Nov 02,2026
A three-step procedure our field engineers run on a customer site in about 20 minutes. It replaces the calendar with a number you can put in a budget.
Two clarifications before the procedure, because most of the optics quotes we are asked to explain start with a vocabulary problem.
The lens is almost never the first thing to fail. The window in front of it fails first, and it fails cheaply. A $28 window left in place for 400 hours instead of 120 will take a $420 focus lens with it. That single sentence is the entire cost story of laser cleaning optics.
And "the lens" is not one part. It is four parts in a beam path, with four different lifetimes and four different prices. Step 1 names them. Step 2 measures your own consumption rate instead of quoting a brochure. Step 3 converts that rate into an interval, a spare-parts kit and an annual number.
| Part, in beam order | Price | Life | Consumable? |
|---|---|---|---|
|
1. Protective window Last glass before the workpiece — takes spatter and fume |
$18–$42 | 40–400 h | Yes |
|
2. Focus lens Converges the beam; sets spot size and energy density |
$180–$620 | 1,500–4,000 h | Wear part |
|
3. Collimator Straightens the divergent beam out of the QBH |
$320–$780 | 3,000–6,000 h | Wear part |
|
4. Galvo mirrors (X/Y) Steers the spot across the scan width at kHz rates |
$520–$1,300/pr | 6,000–10,000 h | Semi-durable |
Table 1 — Spare-part prices from our own list, Q4 2026. Hour figures are field data from our installed base, not warranty terms; your fume load will move them.
Half of all arguments about optics cost come from a service report that says "lens replaced" next to an invoice for $620, when the part that actually died was a $28 window. Fix the vocabulary and the budget becomes predictable.
The protective window is a sacrificial part. It is supposed to get dirty, and it is supposed to be thrown away. It sits in front of the focus optics precisely so that spatter lands on $28 of fused silica instead of $420 of polished quartz. When a supplier advertises that their machine "hardly ever needs the window changed", they are telling you the window is doing less work than it should — and that the lens behind it is doing more.
Open the torch head with the 5 mm key and photograph all four surfaces in the same order every time: window, focus lens, collimator, mirrors. Date the photos. This takes six minutes once, and it gives you a baseline that no service engineer can argue with eighteen months later.
Do not ask "how often". Ask "how many hours", then divide. The window is the only part you need to model, because everything behind it is priced off the window's behaviour.
| Typical job | Avg power | Window life | Windows / 1,000 h |
|---|---|---|---|
| Mould and tool cleaning, light oxide | 200 W | 400 h | 2.5 |
| Light rust, thin section, indoor | 300 W | 265 h | 3.8 |
| Structural steel, medium rust | 500 W | 160 h | 6.3 |
| Heavy scale plus paint, 3 shifts | 800 W | 100 h | 10.0 |
| Continuous 1,000 W, high fume | 1,000 W | 80 h | 12.5 |
Table 2 — Window consumption by application. Two buyers with identical machines can honestly report lifetimes that differ five-fold; this table is why.
That five-fold spread is the reason brochures are useless here and your own log is not. Three lines a day is enough:
After two weeks you own a number. After two months you can predict a replacement two weeks before it happens, which is the difference between a $28 purchase order and a $1,050 emergency.
Now convert hours into a schedule and a purchase order. The inspection cadence matters as much as the replacement cadence: a window is cheap to replace and expensive to inspect too late.
| Part | Replace when | Inspect | Spare & cost |
|---|---|---|---|
| Protective window | Every 80,000 ÷ P hours, or at the first tint | Every shift — 20 s with a torch | 5 in drawer · $120 |
| Focus lens | 2,400 h, or after 3 window burn-throughs, or 15% power loss | Monthly burn test | 1 spare · $420 |
| Collimator | 5,000 h | Annually, with the lens out | On call · $560 |
| Galvo mirrors | 8,000 h, or at a change in scan pitch | Quarterly — air only, never a cloth | On call · $900/pr |
Table 3 — Replacement and inspection schedule. The 5-pack is the only line most shops need to hold as inventory.
Then price it per operating hour, because that is the only figure that survives a comparison between two machines:
| Line item, per 1,000 operating hours | Window kept on schedule | Window run to failure |
|---|---|---|
| Protective windows | $233 (8.3 × $28) | $70 (2.5 × $28) |
| Focus lens | $150 (1 per 2,800 h) | $1,050 (1 per 400 h) |
| Collimator | $112 (1 per 5,000 h) | $467 (1 per 1,200 h) |
| Galvo mirrors | $113 | $113 |
| Optics subtotal | $608 | $1,700 |
| Downtime, 3.5 h per event at $180/h | $0 | $1,890 (3 events) |
| Total per 1,000 h | $608 → $0.61/h | $3,590 → $3.59/h |
Table 4 — Same machine, same parts list, two maintenance habits. Downtime valued at $180/h is deliberately conservative; most job shops quote far higher for a stopped cell.
Put the number in the quote sheet next to electricity and labour. A shop running 1,200 hours a year on a 500 W unit should budget about $730 a year for optics — roughly 0.6% of a $120,000 machine. It is a smaller number than most buyers expect, and it stops being small the moment the window is treated as optional.
You will not always have the log open when something changes. These five observations cover almost everything we see in the field, and each one points at a specific part rather than "the lens".
| What you see | What it usually means | Do this |
|---|---|---|
| Cleaning stripe is dimmer at both edges | Window edge contamination, or fume baked onto the outer ring | Clean and re-test; if unchanged, replace the window — do not chase it with power |
| You keep dropping speed 10–20% to hold the same result | 12–18% transmission loss somewhere in the chain | Replace the window, then burn-test the lens. Never compensate with power alone |
| Gold, brown or milky tint on the window | Coating burned. The glass is now a heat source | Replace immediately. Do not clean it — cleaning removes the last of the coating |
| Dark ring or pit on the focus lens | A window ran past its failure point and dumped energy into the lens | Replace the lens and inspect the collimator in the same session |
| Random dark specks scattered through the stripe | Debris on a galvo mirror | Air only. A cloth on a galvo mirror turns a $900 part into a $900 part plus a realignment |
Table 5 — Field diagnosis. Note that four of the five end at the window or the lens; the galvo is rarely the culprit and always the most expensive guess.
Replacement is not a date on a calendar. It is 80,000 divided by your average power, plus a $120 five-pack in the drawer and a monthly photograph of a burn stripe.
Change the window on hours and the optics line on your P&L settles at about $0.61 an hour. Skip the window and it becomes $3.59 an hour — and the machine spends a shift standing still while somebody drives to a spare-parts shelf that should have been stocked. The lens is not the expensive part. Neglecting the lens is.