Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Hacking the Casio F-91W to Handle 1000 Psi (dvt.name)
198 points by dvt on June 3, 2019 | hide | past | favorite | 80 comments


Some sources suggest heating the mineral oil to ~50° Centigrade, but I’m not exactly sure what the reasoning for that might be, so I stuck with room temperature.

The likely explanation is that heating lowers the surface tension [0] and viscosity of the oil so that it can creep easier between the parts. I also considered heating for a problem where a liquid had to enter very narrow cracks.

[0] https://en.wikipedia.org/wiki/E%C3%B6tv%C3%B6s_rule


Could it also dissolve solved air contained within the oil, as it happens by boiling water?


The liquid should be able to hold less gas at higher temperatures, yes.

What you observe with boiling water though is not solved gases leaving the water but actually water turning into steam. Steam is exactly the same stuff as water, just a lot less dense.

Actually if you increase temperature and pressure enough, there will be a point at which it makes no sense to differentiate between water and steam anymore [1]! Which is a fun thing: you can turn water into steam without boiling it. Raise the temperature and pressure enough while keeping it liquid, reduce temperature while keeping the pressure up and then reduce the pressure as far as you like, as long as it doesn't make the water condensate.

Voilà, steam without boiling!

What is happening there is taking a detour around the critical point instead of crossing the blue line (which is where water boils) in the following graphic[0]:

[0] https://en.wikipedia.org/wiki/Critical_point_(thermodynamics... [1] https://en.wikipedia.org/wiki/Critical_point_(thermodynamics...


I don't know if I would even remotely trust that not to leak. The same watch is notorious for not having a quality seal and quickly fogs up in the shower. I would rather trade some meters and use straight up petroleum jelly injected into the casing at a high pressure; I'd be much more comfortable wearing that (and, yes, it is technically a liquid even at "freezing" temperatures).


How about "potting" it with clear epoxy, or is there some required internal movement?


Can do, although you might be better off using a hard conformal compound (black stuff used for potting electronics) where possible. It has a lower viscosity and is less prone to bubbles. I'm not sure how much small bubbles would actually affect it, though.

There's one other good reason to use conformal coating: it has very little change in size when it dries. There are a few spots on the board that would be affected by that, most importantly the load capacitors for the timekeeping fork. Ceramic capacitors have a piezoelectric effect and change shape ever so slightly (they stretch and squash, alternately contracting around their midsection and then expanding). It's not enough to break the solder or the epoxy but if you pot them it changes their apparent capacitance, which will change the frequency of timekeeping. Probably not enough to matter, but still measurable. Large inductors have similar problems and can eventually break loose from epoxy.


I think ideally, you want something which is non-conductive, incompressible, and solid but soft. A solid material will not flow and potentially leak through the seal. A soft material will not fracture, and transmit stress to the components of the watch. In a hard material, fractures might also appear when thermally shocked (such as getting in and out of the water).

I'd suggest a good candidate would be a silicone gel. It can be applied to the watch in liquid form, allowing all air to vacate, then cure into a soft solid once sealed in the watch. It could also be removed relatively easily, if you decided you don't want it in there any more.


It's all digital from what I've seen. It seems like some clear epoxy would work well. Maybe some mineral oil after it cures, just to help fill little gaps (since epoxy tends to be pretty viscous).

EDIT: Although epoxy makes it a bit difficult to change the battery, depending on how you go about it. Maybe some conformal coating on the PCB + mineral oil.


I've been wearing an F91 for years and everyday I shower with it. Still functioning.


Me too. Swimming down 10m or so in the sea fogged up the screen though.


> The same watch is notorious for not having a quality seal and quickly fogs up in the shower

Casio says the F-91W is only splash resistant, so you shouldn't wear it in the shower:

https://www.casio-europe.com/euro/products/watches/collectio...

If you want one you can wear in the shower try the BD640W or similar:

https://www.casio-europe.com/euro/products/watches/vintage/b...


That's my point, though. If it can't keep water out, I wouldn't trust it to keep mineral oil in (even if it is more viscous and particles are larger).


Hmm...

I would imagine the best case for a seal would be to have equal pressure on both sides.

Since internal gas is compressible it would be easier for water to enter with a pressure differential. Then, because the water would have to be next to the seal with pressurized gas behind it to exit again, the water would probably stay.

Having a non-compressible liquid inside would make the seal work less in most cases.


Wait.

Why are people wearing watches in the shower?


Because if the watch is waterproof, why not wear it in the shower?

You get to forgo taking on/off the watch and you get to wash the watch at the same time, instead of wearing a watch with residue sweat etc back on after the shower...

From your question I take it you don't wear your watch when sleeping either?


But then how do you wash the area under the watch without taking it off?


Do you fasten your watch air-tight? I don't.

You can always jiggle the band a little and let water / whatever flow underneath.


I had a problem of finding an out-of-the-way clock for the shower to keep track of time.

I hung a Casio W96H by its folded strap on the first ring of the shower curtain. It’s been this way for at least 7 years.


> I had a problem of finding an out-of-the-way clock for the shower to keep track of time.

How long are your showers?


I just find it extremely useful to be mindful of time as I make my way through my get-ready-for-work routine.


What do you mean? I'm really tempted to ask "wait, there are people who shower without a watch?" - but I guess it's personal preference. I personally feel weird if I don't have my watch on me in the shower.


The petroleum jelly I know of is fairly opaque. Is there any other?

If not, even with a thin layer, I doubt the result will give you a readable display, as, judging from the red color of the display in the end result, there’s air between the glass and the LCD display in these watches.


Good point. Petroleum jellies can come in different clarities and with different refractive indices, here's an (expensive) version in the form of refractive index matched gels: https://www.ebay.com/itm/The-refractive-index-Matching-Gel-f... (this one is matched to optical fiber, i.e. very clear).


Isn't 1000 psi almost a km deep? Not sure if this would work at all.


1000 psi would be 700 meters under the water.

There are actual diving watches that are rated for 1km or more. The first image on the Wikipedia page for "diving watch" shows one rated for 3km. Rolex famously attached one of their prototypes to the exterior to the Trieste in 1960, and it survived the trip to the Challenge Deep.

[1] https://en.wikipedia.org/wiki/Diving_watch


German manufacturer Sinn makes a watch - the UX - which is rated (if I read the specs correctly) to keep time to 5,000m (16,500ft.) and work after being at the bottom of a km-deep bore hole at the bottom of the Mariana Trench - the case is rated for 12,000m/39,500ft).


I did that with extra virgin olive oil a few years back. No leaks since then. I attribute it to higher oil viscosity.


Mineral oil breaks down latex quickly. What material is used for the seal? Oil expands and contracts with changing temperature. An air bubble will appear soon after if there are no special steps taken to prevent this.

The real advantage of an oil filled face is the greatly increased readability of the watch.


True. There is a Liquid you can put on top of eg marmalade which seals it from air. I think it is a from of liquid silicone (mind the final _e !) which I think is not aggressive to rubber.


Did a little research, and it seems like pure silicone oil is much less aggressive to rubber/latex than mineral oil (and you can get it from Amazon pretty cheaply).

Does anyone have any idea what the chemical differences between the two are that affect how they interact with rubber? I’d love to learn more.


Good point.

Mineral oil, contrary to the remark in TFA, is 100% organic. Unlike olive oil, though, it's saturated. Does that matter? It's not exposed to air, so no. But olive oil goes cloudy when cold.

The watch is unlikely to have latex in it, but quite likely to have silicone parts and seals that might soften in silicone oil.


> Mineral oil, contrary to the remark in TFA, is 100% organic.

Thanks for the correction! From what I was reading, olive oil goes rancid (and acidic; but mineral oil does not).


Olive oil, like any unsaturated oil, may go rancid on exposure to air. But in a sealed watch there will not be enough air.

("Organic", outside of agriculture, means "composed of carbon compounds". In practice, that usually means compounded by something alive, because life has been at the carbon for a long, long time.)


No idea what the stock seal would be made from.

Nitrile is commonly used in hydraulics as natural rubber doesn't play nice with hydrocarbons.


I am frustrated that he didn't have a way to test it and so has no idea if it actually works.


I would bet it did. Here’s an F-91W hydro modded with olive oil taken to 3000 feet.

https://forum.tz-uk.com/showthread.php?330958-The-Little-Wat...


I was just reading yesterday about another case too - https://forums.watchuseek.com/f2/my-oiled-casio-f-91-dives-d...


He seems to mention purposefully creating an air bubble in the enclosure (by letting a few drops of the oil drip out) — why would he want to do that??

> "I slightly loosen two screws, to let out a couple of drops of oil (which would allow forming a bubble inside the box)."


If the oil expands and contracts with temperature then the bubble would allow this without the oil blowing the seal.

I think that’s the idea. Could be wrong.


If you live near an ocean, just tie the watch to the end of 700m of string, and start lowering it over the side of a boat.

Pull it back again and see if it still works.

Obviously you need a deep ocean, so check a map first.


You need a really heavy weight on there as well. Ocean streams and the surface moving will make it not point straight down at all.


A deep ocean indeed.

You’re probably going to have to go off the continental shelf[1] to get 700m it greater depth.

This will make your plan somewhat less trivial.

1. https://en.wikipedia.org/wiki/Continental_shelf


-Off the shelf, or, if you happen to live in a country designed by Slartibartfast, find a fjord.

The view from my office window includes a fjord more than 500m deep; a couple of hours drive to the south, you can get down to 1,300m/4,000ft or so a few hundred meters from the shore.

Edit: Xylakant beat me to it!


Not that hard in some parts of the world. https://en.m.wikipedia.org/wiki/Sognefjord


Fair point, I'll take it out next time I go diving and report back!


Before this became the de factor standard in tech dive computers, advanced divers would often do the same and fill their units with mineral oil.

For anyone interested in how this can go wrong, Raising the Dead - a book about scuba diving at its limits - has a passage about a rescue diver repairing his dive computer in the field, re-filling the oil, and noticing that he'd lost a drop or two in the process, with a resulting tiny air bubble...


How does the piezo buzzer get affected by this?

Is the alarm louder or quieter?

Is the tone any different?

I have no plans for deep sea diving any time soon, however, I do like getting up in the mornings. Would an oil filled Casio help me in this practical regard?

I also wonder about how fragrant the oil could be, if you used some oil based perfume, could the watch that is not waterproof be made to smell nice, if only slightly so, but mitigating the odour of bacteria that accumulate on the back of the watch when tightly worn?


The buzzer is definitely more "muted" but still beeps! I'm not sure if a fragrance would "keep." I used odorless mineral oil so I can't give much insight.


I wish someone would test if it's even water resistant for a period of time in a deep pool. I bet it is. I wish he would at least done the pool test.

If it's waterproof, it's a killer hack for its simplicity and overall cost.


You could pull vacuum over the oil bath to positively coax all the air bubbles out of the watch.


>You could pull vacuum...

Unfortunately, that would also destroy the LCD and quite possibly the crystal oscillator (not sure how strong the metal can enclosure of those things are). Battery wouldn't survive it either but you could easily take it out and insert it after the vacuuming.


The metal can for the oscillator surely doesn't fill with oil, and would therefore crush at 1000 psi?

Filling this with mineral oil might actually reduce rather than increase its pressure withstanding ability.


I have no idea how tightly those things are sealed. Given some of the watch tests in this thread (the olive oil one), it surely can withstand the positive pressure but I'm not certain it could withstand negative pressure. If oil got in there, that oscillator would never keep the correct frequency again.

Hopefully someone does some tests.


Very interesting. I'm intrigued to know if anybody with experience knows why there is a common recommendation of warming the mineral oil, versus room temperature which the author did.


My guess would be that in addition to the viscosity, it preemptively compensates for the thermal expansion of the oil, since on your wrist will likely be warmer than room temperature.

Here is another detailed liquid-filled watch guide [0], which talks more about handling thermal expansion, where rather than using an air bubble, he opts for some sort of foam ring. Another person recommends Flourinert with foam rather than mineral oil [1].

[0]: https://www.christopherwardforum.com/viewtopic.php?f=5&t=496...

[1]: https://forums.watchuseek.com/f21/liquid-filled-citizen-6540...


No experience here. But my guess is warming the oil makes it less viscous and thus easier for small air bubbles to escape.


To be precise, warming makes the oil exponentially less viscous, so it can be a really pronounced effect by warming it 30 degrees (depending on the specific oil).


Author here. Would also love to know if any chemist/physicist knows the definitive answer to this :)

Also: why do some watches purposefully leave an air bubble in the watch? Is this an aesthetic choice or is there any benefit/detriment to it?


Maybe a de facto expansion reservoir?


> I decided to switch to a reddish tint after my first failed attempt.

What happened with the first attempt?


The buttons were misaligned with the contact pads so I had to take it apart again.


What's the problem with the LED, does the liquid filling cause TIR off the watch-face/external air boundary? If that's not the problem, you could try side-mounting the led, as here: https://www.instructables.com/id/Watch-LED-Light-Mod/

Does the oil leak out of the buttons when they're pressed?

How about the buzzer? I think it's attached to the back of the watch so shouldn't be affected too much?

Someone mentioned improved readability. I assume underwater that is definitely the case, as you no longer get TIR reflections off the back of the watch-face/internal air boundary at steep angles, but how is readability affected in air?


Looks like a fun little project to do next weekend. Does anyone know where to get a strap like the one shown in the article?


It's just a NATO strap. Very cheap and functional for people that don't like hard watch straps. It seams the color combo is Navy/Crimson/Ivory

Found this on Amazon

https://www.amazon.com/BARTON-Watch-Bands-Crimson-Standard/d...


> It's just a NATO strap

Which ironically aren't a NATO standard, but a UK MoD spec.


I don't think these even are NATO straps - they don't fold back on themselves like the issue ones do, and these have an extra fabric flap that issued ones don't.

This is a genuine 'NATO' issue strap:

https://www.cwcwatch.com/collections/m-o-d-spec/products/cwc...


Look for an 18mm Nato strap. It looks like Clockwork Synergy makes them with the nylon keeper, if that's important to you.


I have one of these watches and really wanted, but decided not to mod it with such a strap as the metal buckles would potentially start to scratch my laptop.


These straps are like ties - you get one in the colour of your school, university, club, or regiment.


This is super cool! Kinda want to try it next weekend...

Seeing a lot of commenters wondering if the seals on the watch will manage to keep water out/oil in. Anyone know if it's possible to replace the o-rings or anything like that?


The entire watch is 10$.


Yeah, I’m aware. I wouldn’t really care if the watch itself died, but leaking oil would be a bit annoying and making things that should fall apart last longer is often a fun challenge.


The o-ring is pretty flimsy (you can see it in a few pictures), but some have done this mod and have worn their F-91Ws without any leakage for 5+ years, so I expect it to be "good enough."


Good to know. Thanks for the info and for writing the original post about this!


That watch looks amazing with the red tint and that strap! :-)


I wonder whether any mechanical watches could cope with this technique (and maintain any semblance of accuracy).


I'd imagine that any mechanical watch in which the mechanism wasn't designed with the viscosity of oil in mind would be completely incapable of accuracy if one were to add oil.


Well... trying to hack it.

No confirmation if this works.


Does anybody know what dyes he used?


I like the F91W because it refuses to die, keeps good time, has as an alarm and a hour minder. However, to repurpose it into some kind of frankenwatch to behave as a diving watch, is counter-intuitive to the fact that, it will lose most it's useful features, freely admitted to by the author in the article.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: