Last modified: 2023-11-12 08:53:47
< 2023-11-10 2023-11-12 >Well it hasn't stalled yet, that's good news. It seems pretty resistant to stalling if I even press my finger against it, and even once stalled manually it still won't run backwards and starts up again on its own. So the performance in the current state is acceptable.
However it currently has the big electromagnet in series with one of the coils, and it gets even hotter than it did before.
I've ordered a 9v transformer, maybe this will be better. Failing that, I can add a fan like Jay suggested.
In the mean time, it would be good to understand why the electromagnet works, and see if I can come up with some sort of circuit that will give a more convincing phase shift on 6v, and see if the clock will run on 6v after all.
I can't believe there are no decent DuckDuckGo results on how to phase-shift an AC supply by a given angle. Surely this is something other people have wanted to do? ChatGPT says:
To achieve a 90-degree phase shift in a 50Hz AC power supply, you can use a component called a capacitor or an inductor, depending on whether you want the phase shift to be leading or lagging.
Capacitor for Leading Phase Shift:
- Connect a capacitor in series with the load.
- The formula for calculating the capacitive reactance (Xc) is Xc=1/(2.pi.f.C), where f is the frequency (50Hz in your case) and C is the capacitance.
- To achieve a 90-degree phase shift, the capacitive reactance should be equal to the load impedance.
- Xc=Zload
Inductor for Lagging Phase Shift:
- Connect an inductor in series with the load.
- The formula for calculating the inductive reactance (Xl) is Xl=2.pi.f.L, where f is the frequency (50Hz in your case) and L is the inductance.
- To achieve a 90-degree phase shift, the inductive reactance should be equal to the load impedance.
- Xl=Zload
Keep in mind that the actual impedance of the load may include resistance (R) along with any reactance. The total impedance (Zload) is given by:
Zload = sqrt(R^2 + X^2)
Is this right? How can I work out the impedance of my load? I'll just measure the resistance for now.
I measured about 30 ohms across the coil of the stepper motor.
I asked it about a 91 degree phase shift and it said to set Xc to Zload * tan(91 degrees), which implies that for 90 degrees it should be Zload * tan(90 degrees), but it says to use just Zload, and tan(90 degrees) is not 1. Oh well, let's continue.
Zload = 30 ohms
Xc = Zload = 1/(2.pi.f.C) = 1 / (2 * 3.142 * 50 * C)
30 = 1 / (100.pi.C)
100.pi.C = 1/30
C = 1/(3000.pi) = ~0.0001 Farads = ~100uF
So I want to try a 100uF capacitor in series with the load. I think that is pretty big. My "104Z" (big ones) are 100nF, which is a factor of 1000 too small, which explains why I didn't get any different results no matter which capacitors I used. So I'd be looking for some marked "107" I think.
Most of the capacitors that I can find with that much capacitance are electrolytic, which is no good for AC.
There are some available described as "motor start" capacitors. I've ordered one. I think it is bigger than I need because it is rated for 250v, but if it works it works. In the mean time I can try and rummage around and see if I can find a non-electrolytic capacitor a bit bigger than 100nF.
Disheartened by the trouble I am having making these small parts, I want to lay out a new design using a larger escapement and a larger balance wheel.
I still intend to "finish" the escapement prototype I'm working on at the moment, in the sense that I want to make the escape wheel pinion and try running it with a weight. But I don't expect it to work, and I won't put much effort into making it work. Just consider it good experience in making small parts, and move on.
Maybe it would be a good idea to make a 2x scale prototype of the actual watch, but 3d printed where possible. Can I print module 0.8 gears?
Yes, actually, looks reasonably decent: