The unit price of a signage power supply is the number everyone negotiates and the number that matters least. On an outdoor installation with a five-year horizon, purchase price is frequently not even the largest line item — and the gap between the cheapest quote and the cheapest five-year outcome is often very wide, in the opposite direction from what the spreadsheet suggests.
This is a piece about the arithmetic that doesn’t make it into purchase orders. It uses no invented prices, because yours aren’t mine to guess. What it does instead is give you the structure and, more usefully, the break-even point — the single ratio that tells you whether paying more for reliability is the cheaper decision for your specific numbers.
A five-year cost model for installed signage power has five components:
Look at who budgets what. The two components that occur once get budgeted by someone. The three that recur — and that grow with every failure — typically get absorbed as “maintenance” and never modelled.
That’s the structural reason cheap units win procurement decisions and lose five-year ones. The person comparing quotes sees one number. The person paying for failures is a different person, years later, with a different budget line.
To be fair to procurement: they usually don’t have the failure data. Which is why the next two sections are about how to get it.
A power supply failure on a building-mounted sign is not a $40 event. It’s an event that requires:
The key structural point doesn’t depend on the exact figure: access and labour are fixed costs per visit, largely independent of how much the replacement unit costs. Replacing a $30 unit and a $90 unit costs exactly the same truck roll. Which means the cheaper unit’s price advantage is consumed entirely by the first failure, and then some.
This is also why mounting location is a procurement decision, not just an installation decision. A supply mounted where it can be swapped in ten minutes from a ladder has a completely different five-year cost profile than the identical supply mounted where it needs a bucket truck and a lane closure. If you can influence placement, that is worth more than most unit-price negotiations.
” What’s your failure rate?” is a bad question, and not because suppliers lie. It’s bad because the term has no shared definition.
One supplier’s failure rate counts units returned under warranty. Another counts units that failed on the test bench before shipping. Another counts field failures per thousand per year. Another counts anything a customer complained about, including units that were fine. You can get four honest answers to the same question and they will not be comparable.
Ask these instead, because they’re harder to fudge:
1. What percentage of production gets burn-in, at what load, for how long? This is a process question with a verifiable answer. Our published position is 100% full-load burn-in — every unit, at full load, before it ships. A supplier running burn-in on a sampled basis has a fundamentally different outgoing quality profile than one running it on everything, and that difference shows up in the first ninety days of field life, which is exactly when early-life failures cluster.
2. What’s your DPPM or PPM figure on outgoing product? Defective parts per million, measured against a defined window. If they can’t produce one, ask what they do measure.
3. What does your warranty claim rate actually look like by product family? Not overall. By family. Aggregate rates hide the one product line that’s causing the problems.
4. Can I see the QC gates? A supplier who can enumerate their test stages — and ideally walk you through them — is a supplier with a quality system. We’ve documented ours at eleven gates; see 11 QC gates including burn-in.
5. What’s the typical field life in an application like mine? Not the MTBF number. MTBF is a statistical construct that doesn’t mean “how long will this last.” Ask what they see in the field for comparable installations.
Two answers should raise your suspicion immediately: a supplier who quotes MTBF in the hundreds of thousands of hours as if it’s a life expectancy (it isn’t), and a supplier who can’t tell you their burn-in coverage.
A five-year warranty is not a five-year cost guarantee. It’s a five-year commitment with terms, and the terms are where the money is.
Our published warranty is 5 years. Here’s what you should establish in writing before relying on that number:
Does it cover the replacement unit only, or labour too? The common structure is: the unit is replaced free, the labour to swap it is not. On a building-mounted sign, that means the warranty covers the cheapest part of the failure and you absorb the expensive part. This single distinction is often worth more than the difference between a 3-year and a 5-year warranty.
Does it cover consequential damage or access equipment? Almost never. If a failed unit needs a bucket truck, the truck is yours.
What voids it? Water ingress is the usual exclusion that matters most in signage — which is awkward, because water ingress is the most common cause of failure in outdoor installations. Read the exclusion list specifically for ingress, condensation, and improper installation.
What’s the claim process and turnaround? A warranty that takes eight weeks to process is a warranty that doesn’t help you keep a sign lit. Ask how replacements are shipped and how fast.
The counter-intuitive summary: a longer warranty with narrow coverage is frequently worse than a shorter one with broad coverage. Compare what’s excluded, not how many years.
No invented prices. Instead, the break-even — which is more useful, because you can apply it to your own quotes.
Set up two options for N units over Y years:
Over Y years, expected failures are N × f × Y for each option. The difference in the number of truck rolls is N × Y × (f_A − f_B).
N × f × Y
N × Y × (f_A − f_B)
Now the key variable. Let K = the cost of one replacement event (labour, access, travel, downtime) divided by the unit price. K is dimensionless, which is what makes this useful — you can estimate it without knowing any absolute prices.
Option B’s premium is worth paying when:
The N cancels. So the test simplifies to:
Plug in something conservative. Say Option B costs 30% more (premium fraction = 0.30), over 5 years, and the failure rates differ by 2 percentage points a year (f_A − f_B = 0.02):
So: if one replacement event costs more than three times the price of a power supply, paying 30% more for a unit that fails 2 percentage points less often per year is the cheaper choice. On any installation requiring a bucket truck, K is comfortably above 3. On a ground-level, walk-up installation, it may not be — and there the cheap unit may genuinely be correct.
That’s the whole point of the model. The answer depends on your access cost, not on the unit price. Which is why “which is cheaper” is unanswerable from a price list alone, and why the right question to ask a supplier is not their price but their failure data.
For a detailed walkthrough of diagnosing field failures once they happen — which reduces the diagnosis-time component of K substantially — see field failure diagnostics.
Between “replace it when it fails” and “keep a warehouse full” there’s a right answer, and it’s calculable.
If you run N units with an annual failure rate f, expected failures per year are N × f. For a first-order quantity of 50 at a 1% annual rate, that’s 0.5 failures a year — which sounds like “don’t bother holding spares” until you remember that failures aren’t evenly distributed and the lead time on a reorder is 15 days plus shipping.
N × f
The two failure modes of spares planning:
Too few spares means a dark sign for two to three weeks waiting on a reorder. The cost isn’t the replacement unit, it’s the downtime — and on revenue-generating signage that dwarfs everything else in this article.
Too many spares ties up capital and, worse, risks obsolescence. Power supply designs get revised. A pallet of units bought as spares in year one may not match what’s installed in year four.
A workable approach:
That last point belongs in the quote comparison, not the maintenance plan. It’s item 10 on the list below precisely because it’s the one procurement never asks and maintenance always regrets.
Ten things to put on a single comparison sheet. Vendors who can’t answer one of these are telling you something.
Item 10 is the one that gets forgotten and then hurts. Signage projects get extended, phased, and repaired for years. A supply that can’t be reordered identically forces a redesign — and a redesign costs engineering time, not just unit price. Ask about product continuity and EOL policy.
Item 3 has a hidden trap too: a great first-order lead time and a terrible reorder lead time is a common pattern, because the first order got attention.
Our published baseline, for reference: 5-year warranty, MOQ 50 pcs, 15-day lead time, 100% full-load burn-in, CE/FCC/CB/UL certification, and 57 models across IP20 to IP67 shipped to 46 countries since 2009. Confirm any of these against your specific build before you rely on them.
Ready to run this properly? Send us your quantity, mounting conditions and access cost and we’ll work the five-year model against your actual numbers rather than averages — including the warranty exclusions in writing.
What percentage of total cost is the power supply itself over five years? It varies with access cost and failure rate, but on installations requiring elevated access equipment the purchase price is frequently not the largest component. Use the break-even model above with your own K value.
Is a 5-year warranty the same as five years of no cost? No. Most warranties cover the replacement unit and not the labour, access equipment, or consequential loss. On a building-mounted sign, labour and access are usually the larger cost. Always read the exclusions.
Why won’t suppliers just tell me their failure rate? Many will, but the term is defined inconsistently across the industry. Ask instead for burn-in coverage, outgoing DPPM, and warranty claim rate by product family — those are harder to define differently.
Does MTBF tell me how long a unit will last? No. MTBF is a statistical reliability parameter, not a life expectancy. A high MTBF does not mean a unit will run for that many hours.
When is the cheap unit actually the right call? When access is easy and the failure consequences are low — ground-level installations, easy swap access, non-critical signage. The break-even model above tells you where your situation falls.
Published references: 5-year warranty, MOQ 50pcs, 15-day lead time, 100% full-load burn-in, CE/FCC/CB/UL, 57 models, 46 countries, factory operating since 2009 with 12,000 m² ISO9001 facility and 8 SMT lines. Warranty exclusion terms, DPPM/failure rate data, burn-in duration and typical truck-roll cost pending confirmation — do not publish the worked example as fact until the failure-rate figures are supplied.
Looking for heavy-duty signage illumination? Explore our top-performing waterproof power solutions engineered for reliable outdoor performance.
“We’re UL certified” is one of the least informative sentences in power supply sourcing. UL publishes hundreds of standards. A power…
An IP65 rating does not mean a power supply is waterproof. It means it survived a specific test involving a 6.3mm nozzle, 12.5 litres per minute, from…
“Non-standard” is one of the most expensive words in power supply sourcing, mostly because nobody defines it. Two buyers can use the term …
A 600W 24V switching power supply pulls 25A at full load. That number is the whole problem. It is also the number most spec sheets bury somewhere belo…
The 20% headroom rule is not about future expansion. It is about keeping the driver between 60% and 80% load, which is where both the efficiency curve…
24 V is not automatically the better choice. Below about 3 metres and 30 W per run, 12 V modules cost less and you gain nothing from 24 V. The math be…
