Walk up to a used CNC machine and the first number anyone offers is the hour meter. "This VF-2 has 8,500 hours." You do the arithmetic against its age, decide whether the ask feels fair, and move on.
That number is worth having. It is not worth what most buyers pay for it, because a single hours figure conflates two different clocks and tells you nothing about the variable that actually consumes bearing life.
The two clocks
Machine controls track power-on hours and spindle hours separately. Ask for both, every time.
Power-on hours count how long the control has been energized. That clock runs while the machine is cutting, while it sits between jobs, while somebody is at lunch, and while an operator is fixturing the next part. It is closer to "how long has this thing been plugged in and switched on" than to "how much work has it done."
Spindle hours count how long the spindle has actually been turning. That is the meter tied to the component whose failure defines the economics of the whole machine.
The gap between them is the useful part. A machine showing 8,500 power-on hours and 4,000 spindle hours has spent more than half its energized life not cutting, which is entirely normal for job-shop work and tells you something real about how it was used.
If the seller does not know the spindle hours, ask them to pull both off the control. If they cannot find the readout, note it. That might mean the control has been reset, or it might mean the machine predates useful spindle tracking. Either way it is a fact about your information, and it should show up in the price.
Why the spindle is the whole argument
The reason to care this much about one meter is that on these machines, one component carries most of the financial risk.
Specialist-published spindle economics for the Haas VF-series, which is the best-documented line in our records: a rebuild runs $4,800 to $7,200, and a full replacement cartridge runs $16,000 to $24,000 before install labor. Set that against the market. Our VF-2 record shows 1994 to 2008 machines trading at $8,900 to $12,500. On an older machine at those prices, a spindle event is the machine's entire value, and possibly more than it.
So the hour meter is not a general condition score. It is your cheapest available proxy for the odds on that one repair. Read it that way and the follow-up questions get obvious.
Hours are the wrong unit. Work is the right one.
A bearing does not track hours. It accumulates damage from load and speed and heat, and those vary enormously between two machines showing identical meters.
A machine that ran 12,000 hours cutting aluminum at moderate loads in a clean production shop can genuinely have more life left in it than one that ran 5,000 hours hogging titanium at the top of the spindle range. This is the consistent testimony across our evidence set, and it is why "low hours" on its own is not a selling point.
We want to be careful about how far we push that, because we have not sourced a published wear curve and we are not going to invent one. What we can say is that the relationship is not one-to-one with hours, that load and speed drive it, and that no honest source we found publishes a formula you can apply to a specific machine.
One directional data point from our video evidence, and we are flagging it as directional rather than as a life expectancy: a shop technician working on a VF-SS reported that the spindle bearings failed at roughly the one-year mark. On a high-rpm spindle variant, bearing service life can be dramatically shorter than the structural life of the machine around it. Which means a low-hours machine with a high-speed spindle and an aggressive work history is not automatically the safer buy.
So the second question, right after the two meter readings: what was this machine cutting, at what speeds, and how hard? Aluminum or steel or titanium? Full production shifts or intermittent job work? The meters give you a denominator. The work history gives you the numerator.
What the meters cannot tell you at all
Here is the limit worth internalizing. No hour reading tells you what parts are inside the machine.
Our best-documented failure case in this vertical is a Fadal thrust bearing repair where the original part number is 7304B and a cheap aftermarket substitute, part 76034, turned out to be a general ball-screw bearing with a 45-degree contact angle where the correct spec is 60 degrees. The technician caught it on camera before installing it. A different technician, on a different machine, five years ago, might not have. A machine carrying a wrong-spec bearing reads exactly the same on the hour meter as one carrying the right part.
The same evidence set shows failures arriving in clusters rather than singly. One Fadal spindle teardown turned up an unexpected bearing failure plus two seals that had lost their shape, all in the same disassembly. Hours did not predict any of that, and no inspection short of opening the housing would have found it.
That is the honest boundary of hour-meter analysis. It is a filter, not a diagnosis.
Reset meters and gaps you cannot explain
Control boards get replaced and meters sometimes restart. That is not automatically suspicious, and on an older machine it is fairly common.
What matters is whether the seller can account for it. A machine of known age showing implausibly few hours needs an explanation, and a documented one is fine: the control was replaced in such-and-such year, here is the invoice, the previous meter read roughly this much. If there is no explanation, you are not buying a low-hours machine. You are buying a machine with unknown hours, and the price should say so.
Ask when the control was installed, and whether it is younger than the machine. That single question resolves most meter mysteries.
Why this matters more than it should
You might reasonably ask why so much weight lands on inspection rather than on a repair budget. The answer is that a repair budget is hard to build in advance on these machines.
Haas parts are dealer-quote only. There is no public OEM price list, and the parts site routes you into the Factory Outlet dealer network rather than a catalog. We checked that specifically, and it is a stable fact rather than a page we failed to find. So you cannot price a parts scenario before anything breaks. You anchor to the one number that is publicly published, which is the spindle band above, and treat the meters and the work history as your only real pre-purchase lever on the odds.
The inspection order
- Pull both readings, power-on and spindle, off the control display. Not from the listing. From the control.
- Cross-check them against the machine's known age and against each other. Does the ratio match the story you have been told about how it was used?
- Ask what it cut, at what speeds, and what the duty cycle looked like. Get specifics, not "general job work."
- Warm the spindle and measure runout at the taper. The spindle-economics source used across our records treats under 0.0002 inch as the rebuild-candidate threshold and over 0.0005 inch as a replacement signal, on Haas VF-series machines. Borrow those figures for other brands knowingly, since we have not found equivalent published thresholds elsewhere.
- Listen at speed, across the range. Grinding, whistle, or scraping under load is worth more than any meter reading.
- Ask what has been replaced and what parts were used. This is the question that covers the ground the meters cannot reach.
- If the hours story and the spindle behavior disagree in either direction, stop and find out why before you negotiate.
The short version
Hours are not age, spindle hours are not power-on hours, and neither one is a condition report. A 12,000-hour production machine can beat a 5,000-hour one, and the reason is that the meter counts time while the bearing counts work. Use the meters as the first filter, then spend your remaining leverage on the work history, the spindle sound, and the runout number. Those are the story. The hour meter is the first sentence of it.