Yes, you draw them. A loop diagram is the one sheet that walks a technician from a PLC point back to the signal source, and that path exists whether the wire leaves the cabinet or not. Take the classic panel: a 2-wire transmitter reading cabinet air temperature and two UPS status contacts, all landing on the PLC in the same enclosure. That is three loop sheets, each marked “cabinet internal”, each pointing to the construction drawing for the wire numbers. The loop sheet owns what the signal is — tag, scaling, PLC address — and the construction drawing owns how it is wired. Two documents, no duplicated facts, and a night-shift trace that takes about 5 minutes instead of a 30-minute search through a 60-page package.
Do you draw loop diagrams for loops that never leave the cabinet?
Yes — one loop sheet per internal signal, marked “cabinet internal”, carrying the tag, the PLC address and the scaling, while the construction drawing keeps the terminal and DIP detail; a transmitter plus two UPS signals makes three sheets.
The test that decides it: start at the PLC point and ask whether one document gets you to the source. When your construction package already passes that test for a loop, an index line is enough. Drawings organized by terminal number rarely pass, because they answer “where does this wire go” and the question at 2 a.m. is “what does this signal mean”. So draw the sheets, and keep the two ideas separate: “loop” names a signal path, “field” names where part of it runs. A loop diagram is a control-loop document, the field is one section of it, and on these three loops that section is empty.
What is a loop diagram for when the construction drawing already shows the wiring?
It is a one-page route from the PLC point to the signal source, terminal by terminal, so one loop can be traced without the rest of the package; ISA-5.4 built the format around exactly that job, down to terminals, junction boxes and cable tags.
The two documents are organized differently, and that difference is the whole point. A construction drawing is organized by location: this cabinet, this terminal strip, this wire number. A loop sheet is organized by signal: this measurement, from that device, through these terminals, into that channel, scaled to these units. ISA-5.4 is the standard behind instrument loop diagrams, and its sheet follows one signal through every junction box and terminal block to the controller. Most loops you draw carry a field device, a field junction box, the panel and the controller — loops that run out through a junction box fill all four sections. A cabinet-internal loop fills the panel section and stops, and that is a complete, correct loop diagram.
What counts as a loop with no field wiring?
Any signal path a technician may have to trace: a source, its connection points, the PLC input, and the interpretation of the value — the cabinet temperature loop covers all four in about 600 mm of wire and earns its sheet the same way a 600 m field run does.
A wrong value has three places to originate, and the sheet names all three. The source: a transmitter ranged wrong, or a DIP switch still at its factory setting. The connections: a terminal left finger-tight at commissioning. The point: a channel wired one slot over from its address in the program, or a scaling block fed 0 to 100 when the device sends 0 to 60.
A UPS discrete signal carries the same chain in miniature — contact state, terminal pair, DI channel, point assignment — and you trace a disagreeing “on battery” indication exactly the way you trace a bad temperature reading. What the sheet leaves out matters as much as what it carries: a loop with no field section is drawn with no field section, so nobody goes hunting for a junction box that was never built.
What goes on the sheet for a cabinet-internal signal?
Six entries: tag and service, the source device, the terminal pair it lands on, the PLC card and channel, the signal range and scaling, and the drawing number that owns the wire detail — with the words “cabinet internal” printed on the sheet.
| What | Loop sheet | Construction drawing |
|---|---|---|
| Tag, service, function | owns it | shows the device |
| Scaling, units, expected values | owns it | — |
| PLC card, channel, address | owns it | shows the card layout |
| Terminal numbers (say TB4-13/14) | names the pair | owns the terminations |
| Wire numbers and ferrules | points to the drawing | owns them |
| DIP switches, device setup | points to the drawing | owns them |
The sheet only works when the metal agrees with the paper: the terminal labels and device tags in the cabinet must match what the sheet names, and the label checklist before the panel ships is where that gets caught. Scaling lives on the loop sheet because it is the number a technician computes against — 0 to 60 °C on 4 to 20 mA, and the expected current at today's temperature.
How do you keep the loop sheet from duplicating the construction drawing?
Give every fact exactly one owner and have the other document point to it by drawing number and revision: the loop sheet owns what the signal is, the construction drawing owns how it is wired, and no fact stands as the authority in both.
Two copies of one fact drift apart at the first field modification — one gets the red pen, the other stays at revision B, and from then on both are suspect. So split the facts once, at the start. A wiring change updates the construction drawing; a scaling or point-assignment change updates the loop sheet; each carries a note naming the other's new revision. If you keep a loop database as the plant's index, put every loop in it, internal ones included, and let each entry link to the current drawing. The drawing package scope settles that split before the first sheet is drawn, and what the training schematic includes answers the same question for the schematic side.
What does a real cabinet-internal loop sheet look like?
Take TT-101, a 2-wire transmitter on the cabinet wall reading 0 to 60 °C as 4 to 20 mA into a 24 V DC analog input: at a 35 °C interior, the scaling line on its sheet should read 13.3 mA at the channel.
The sheet, line by line:
- Tag: TT-101, cabinet interior temperature, 2-wire, powered from the 24 V DC supply
- Path: transmitter terminals 1-2, to TB4-13/14, to the AI card channel 5
- Scaling: 0-60 °C = 4-20 mA, which is 16 mA over 60 °C, or 0.27 mA per °C
- Live check at 35 °C: 4 + (35 / 60) × 16 = 13.3 mA
- Field section: none — cabinet internal
- Wire numbers and transmitter DIP switches: drawing E-102, rev C
The two UPS loops take the same shape with a state table in place of the scaling line: a 16-channel DI card, channels 9 and 10, dry contacts closing to 24 V DC, “UPS on battery” and “UPS on bypass”. The state table says contact open reads 0 and contact closed reads 1, so a disagreement between the UPS front panel and the PLC points at the contact or the point assignment in one look.
How do you check the sheet works during a night shift?
Run the trace test with a stopwatch: from the loop book alone, a technician should reach TT-101's terminal pair and its expected current in under 5 minutes, where the same question through a 60-page construction package costs about 30.
Then check the paper against the panel. At a 35 °C interior the AI channel should read 13.3 mA and the PLC should display 35 °C; a thermometer in the cabinet settles which side is wrong. On each UPS loop, compare the front-panel state with the PLC state — agreement points you at the UPS, disagreement at the contact or the point. Record where the chain diverges, source, connection or point, before you change a single setting. Lock out the panel before pulling any terminal; the 24 V DC loop you can measure in place with insulated probes. After any change, confirm the sheet still names the current revision of the construction drawing, because a pointer to rev C after the package moved to rev D is how good documentation quietly dies.
What belongs in the handover package?
Every loop in the panel, the three internal ones included, one sheet each, stored with the construction package they point to — and mounted where the night shift finds them, in a document pocket inside the cabinet door.
Split packages are how loop books go stale: the drawings live in one binder, the loops in another, and neither collects the field updates. Keep them as one package and give the loops a home at the panel — document pocket attachment methods put the book inside the door of a NEMA 4X stainless enclosure (rated by the manufacturer, not UL listed), dry and at hand. Counting the loops early also sizes the cabinet, since every loop is a terminal pair and a channel, and the enclosure selector turns that count into inside height and width. When you order, settle the terminal plan and cutouts up front with custom stainless enclosures; our custom boxes are delivered in 10 business days, counted from your approval of the drawing, with every opening already in place for the loops you documented.
Enclosures for this job
Catalog sizes that suit the job this article answers, picked from its subject; open one for its full specifications.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 60 × 36 × 16 in | 304 stainless | 60 x 36 x 16 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 48 × 36 × 16 in | 304 stainless | 48 x 36 x 16 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 60 × 36 × 12 in | 304 stainless | 60 x 36 x 12 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 12 × 8 in | 304 stainless | 16 x 12 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Should cabinet-internal loops go into the loop database or instrument index?
Yes. Keep every loop in the one index technicians search, internal ones included, so a point lookup never sends them digging through a separate PDF package. An entry can be three lines — tag, PLC address, “cabinet internal, see drawing E-102” — and it still saves the search. Index all 3 loops even though the wiring detail lives elsewhere.
How do you show an internal loop on the diagram itself?
Draw the panel run as a solid line and label the field section “none — cabinet internal” instead of leaving it dashed. A dashed line conventionally means a field conductor, and an internal run drawn dashed sends a technician hunting a junction box that was never built. One note on the sheet settles the convention for every reader who opens it.
Who owns the DIP switch settings when both drawings exist?
The construction drawing owns them, and the loop sheet points to that drawing by number and revision. Split the rest the same way: the loop sheet names the terminal pair, the construction drawing owns the terminations and wire numbers. A wiring change updates one document; a scaling or point-assignment change updates the other, and each notes the other's new revision.
How long should tracing one loop take with a good sheet?
Under 5 minutes from PLC point to source, working from the loop book alone. Without one, the same trace through a 60-page construction package runs about 30 minutes; across 40 panels that is 40 × 30 = 1200 minutes, 20 hours of night-shift searching one book removes. That gap is the reason the sheets exist, not construction convenience.
What current should a 0-60 °C transmitter loop read at 35 °C?
Expect 4 mA plus 16 mA spread over 60 °C, which is 0.27 mA per °C. At 35 °C, 4 + (35 / 60) × 16 = 13.3 mA; at 40 °C, about 14.7 mA. The scaling line on the loop sheet carries exactly these numbers, which is why the technician checks the sheet before the construction drawing.
