Need help pinpointing the problem in a mechanical timer
I have a cheap mechanical timer in my kitchen that I recently over-turned and broke. As a means of training and out of mechanical interest (and obviously for the repair ethics), I'm trying to figure out if I can get it to work again.
I disassembled the entire thing and reverse-engineered how it works (I had never seen a disassembled clock with my own eyes before) and am pretty confident in understanding most of it now (I can reliably make specific functions happen or prevent them depending on what I wanna do). The broken part is a tiny plastik ring with a tip that transmits the turning of the main clock to the other half of the spring that's responsible for the vibration:
The black part is the replacement I designed in Blender and 3D-printed:
I have two basic problems:
- The 3d printed part doesn't stick perfectly and eventually stops transmitting the power, disconnecting the timer from the vibration
- The clock only ticks for seconds, up to a minute
The first one isn't really gonna be fixable, I think. My 3D printer is old and cheap, so I won't be able to achieve dimensional accuracy, and the filament is always gonna be vulnerable to wear on that size scale. Once I pushed the printed part on and removed it again often enough, it's not firm enough anymore to hold the steel bar that's connected to the spring and that half of the spring just completely discharges. I'm always able to print more though, and I'm hoping that if I use a freshly printed part once I solved the second problem, there won't be too much wear from pushing it onto the device once so that it never disconnects once reassembled. But if someone has advice on a better design, I can always get back to this! The printed part is much higher than the original one because movement of the gear seems to push the printed part up a lot, so by making it this high, it will then push against the lid that I screw on top of all this once I reassemble the clock and it won't be able to be pushed off.
The second one is what I need much more analysis and advice for. I'm sure the problem lies in the distribution of the spring on the backside, in that I have no idea how much spool is supposed to be on either side. As I understand it, the balance-wheel works with a specific amount of tension in the spring (the one in the middle) that's so high that it translates through all the gears with enough force left for the curb-pin to move and the balance-wheel to do its thing and move the gears. At the same time, some amount of force needs to be on the other half of the spring because that's where the vibration comes from. This can be much less force though because there's way less gears to move for a much shorter amount of time.
The problem is that I don't understand springs and clocks enough to know how exactly the spring needs to be positioned to provide the right amount of force. I tried many different versions with 50/50 distribution, most of the spring being on the left or more being on the right, but none made the balance-wheel do its thing for longer than a few seconds once I winded up the clock as much as possible, often only 1-3 ticks, at most a minute (when most of the spring is on the left side). Either I misunderstood some of the mechanisms, or I actually assembled something wrongly, or I just didn't find the correct setup for the spring yet.
I would be grateful for any kind of advice on this! I realize that the clock is cheap and meant to be thrown away and I'll never get it back to full original function, but I'm using this as a learning opportunity. I wanna know how this works and how to fix it!
I'll try to describe a little more specifically how the mechanism works. Here's a photo of the side-gear from photo 3 disassembled; it's 3 individual pieces:
The lowest piece that I left on top of the steel rod (which goes through to the side-section of the spring on the other side of the device) is the mechanism that prevents the vibration hammer below it from triggering; it's under constant tension as long as that section of the spring is wound up to any degree. Right now it's wound up completely (the corner prevents me from being able to wind the clock beyond 60 minutes); when the timer hits 0, the hammer reaches the gap and can trigger.
While it does so, the side-section of the spring unwinds, making the steel rod turn, until the top piece, the 3d-printed one that broke, hits the nose that goes through the hole of the middle gear. At that point the rod can't turn any further and the spring retains the rest of the tension (which means that the 3d-printed part is under permanent tension at all times, which is probably why it's the part that broke).
The middle gear's job is to limit the winding to one full turn (until the gear hits the extended section / nose, a second mechanism preventing an over-wind of the clock). The hole is a little wider than the nose from below to allow for a little bit of extra vibration time (I assume).
The gear below the hammer is what creates the beloved "winding" sound (through a mechanism inside the gear) and transfers movement from the steel rod to another gear which powers the hammer.
The unintuitive part about this entire mechanism is that there's truly no reason for the two spindles to be connected. When one side unwinds, it doesn't push steel to the other side, it just pushes against the inner walls. That's why there's a metal plate laying on top of the spools when you reassemble it.
I usually try to keep that plate on, even though I have to hold it with one finger, because the force of the spring against the walls is so strong that it can easily jump out of its cavity. I spent an entire evening trying to get it back inside when it did that once. Because of that, I can also say that flipping the spindle around doesn't change anything - removed from the cavity, it's basically symmetrical. But, even though either spindle unwinding doesn't push anything over, with enough force, I can pull parts of the side towards the middle - as long as I remove the middle gear with the nose so that I can wind more than one full round and prevent the vibration from triggering, to prevent the side-spindle from unwinding. (If I want to give the side more spool, I just let the vibration trigger and manually unwind the middle section, the side-spindle then pulls on its own.) This does mean I can relatively easily change the distribution of how much of the spring is wound up on either side, which I think is the solution to this - in the worst case the distribution is just extremely specific and I won't be able to find it just by experimenting.
I think the two winding mechanisms, the two spindles, are only connected because that's cheaper (one part instead of two). Also, the spring is much stronger than what can be loosened or tightened completely with just one full wind, so unfortunately it's not as simple as making it loose and tightening it with one wind (also because the spring's natural form is much larger in diameter than its cavity, so it's under constant tension anyway).
After recommendation from @notbroken, I applied epoxy with some threads to try to bridge & heal the original piece. Let's see how this turns out:
Cette question est-elle utile ?
@nikkiflausch one of the best questions we have had on here in a long time. Outstanding! Don't you dare throw it out :-)
There are some outstanding fixers on here and let's see if we can get their attention to this problem @geirandersen and @dadibrokeit what do you think?
@oldturkey03 I’m making this comment to remind to get to this in the morning as I’m currently in bed. This is a very interesting repair!!
par Geir Ove Andersen
If the cracked white plastic is only split where it fastens onto the hex head nylock nut resulting in a loose fit, there may be a simple solution; epoxy repairs after cleaning the nylock nut and hex portion of the plastic gear. Plastic injection molded parts eventually dry out and become brittle, hence cracking. Epoxying on both surfaces of the crack, held snug with a clamp should allow experimenting on placement in correct position to restore full timer function. Once determined of position, epoxying this piece to the steel nut should make this a permanent repair.
par F Dryer
@notbroken The cracked plastic part needs to be able to move independently from the gear, a (nearly) full turn eventually stopped by the tooth (is that the term?) on top of the plastic gear below is what makes the alarm ring for a few seconds once the timer ran out. The cracked plastic can only be fixated to the steel if that function is supposed to be retained.
How would I make the epoxy repair happen?
@nikkiflausch I've disassembled a few clockwork devices, and generally when there's a spring for a wind-up mechanism it's usually loosely wound when the device is wound down and tightly wound after you wind it up. I'm with you as far as not quite knowing what to make of that mechanism you have that seems to be transferring the spring from one spool to the other. I assume one of the spindles is the winding mechanism and the other goes to the timer gears, is that right? If it works like others I've seen, then most of the spring should be wound up on the timer side, then the action of winding it should pull it over to the other spool. Also, is there any chance it's reversed? As in, will it make any difference to flip it over and swap which end goes on which spool?
I have to agree with @oldturkey03 that this is a very curious repair and like him I'm interested in seeing how it comes out.
par Jerry W
Voir 9 commentaires en plus