Accutron 218 transistors

[ManSkirtBrew]

Having had success replacing the resistor and capacitor in a few 218 and 219 circuits, I've gone down the rabbit hole trying to find a modern replacement transistor. I've searched everywhere I know to look for specs on the original transistors and come up empty.

My last digital design class was too many decades ago, so I was hoping one of y'all out there is a EEE (Electronics Engineer or Expert) and can offer some advice.

For example, let's say I'm considering an NPN digital transistor. When I look up the part, it's available with a built in 47k, 22k, or 10k ohm input resistance.

How does one decide? I assume that's what the starred *R resistor in the schematic below is for.

And then how do we make the decision between a digital and bipolar transistor?

[nevenbekriev]

No, You don't need 'digital' transistor. What You need is small comon use transistor, where small usually means Ic max about 100 mA. The other thing important is to know the material of the transistor - well, looking on the schematic we see it is NPN, which usually means it is silicium transistor. But You have to check to confirm, as it is possible that they may have used germanium one. So, use digital multimeter and check the Uf of the PN junctions, directly on the PCB (remove the baterry). If it is about 650 mV it is silicium and if about 200 then germanium. 

If it is, as I suppose, silicium, then use BC548C or similar. If the H21 is much bigger than needed, then the index wheel will move twice faster, and to avoid this, You will have to increase the '*R' on the schematiic.

[praezis]

1 hour ago, nevenbekriev said:

If the H21 is much bigger than needed, then the index wheel will move twice faster, and to avoid this, You will have to increase the '*R' on the schematiic.

… or avoid the _C type and use _A or _B transistor.

Frank

[JohnR725]

4 hours ago, nevenbekriev said:

No, You don't need 'digital' transistor.

 

what? We live in the digital world everything is digital today except I believe this is a linear circuit seems like it should need something that's not digital. then I suspect pure digital transistors designed for high-speed switching might not actually be happy to biased it into to a linear mode.

4 hours ago, nevenbekriev said:

The other thing important is to know the material of the transistor - well, looking on the schematic we see it is NPN, which usually means it is silicium transistor.

I wondered about this because for instance the 214 came out in the early 60s and I don't know how plentiful silicon transistors were? Then yes the 218 is later but the circuitry is basically very similar.  

out of curiosity as doing a quick patent search the see if it tell us what the transistors they used our and right this instant know they have not been to have schematics though. Then unlike general concept patents Bulova was very nice and made exact patents for a variety of their watches because it wanted the mechanical + aspect patented

.then I did remember something about the transistor so I searched for a transistor number and Bulova this came up first all snipped out an image. Now you have at least one transistor that was used in tuning fork watches. So this will give you an idea of specifications. Then yes I really like where the meaning of Accutron comes from you can see it in the title down below. Which works out nice for them now that they have all kinds of electric watches including one powered by static electricity and it does not go against the name which everyone thinks means tuning fork watch. Yes you really should check out the nifty static electricity watch it's really quite amusing.

https://www.hodinkee.com/articles/five-star-general-omar-bradley-special-bulova-accutron

 

 

 

 

 

hadn't quite closed out my patent search basically the patents for the 214 we have this.

then before I closed out my Google patent search added in NPN and it only came up three entries of attaching one of them as it says silicon.

US3421309 Bulova NPN silicon.pdf

[rjenkinsgb]

So-called "Digital transistors" are just transistors with appropriate resistors included in the package to allow them to be driven directly from digital logic ICs & microcontrollers etc.

You definitely need a normal transistor for the Accutron.

[ManSkirtBrew]

Thanks everyone. I've got some parts on order to do some breadboarding tomorrow. I'll keep you posted!

[ManSkirtBrew]

Well, the first attempt was a wash because I bought 22nF capacitors instead of 220nF. So after a second order, I soldered everything onto my SMD breadbaord.

I de-soldered the electronics from a 218 movement that I had previous tested and hooked my little circuit up in its place, and hooked the whole thing up to 1.4v from my power supply.

...and realized I'd gotten the pinout of the transistor wrong. After I removed and re-soldered it in the correct orientation, I realized I had the connections to the coils backwards.

Once I fixed that, THE FORK STARTED VIBRATING! I could not have been more excited!!!

If you haven't guessed, my somewhat evil plan is to make a surface-mount PCB replacement for the resistor, capacitor, and transistor in the 218s. Why? Because why not?

The transistor is in a SOT-523 package, the resistors and capacitor are 0402 (1005 metric) packages. I have to redo the circuit to get the transistor pinout correct, but here's what I have in mind:

[ManSkirtBrew]

This sure is fun Now that the proof of concept works, it's time to prototype.

Looks like I've got just about 1.6mm of height from the plastic component tray to the top of the main bridge. With an 0.6mm thick board and the 0.9mm tall transistor, this should fit without any modifications other than removing the existing components.

For anyone wondering, the transistor is CMUT2222A bipolar NPN. R1 is the bias resistor, which is just a 0 ohm jumper for this setup.

R2 is the 2.2M ohm resistor, and C1 is the 220nF capacitor. DR goes to the drive coils, FB goes to the feedback coil.

[ManSkirtBrew]

They showed up! It's a perfect fit. Can't wait to populate it and try it out.

 

[ManSkirtBrew]

As is tradition, one step forward, two steps back.

Got the board populated and soldered into place without any issues.

 

But no hum. So I started testing the coils with an ohmmeter. I got 5.84k ohms across D1 (from red to red in the picture below), which is as expected. But I'm getting an open circuit for the other drive coil and feedback coil, D2 and F1 (from green to each of the two yellows).

 

Since the movement was working with my breadboard setup, it implies I somehow broke the connection between the coils and the solder lugs. They're all the way at the bottom of the lugs, but maybe the heat migrated down and broke the connections?

I guess it's possible it happened while cleaning the flux off, but I used a soft artist's brush and isopropyl alcohol.

I did a lot of high magnification examination, and I don't see any issues, but let me know if you see anything I missed or if you can think of anything else I should check.

[ManSkirtBrew]

For clarity and ease of testing, I redid that diagram, flipping the circuit to match the layout of the physical watch.