Switch a transformer on and it draws an enormous current for a few milliseconds while the magnetic field establishes — many times its running current. An ordinary fuse would blow every time you turned the machine on. That is exactly what aM exists for.

Transformer secondaries — `gL`, because now it is just cable¶

Fuse	Rating	Protecting	Sheet
`-F9`	2 A `aM`	The 110 V control circuit	3.2
`-F10`	6 A `gL`	The 18 V feed into `-G2`, the bridge rectifier	3.2
`-F15`	6 A `gL`	The 220 V rail — and almost certainly the feed into `-G3`, the Cabur	3.6
`-F13`	6 A `gL`	🔴 The 24 V AC to the old controller, wires 9 and 10	3.7
⚠️ `-F14`	1 A `gL`	Purpose not recorded. Sits at sheet 3.7, beside `-F13`	3.7

On the Housing panel¶


`FUSE 1`, `FUSE 2`	1 A, on `FILI VIOLA` — violet wires. ⚠️ Purpose not recorded either

🔴 `QFRI` and the spindle brake resistor — the cleverest wiring on the machine¶

First, why there is a resistor at all¶

When you tell a spinning motor to slow down, the motor turns into a generator.

The spindle and its tooling have real kinetic energy. That energy cannot vanish — and if you are slowing the motor electrically rather than letting it coast, the motor pushes it back into the inverter.


The inverter stores it in its DC bus capacitors
🔴 The bus voltage climbs
Too far, and the drive trips on over-voltage or damages itself

So you fit a braking resistor. The drive switches it across the DC bus when the voltage rises, and the energy becomes heat in the resistor instead of voltage in the capacitors.

⭐ That is why -R2 is 30 Ω, 1300 W. It is not a signal component. It is designed to turn kilowatts into heat, which is also why it has its own cooling fan (-M5).

Now the problem the builder had to solve¶

🔴 A resistor failing open is invisible. Nothing about it announces a fault. The drive keeps running, the spindle keeps spinning, everything looks normal — right up until the next time you ask it to stop.

And the drive has been told it has a resistor. n092 = 1, whose manual entry reads:

"Not provided (set this when a braking resistor is mounted)."

That setting turns OFF stall prevention during deceleration.

What "backing off the ramp" actually means¶

The ramp is the deceleration slope — how fast you asked the drive to slow the spindle. On this machine n020 sets it to 2.5 seconds.

When the drive starts slowing the motor, the motor becomes a generator and the DC bus voltage climbs. The drive now has a choice, and there are only two options.

	Stall prevention ON — the factory default	Stall prevention OFF — `n092 = 1`, this machine
What the drive does	🔴 Stops decelerating and waits. It watches the bus voltage, pauses the ramp until it falls back, then continues	Holds the ramp you asked for and dumps the surplus into the braking resistor
The energy goes	Nowhere — the drive simply refuses to generate more of it	Into the resistor, as heat
Result	The spindle takes longer to stop than you asked, and how much longer depends on the load	✅ The spindle stops in 2.5 seconds. Every time

⭐ "Backing off the ramp" is the drive making its own deceleration slope shallower to protect itself. It is not a fault. It is the drive saying "I cannot absorb energy that fast, so I will slow down more gently."

⭐ The analogy, and it is exact rather than loose¶

A car coming down a long hill.


No brakes, only engine braking	You must take the hill slowly. You get down safely, but not on your schedule, and how slowly depends on the load in the boot
Proper brakes	You choose the speed. The brake discs turn the car's energy into heat and throw it away

🔴 The braking resistor IS the brake disc. Not a metaphor — the same physics. A disc turns motion into heat by friction; the resistor turns it into heat by resistance. -R2 is rated 1300 W because that is genuinely how much heat it has to shed.

So n092 = 1 says: "I have brakes fitted. Use them. Do not coast down the hill to protect yourself."

✅ Why the builder wanted the hard stop¶

Predictability. The drawbar cannot open until the spindle has actually stopped, so the tool-change cycle has to know how long that takes.

⚠️ With stall prevention on, stop time varies with whatever is in the spindle — a heavy cutter takes longer than a light one, because there is more energy to get rid of. A tool change sequenced on a fixed dwell would sometimes fire early.

🔴 And that is exactly why losing the resistor is dangerous. The drive has been told the brakes are fitted. If they are gone and nobody tells it, it brakes hard into nothing — and the energy that should have become heat becomes voltage in the capacitors instead.

⚠️ So if the resistor is gone and the drive does not know, the next stop command sends the energy into capacitors that cannot take it.

🔴 The solution: the breaker tells the drive¶

-QFRI is a motor breaker, and motor breakers have auxiliary contacts — small extra contacts that follow the main ones. The main contacts carry the resistor current. One auxiliary contact is wired to the drive's S3 terminal on wire 56.

And S3 is programmed n052 = 21 — emergency stop, normally-closed contact.

   QFRI healthy  -->  aux contact CLOSED  -->  S3 held closed  -->  drive runs
   QFRI TRIPS    -->  aux contact OPENS   -->  S3 opens        -->  drive STOPS

⭐ The moment the resistor is disconnected, the drive is told and shuts down — before it can ever attempt a braked deceleration into nothing.

✅ This is why n052 = 21 was the single non-default terminal assignment on the whole VFD. Six of seven multi-function inputs are factory settings. Somebody deliberately changed that one, and this is what for.

⚠️ The operational consequence you would otherwise chase for hours¶

🔴 A QFRI trip presents as "the spindle is dead and will not reset."

Two things stack up:


The drive is still powered	`QFRI` disconnects the resistor, not the drive's supply — that comes from `QFINV`. So the keypad lights up and everything looks alive
🔴 `Option = 1` on the VFD means an emergency-stop fault clears by POWER CYCLE ONLY	A reset command will not do it

So: keypad on, drive faulted, reset button does nothing, and the actual cause is a tripped breaker across the cabinet with nothing obviously wrong with it.

⭐ If the spindle ever behaves like that, check QFRI before anything else.

What is certain and what is reasoning¶


✅ Established	`QFRI` is a `GV2-ME14` set at 6.5 A, feeding `-R2` via the VFD's `B1`/`B2`. An aux contact runs on wire 56 to `S3`. `n052 = 21` makes `S3` an emergency stop on an NC contact. `n092 = 1` means the drive expects a resistor
⚠️ My interpretation	That the purpose is specifically to prevent a braked deceleration into a missing resistor. It fits every fact and it is the standard reason — but the builder may simply have applied a general rule: "if anything in the drive's power circuit trips, stop the drive." Both give the same wiring

⚠️ And a correction to something said loosely earlier: this was described as protecting "a 24000 rpm spindle". The spindle's actual speed is not established — it depends on the motor pole count, which is still a nameplate photograph on the to-do list. What is certain is that the drive runs to 400 Hz, and that a spindle at 400 Hz stores enough energy to need a 1300 W resistor.

✅ For the retrofit: leave it entirely alone¶

The VFD, the resistor, QFRI and this interlock all stay exactly as they are. LinuxCNC never touches any of it.

🔴 But whoever commissions the machine needs to know it exists, because "spindle dead, will not reset" is otherwise a very confusing morning.

🔴 Is wire `014` really unfused?¶

Grant's question: "why is wire 014 left unfused, or am I reading the wiring diagrams incorrectly?"

✅ You are reading them correctly¶

As far as everything read so far shows, the 24 V DC output of -G3 has no fuse in it. 014 leaves the Cabur and goes to the machine.

⚠️ One caveat, stated honestly: -F14, a 1 A gL fuse at sheet 3.7, has no recorded function. It is small enough to be a DC branch fuse. It is the one thing that could contradict this, and settling it means looking at sheet 3.7 rather than at anything already read.

Why that was a defensible design¶

🔴 Because -G3 is a switching power supply, and a switching supply is its own fuse.


A transformer will happily deliver ten times its rated current into a short, until something melts. It needs a fuse
A switching supply electronically limits its own output. The Cabur `CS6` cannot deliver much beyond 6 A — on a short it folds back or shuts down and tries again

⭐ So a fuse on the output would be in an awkward position. Rate it above 6 A and it can never blow, because the supply will not pass that much. Rate it below and it nuisance-trips on ordinary inrush. The supply is already doing the job.

✅ And the input IS protected — -F15, 6 A gL, on the 220 V feed, same sheet as -G3. That covers the supply itself and the cable feeding it.

⚠️ But it is not what anyone would do today, for two reasons¶


🔴 No discrimination	One shorted branch takes down the entire 24 V system. Every sensor, every drive's logic, everything, all at once — and nothing tells you which branch caused it. You are left unplugging things one at a time
⚠️ The limit protects the supply, not the wiring	6 A into a thin conductor is a lot of heat. The supply is happy; the wire may not be

Modern practice is a small MCB or an electronic fuse module per branch, so a fault drops one circuit and lights an indicator.

🔴 What this means for the retrofit¶

The Mesa build is about to hang 41 relay coils and three cards off this same unfused rail — the one the servo drives' logic runs on.

⭐ Put a fuse or a small MCB in the new 014 feed into the Housing. It is a few dollars and one terminal, and it buys:


✅ A fault in the new wiring stops at the Housing	It does not take the servo drives' logic down with it
✅ You know where the fault is	The Mesa fuse is blown, or it is not
✅ You can isolate the new work	Pull one fuse and the retrofit is dead while the rest of the machine still powers up

Size it for the actual load — about 1 A of coils plus the cards — so 2 A or 3 A with margin, well under the Cabur's 6 A limit so it blows first.

⚠️ 🔴 One thing it must NOT do: sit in the Mesa VFIELD path as a switch. Mesa require field power to ramp in under 10 V/ms and connect directly to the source, with no switches, breakers or relay contacts. A fuse is explicitly fine; a breaker used as an on/off switch is not.

power-in-plain-english.md — the whole power system, same style
power-distribution.md — the same devices in the diagram's own terms
housing-power.md — feeding the Mesa cards, and the checks to do first