How to Use This Page
The thirty terms run in the order the old technology was actually assembled: the lamps themselves, the gear that fed them, the units people measured them with, the hardware that tamed their heat, the technologies that already left the building, and finally the sales language that outlived all of it. Start with the five that still do real work in a purchase conversation, then take the groups in whatever order your own room demands. Where this series has already settled a number, the entry links there instead of repeating it.
The 30 Terms at a Glance
The table below is the fastest way in: scan all thirty, then jump straight to the entry you need. Each row gives the full name or context, the unit it lived in, and the one thing the term settles for a buyer reading an old sheet today.
| Term | Full name / context | Unit | What it tells you |
|---|---|---|---|
| HPS (Single-Ended) | High-Pressure Sodium, mogul base | W · µmol/J | The workhorse lamp; 1.0–1.4 µmol/J in single-ended form |
| Double-Ended HPS | DE HPS, 2010s flagship | W · µmol/J | The 1.7 µmol/J peak that 2014's best LED only tied |
| Metal Halide (MH) | quartz arc tube lamp | K · hours | The blue-heavy veg lamp with the shortest service life |
| CMH / LEC | Ceramic Metal Halide | W · CRI | The 315W lamp with 90+ CRI and mid-era efficiency |
| Mercury Vapor & LPS | oldest HID family | none | The reason "HID" needed a quality warning label |
| Ballast | lamp-feeding gear | W loss | The 5–20% power tax the lamp rating never showed |
| Restrike Time | hot-restart delay | minutes | Why one outage darkened a room for a quarter hour |
| Warm-up Period | strike to full output | minutes | Why HID rooms never tested right after switch-on |
| Re-lamping Cycle | planned lamp replacement | hours | The maintenance line LED deleted: 10,000 h for HPS |
| Spectral Shift | aging lamp drift | none | Full watts, shrinking and drifting light |
| Footcandle | imperial illuminance | lm/ft² | The old room-sizing unit; 1 fc = 10.76 lux, eye-weighted |
| Candela | luminous intensity | cd | A spotlight metric; already on the ignore list in Part 1 |
| Lumens per Square Foot | HID-era sizing | lm/ft² | 2,000 minimum, 7,500 optimal, and wrong for plants |
| Watts per Square Foot | power rule of thumb | W/ft² | The 50 W HPS rule that became 30–40 W under LED |
| Grams per Watt | yield rule of thumb | g/W | The 1 g/W era benchmark that efficiency retired |
| Air-Cooled Hood | ducted sealed reflector | none | How HID rooms moved lamp heat out before LEDs ran cool |
| Cool Tube | glass cylinder hood | none | The close-mount answer that traded efficiency for distance |
| Sealed Reflector | enclosed hood | % | Caught more light, held more heat |
| Parabolic Reflector | wide spread hood | none | Soft wide coverage with a height penalty |
| Light Mover | motorised rail | m/min | Uniformity by motion, before arrays did it statically |
| T5 / T8 / T12 | fluorescent tubes | W · µmol/J | 0.9–1.3 µmol/J; the clone bench's last stand |
| CFL | compact fluorescent | W | The household spiral that taught a generation to buy photons |
| Induction & Plasma | electrodeless lamps | hours · W | Long-life and full-spectrum promises the market declined |
| Blurple | early red-blue LED | µmol/J | 1.0–1.5 µmol/J; the colour dates the fixture better than the label |
| Incandescent / PAR38 | grow bulbs | W | Heat lamps with a coloured coat; history only |
| Mogul Base | E39/E40 screw socket | none | The connector that identifies an HID fixture at a glance |
| Conversion Bulb | cross-ballast lamp | none | Running HPS light on a metal halide ballast, briefly |
| Ballast Bypass | retrofit wiring step | none | What the electrician does to the old gear before LED hangs |
| "1000W Equivalent" | marketing wattage | none | A comparison to a lamp, not a measurement of a fixture |
| Boost Mode | ballast overdrive | % | The old way to dim up: buy lamp life with output |
The Five Terms That Still Do Real Work
| Term | Unit | One line | Where it bites |
|---|---|---|---|
| HPS (SE / DE) | µmol/J | 1.0–1.4 single-ended, 1.7 the double-ended peak | The efficiency gap any retrofit quote must beat |
| Ballast | W of loss | 5–20% of the lamp rating spent as heat and hum | Hides behind every old lamp wattage line |
| Re-lamping cycle | hours | 10,000 h HPS, 8,000 h MH/CMH, drift before death | The running cost line LED quietly deletes |
| Watts per square foot | W/ft² | A thermal proxy from the HPS decades | Still quoted; convert to PPFD before believing |
| Ballast bypass | n/a | Disconnecting legacy gear so LED feeds clean | The electrical scope of a retrofit, priced from photos |
Everything else in this glossary is either the technology those five describe or the paperwork that sold it. The pairs most often confused, and wrongly interchanged, get their own table near the end.
The HID Lamps: What Hung Over Every Room
HPS, Single-Ended High-Pressure Sodium · mogul screw base
The lamp that built the industry. Sodium vapour under pressure emits a yellow-orange flood with spikes near 570–640 nm, and a single-ended bulb screwed into a mogul socket delivers it at roughly 1.0–1.4 µmol/J depending on age and ballast. The 2014 fixture testing cycle measured the common mogul-base lamps at about 1.02 µmol/J, which is the honest floor of the commercial era.
Reading it today: an SE HPS number on an old sheet describes photons per lamp-watt, before the ballast's share. Add 10–15% to the wall draw before comparing it with any LED figure: what a replacement actually involves.
Double-Ended HPS DE HPS · 2010s flagship
The lamp that held the efficiency crown for half a decade. Clamping the arc tube at both ends under higher pressure pushed double-ended lamps to 1.66–1.70 µmol/J, and in the 2014 testing cycle that figure tied the best LED on the bench. The tie is the hinge of the whole market: every year since, LED improved and the DE design did not, so the 1.7 peak is now roughly 40% behind the mainstream.
Reading it today: a quote praising "1.7 µmol/J, same as the old Gavita" is quoting a 2014 number as if time stopped. The same wall wattage buys 2.8–3.1 µmol/J in current mainstream LED, and the efficacy page shows the audit trail.
Metal Halide (MH) quartz arc tube · blue-heavy
The vegetative lamp of the HID pair. Metal halides in the arc tube emit a broader, bluer spectrum than sodium, and growers switched lamps at the flip to 12/12. The costs were steep: service life around 8,000 hours with faster output decay than HPS, a hot-restrike delay of 4 to 15 minutes, and spectrum that drifted as the lamp aged.
Reading it today: the "blue for veg" instinct survives, but the delivery changed. A dimmable white LED or an adjustable-spectrum bar covers both phases on one hanging structure, and the Part 2 entries on R:FR and stretch describe what blue actually steers.
CMH / LEC Ceramic Metal Halide · 315 W / 630 W
The last HID lamp worth taking seriously. A ceramic arc tube holds salts at a higher, more stable temperature, delivering 90+ CRI, meaningful UV content and a claimed 1.9–2.0 µmol/J, though the best independently measured fixture in the era testing read 1.46. LEC is a brand name that became a generic synonym; the lamp is the same either way.
Reading it today: growers who loved CMH loved the light quality, which is a real property and not a yield property. If a current quote leans on CMH, ask for a photon measurement rather than the CRI figure, and compare it against the DLC floor on the Part 1 PPE entry.
Mercury Vapor & LPS the oldest HID family · retired
The ancestors. Mercury vapour lamps lit the first generation of indoor grows with low efficacy and a spectrum plants use poorly; low-pressure sodium went the other way, extremely efficient at producing almost exclusively 589 nm yellow that photosynthesis treats as a narrow snack. Both are extinct in cultivation, and their presence on any current document is a museum label.
Reading it today: if a sheet still says mercury vapour or LPS, it is describing either a streetlight retrofit or a very old room. Either way the fixture is a candidate for full replacement rather than comparison: retrofit economics.
Ballast and Lamp Life: The Gear Behind the Lamp
Ballast magnetic vs digital · W of overhead
The component an HID lamp cannot live without: it strikes the arc with a high-voltage pulse and then strangles the current to a level the lamp survives. Magnetic ballasts did this with an iron core and coil, heavy, humming, and spending roughly 10–15% of the lamp rating as heat; digital ballasts do it with high-frequency switching at roughly 5% overhead, lighter and dimmable. The lamp wattage on an old sheet was never the wall draw.
Reading it today: when an old quote says 1,000 W, the wall often saw 1,100 or more, and the LED comparison must be made against wall watts, not lamp watts. The Part 1 watt entry covers the same trap on the LED side.
Restrike Time hot-restart delay · minutes
The waiting period before a hot HID lamp will strike again after losing power: roughly 4 to 15 minutes for metal halide while the arc tube cools toward a pressure the ballast can arc through. A momentary grid blink therefore cost an HID room its entire photoperiod segment, which is why legacy facilities invested in seamless transfer switches.
Reading it today: the term survives as a reminder of what was normal. An LED room restores instantly after an outage, and photoperiod integrity now depends on the controller's recovery behaviour rather than on lamp physics.
Warm-up Period strike to stable output · minutes
The ramp from ignition to full, stable output, several minutes for most HID lamps as the arc tube reaches operating temperature and pressure. Readings taken during warm-up sit well below the lamp's rating, which made quick spot checks systematically pessimistic and made "the room seems dim" a diagnosis that needed a quarter hour of patience.
Reading it today: LEDs reach full output instantly, so a modern PPFD map is valid the moment the room lights. If an old room's measured numbers disagree with its paperwork, warm-up and lamp age are the first two suspects.
Re-lamping Cycle planned lamp replacement · hours
The maintenance schedule that defined HID ownership: HPS lamps replaced at about 10,000 hours, metal halide and CMH near 8,000, on the industry rule that a 1% loss of light is a 1% loss of production. A 12-hour flowering room hits the HPS interval in 12 to 14 months. The lamp still lights at that age; it just no longer earns its electricity. Catalog rated-life figures for the same lamp family can print up to 24,000 hours, and the gap between that physics number and the 10,000-hour rule is economics: decayed output costs more than a fresh lamp saves.
Reading it today: this is the cost line LED deletes rather than reduces. When comparing quotes, the LED fixture's L90 figure replaces the whole re-lamping budget, and the Part 1 flux-maintenance entry explains how to read it.
Spectral Shift aging lamp drift · none
What an aging HID lamp does besides dim: the arc tube's chemistry changes, output drifts toward the far end of the spectrum, and the crisp blue and yellow peaks of a new lamp smear. The room draws identical watts while the light it buys quietly changes character, which is why "the lamp still works" was never a maintenance standard.
Reading it today: spectral shift is the HID cousin of the lumen-depreciation problem LED solved with published L90 data. On any fixture you buy today, the projection belongs on the sheet: how L90 and Q90 are earned.
Old Units and Rules: How Rooms Were Sized
Footcandle imperial illuminance · lm per ft²
The imperial sibling of the lux: one lumen falling on one square foot, equal to 10.76 lux. It measures brightness as a human eye experiences it, weighted hard toward yellow-green, so a footcandle figure built under an HPS lamp counts the yellow the eye loves and the plants merely tolerate. Whole rooms were sized on the 2,000 / 5,000 / 7,500 footcandle ladder.
Reading it today: there is no spectrum-independent conversion from footcandles to PPFD, so treat any offer to convert as an approximation tied to one lamp type. The room deserves a photon measurement: PPFD and DLI targets by stage.
Candela luminous intensity · cd
Eye-weighted intensity in a single direction, the unit of spotlights and flashlights. It earned its place on grow sheets in the reflector decades, when aiming light was the product. Part 1 of this series already sentenced it to the ignore list; it appears here only because old spec sheets and forum arguments still carry it.
Reading it today: a candela figure on a horticultural document is a signal about the document's age, not about the fixture. Skip to the photon numbers: Part 1, the ignore list.
Lumens per Square Foot HID-era sizing · lm/ft²
The rule that sized a decade of rooms: about 2,000 lumens per square foot as a minimum, 5,000 as workable, 7,500 and up as optimal for flowering. The numbers were internally consistent for HPS, which is why they worked, and meaningless across technologies, which is why they died. Lumens count light as eyes use it, and a blue-rich LED and a yellow HPS with equal lumens feed plants very differently.
Reading it today: any figure in lumens per square foot is describing the human experience of a room that no longer exists. The plant-side replacement is DLI: Part 1, lumen and lux.
Watts per Square Foot power rule of thumb · W/ft²
The most durable fossil in the industry. Fifty watts per square foot described a flowering HPS room accurately when lamps sat near 1 µmol/J; the same PPFD today takes 30–40 watts of LED. The rule was never about plant biology, it was an efficiency ratio frozen in time, and quoting it today without the technology context is quoting a 1990s room.
Reading it today: use it only as a sanity check on a quote, never as a sizing method. Pick the PPFD target, verify with a map, and let efficacy set the wattage: dim it or raise it.
Grams per Watt yield rule of thumb · g/W
The grower's scoreboard from the HPS decades: one gram of dried flower per wall watt as the benchmark of a competent room. It folded light, climate, genetics and skill into a single number, which made it socially powerful and scientifically blurry, because a gram earned under 1 µmol/J light and one earned under 3 µmol/J light are entirely different achievements.
Reading it today: efficiency retired the rule from the buying conversation. The modern equivalents sit in grams per joule of light delivered and the DLI the room banks, both traceable on the efficacy page.
Reflectors and Heat: The Hardware That Tamed the Lamp
Air-Cooled Hood ducted sealed reflector · n/a
A sealed glass-fronted reflector with duct ports, through which a fan dragged lamp heat out of the room before it ever mixed with the canopy air. It was the defining piece of HID plumbing: the room's cooling problem was so large that every fixture became part of the ventilation system, glass and all.
Reading it today: the hood retires with the lamp, and its ducting may still be useful for dehumidification or airflow. But LED changed the equation: the same photon delivery now needs a fraction of the exhaust, which is where the freed fan capacity gets reassigned: Part 2, air exchange.
Cool Tube glass cylinder hood · n/a
A bare glass tube around the lamp, with air pulled through it, that let growers hang the lamp close to the canopy without burning it. It solved the distance problem of a 600 or 1,000 W point source at the cost of trapping the light inside a curved, hard-to-clean cylinder, and it became the emblem of HID rooms that had run out of headroom.
Reading it today: close mounting is no longer a hazard budget but a design decision, because bar fixtures spread their light across an extended source. The Part 1 form-factor entry covers why bars behave differently at low height.
Sealed Reflector enclosed hood · capture vs heat
The enclosed reflector traded airflow for light capture: glass in front keeps every photon inside the optical system, a few percent better on paper, and keeps the lamp's heat there too. In practice the choice between open and enclosed hoods was a choice about which problem the room could afford, spill or heat, and both problems vanished with the lamp.
Reading it today: reflector capture percentages belong to a procurement conversation that no longer exists. Modern fixtures publish distribution curves instead: Part 1, beam angle.
Parabolic Reflector wide-spread hood · n/a
The deep, curved hood that spread a point source across a wide, soft footprint, at the price of mounting height and a few percent of capture to the lamp's own hardware. It was the honest answer to a real problem, one lamp per large area, and its geometry lessons survive inside modern LED secondary optics.
Reading it today: the problem it solved, wide soft coverage from one source, is now solved by arrays of small sources. Uniformity moved from the shape of the hood to the layout of the map: the nine-point grid.
Light Mover motorised rail · m/min
A rail and carriage that carried lamps slowly back and forth over the canopy, smearing each fixture's hot spot into an average. It was uniformity by motion: fewer lamps covering more area, with moving parts, cables and a maintenance schedule in the canopy zone. The sunlight-simulation argument around it was marketing; the hot-spot-smearing argument was real.
Reading it today: static LED layouts deliver better uniformity with nothing that moves, which retired the category. The modern version of the same purchase is fixture count versus uniformity grade: reading a PPFD map.
Retired Technologies: The Museum Wing
T5 / T8 / T12 Fluorescent tube lights · 0.9–1.3 µmol/J
The high-output tube, and specifically the T5, held the clone bench and the seedling tray for two decades: cheap, cool, even, and hopelessly inefficient at scale, the era testing reading fluorescent at about 0.95 µmol/J. Tube number was the diameter in eighths of an inch, a labelling system only a lighting historian could love.
Reading it today: the last legitimate fluorescent job, low-intensity propagation, is now covered by dimmable LED run low, on the same fixtures that will light the room later. The seedling targets are low enough that one fixture family can serve the whole cycle.
CFL compact fluorescent · W
The household spiral in a grow-shop box. CFL taught an entire first generation of growers to think about light in watts and colour temperature, and it delivered both badly: low photon flux, declining output, and a form factor that shaded everything it did not touch. Its cultural footprint outlived its horticultural one by a decade.
Reading it today: a "CFL-friendly" height chart is an artefact. If a current document still sizes anything in CFL terms, its PPFD numbers deserve the same suspicion as its units.
Induction & Plasma (LEP) electrodeless lamps · hours · W
The two electrodeless experiments, grouped because they lost for the same reason. Induction excited a fluorescent phosphor by magnetic field instead of electrodes, removing the filament that normally dies first: life was genuinely long, efficacy genuinely modest, and the price pointed genuinely skyward, so a decade as the "reliable alternative" never turned into market share. Plasma went further out: a tiny quartz capsule pumped by microwaves into a sunlike emitter, full spectrum and genuinely so, but expensive, hot, RF-noisy and thermally unforgiving. Both promised to leapfrog LED and both became cautionary tales about physics that demos well and scales badly.
Reading it today: nothing to read beyond the names. LED won induction's longevity argument so completely that the category folded, and every problem plasma promised to solve was solved by LED before it shipped at scale. A current sheet leading with either heritage is borrowing credibility from technology that no longer ships.
Blurple early red-blue LED · µmol/J
The slang for the first generation of horticultural LED panels, which mixed only red and blue diodes and glowed magenta, at roughly 1.0–1.5 µmol/J in the era's testing. The two-colour diet was a cost trick, not a spectral ideal: red and blue were the cheapest photons per mole, so early boards bought only the wavelengths the textbooks of the day praised, and the purple glow became the era's signature.
Reading it today: blurple is a date stamp, not a spec. Modern boards are white full-spectrum and are judged by measured SPD and PPE, not by how violet they glow; a listing still leaning on purple is telling you its age, and the efficacy page shows what replaced it.
Incandescent & PAR38 Grow Bulbs heat lamps in costume · W
The origin artefact: ordinary incandescent bulbs, sometimes with a coloured coating, sold as grow lights. Their photon contribution was a rounding error next to their heat contribution, but they established the retail pattern, a bulb in a box with a plant on the label, that the industry has been unlearning ever since.
Reading it today: purely historical, and useful as the bottom anchor of the spectrum: everything in cultivation lighting since has been an argument against measuring light in watts of heat.
Legacy Sales Language: What the Quote Still Says
Mogul Base E39 / E40 screw socket · n/a
The oversized screw socket that identified HID hardware at a glance, sized for the lamp's weight and heat rather than for anything optical. It is the fastest visual diagnostic in a legacy room: mogul sockets mean lamps, and lamps mean ballasts, reflectors and a re-lamping schedule somewhere in the budget.
Reading it today: LED deletes the socket along with the lamp, so a "compatible with mogul" claim on a retrofit product is nostalgia, not engineering. What matters is the mounting structure the old socket lived in: what gets reused and what does not.
Conversion Bulb cross-ballast lamp · n/a
A lamp engineered to run on the wrong ballast: HPS light from a metal halide ballast, or the reverse, so a grower could change spectrum without changing gear. It was a clever bridge across the lamp-ballast lockstep of the HID era, sold with the caveat-dense packaging that clever bridges usually carry.
Reading it today: the bridge is obsolete because the lockstep is: LED drivers accept their own power and are dimmed by signal, not by ballast chemistry. Spectrum changes now happen on the controller, not in the socket.
Ballast Bypass retrofit wiring step · n/a
The electrical work of taking the legacy ballast out of the circuit, by removal or disconnection, so the new LED fixture feeds from clean line voltage through its own driver. It is the standard first line of a retrofit electrical scope, and the reason retrofit quotes ask for photos of the panel: what the bypass costs depends on the room's wiring, not its floor area.
Reading it today: expect the phrase in any retrofit quote and price it as scope, not as padding. The freed electrical capacity after a bypass is often the quiet win: the electrical scope of a retrofit.
"1000W Equivalent" marketing wattage · none
The claim that a fixture "replaces" a 1,000 W HPS, quoted without photon numbers behind it. The comparison lamp was itself wildly variable, single-ended or double-ended, fresh or 10,000 hours old, and the replacement claim was never audited by anyone. Part 1 carries this phrase in its FAQ; it appears here because the legacy vocabulary is what made the claim uncheckable.
Reading it today: answer marketing units with measurement units: PPF, PPFD at a stated height, and a map. The translation is one line: Part 1, watt.
Boost Mode ballast overdrive · % above rating
The digital ballast's party trick: overdriving the lamp 10% or so beyond its rating for more output, at the cost of accelerated lamp decay. It inverted the logic that LED later made normal: HID dimmed down reluctantly and boosted up destructively, while a modern fixture dims to 10% or below without shortening anything but the electricity bill.
Reading it today: if an old sheet advertises boost, read it as a lamp-life surcharge, not a feature. On a modern quote the equivalent conversation is dimming range and efficacy at part load: Part 1, dimming.
The Terms That Get Confused
Most legacy-word arguments are two people using one word for two things, or one word across two decades. These five pairs cover most of them:
| The pair | The difference in one line |
|---|---|
| Candela vs footcandle | Candela is intensity in one direction; a footcandle is what lands on a square foot. Both are eye-weighted, and neither is a plant number. |
| CMH vs LEC | The same ceramic metal halide lamp; LEC is a brand name that became generic. The confusion is marketing, not physics. |
| Single-ended vs double-ended HPS | Same gas, different engineering: the double-ended format reached 1.7 µmol/J where common single-ended lamps sat near 1.0–1.4. |
| Restrike vs warm-up | Restrike is the delay before a hot lamp restarts after an interruption; warm-up is the ramp to full output after a successful strike. |
| Lumens per square foot vs PPFD | One counts light as human eyes spend it; the other counts photons as leaves spend them. No spectrum-independent conversion exists. |
Reading an Old Spec Sheet Today
This is where the vocabulary earns its keep: five lines that still appear on legacy paperwork, and what each one means in the units a modern quote should speak:
| The legacy line | What it means | What to ask for instead |
|---|---|---|
| "90,000 initial lumens, 2000K" | Eye-weighted brightness and a colour note for a lamp whose spectrum is fixed anyway | PPF in µmol/s and an SPD graph; the Part 1 lumen entry explains the gap |
| "1000 W double-ended, digital ballast" | About 1,700 µmol/s of light for roughly 1,050–1,100 wall watts once the ballast overhead is paid | The same wall watts in LED: 2,800–3,100 µmol/s at mainstream efficacy |
| "Coverage: 4 × 4 ft at 24 in" | A footprint valid at one height, one dim level and one reflector, none of them yours | A PPFD map run at your mounting height and dimensions |
| "Replace lamps every 12 months" | An admission that output decays, priced as an annual line item forever | An L90 projection at stated temperature, and no lamp budget at all |
| "Replaces a 1000W HPS" | Marketing wattage with no audited photon number behind it | PPF, efficacy in µmol/J and DLC listing: how to verify the claim |
A fair question follows: if these units are dead, why do product pages still print them? Because catalogs, marketplace filters and brochure copy move slower than the physics. A 2014-era template still sorted fixtures by watts and lumens, marketplace listings keep the search fields shoppers type into, and makers recycle old brochure lines onto new boxes. None of it means the legacy number drives the decision; the test is whether a photon-based figure sits beside it. Where watts and lumens stand alone, they are decoration, and the filter that matters is PPFD, efficacy and the certification marks.
One related check belongs to the modern sheet rather than the old one: the laboratory methods and standards behind a fixture's marks (LM-80 data, TM-33 files, UL 8800 testing, DLC listing) are the manufacturer's proof, not reading you owe time to. Skip the clause and check the mark: the certifications page lists which labels a cultivation fixture should carry.
What Comes Next
This closes the trilogy. Part 1 is the language of the light, the thirty terms a modern purchase runs on. Part 2 is the language of the room, the environment and engineering words from VPD to sensible heat ratio. This part cleared away the inherited vocabulary so the new numbers can be read without translation. The deep dives already published in this series cover PPFD and DLI targets by stage and grow light efficacy.
FAQ
Is HPS still worth buying in 2026?
Only in narrow cases: very short leases where the low upfront cost matters, or cold-climate rooms that genuinely consume the waste heat. A modern LED fixture converts roughly 60 to 80 percent more of each wall watt into photons than the best double-ended HPS, and unlike the lamp, the LED keeps improving while the HPS design has been static for a decade. For any facility with a horizon beyond three years, the arithmetic ends quickly.
What is a ballast bypass, and when does a retrofit need one?
A ballast bypass is the electrical work of disconnecting or removing the HID ballast so the circuit feeds the new LED driver directly. It matters when the LED fixture replaces the lamp in the existing position rather than hanging beside it: an old ballast left in line wastes power, generates heat and can damage the new driver. Whether a given retrofit needs a full bypass or a clean new circuit depends on the panel and the fixture layout, which is why the electrical scope is priced from photos of the room, not from its floor area.
Are CMH and LEC the same lamp?
Yes. Both names describe the ceramic metal halide lamp, which uses a ceramic arc tube instead of the quartz tube of a standard metal halide. LEC is a brand marketing name that stuck as a generic term. The lamp itself is real technology with strong colour rendering, but at roughly 1.5 to 2.0 umol/J it sits far behind modern LED, and its best independently measured fixture result in the 2014 era testing was 1.46 umol/J.
How do I convert footcandles to PPFD?
You cannot convert them exactly, because a footcandle is weighted for the human eye and PPFD counts photons between 400 and 700 nm equally. The ratio between them depends on the spectrum of the source, so a conversion factor built under an HPS lamp is wrong under white LED. Treat any footcandle figure as a hint about brightness for people, and ask for a photon measurement at the canopy instead.
How often should HPS lamps be replaced?
Commercial practice replaces HPS lamps at about 10,000 hours, which is 12 to 14 months in a 12-hour flowering schedule; metal halide and CMH run shorter at around 8,000 hours. The lamp keeps lighting long after that, but its output has already fallen 7 to 8 percent and its spectrum drifts, so the room pays full electricity for a shrinking crop response. Group re-lamping on a schedule replaced per-lamp replacement in serious facilities for a reason.
Is the 50 watts per square foot rule still valid?
It is a fossil with a history, not a specification. It described how much HPS wattage a canopy needed when the lamp converted about 1 percent of electricity into photons a plant can use. Modern LED needs roughly 30 to 40 watts per square foot to deliver the same PPFD, and the honest method no longer needs the rule: pick the PPFD target for the stage, verify it with a map, and let efficacy decide the wattage.
What is restrike time?
The delay a hot HID lamp needs before it can strike again after a power interruption: roughly 4 to 15 minutes for metal halide while the arc tube cools. It is why a brief outage darkened a whole HID room for a quarter hour, and why backup power for HID rooms had to be seamless. An LED room simply comes back on when the power does, which quietly removed an entire category of contingency planning.
What should I do with the old reflectors and ballasts after a retrofit?
The reflectors retire with the lamps: LED optics aim the photons at manufacture, so a legacy hood adds nothing and its reflectivity has been degrading anyway. The ballasts come out of the circuit either by removal or bypass, and the freed electrical capacity often lets the same panel feed more fixtures. Scrap value aside, the useful act is reclaiming the circuits, not the hardware.
Why do old quotes and new quotes use different numbers?
Because the industry changed how it counts light. Quotes written before roughly 2015 speak lumens, Kelvin, watts per square foot and lamp life; quotes written after speak PPFD, DLI, umol/J and L90 hours, with the DLC list as the verification anchor. The 2014 testing cycle is the hinge: the best double-ended HPS and the best LED of that era measured equal, and every year since has moved in one direction only. Ask any supplier holding the old vocabulary for the new numbers, measured at your canopy.