Skip to content
Level 3 · AdvancedLessonPart 17 · page 2 of 760 minSafety level B · Advanced home laboratoryScienceCraft
60Minutes
10Sources
BSafety level

Safety level B, advanced home laboratory. Needs additional controls, experience and precautions beyond the standard darkroom: stronger ventilation, splash protection, careful handling of concentrated reagents or of energies such as UV and low-voltage electronics.

Switching Lamps Safely: Isolation, Certified Modules and the Mains Boundary

Somewhere in every darkroom-timer project there is a moment when a microcontroller has to make a lamp come on. What happens at that moment is the only genuinely dangerous decision in this whole part of the course, and this page makes it once, for every page that follows, so that no later page has to argue about it in passing.

The answer has two halves. The first has been the course’s rule since Part XIV and does not change: the student builds nothing at mains potential. The second half used to be “and where mains must be switched, a purchased certified appliance switch does it” — and on 5 September 2026 a market survey found that no such product could be bought. What happened next is the substance of this page, because the rule that governed it was written down first: if no acceptable product exists, the build changes, not the safety rule.

By the end you should be able to say what an isolation barrier actually protects, read a switching device’s datasheet for the four numbers that decide whether it can time an exposure, judge a product against criteria rather than against its photograph, and explain why the Pure Silver timer has one output instead of two.

Level B, on the rubric’s electronics criterion: low-voltage electronics builds where a certified power supply provides isolation from the mains, with no exposed mains conductors at any stage. Nothing on this page is a procedure for wiring, opening or modifying anything at mains potential, and the course describes live working nowhere.

What is not a hazard here, and why. This page itself puts you in front of nothing: it is reading and arithmetic. Where a reader does own a mains enlarger, the hazard is not created by anything the course asks them to do — the appliance was certified by its maker, it arrives with a moulded plug, and the instruction throughout is to leave it exactly as it is. The reason the page is Level B rather than Level A is what it discusses: a reader who took a sentence here as licence to open an enclosure would be doing Level C work, and the level is set high enough that the boundary cannot be crossed by accident.

HSE state the control in one line: limit the supply voltage to the lowest needed to get the job done. They also state, without qualification, that normal mains voltage — 230 V ac — can kill, and that the risks are greatest in wet surroundings.

A darkroom is a room with open trays of liquid in it, worked in the dark, by somebody whose hands are wet. It is close to the worst environment the guidance describes, and the course’s design answer is therefore not “be careful near the mains” but “have no mains in the instrument at all”.

Part XIV established what that means for the electronics and this page does not repeat it. What this page owns is the line itself: everything the student builds stays at extra-low voltage, and anything at mains potential is a purchased, certified, sealed appliance that is plugged in, used as its maker directs, and never opened.

The voltage bands, so that “low voltage” stops being a loose phrase

Section titled “The voltage bands, so that “low voltage” stops being a loose phrase”

In ordinary speech “low voltage” means “probably will not hurt you”. In the standards it is a band that includes the mains, which is why this course writes extra-low voltage every time and never uses “low voltage” as a reassurance.

HSE’s guidance on the Electricity at Work Regulations gives the figure. An extra-low-voltage system is one operating at or below 50 V ac or 120 V dc, and it is named there alongside the 55-0-55 V site supply as an example of a reduced-voltage system — the guidance recommending reduced voltages particularly for portable equipment and for highly conducting locations where the body may be damp and in large-area contact with the surroundings. Set that beside the same organisation’s statement that 230 V ac can kill and the two bands are plainly not the same thing.

Two more numbers from the same guidance are worth carrying. In a conventional dry working environment, a fault current limited preferably to 1 mA and certainly to no more than 3 mA will not usually present a risk of injury to a healthy person exposed occasionally and briefly — but the guidance adds, in the same breath, that even that may give a perceptible shock, and that a shock can cause a consequential injury such as a fall. On a stepladder aligning an enlarger head in the dark, the consequential injury is the one to think about.

What isolation is, and the thing everybody gets wrong about it

Section titled “What isolation is, and the thing everybody gets wrong about it”

An isolation barrier is a designed separation between the side of a circuit that may be at mains potential and the side a person or a microcontroller may touch. It is made of two things: material, which is what basic and reinforced insulation mean, and distancecreepage measured across a surface and clearance measured through the air.

This page prints no creepage or clearance figures. The equipment safety standard that gives them has not been consulted, and a millimetre invented here would be worse than none, because the required distance is a designed quantity rather than something you assess by looking at a board. What can be checked is whether the maker states one.

Two components make the crossing. A transformer passes power across the barrier through a magnetic field, with no conductor between the windings. An optocoupler passes a signal across it as light, with no conductor between the emitter and the detector — the glossary’s opto-isolation entry has the mechanism.

Here is what people get wrong, and it is the single most consequential misunderstanding in hobby mains switching.

Where the barrier is, and where it is not

The isolation barrierControl input1Opto2Side a person may touchCoil and contacts3L N4Mains potentialopen screwsNothing here stands between a hand and the terminals at 4.
  1. Control input, 3 to 5 V — the side a person may touch
  2. Optocoupler — the only crossing, and it carries a signal rather than power
  3. Relay coil and contacts — inside the barrier, at mains potential when the load is live
  4. Screw terminals, live and neutral — accessible, at mains potential, and outside every barrier on the board
The barrier is between two circuits. Adding a person to the drawing does not add them to the protection.

The four switching devices, compared for this one job

Section titled “The four switching devices, compared for this one job”

The job is specific and it is not “turn a lamp on”. It is: hold a lamp on for an interval computed somewhere else, repeatably, to a stated tolerance, in a wet room. Judged against that, the four candidates behave very differently.

An electromagnet pulls a contact closed. The contact has a current rating, it arcs as it opens and closes, and it has a finite life measured in operations. It is the cheapest thing that will switch a mains lamp and the least suitable for timing one, for a reason that has nothing to do with safety: the armature takes a few milliseconds to move and the contacts bounce on arrival, so the moment the lamp comes on is neither instantaneous nor exactly the same twice.

No mechanical relay datasheet is in this course’s corpus, and no contact-life or bounce figure is stated here. What is stated below, from datasheets that are in the corpus, is what a filament lamp does to a contact rating, and it applies to a mechanical contact more sharply than to anything else.

No moving parts: a semiconductor switch, controlled across an optical barrier. Nothing arcs and nothing wears out mechanically. Four numbers out of Omron’s datasheets decide whether one can time an exposure, and three of them are unwelcome.

Zero-crossing costs you timing resolution. A zero-cross type waits for the alternating supply to pass through zero before it turns on, which keeps the switching electrically quiet. Omron quote the price directly: operate and release times of one half of the load power source cycle plus 1 ms maximum. At 50 Hz a half cycle is 10 ms, so each edge can be up to 11 ms late, and since your command is not synchronised to the mains, the delay is different every time.

Off-state leakage means switched off is not disconnected. An SSR passes a small current when it is commanded off — 1.50 mA at 200 V ac for the G3MB, and 5 to 10 mA for the larger G3NA. Omron’s own precaution is the part that matters more than the number: the leakage flows through the device’s snubber circuit even when there is no power at the input, so the load side is treated as live until the supply is proved off.

Whether that leakage lights an enlarger lamp is a separate question, and it is worth doing the arithmetic because the claim is repeated everywhere. A 150 W lamp at 240 V draws 0.63 A and has a resistance at its operating temperature of 384 Ω. A leakage of 1.5 mA through it develops 0.86 mW; 10 mA develops 38 mW. Against 150 W those are 6 and 256 parts per million, and a cold filament’s resistance is lower still, which makes the dissipation smaller rather than larger.

And it gets hot. The G3MB’s on-state voltage drop is quoted at 1.60 V RMS maximum. At the 0.63 A a 150 W lamp draws, that is 1.0 W dissipated inside the relay, continuously, for the whole exposure — which is why panel-mounting SSRs are sold with heatsinks and why the small PCB parts are derated when several sit side by side.

A sealed device with a moulded plug on one end, a socket on the other and a low-voltage trigger input: you plug the enlarger into it, plug it into the wall, and close the trigger from the timer. Nothing is opened, nothing is terminated, and the whole mains path stays inside a product somebody certified.

This is the category the survey went looking for. Keep reading.

A sealed plug-in body that switches its socket on a command from a phone or a hub. The form is right and the function is not: the two makers’ documentation read in the survey lists no dry-contact or low-voltage input, and — separately fatal — publishes no switching latency at all. A device meant to time an exposure whose delay from command to contact is unpublished and unmeasured cannot be trusted with an exposure, and no measurement of that latency exists in this course. It remains on the part’s must-verify list as unmeasured rather than being asserted either way.

A cold filament is a short circuit that gets better

Section titled “A cold filament is a short circuit that gets better”

A tungsten filament’s resistance rises steeply with temperature, so at the instant of switch-on a cold lamp draws far more than its running current. That inrush is what welds an under-rated contact shut, and it is the reason a relay rated for 10 A of resistive load is not rated for 10 A of lamp.

The multiple and its duration are not sourced. No figure for the inrush of a cold tungsten filament at enlarger wattages was found in this course’s corpus, and none is stated here — it stays on this part’s must-verify list. What is sourced is better, because it is the number a designer actually uses: the manufacturers’ own derating for a tungsten load.

Device Rating, general purpose Rating, tungsten Derating
Omron G3MB-202P 2 A at 240 V ac 1 A at 240 V ac
Omron G3NA-210B 10 A device 4 A general use and tungsten 2.5×
Omron G3NA-240B 40 A device 6 A general use and tungsten 6.7×

A maker who halves — or sixths — their own headline rating when a filament is on the end of it is telling you what the inrush does, in the only units that matter. Read a switching device’s rating for your load type, not off the front of the box, and then leave margin on top of that.

An incandescent lamp is not a fast shutter

Section titled “An incandescent lamp is not a fast shutter”

A filament has thermal mass. It takes time to reach full output when switched on and time to go dark when switched off, and during those intervals the paper is receiving light at some intermediate rate.

No published rise or fall time for an incandescent lamp at enlarger wattages was found in this course’s corpus. That gap is why the calibration page measures light against time at the easel with the photodiode head rather than quoting a figure, and why this part’s claim of tenth-of-a-second resolution is written as a specification of the timer and not a promise about the print.

An LED head does not have the problem in the same form: its output follows its drive current, and Part XIV computed the low-side switch’s own limit at a gate time constant of 0.56 µs — six orders of magnitude below a tenth of a second. Whatever limits an LED head’s shortest honest exposure, it is not the transistor.

Discharge sources, and a claim the course will not make

Section titled “Discharge sources, and a claim the course will not make”

ILFORD state that MULTIGRADE RC papers suit tungsten and tungsten-halogen sources, cold cathode sources and LED heads designed for variable-contrast papers, and that other cold cathode and pulsed xenon sources may give a reduced contrast range. That is a statement about contrast, not about switching.

The course has no source for how a discharge lamp’s output behaves in the first fraction of a second after it strikes, and therefore makes no claim about the shortest exposure such a head can give. If you print with one, the honest procedure is the same one the calibration page uses for everything else: measure the light against time at the easel and let the measurement decide.

Certification marks, and what they are worth

Section titled “Certification marks, and what they are worth”

The marks on an appliance assert that somebody believes it conforms to something. That assertion is only checkable if you can find out what, which is why the survey’s second criterion asks the manufacturer to name the standard in literature a buyer can read before buying, rather than accepting a mark in a marketplace photograph.

Electrical Safety First give a householder two checks that need no literature at all, and they are the two the survey used:

  • The plug meets BS 1363, marked on the back of it.
  • The fuse is the correct size and meets BS 1362, marked on the fuse body.

They add a third worth knowing, because it catches most of what arrives from an overseas marketplace: an appliance that is not a shaver and is not meant to be hardwired should come with a standard United Kingdom plug or a conversion adaptor rather than a travel adaptor, and if it does not, it may not meet United Kingdom safety requirements. Their fuse guidance is that manufacturers have standardised on 3 A and 13 A, with 3 A named for a table lamp, a television, a computer or a soldering iron.

On recognising a counterfeit, the course found no authority beyond those checks and states so. What it will say is a matter of evidence rather than of forensics: a bare mark on a listing, with no named manufacturer behind it and no standard cited, is a Tier 3 source. Tier 3 is where you go to find something to look into, and it is never evidence for a safety claim.

Shock. Current through the body, and the path matters as much as the magnitude: hand to hand crosses the chest, and a wet hand makes a far better contact than a dry one. HSE’s figures above give the scale — 1 mA preferably, 3 mA at most, in a dry environment, for a healthy person occasionally exposed — and their own warning is that even a perceptible shock can cause a fall.

Fault current, and what an RCD does. A residual current device compares the current going out along the live conductor with the current coming back along the neutral and disconnects when they differ, because the difference is going somewhere else — through insulation, through earth, or through a person. What it does not do is protect you from a current that flows out and back correctly, which includes a shock taken between live and neutral, and it does not make a badly built circuit acceptable. It is a last line, and the course’s darkroom recommendation is that mains sockets in the room are on an RCD-protected supply as a condition of everything else rather than as a refinement.

Fire. A loose terminal is a resistance, a resistance carrying current is a heater, and the heater is inside an enclosure full of plastic. Electrical Safety First’s list of what not to ignore is the practical version: burn marks, arcing sounds, fuses blowing, breakers tripping, or a plug that feels hot. In a darkroom, add that the smell will reach you before the sight does, because the light is off.

Water. Everything above is worse in a room with trays in it. Keep mains equipment on the dry side of the room, off the floor, out of the splash zone, on a surface that does not get wet, with cables routed so that a drip runs away from a connector rather than into it.

Domestic fixed wiring — the sockets in the wall, the circuit that feeds them — is regulated almost everywhere. In the United Kingdom the national standard for electrical installations is BS 7671, named here by number and neither quoted nor summarised. Plug-in appliances are a different matter: nobody inspects them, and a person who builds one has taken on the responsibility that a manufacturer would otherwise carry.

That asymmetry is exactly why this course gives no mains wiring procedure. It is not that the work is beyond a competent person; it is that “competent” is a defined role with a scope, the course cannot assess anyone’s, and a procedure published to thousands of readers would be followed by the ones it was not written for.

A caveat about the sources on this page. HSE’s guidance on the Electricity at Work Regulations is addressed to work activities, and this course cites it for its technical statements — the voltage bands, the current thresholds, the note about coupling and leakage — rather than to suggest that those Regulations govern a home darkroom. Regulation differs by jurisdiction; readers outside the United Kingdom should substitute their own, and the substance of the rule survives translation because it is a design decision rather than a legal one.

The survey: what was looked for, what was found, and what changed

Section titled “The survey: what was looked for, what was found, and what changed”

Open question 1.4 asked whether a sealed, certified, low-voltage-triggered mains appliance switch could be bought in the launch market at a sensible price. It was answered on 5 September 2026, and the answer was no.

The five criteria, written before anything was looked at

Section titled “The five criteria, written before anything was looked at”

Each traces to a commitment the course had already made, so that the survey was a test rather than a shopping trip.

  1. A sealed, finished appliance the reader never opens. A moulded plug and a socket or IEC outlet, with no mains conductor to terminate, no screw terminals, no cover to remove.
  2. A genuine certification mark against a named standard, named by the maker in literature a reader can read before buying, plus the BS 1363 and BS 1362 checks above.
  3. Isolation between the trigger and the mains side, stated by the manufacturer — a figure or a named clause, never inferred from a photograph of an optocoupler.
  4. A current rating that covers an enlarger lamp with margin, stated for a lamp or resistive load. The load was established first: a 240 V, 150 W screw lamp is 0.63 A steady, and RH Designs put the same range from the other side, saying most enlargers are 100 to 250 W.
  5. A trigger that carries the timer’s own interval — a documented dry contact or logic-level input which holds the output on while asserted. An input that starts a program the other device times does not meet this, and that distinction is the whole point.

And a price test, relative rather than invented. The mains route is worth building only if it costs materially less than the cheapest certified route that already works without it — which on the day was a commercial enlarger timer with its own switched socket at £112.20. A switch priced near that fails, because the honest instruction would then be to buy the timer.

Category What was found Criteria failed
Relay modules with optional cases Boards with screw terminals; the case is an accessory, no standard named, no isolation published 1, 2, 3
The right shape, wrong market A North American extension-cord relay of exactly the correct form, discontinued, 120 V, no certification claimed on the maker’s page; and its named replacement, likewise 2, 3, and 1 for a British reader
Commercial darkroom timers Plentiful, sealed, moulded plug, “do not open”, 500 to 750 W of switching — and no trigger input that carries an outside interval 5
Solid-state relays in IP-rated housings A component, not an appliance. An IP rating describes ingress, not a plug, a socket or a fuse 1
Smart plugs Sealed and plug-in, but no dry contact and no published latency; the one model that does have a potential-free input arrives on screw terminals for an installer 1 or 5, depending on the model
Stage-lighting relay pack The nearest miss: sealed, metal-housed, commanded entirely over a low-voltage data link, £95.00 — and on Schuko connectors, with no standard named and no isolation stated 1, 2, 3, and 5 in the form needed

The commercial timers are the row that decides it, and the manufacturers’ own manuals are what refuse them. RH Designs describe their optional foot switch as one that exactly replicates the operation of the Start/Stop key; on their Timer 3 it has precisely the same effect as pressing the Control Dial (except setting exposure). That parenthesis is the finding. A foot switch starts the host timer’s own programmed exposure; it cannot carry an interval computed anywhere else. Kaiser’s sheet for their 4030 lists a 500 W maximum switching capacity, 0.1 to 99 seconds and a focusing light switch, and no external trigger input of any kind.

The safety rule did not move. The build did.

What light source have you got, and what switches it

  1. A low-voltage LED head, safelight, contact printer or UVA arraybuilt in Part XVI. The timer switches it directly through a logic-level MOSFET. Nothing is at mains potential.
  2. A legacy mains enlarger, route one: convert itreplace the lamphouse with the low-voltage LED head of Part XVI. The mains problem is abolished rather than solved.
  3. A legacy mains enlarger, route two: switch it by handleave the enlarger on its own switch and count the exposure by ear on the metronome. Costs resolution, costs nothing else.
  4. A legacy mains enlarger, route three: buy the timera certified enlarger timer with its own switched socket owns the mains, £112.20 to £429.00 on 5 September 2026, while the Pure Silver timer keeps the low-voltage heads.
  5. There is no fourth routeno relay board, no smart plug, no home-built appliance, and no procedure for any of them anywhere in this course.

The instrument this part builds therefore has one output. Its cost is real and worth naming: a reader with a tungsten enlarger and no wish to convert it cannot use the Pure Silver timer to time their prints, and must either count by ear or spend at least £112.20. That is a worse outcome for that reader than the two-output design would have been — if the two-output design had been buildable to the standard the course set, which it was not.

  • The course’s line: everything the student builds stays at extra-low voltage; anything at mains potential is a bought, certified, sealed appliance, used as directed and never opened.
  • HSE put extra-low voltage at 50 V ac or 120 V dc and below, and 230 V ac in the band that kills. The everyday phrase “low voltage” belongs to the second band, not the first.
  • An isolation barrier separates two circuits. An optocoupler on a relay board protects your microcontroller and does nothing at all for your hand.
  • A zero-cross solid-state relay quantises every edge by up to 11 ms at 50 Hz, which is 0.30 stops on a tenth of a second and 0.0016 stops on twenty seconds. Suitability depends on the exposure.
  • Manufacturers derate their own products by two to nearly seven times for a tungsten load. That derating is the usable evidence about inrush; the inrush multiple itself is not sourced in this course.
  • No incandescent rise-and-fall time and no paper reciprocity figure are quoted here, so tenths of a second remain a specification of the timer until the calibration page measures the system.
  • The survey of 5 September 2026 found no acceptable product, so the mains half of the build was removed. Three routes remain for a legacy enlarger, and there is no fourth.

Check your understanding

Question 1. A relay board is advertised as opto-isolated and a photograph shows the optocoupler clearly. What does that establish about using it to switch an enlarger lamp?
Show the answer and why

Answer: That a fault on the mains side is unlikely to travel back into the microcontroller — and nothing at all about the exposed screw terminals, which are outside every barrier on the board

Isolation is a property of a barrier between two circuits. The optocoupler stands between the control input and the relay coil, so it protects the controller and the USB cable behind it; the mains terminals on top of the board are on the far side of that barrier and are accessible to a hand. A photograph is also not a certification: the survey's second and third criteria ask the maker to name a standard and state an isolation figure, and a hobby board's listing does neither. The 2500 V ac figure in the last option is real but belongs to a named Omron part with a datasheet, not to an unidentified board.

Question 2. You are offered a zero-cross solid-state relay for a timer that is meant to resolve a tenth of a second on a 50 Hz supply. What is the objection, quantitatively?
Show the answer and why

Answer: Each edge can be up to a half cycle plus 1 ms late — 11 ms — and because the command is not synchronised to the mains that is up to ±11 % of a 0.1 s exposure, or about 0.30 stops of spread

Omron quote operate and release times of one half of the load power source cycle plus 1 ms maximum for their zero-cross models, which at 50 Hz is 11 ms; the same datasheet quotes 1 ms for the non-zero-cross variant, and their precautions say to choose that variant where the switching instant matters. Eleven milliseconds is 0.0016 stops at twenty seconds and 0.30 stops at a tenth of a second — nine tenths of a third-stop test-strip band — so the device is fine for long exposures and unusable for short ones. Leakage is a separate issue and the arithmetic on this page says it does not light a 150 W filament.

Question 3. A solid-state relay is rated at 10 A. Why is a 150 W tungsten enlarger lamp, drawing 0.63 A, not obviously comfortable on it?
Show the answer and why

Answer: Because a cold filament draws far more than its running current at switch-on, and Omron themselves rate that 10 A device at 4 A for a tungsten load — a derating of 2.5 times, with their 40 A part derated to 6 A

A filament is resistive, not inductive, but its resistance rises steeply with temperature, so the inrush at switch-on is far above the running current. This course could not source the inrush multiple or its duration at enlarger wattages and states no figure for it — but the manufacturers publish their own answer in the form that matters, a separate and much lower rating for a tungsten load. The G3NA-210B is a 10 A device rated 4 A for tungsten and the G3NA-240B a 40 A device rated 6 A, so the derating runs from 2.5 to nearly 7 times. Read the rating for your load type.

Question 4. Every commercial darkroom timer surveyed has a foot switch, which is a low-voltage input on a sealed certified appliance that switches mains. Why did none of them satisfy the fifth criterion?
Show the answer and why

Answer: Because the foot switch starts the host timer's own programmed exposure rather than holding the output on while it is asserted, so the interval delivered is the host's and not the Pure Silver timer's

RH Designs answer this in their own manuals: the optional foot switch exactly replicates the operation of the Start/Stop key, and on the Timer 3 it has precisely the same effect as pressing the Control Dial (except setting exposure). Pressing start is not the same thing as holding an output on for an interval computed elsewhere — with a foot switch the Pure Silver timer would be a button, and the exposure would be the commercial timer's. Bounce is a real but solvable problem, and isolation and pinout are not what the criterion asks about.

Question 5. The survey found no acceptable product. Which of these is the course's response, and why?
Show the answer and why

Answer: Remove the mains half of the build, publish the survey with its criteria and its refusals, and give a legacy enlarger three routes — because the rule that if no acceptable product exists the build changes rather than the safety rule was fixed before the search began

The order of events is the whole argument. The criteria and the consequence were written down before anything was looked at, precisely so that an unwelcome answer could not be met by relaxing the test. Deferring would leave a published design promising a feature that cannot be built, which is why the manifest's own wording was corrected rather than left standing. And building one is Level C work the course has refused everywhere else: assembling a mains appliance and calling it certified is exactly the thing the first criterion exists to prevent.

Sources for this page

10 cited · checked 2026-09-05

  1. 01Electrical safety and you: A brief guide, INDG231(rev1)Health and Safety Executive, 2012§ What are the hazards - contact with live parts causes shock and burns, normal mains voltage of 230 volts AC can kill, and the risks are greatest in wet surroundings; Reducing the risk, Reduce the voltage - limit the supply voltage to the lowest needed to get the job donehse.gov.uk/pubns/indg231.pdftier 1, primary2026-09-05
  2. 02Memorandum of guidance on the Electricity at Work Regulations 1989, HSR25, third editionHealth and Safety Executive, 2015§ Paragraph 142, use of reduced voltages, which are particularly appropriate for portable equipment and in highly conducting locations where the body may be damp; paragraph 143, which gives an extra-low-voltage system as one operating at or below 50 V ac or 120 V dc and notes that such systems are earth-referenced special cases; paragraph 146, current limitation, where a current limited preferably to 1 mA and certainly to no more than 3 mA in a conventional dry environment will not usually present a risk of injury but may still give a perceptible shock and a consequential injury such as a fall; paragraph 147, separated or isolated systems, that all systems are to some extent referenced to their environment by capacitive or inductive coupling or by leakagehse.gov.uk/pubns/priced/hsr25.pdftier 1, primary2026-09-05
  3. 03Plugs and fuses: usage and safetyElectrical Safety First§ Checking the plug and cable, and Checking inside the plug - check the plug meets British Standard BS 1363, marked on the back, and that the fuse is the correct size and meets British Standard BS 1362, marked on the fuse body; appliances other than shavers and hardwired items should come with a standard UK plug or a conversion adaptor rather than a travel adaptor, and otherwise may not meet UK safety requirements; the signs of overheating and the instruction not to ignore burn marks or arcing sounds; and the standardisation of plug fuses on 3 A and 13 Aelectricalsafetyfirst.org.uk/guidance/safety-around-the-home/plugs-and-fusestier 2, specialist2026-09-05
  4. 04BS 7671, Requirements for Electrical Installations, IET Wiring Regulations, 18th editionInstitution of Engineering and Technology and BSI, 2018§ Cited by number only, for the publisher's statement that BS 7671 is the national standard for electrical installations in the United Kingdom; no requirement is reproduced or paraphrasedelectrical.theiet.org/bs-7671tier 1, primary2026-09-05
  5. 05Solid State Relay G3MB, PCB-mounting SSR, data sheetOmron Electronic Components LLC§ Load ratings by load type for the G3MB-202P - 2 A at 240 VAC general purpose against 1 A at 240 VAC tungsten; operate and release times of one half of the load power source cycle plus 1 ms maximum for the zero-cross models against 1 ms maximum for the non-zero-cross model; leakage current 1.50 mA at 200 VAC; output ON voltage drop 1.60 V RMS maximum; insulation resistance 1000 megohms minimum at 500 VDC and dielectric strength 2500 VAC for one minute; and the precaution that a model without a zero-cross function must be chosen for phase controlomronfs.omron.com/en_US/ecb/products/pdf/en-g3mb.pdftier 1, primary2026-09-05
  6. 06Solid State Relay G3NA, panel-mounting SSR, data sheetOmron Electronic Components LLC§ Load ratings by load type - the G3NA-210B rated 4 A for general use and tungsten against a 10 A device rating, the G3NA-220B and G3NA-225B likewise at 4 A, and the 40 A G3NA-240B at 6 A; leakage current 5 mA maximum at 100 VAC and 10 mA maximum at 200 VAC; and the precaution that leakage current flows through the snubber circuit even when there is no power input, so the supply must be proved off before wiringomronfs.omron.com/en_US/ecb/products/pdf/en-g3na.pdftier 1, primary2026-09-05
  7. 07StopClock Professional and StopClock Vario instructions, issue 7RH Designs, 2019§ Package contents; the statements that StopClock functions as a mains switch and can control loads of up to 750 W, with most enlargers at 100 to 250 W; the warning that the appliance must be earthed, the 5 A plug fuse and internal T4A; the instruction not to open the unit or remove any covers; and the description of the optional foot switch as exactly replicating the operation of the Start/Stop keyrhdesigns.co.uk/wp-content/uploads/2020/04/SC-Pro-v85.pdftier 1, primary2026-09-05
  8. 08RH Designs Timer 3 instructionsRH Designs§ Package contents - the Timer 3, a mains lead fitted with a plug and one IEC connector; and the description of the foot switch as having precisely the same effect as pressing the Control Dial, except setting exposurerhdesigns.co.uk/wp-content/uploads/2020/10/Timer3-Manual.pdftier 1, primary2026-09-05
  9. 09'digital timer' Electronic Exposure Timer for enlargers, article 4030Kaiser Fototechnik GmbH und Co. KG§ Specification - exposure times 0.1 to 99 seconds, maximum switching capacity 500 W, focusing light switch and 2 m power cable, with no external trigger input listedkaiser-fototechnik.de/en/produkte/2_1_produktanzeige.asptier 1, primary2026-09-05
  10. 10MULTIGRADE RC Papers, technical informationHARMAN technology Limited (ILFORD Photo), 2020§ Exposing light sources - designed for enlargers with tungsten or tungsten-halogen sources, also suitable for cold cathode sources and for LED exposing heads designed for variable contrast papers, while other cold cathode and pulsed xenon sources may give a reduced contrast rangeilfordphoto.com/wp/wp-content/uploads/2021/01/MULTIGRADE-RC-Papers-J20.pdftier 1, primary2026-09-05

Formulas, hazard statements, historical dates and process descriptions on this page were checked against the sources above on the date shown. Safety data changes: obtain the current safety data sheet for the product you actually buy before you open it.