Regulating the unregulated
Throughout the microbrand watch space you’ll hear many brands tout that their movements are “regulated”. While this phrase initially strikes confidence that your watch is being adjusted finely and has a closer set of eyes on it, there is much more to this unregulated claim than an un-critical take may lead you to suspect. To break this apart we’re going to need to understand a few things and those things will be, chiefly:
What is regulating a watch?
Why do watches need to be regulated?
What are the industry standards?
Regulating a movement is a “value added” phrase which is challenging for the average watch consumer to verify with the same technical veracity as a watchmaker or technician equipped with proper technique and tools. This is also an unregulated and arbitrary claim, meaning there wont be a legal issue if someone isn’t entirely honest about when and where their movements are regulated or adjusted. The combination of these two things, an unregulated phrase used to imbue value to a product, and a technique that requires specialized tools and some degree of training to interpret the data, means that the average consumer is ill-equipped to challenge the claim a brand will make about the further adjustment of their watch, not only are they ill-equipped to challenge it they are in the dark on how a brand providing additional regulation beyond the factory may in fact, be a hint at subpar quality and cost cutting. Before we get into that, lets understand regulation first.
What is regulation?
Regulating a movement, in its simplest execution means that the timing rate has been adjusted. This, most commonly is done by changing the effective length of the hairspring by moving the regulating arm along the hairspring. On freesprung balances, that is, movements that do not have regulating pins, the moment of inertia for the balance is changed to alter the rate - this is done by moving weighted screws on the edge of the balance in or out. Much like when you spin around in an office chair, if you stick your arms out you will slow your speed, tuck them in and you will spin faster. Freesprung balances and hairsprings are generally reserved for higher end watches, since the claims of regulation are generally a microbrand phenomena we will focus on movements that use regulating pins. The two minute video below succinctly shows you (in an abbreviated manner) how a regulation may be carried out with such a movement.
So in its simplest terms, it is adjusting the regulating pins or timing screws/nuts to alter the rate. Regulating pins alter the effective length of the hairspring - they are two pins that the hairspring goes between, as the balance oscillates it bounces between them. Making it longer will slow the rate, while shorter will increase the rate. In the video or elsewhere you may have heard of “dial down” or “dial up” perhaps “horizontal and vertical positions”. A watch is a mechanical machine, the tiny pinions, pivots and gears are subject to gravity, friction and other forces that affect how the machine keeps time. A horizontal position would be when the dial is facing up, if you rested the watch on its caseback on your table, it would be in the dial up position. Conversely, if you put it dial down, it would be resting on its crystal. These are the horizontal positions. Watches, generally, are regulated in the dial down position and run better in the horizontal positions. There are a few reasons for this, but it mostly has to do with the pivots of the gears are exposed to less friction in these positions. This reduction in friction means higher amplitude and generally better timekeeping. Amplitude, simply, is the degrees of rotation of the balance wheel. This is used as a proxy for the health of a movement as a watch with higher amplitude generally means more energy is moving through the gear train. If the flow of energy through the gear train is homogeneous and uninterrupted, your timekeeping will also be better. So if you’re timing a watch and have erratic amplitude it would be suggestive of a problem. Stable amplitude is a good sign and amplitude is generally very good in your horizontal positions. The vertical positions would be when the watch is in a position that isn’t flat on a table. Imagine a pocket watch in the pocket of your vest or perhaps when you bring your wrist up to your eyes to check the time - these would be vertical positions. It is normal and expected to see a drop in amplitude in vertical positions when compared to horizontal positions, again, there are many influences but the cardinal influence here is that the surface area that the pivots of the watch movement is exposed to increase. This increase in surface contact increases the friction and consumes some of that energy that would be converted into amplitude.
An example of a Rolex balance with timing screws.
Watchmakers are constantly hunting ways to battle the disturbances of isochronism, which is the ability for the oscillator to maintain the same rate regardless of the amplitude. Which if people pay attention to fancy new materials in watch movements - these are generally things that aid in the battle against those disturbances of isochronism. Many finely made and adjusted movements can do very well in the various positions but the reality is that change in amplitude affects the rate. In lower cost or older movements, those changes may be more noticeable. Again, this is something a watchmaker knows and expects. Regulating a watch means anticipating this. Remember amplitude affecting rate as well as horizontal and vertical positions affecting rate and amplitude going forward - this is an important principle.
Why do watches need to be regulated?
Movements require regulation for various reasons. When a watch is first manufactured the balance wheel needs to be trued and poised - microscopic amounts of material may need to be removed to ensure the weight is consistent all around the balance - this will affect its rate. The length of the hairspring needs to be perfect or it will not vibrate at the correct rate. Regulating pins may need to be centered, opened or closed. Stud length may need to be adjusted - the list goes on. Once a movement is made and the watch is assembled all the forces of isochronism need to be accounted for and the tolerances of machining come to light. Watches need regulation to keep time, otherwise they lose all utility as a tool and become a meaningless mechanical bauble. The concept of a watch needing to be regulated after it is manufactured and unadjusted is not a hard one to grasp, it is starting from nothing, after all. After being worn, a watch may require additional regulation. Things like impacts, wear and tear, loss of lubrication and simply existing outside of a vacuum and a theoretically perfect world will cause timing metrics to drift. The amount the regulating arm needs to move to create a significant change in timing is a very small amount - often fractions of a millimeter in distance, it shouldn’t be hard to imagine that a life on the wrist may introduce small shocks (even imperceptible to the wearer) that could slowly bump the regulating or stud arm in either direction, causing a not insignificant change in rate. It’s a mechanical machine, metal wears out, lubricants oxidize or disappear - as this happens metal dust appears in the system or tolerances change further, worsening the timepieces ability to carry out its duty. Sometimes age alone is enough to cause lubricants to wither and wear to set in. It’s normal for watches to need adjustment as they are worn or age.
A ruby bushing. The pivot of the wheel goes inside the ruby bushing. You can see the hole on this one is no longer round and there is a groove cut into one side of the jewel. This happens when wear or mis-adjustment causes the wheel to lean and it no longer rotates concentrically. Cutting into the jewel over time.
Watch movements are regulated at the factory. Movement manufacturers sell their movements in tiers of fineness. Each tier of fineness will be accompanied by things like higher grade finishing and timing. The factory will then define standard operating parameters and tolerances for their movements, they often ensure that their stated numbers for things like rate and power reserve have enough of a buffer built in so that the movements they sell can outperform their stated tolerances on performance metrics. This is important.
What are the industry standards?
For watchmakers and service jobs, the standard for timing and regulating a watch is 5 positions. Dial up, dial down, crown up, crown down, and crown left. The 6th position, crown right, is an awkward position that often has the worst timing metrics and isn’t experienced with reasonable frequency when worn. Some brands include the 6th position. When I timed out and regulated all the SW300s and 2892s for the Calamity, I timed and regulated them to 6 positions. Going back to movement tiers and stated manufacturer tolerances for a moment. Lower grades of movements or large service centers may only time 2 or 3 positions. Since comprehensive timing takes, well, time, a way for them to reduce costs is to reduce the time spent timing and adjusting a movement. If they’re only timing for 3 positions and their tolerances only account for those positions, you’re omitting a large potential data set of rate averages. Still, money is king. As the grade of the movement increases you may see the number of positions timed in, increase to 4 or 5. Additionally, these factories sort their balance wheels. Some balance wheels just keep better time. The best ones will be reserved for chronometer grade or top grade which undergo the most comprehensive timing tests and have the tightest tolerances. That’s why these also cost the most. Though, when you see a brand slyly remark that “their movement is keeping COSC time” it is a nefarious remark designed to exploit ignorance on what timing and regulating is. The COSC chronometer tests are extensive, 5 positions over 15 days at various temperatures and conditions. COSC is also an independent organization, such independence removes a conflict of interest and bias. I also have serious doubt about brands that make dubious claims would take 3 weeks time (of working days) to time their watches and not want something more objective to show for it. There is another timing metric not yet touched on called ‘delta’. The delta is the difference between your fastest and slowest positions. The delta tells a different picture than the average rate. You can imagine it’s possible to average 5 huge numbers to achieve 1 small number, this could be a way of saying “your average rate is good!” when in reality, the rate of any one position could be massively out of tolerance. This is where delta comes in, the average rate is only meaningful if the delta is small. The smaller the delta, the more impressive the timekeeping.
With all this information you may start to see where I’m going. Are brands providing timing sheets? Is this evidence of regulation or simply timing? Was it only 1 position? Perhaps 2 or 3? Again, this omits a significant amount of the data. Full wind & half wind?
A Rolex Timing Sheet
You can see the 5 positions here as well as slots for amplitude and rate at fully wound and at half wind. A comprehensive timing test includes half-wind metrics to ensure isochronism is not sufficiently disturbed when the mainspring is not fully wound. Since at full wind, the mainspring will deliver the highest amplitude and best timing - if a watch is incapable of providing stable timing at half-wind, it would fail this timing test. Additionally, you can see spaces for the rate to be averaged. When a watch completes its timing test, the average rate (of the 5 positions, or 5 positions at full wind and half wind) would be given to the customer as the rate of their watch.
Many people may be familiar with the given tolerances for movements like Seiko’s NHxx family or the Miyota 9000 family. Perhaps even the different grades of entry level ETA and Sellita movements. Now, remember when I said they add a buffer to stated tolerances, especially on lower cost movements? This is where the whole “We regulate our movements!” craze came from. Seiko states the tolerance for most of their NH movements to be -20 to +40 seconds a day. Miyota states -10 to +30. Now, I’ve dealt with hundreds, if not thousands of these movements and they are never this bad as a batch. Sure, every now and then you might get a bad one, but for me, problematic movements, we’re talking outside of 30s/d has been less than 1%. I’ve found that Seiko movements dont have as great rates as the Miyotas but even more specifically, Seiko NH movements have large poise errors and big deltas where the Miyotas do not. Additionally, I’ve found that many of my Miyota movements come to me keeping +5 to +10s/d accuracy. Which is spectacular. I went through, timed them all and discovered the rates were so good that I (within reason) couldn’t improve them. When you adjust the regulating arm on these watches it adjusts the rate of all positions. Omitting positions can be either a factory simply selling a lower tier movement, or a company intentionally hiding a significant percentage of data. If you, for example, only regulate it in dial down it is not hard to get it to 0 s/d. What did that do to the other positions? The horizontal positions usually mirror each other in metrics, so two positions, dial up and dial down, are not sufficient for a representation of the timekeeping of a movement. Going back you may recall when I mentioned the poise of a balance wheel - many of these more affordable movements may regulate a movement around a heavy spot in the balance. This means one position will stick out like a sore thumb, either much faster or slower than the rest of the positions, the opposite position of the poise error will have the inverse reading. For example, a poorly poised balance may show up as a very fast rate in the crown at 6:00 position, with the crown at 12:00, if it is truly a poise error, would read as the slowest position. Now if you’re timing in 3 positions, dial up, dial down and crown left, in this scenario, you’ve very conveniently hidden a poise error that would not only alter your average rate, but give you a massive delta. The only way to correct a poise error is to remove the balance from the watch, identify the heavy spot, and cut away a nearly microscopic amount of metal, reinstall it and re-test it (there are some balances with weights that you can use to adjust poise, but we generally don’t see these in microbrands). Poising a balance is a very delicate operation. Manipulating the escapement of the watch, that is things like the balance, hairspring, pallet fork and escape wheel, require great skill and many, many hours of training. Poising the balances of a Miyota, Seiko or entry level ETAs or Sellitas isn’t a very sustainable practice on a large scale - and I would know, I’ve done it. You can get the Miyotas and ETAs/Sellitas to really sing and have incredible rates if you’re determined, but the truth is, if you’re getting them factory fresh they’re already way better than their tolerances would suggest and barring advanced timing techniques that take considerable skill and time, they really wouldn’t see any, or much, benefit from regulating.
There are multiple ways to check the poise of balances, this is one. On razor polished stainless steel or ruby jaws on a perfectly leveled special vise you rest the balance staff and very gently, with a puff of air or a tickle from a cat whisker you roll the balance. It should stop without rolling to a heavy spot. Check multiple times and locations to see if it rolls towards a heavy spot.
I think that there is some hysteria around seeing the tech sheets from the movement brands then assuming that all their movements are going to come out running -10 to +30, but they don’t and brands know this and it’s really easy to say that you were the one that kindly regulated it. Unless, the movements actually do need regulation, which raises more questions.
In the watch industry, movement purchasing can often represent the bulk of the cost of the watch batch. The factories that produce movements generally want high minimum order quantities before bulk discounts kick in. In many places, watch factories (the ones making cases, bracelets, dials, etc) may have a deal with a movement factory to purchase, say, 10,000 movements which get sold at a very nice discount, they’re able to pass that onto their clientele (with a markup of course) but in the end it still saves a microbrand from spending the thousands extra from an order on the scale of hundreds of units (or less). That is, however, in an ideal and perfectly honest world, which this is not. The reality of movements, especially in demand models or more expensive ones is that refurbished units make their way back into circulation frequently. Unscrupulous or aloof brand owners may find themselves looking for the best possible deal on movements and this is where refurbs come in.
The macro brands make watches in batches of thousands. They often have designs and releases planned years in advance. Sometimes their designs simply suck, perhaps tastes change or there’s a bump in the market. Doesn’t matter. Sometimes the result is watches that are already assembled sit, unsold, for years. Assembly houses and distribution centers all cost something and these big brands often find themselves losing more than the watch may be worth if it’s not selling for whatever reason. To recoup costs, they may find a movement broker, a professional who sells and resells movements to move their idle inventory. These people are happy to sell at low prices and low minimum order quantities. Bear in mind, these are also movements that have been presumably sitting for years and now they’re being quite literally ripped out of watches by a company looking to save as much money as possible to pass onto the next person.
An example of testing a “vertical” position.
As I mentioned earlier, a watch sitting for a long time is enough to, in many cases, warrant regulation if not an outright service. A movement brokers job is to sell movements. Not service them. Not inspect them. Just sell them. This broker is happy to sell to factories and individuals alike - these are both entities interested in saving money. This is how movements that aren’t fresh enter circulation, while some of these movements have been cared for and properly decased and perhaps serviced, just as many, if not more are neglected. The result is many brands, perhaps to their knowledge or not, are now selling used or, if not quite used then old stock, movements as new and these movements will often have abysmal timing and strange defects, this is simply by virtue of rough and hasty handling, aging on a shelf somewhere, an unskilled service or perhaps no service at all.
Then imagine you get your batch of movements and the timing is closer to what that tech sheet from the factory says. You see 30 or 40 seconds a day and it confirms some suspicion or bias that the tolerance sheet was simply accurate and that’s how these movements perform. So you get to work regulating them and you pass that onto your customer, that your movements were loving regulated and brought into much tighter tolerances than when they arrived. Other brands see this and buy a Chinese timegrapher and printer and start printing out timing sheets, or maybe just writing on a card in 3 positions, but their movements are fresh and don’t need to be regulated - not wanting to seem like a slacker, they say they regulate them too, surely a card with timing values is enough to convince someone their movement was regulated. Then you have a runaway issue of brands claiming regulation, with little or dubious evidence and perhaps little to no training and what they’re calling regulation is in fact very far from what a watchmaker would call regulation or a timing test. There’s no way to reasonably poise 500 Miyota or Seiko movements, the time cost would be insane, additionally, this is a skill so few people have, generally left to trained watchmakers. I think we can easily rule out that microbrands, with the odd exception of the occasional experiment or observatory submission, are not dynamically poising their watches en masse. Still, we can also assume that many are not doing full timing tests either, let alone power reserve tests (each Calamity I did was timed in 6 positions and had a 3 day timing/power reserve test). The pricing and the language around it speaks volumes.
Part of a correspondence from an undisclosed movement brand - note that the rates & number of positions are mentioned.
And I’m not saying everyone has to regulate or submit all their watches to a timing test, that’s what the factory should be doing. You should presumably have 1) the movement factory providing its checks then 2) the assembly house or watch factory providing its checks and finally 3) the brands final QC and checks which should include simple timing, which as I hope you can see now, timing a watch is different than regulating it or having a full timing test. If somehow, you, as a brand, are ending up with movements and watches that after multiple presumed steps of QC and adjustment are still failing or performing so poorly you need to intervene, it may be time to start asking about the provenance of your movements. Sometimes its an active cost cutting decision, sometimes someone in the chain tricks you and you have no idea.
Regulating as a marketing phrase and value added proposition in the microbrand space, for me, raises more questions than it answers. I may be a purist but I like when trained and competent people work on my watches and, much like human medicine, we shouldn’t really be intervening needlessly - there should be a sufficiently good reason. I purposely became a certified watchmaker to understand these things at a foundational level, I see many brands that have offsite watchmaker support or no watchmaker support making these claims and I hope that I can arm you, the reader, with some critical knowledge as to what regulation is, why it would be needed, and how to ask appropriate questions that may support the veracity of the claims of a presumed timing test carried out on your watch.
To very finely regulate a watch (i.e. poise, escapement and hairspring adjustment) takes considerable time and skill, to the degree you need sufficient training. To generally regulate a watch takes quite a bit less work, this is where I think most of the claims would fall. So I invite you to inquire on the details of regulation. Why is it necessary? What’s the delta? How many positions? Was power reserve checked? What about timing it at half wind? Even the equipment used can have a bearing on it. Any company that is honestly flaunting regulation as a testament to their skill should be happy to share the information with an honestly engaged and curious customer, while I suspect a less than honest company may provide murky and vague answers. So is that brand regulating your watch? Maybe. And maybe it’s for a good reason and maybe it’s for a bad reason.
In summary and using reductive logic we can imagine a few scenarios:
1) A company buys movements directly from the factory. The factory regulates them, provides timing sheets as proof for each movement which is then verified by the purchasing company upon receipt. Any movement outside agreed upon tolerances is returned and replaced under warranty. The customer buying the movements in this case does not need to regulate them, unless they wanted to save money by buying a lower grade timing spec and spend more time personally adjusting them. At which point proof should be readily deliverable of said adjustments.
2) A brand owner buys movements through their watch factory, who then sources them through a broker or the movement factory. This scenario is like the first one with extra steps. There are 3 parties now, the party selling the movements, the watch factory building and assembling the watch with said movements, and the brand owner who will eventually receive the watches. In this case, party 1 can either be the factory or a broker. In which case, the factory will provide timing sheets as proof of performance to their customer, which is, in this case, the watch factory. The brand owner will not see the timing sheets, but they existed and regulation was performed. Now, the factory has to QC and ensure proper functionality after the watches are assembled and before they are sent to their client, the brand owner. This will include timing tests. At this point, the watches have been regulated at least once and checked at least twice. When the owner receives them, they should perform a final QC which would include assessment of the rate. Now repeat this scenario but instead of the watch factory purchasing from a movement factory, they purchased from a broker. The broker might include timing sheets, but they are not obligated to and many times they will not. Quality and performance may not be guaranteed. In the scenario from the broker, there exists the possibility of the movements being in disrepair, in need of service or regulation. At which point, the watch factory should catch that in their QC and there is still the final QC of the brand owner. As movements stray further from service intervals, it is common for their timing to become less stable. This is a scenario in which a final regulation may be warranted and even beneficial, though, it could end up being a band aid on a larger problem.
3) A company buys directly from a broker. Like scenario 1, a company buys movements directly. Brokers, however, are not obligated to provide the same warranties and tolerance adherence as a factory is. There is a very likely chance these could be old stock or otherwise used or refurbished movements. The owner now has movements that do not carry a guarantee of rate and likely did not come with timing sheets to verify the performance of the movement. This has had virtually 0 eyes on it in regards to regulation or QC at this point and all of that will need to be either carried out by the owner or, the assembly house/watch factory if the purchaser of the movements decides to forward them. In this scenario, you can imagine the necessity for a regulation, let alone a full service is probably necessitated (Sometimes brokers deliver great product! Sometimes they don’t, it’s a gamble and with gambles there are no guarantees). A scenario like this is more likely to see or require additional regulation further downstream in the process.
Then, there are the market pressures, where brands see other brands offering something of perceived value so they want to offer it as well. If, however, whatever they’re offering is not only disprovable but in their interest to make said claim, you can see where you may have an issue of runaway claims. Honesty and deception mixing in a market. If the consumer has no way to readily ask for evidence of said claim, how are they able to discern an honest claim from a deceptive one? The intent of this article is to arm you, the reader, with some knowledge to provoke questions that would reveal an honest claim and a deceptive claim. Since stating things like “we regulate our movements” is a vague term that is unregulated, there is little downside to lie about it if your assertions will never be tested - and the liars extoll an additional cost on those who are honest - those who put in the extra time (and time is money) to either correct issues or demonstrate their competency at regulating a movement now are competing with those who do not expend that additional cost yet make the claim they do. A failure to hold claims accountable, will eventually lead us to a market of wondrous and fabulous statements of regulation and intervention and all sorts of rarified horological adjustments that don’t actually mean anything. Which, is why accountability is important, to ensure that we get what we pay for and to ensure that the words that are being used accurately convey the service and product being provided. There are many ways this claim can go, from honest infusion of competent watchmaking, to a lie simply to keep up with those doing the actual work - there is value in people taking time to create a better product but I know, personally, there are many people who have no issue bending words or finding loopholes to make it sound like they’re doing work they never did.
