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.

Housing connectors - the complete reference¶

2 September 2026. Everything the wiring diagram says about every connector on or at the Tecnos Housing, gathered into one place.

This file is self-contained. A session that reads only this file has the whole picture on Housing connectors and does not need to open any other document.

Naming, first (Grant, 2 Sep 2026)¶

Term	Means
Housing (Tecnos / Monoscheda / Mesa Housing)	The rack enclosure holding the Monoscheda card and its Syncro daughterboard. Its front panel is what this document is about
Cabinet	The outer electrical cabinet (`Q.E.`) the Housing sits inside - drives, contactors, breakers, transformers, the DIN-rail I/O modules, terminal strips
Field	Everything past the cabinet's Harting connectors - machine junction boxes, pendant, sensors, valves, switches

Sources¶

Every fact below traces to one of these, all already read in full:

Source	Gives
SGM section, pages 6-8	The definitive pin-by-pin schedule for `CONN.X`, `CONN.Y`, `CONN.Z`
SGM page 5	`CONN.S`, the spindle connector
SGM pages 11-12, `COLLEGAMENTI SERIALI`	The four serial cables
Q.E. sheet 35	The Monoscheda front panel layout and the CNC power sheet
Q.E. sheets 5, 7, 8, 9, 22	The axis drive sheets and the home-switch sheet
Q.E. sheets 33, 34	The I/O module sheets and the ribbon cross-references
`LISTA DISPOSITIVI` (14-page device list)	Part numbers for every connector and cable

1. Every connector on the Housing, at a glance¶

Connector	Type	Carries
`AXIS 1`	DA-15 15-pin D-sub, 2 row	X axis - analog command, encoder, servo enable, home switch, 24 V
`AXIS 2`	DA-15	Y axis - same set
`AXIS 3`	DA-15	Z axis - same set
`AXIS 4`	DA-15	Spindle (`CONN.S`) - analog speed reference only, 2 pins used
`CN1` (lower left)	26-way IDC	Inputs `IF.1`-`IF.24`
`CN4` (upper left)	26-way IDC	Inputs `IF.25`-`IF.48`
`CN7` (lower right)	26-way IDC	Outputs - one of the two output ribbons
`CN10` (upper right)	26-way IDC	Outputs - the other
`COM1`	DE-9 9-pin D-sub	Serial to the PC. The motion link
`COM2`	DE-9	Auxiliary serial. Configured but unused
`CN5P`	5-pole power	24 V AC supply to the Monoscheda
`FUSE 1` and `FUSE 2`	2 × fuseholder	1 A, on `FILI VIOLA` (violet wires), CNC supply. ✅ Two, confirmed from the panel photo 2 Sep. Sheet 35 was read as naming only `FUSE 1`
`RUN`, `MOD`, `COM1`, `COM2`	LEDs	Status
Two LED blocks `A Z I Z A / B E I E B`	LEDs	One between `AXIS 1`/`AXIS 2`, one between `AXIS 3`/`AXIS 4`. Each serves the two axes either side

🔴 The naming wrinkle - read this before touching a ribbon¶

The sockets on the panel and the cables in the drawing are numbered differently.


Sockets on the Monoscheda panel (sheet 35)	`CN1`, `CN4`, `CN7`, `CN10`
Cables in the device list and on sheets 33-34	`CN1`, `CN4`, `CN10`, `CN13`

The two input names agree. The two output cables land on sockets CN7 and CN10, and the drawing does not make the order explicit.

✅ Partly resolved 2 Sep from a clear panel photo. The four marker-pen labels read CN4 IN 25-48 (upper left), CN1 IN 1-24 (lower left), CN13 OUT 25-48 (upper right) and a fourth lower right obscured by ribbons. So:

The earlier reading of "CN3 OUT 25-48" was CN13, as suspected
The marker labels use the CABLE names (CN1, CN4, CN13), which agree with the device list and sheets 33-34 - they do not use the CN7 / CN10 socket names from sheet 35
The labels are therefore more useful than previously thought, but they are still hand-written by an unknown hand at an unknown date

⚠️ Still: confirm every ribbon by the module it lands on - IF1, IF4, UF1, UF4 - before trusting any label. The fourth label has not been read, and a machine that has been serviced can have labels that no longer match.

2. The four axis connectors - `CONN.X`, `CONN.Y`, `CONN.Z`, `CONN.S`¶

What the connector is called¶


Proper name	DA-15 - D-subminiature, A shell, 15 contacts in two rows (8 + 7)
Catalogue name	"DB15". Technically wrong (`DB` is the 25-way shell) but near-universal, so search both
⚠️ Not	HD-15 / DE-15, the 15-pin three-row high-density VGA connector. Same contact count, different part, and it is what a plain "15 pin D-sub" search returns
Loom end (at the Housing)	🔴 MALE / pins - `DA-15P` - verified physically by Grant, 2 Sep 2026, and matching `MASCHIO` in the device list
Housing panel end	🔴 FEMALE / sockets - `DA-15S`. This is what Mesa must present
Shell over it	Backshell or hood
Part, connector	Zaniboni `1402005`, `025ZAN0020`, `CONNETTORE VASCHETTA 15PIN, MASCHIO`
Part, cover	Zaniboni `1413014`, `025ZAN0026`, `COVER X 15PIN`

VASCHETTA is the Italian trade term for a D-sub shell - literally "little tray".

⚠️ Sheet cross-references from the device list put CONN.S at 5.8, CONN.X at 6.8, CONN.Y at 7.8, CONN.Z at 8.8 - one sheet earlier than each axis's AZION. ASSE sheet (X = 7, Y = 8, Z = 9). The connector symbol appears at column 8 of the preceding sheet. Not resolved further; the SGM schedule is the authority for pins.

`CONN.X` / `CONN.Y` / `CONN.Z` - identical on all three axes¶

Definitive, from SGM pages 6, 7 and 8.

Pin	Conductor	Wire X	Wire Y	Wire Z	Signal	Drive CN1
1	-	-	-	-	unused	-
2	-	-	-	-	unused	-
3	YE giallo	64	78	92	Encoder B	35 `PBO`
4	BL blu	63	77	91	Encoder A	33 `PAO`
5	GR grigio	65	79	93	Encoder Z / index	19 `PCO`
6	-	137	138	139	Home switch `BOX`/`BOY`/`BOZ`	none - stops at the Housing
7	-	`014`	`014`	`014`	+24 V DC out to the drive	47 `+24V-IN`
8	BN marrone	69	83	97	Velocity command + (`V-REF`), ±10 V	5 `V-REF`
9	RD-BN rosso-marrone	62	76	90	Encoder common	1 `SG`
10	PK rosa	67	81	95	Encoder /B	36 `/PBO`
11	GN verde	66	80	94	Encoder /A	34 `/PAO`
12	WH bianco	68	82	96	Encoder /Z	20 `/PCO`
13	-	71	85	99 (0.5 mm²)	Servo enable `/S-ON`	40 `/S-ON`
14	-	-	-	-	unused	-
15	WH bianco	70	84	98	Velocity command − (analog reference 0 V)	6 `SG`

Twelve of fifteen pins used. Pins 1, 2 and 14 are spare - only three, which is why a step/direction pair (four conductors) could never have been fitted.

Cables inside each loom¶

Cable	Spec	Used
Encoder	`SCH 8x0.25` - 8 core, screened, single overall screen, not screened pairs	7 of 8. One spare conductor
Analog reference	`SCH 2x0.50` - individually screened pair	2 of 2
Home switch, enable, 24 V	Single conductors, 0.5-1 mm²	-

⚠️ Sheet 7 prints the encoder cable as 8X0.22. The SGM schedule says SCH 8x0.25. Prefer the schedule.

⚠️ The single overall screen on the encoder cable is the known weak point if noise appears in commissioning. Land the screen at one end only, at the Housing.

Three grounds, deliberately kept separate¶

Ground	Pin	Belongs to
Encoder common	9	Encoder circuit (`CN1-1 SG`)
Analog reference	15	Analog command circuit (`CN1-6 SG`)
24 V return	via wire `12` at the cabinet	Drive supply

TEA separated them. Do not common them. There is an 8 kW VFD in the cabinet.

Home switch routes - pin 6 in detail¶

Axis	Switch	Wire	Route from the Housing outward
X	`-BOX`	137	`CONN.X` 6 → `-X:6` → Harting `-B:5` → `=BM+X43`, `X43:137` → `BOX` (1-2) → `X43:14`
Y	`-BOY`	138	`CONN.Y` 6 → `-Y:6` → Harting `-C:19` → `=BM+X4`, `X4:138` → `BOY` → `X4:14`
Z	`-BOZ`	139	`CONN.Z` 6 → `-Z:6` → Harting `-C:20` → `X4:139` → `BOZ` → `X4:14`

🔴 Pin 6 is the only pin on this connector with no drive destination. The home switches come back to the Housing and stop there, at the Syncro axis card. That is why they appear nowhere in MW310.EQU, and it proves the drives are not self-homing - the Tecnos homes them.

They return to rail 14, a 24 V control rail, so they read as 24 V dry contacts or proximity switches.

`CONN.S` - the spindle, on `AXIS 4`¶

From SGM page 5. Only two pins used.

Pin	Conductor	Signal	Goes to
8	WH bianco	Spindle speed command	`-R6:1`
15	BN marrone	Spindle command return	`-R6:4`

Cable: SCH 2x0.50 screened pair.

⚠️ The colours are reversed relative to CONN.X/Y/Z - WH on 8 and BN on 15, where the axes have BN on 8 and WH on 15. Go by pin number, not colour.

🔴 AXIS 4 is the spindle, not the carousel. Confirmed by "POS. ASSE S" 4 in Param/pargen.prm and the S-axis block in Syncro.prm. The Tecnos treats the spindle as its fourth axis and commands it by voltage, but with only 2 pins used the spindle loop is open - no encoder comes back.

The command goes into the -R6 pot network, which then feeds the VS-606V7 inverter. -R6 is Tecnos part 025TEC0027, KIT SPINDLE LUNGO + CAVI. It is a speed override in the controller's path, the counterpart of -FR1 (COMANDO OVER-RIDE, 025TEC0010) on the feed.

3. The four 26-way ribbon headers¶

Cable ref	Part	Type	Description	Sheet
`CN1`	`025TEC0005`	TC26/225	`COLL. IN OUT A CNC`	33.1
`CN4`	`025TEC0005`	TC26/225	`COLL. IN OUT A CNC`	33.1
`CN10`	`025TEC0005`	TC26/225	`COLL. IN OUT A CNC`	34.0
`CN13`	`025TEC0005`	TC26/225	`COLL. IN OUT A CNC`	34.0

TC26/225 reads as 26-way, 225 mm long. All four are the same part.

✅ The headers are latched - confirmed from the panel photo, 2 Sep¶

The four headers are white shrouded box headers with integral ejector latches - the classic hinged white arms that swing out to eject the socket and lock it in when closed. Gold-plated dual-row pins. They sit in rectangular cutouts in the metal front fascia, soldered to the Monoscheda board (Grant, 2 Sep) rather than mounted to the panel.

Buy latched breakout modules. Many cheap IDC-to-terminal modules use plain shrouded headers with no ejectors. A plain header still accepts the socket and holds it by friction, but this is a machine that vibrates and the latch is worth having.

The socket photographed separately therefore does carry the matching latch cut-outs, and an earlier note doubting that is withdrawn.

✅ Measured 2 Sep 2026 (Grant): the pitch is 2.54 mm (22.4 mm across pin 1 to pin 10, which is 9 gaps → 2.49 mm; 2.24 mm is not a standard pitch). Worth one confirming span across pin 1 to pin 13 = 30.5 mm, or the shroud width ≈ 35.5 mm, before ordering.

✅ The four headers are soldered to the Monoscheda board, not panel parts. They leave with the card. There is nothing on the Housing panel to reuse or match, and a new front panel has to be made regardless - the existing one is the Monoscheda's own front bracket.

✅ The ribbons are comfortably long enough to reach DIN-rail breakouts mounted inside the Housing.

What each carries¶

Cable	Points	Lands on modules
`CN1`	Inputs `IF.1`-`IF.24`	`IF1` (head), `IF2`, `IF3`
`CN4`	Inputs `IF.25`-`IF.48`	`IF4` (head), `IF5`, `IF6`
`CN10`	Outputs `OF.1`-`OF.24`	`UF1` (head), `UF2`, `UF3`
`CN13`	Outputs `OF.25`-`OF.48`	`UF4` (head), `UF5`, `UF6`

Sheet 33 carries all six input modules; sheet 34 carries all six output modules. Confirmed from device-list page 12.

The modules at the far end - in the cabinet, not the Housing¶

⚠️ The IF1-IF6 / UF1-UF6 modules are DIN-rail units on the cabinet backplate (sheet 38 layout), not inside the Housing. The ribbons run from the Housing's front panel out to them.

Module	Part	Type
`IF1`, `IF4`	`025TEC0001`	`MODULO INPUT CNC`, IN C3 - head module, takes the 26-way
`IF2`, `IF3`, `IF5`, `IF6`	`025TEC0002`	`MODULO INPUT CNC`, IN C1 - chains off the head on 10-way ribbons
`UF1`, `UF4`	`025TEC0003`	`MODULO OUTPUT CNC`, OUT C3 - head module
`UF2`, `UF3`, `UF5`, `UF6`	`025TEC0004`	`MODULO OUTPUT CNC`, OUT C1

Eight points per module. Three head modules feeding chains is what "N.SCHEDE I/O" 3 in pargen.prm refers to.

⚠️ CORRECTED 7 September — the `TD62783AP` is in the HOUSING, not on the output modules¶

An earlier version of this page said "the output modules use TD62783AP source drivers - read off the chips in the rack photos", and concluded from it that the ribbon is the logic side of the opto-isolators.

🔴 Both halves were wrong. Grant, at the machine: "The Toshiba is in the Housing. It is getting replaced by Mesa. There is no chip on the output cards."

That was a photo interpretation on my part against direct observation on his, and the misattribution inverted the conclusion. The driver sits on the Tecnos output board inside the Housing; the output modules in the cabinet are passive.

✅ What follows from the driver being on the Housing side¶


🔴 The output ribbon conductors carry the DRIVE, not logic	They are the output side of the source driver — up to 50 V and 500 mA per the part, in practice the machine's 24 V
✅ The `TD62783`'s 15 V input limit is irrelevant to the ribbon	That limit is on the chip's input pins, which face the Monoscheda inside the Housing. It is being removed
🔴 The Mesa outputs must be configured SOURCING	A source driver on the Housing side means the modules are commoned to 0 V and expect to be fed +24 V. All three Mesa cards' outputs are universal, so this is a software setting, not a wiring one
✅ Driving a ribbon output conductor with 24 V is what the Mesa will do	So doing it by hand to identify which relay is which is safe, and representative

✅ CONFIRMED 7 September — the modules are 24 V, and 12 V will not do¶

Grant, at the bench: "I've just driven a relay with 24 V. Note it did not move with 12 V."

🔴 So the output field supply is 24 V, not a choice. A 12 V rail does not pull these relays in, so the Mesa's output-carrying blocks must run VFIELD at 24 V.


✅ Mesa outputs are 5-28 V	24 V sits comfortably inside, on all three cards
✅ Sourcing, as deduced	Confirmed in practice — feeding +24 V into the module pulls the relay in
🔴 `VFIELD` = 24 V on every block carrying outputs	Settled. It was an open question

⚠️ Consequence for the inputs. Mesa input thresholds are 50 % of that block's field supply. On the 7I77U and 7I84U, VFIELDA and VFIELDB are independent, so an input block could run at a different voltage — but with 24 V field sensors there is no reason to, and a single 24 V rail keeps the thresholds at 12 V where they belong.

✅ MEASURED 8 September — 18.5 mA per relay, and the allocation is unconstrained¶

Grant, at the boards: "The relay in the output cards draw 18.50 mA so let's call it 20 mA. Max of 960 mA or 1 A max with some tolerance for voltage drop."

20 mA is the working figure, and 960 mA is the honest worst case — 48 points, not the 41 that are real.

🔴 Against every limit, this is nowhere near.

Limit	Value	Worst case	Used
Mesa, per output	500 mA	20 mA	4 %
🔴 Mesa, per driver chip (8 outputs)	2000 mA	160 mA	8 %
`-G3` supply, whole rail	6 A	~1 A	~16 %

✅ So the per-chip limit stops being a design constraint. Two earlier notes warned that the seven spare outputs were "not freely placed" because each Mesa chip serves eight outputs at 2.0 A. At 20 mA a chip's eight outputs draw 160 mA — eight per cent of its budget. Outputs can be allocated wherever is tidiest.

✅ And the existing 24 V supply is ample. About 1 A of coils against 6 A, leaving headroom for the drives' +24V-IN, the field sensors and the Mesa cards' own VIN (~25 mA each). No second supply.

🔴 Ribbon ways 25 and 26 are the shared negative, doubled up¶

Grant, 8 September: "On the ribbon cable the last two wires are connected together on the board — checked with an ohm meter, zero resistance. They act as the negative side of the circuit. They are paired to increase current carrying capacity of the return wires."

✅ That closes the "which two ways are not signals" question, and it confirms the shape: 24 signal ways plus a doubled return.

🔴 And the pairing is explained by the arithmetic above. Every energised output's current comes back through those two conductors. With all 24 points on one ribbon at 20 mA, that is 480 mA through two paralleled ribbon conductors — comfortable for two, marginal-to-warm for one. Whoever built it doubled them for exactly that reason.

⚠️ Carry this into the Mesa build: those two ways land on the field-power GROUND, and the breakout terminal has to take the sum of every output on that ribbon, not one point's worth. Do not land the pair on a single screw unless the terminal is rated for it — keep them on two.

Mesa output rating
Voltage	5-28 V — 24 V is squarely inside
Per output	500 mA
⚠️ Per driver chip	2.0 A, and each chip serves eight sequential outputs

🔴 The per-chip limit is the one to check, not the per-output one. If eight points on one Mesa chip can be on together, the sum has to stay under 2 A — 250 mA average. A relay coil will be nowhere near that, but it is worth one measurement rather than an assumption.

⚠️ The input side is not settled by this. Where the input receivers live — Housing or module — has not been established, so nothing here says what the input ribbon conductors carry.

Address to module terminal - the formulas¶

Proven against the SGM schedules on thirteen sample points.

For	Module	Terminal	Wire tag
Input `IF.n`	`IF[⌊(n−1)/8⌋+1]`	`((n−1) mod 8) + 2`	`E1nn`
Output `OF.n`, with `k = ((n−1) mod 8) + 1`	`UF[⌊(n−1)/8⌋+1]`	`3k−1` (the group is `3k−2`, `3k−1`, `3k`)	`A2nn`

The output wire always lands on the middle terminal of each group of three: 2, 5, 8, 11, 14, 17, 20, 23.

The arithmetic that closes¶

4 ribbons × 24 points = 96 = 48 inputs + 48 outputs, matching the EIN1-48 / AUS1-48 numbering on sheets 16-28. Of those, 41 inputs and 41 outputs are real.

	Unused addresses
Inputs	5, 6, 19, 41, 42, 47, 48
Outputs	4, 5, 31, 32, 33, 34, 39

✅ One ribbon wire per point - tested at the boards by Grant, 31 Aug 2026. No multiplexing, no scanned bus, no strobed addressing.

✅ SOLVED 9 September — the ribbon way map¶

Grant mapped all four ribbons by colour. This was the single biggest gap in the project and the only thing blocking the Mesa build.

🔴 The ways are NOT in address order¶

The guess on this page was "likely ways 1-24 in order". Wrong. Each ribbon carries three modules of eight points, and the module blocks sit on the ribbon in reverse order:

Ways	Carries
1-8	the third module — `IF3`, `IF6`, `UF3`, `UF6`
9-16	the second module — `IF2`, `IF5`, `UF2`, `UF5`
17-24	the first module — `IF1`, `IF4`, `UF1`, `UF4`
25, 26	joined on the board, the shared negative

Inside each block of eight the points ascend normally. So it is the blocks that are reversed, not the points.

The formula¶

For an address n, with p its position within its own ribbon:

p = ((n - 1) mod 24) + 1        position on this ribbon, 1-24
b = ((p - 1) div 8) + 1         which module block, 1-3
k = ((p - 1) mod 8) + 1         position within that block, 1-8

way = (3 - b) x 8 + k

Ribbon	Addresses
`CN1`	`IF.1`-`IF.24`
`CN4`	`IF.25`-`IF.48`
`CN10`	`OF.1`-`OF.24`
`CN13`	`OF.25`-`OF.48`

✅ Verified against all 96 points from Grant's capture: zero mismatches, and each ribbon uses ways 1-24 exactly once. Worked examples: IF.1 is way 17, IF.17 is way 1, OF.41 is way 1, OF.48 is way 8.

Reading a way by colour¶

The rainbow repeats every ten from way 1, so a colour appears at most three times:

	1st	2nd	3rd
brown	1	11	21
red	2	12	22
orange	3	13	23
yellow	4	14	24
green	5	15	25
blue	6	16	26
purple	7	17	—
grey	8	18	—
white	9	19	—
black	10	20	—

✅ The joined return pair is green 3 and blue 3 — ways 25 and 26, exactly where the arithmetic puts them, and matching Grant's correction that the pair is green and blue rather than green and purple.

🔴 Every point now carries its ribbon, way and colour on the I/O lookup page, generated from this formula rather than transcribed, so it cannot drift.

What is still not known about the ribbons¶

One left.

Unknown	Note
🔴 The input side — voltage in both states, and sourcing or sinking	Sets the Mesa input configuration. ⚠️ An earlier guess of "probably 5 V TTL" rested on the `TD62783` mislocation corrected above, and only ever concerned outputs. Where the input receivers live is still unestablished

✅ Closed, in order:


Which of the 26 ways carries which point	9 Sep — module blocks reversed, formula above, verified on all 96
Which two ways are not signals	8 Sep — ways 25 and 26, joined on the board, the shared negative, doubled for current capacity
Current drawn by one output module point	8 Sep — 18.5 mA. Every Mesa and supply limit has an order of magnitude of headroom
Sourcing or sinking, outputs	7 Sep — sourcing, fed +24 V against a 0 V common, confirmed by a relay pulling in at 24 V and not at 12 V

⚠️ The one remaining unknown still needs the machine live with the Monoscheda in place. It cannot be answered after strip-out.

The eight 10-way ribbons - NOT on the Housing¶

They chain module to module inside the cabinet. Listed here only so nobody looks for them on the panel.

Ref	Part	Type	Description
`CN2`, `CN5`, `CN12`, `CN15`	`025TEC0008`	TC10/050	`COLL. OUT C3-OUT C1 E IN C3-IN C1`
`CN3`, `CN6`	`025TEC0006`	TC10/025	`COLL. IN C3-IN C1`
`CN11`, `CN14`	`025TEC0007`	TC10/035	`COLL. OUT C3-OUT C1`

4. `COM1` and `COM2` - the serial ports¶

Both are 9-pin D-sub (properly DE-9; universally miscalled "DB9").

	`COM1`	`COM2`
Use	The PC-to-Monoscheda motion link	Auxiliary. Configured but nothing uses it
Baud	19200	38400
Protocol	`ACKNACK` (`TCBIN` for the service tools)	`XONXOFF`
Settings live in	`C:\Windows\Mw310.ini` - `Porta COM=1`, `COM1 Syncro=ACKNACK`, `B.R. COM1 Syncro=19200`	`Porta COM AUX=2`

The `COM1` cable - `SERIALE SYNCRO-P.C.`¶

Part CAV90240 / 025TEC0028. From SGM pages 11-12, COLLEGAMENTI SERIALI.


Connectors	DB-9F to DB-9F
Wiring	Null modem, three wire: 2↔3, 3↔2, 5↔5
Shield	To both shells
Protection	Two 13 V `F-126` suppressor diodes
Handshake lines	None. RTS, CTS, DTR, DSR, DCD all unused

🔴 A plain three-wire null modem with no handshake, so a passive listening tap is trivial and cannot disturb the link. No com0com, no hub4com, nothing to roll back.

⚠️ Data bits, parity and stop bits are recorded nowhere - not in any INI, not in the registry. Comdrv.exe sets the framing programmatically. They must be observed from a capture.

The other three serial cables on the same drawing - none of them at the Housing¶

	Name	Connectors	Route
B	`SERIALE ARMADIO-P.C.`	DB-25F to DB-9F	Cabinet `PULSANTIERA` to PC
C	`SERIALE PC-YASKAWA`	DB-9M to DB-9F	`3CN` on the Yaskawa drives. ⚠️ Needs Yaskawa's own software; `3CN` is not a generic serial port
D	`SERIALE PC-PANASONIC`	mini-DIN to DB-9F	`CN SER` on the Panasonic

5. `CN5P` - the Housing power connector¶


Part	`025TEC0009`, `CONN. ALIMENT. CNC`, `CN5P`
Type	5-pole pluggable terminal, the orange connector seen on the rack
Supply	🔴 24 V AC, on pins 1 and 2
Source	`-T3`, a 1500 VA transformer, `0-400` primary, `0-24 V` 250 VA secondary
Protection	`-F13`, 6 A GL
Also on the panel	`FUSE 1` and `FUSE 2`, 1 A, on `FILI VIOLA` (violet wires)
Sheet	35.1

⚠️ CORRECTED 9 September from a photograph — pin 5 IS used¶

The drawing note said "pins 3, 4 and 5 are drawn but not obviously used." Pin 5 is used.

Pin	Wire	Colour
1	9	red
2	10	red
3	—	empty
4	—	empty
5	🔴 12	🔴 see below

So CN5P carries three conductors, not two. The 24 V AC pair plus wire 12 on the end position — which on a five-pole power connector is the conventional earth position.

The device list entry for the controller itself reads -CNC. 025TEC0022 POSIZIONATORE M68K, M68K, 24 Vca - 24 Vca is 24 V AC, confirming it twice.

Note on sheet 38: USCITA CAVO ALIMENTAZIONE M68K (-CNC) L=60cm DA CANALA - the supply cable leaves the trunking with 60 cm free.

⚠️ Nothing in a Mesa build wants 24 V AC. This feed goes away; Mesa takes 24 V DC from the existing -G3 rail (wires 014 positive, 12 return).

6. What is NOT a Housing connector¶

Listed explicitly, because each has been mistaken for one at some point.

Not on the Housing	Where it actually is
`-SIG2`, the carousel encoder	On the Panasonic `MSS083A1XP` drive, closing its own indexer loop. Device list: `-E.561`, `025AMP0001`, "Amphenol, Panasonic encoder", beside `-M560` "Amphenol, Panasonic motor". Sheet 32's header reads `CONNETTORE AZIONAMENTO ASSE 561` - drive connector. The Tecnos never sees a carousel encoder; it gets position from binary bits `IF.35`-`IF.40`
The `IF`/`UF` I/O modules	DIN rail in the cabinet (sheet 38 layout)
The eight 10-way ribbons	Module to module, in the cabinet
Drive fault `ALM`	Wires 72/86/100 go from the drives straight to the `KAX`/`KAY`/`KAZ` relays in the cabinet. Not through any axis loom
Overtravel `P-OT` / `N-OT`	Limit switches straight to drive `CN1-42/43`. Never reach the Housing
Alarm reset `/ALM-RST`	Pushbutton `-SB15` to drive `CN1-44`. Never reaches the Housing
The motor encoders	Serial `PS`//`PS` only, drive to motor (sheets 29-31). LinuxCNC can never read them directly

7. Summary of what is still unknown¶

#	Unknown	Blocks
1	Ribbon way ↔ I/O address	The whole I/O half of any Mesa adapter
2	Ribbon logic voltage and polarity	Mesa field-power domains and input/output configuration
3	Where the I/O modules get their logic power	Whether Mesa adapters must feed the module supply. Two spare ways on a 26-way suggests they do
4	Serial framing bits on `COM1`	Only the serial tap. Not the wiring
5	Which output cable lands on socket `CN7` vs `CN10`	Label by module, not by socket, and it does not matter

🔴 Items 1-3 all need the machine live with the Monoscheda in place, and all three can be answered in one session.

mesa-wiring-schedule.md - where each of these signals goes on the Mesa cards
connector-schedules.md - the raw SGM transcription this file draws on
io-map.md - all 82 I/O points with field devices and Harting pins
servo-drive-interface.md - the drive end of the axis looms
homing.md - why pin 6 matters
serial-port.md - the full COM1 write-up
device-list.md - every part number
retrofit-boundary.md - scope