Accutron woes after a string of good luck

[LittleWatchShop]

After servicing about six with minimal issues, I have run across a couple that stump me.

I think I mentioned one before that seemed electrically sound but did not hum.  I have another.

All components check to the extent that I have measured them (cannot measure transistor Beta).

When measured on the test set, I get a little over 25uA.  This is about right if the unit is not oscillating.  1.5V and a 2.2M bias resistor with a beta of 50 would get in the ball park of 20uA.  A little more beta and you are north of 25uA.  This further confirms that the circuitry is functioning.  The cap measures about right.  I do not believe its value is too critical anyway.  It mostly serves as dc blocking and not establishing resonance.

I tried another tuning fork and...no luck.  I am going to drive this one to the end so that I can understand root cause.

[HectorLooi]

Thank you for your persistence. I'm sure you'll solve it in the end.

Just out of curiosity, when you measure the coils, are you measuring DC resistance or AC inductance? 

Is this the 218 or 214 circuit?

[LittleWatchShop]

36 minutes ago, HectorLooi said:

Thank you for your persistence. I'm sure you'll solve it in the end.

Just out of curiosity, when you measure the coils, are you measuring DC resistance or AC inductance? 

Is this the 218 or 214 circuit?

dc resistance.  I believe that the ac impedance will be intrinsic to the physical structure of the coil.  I do have the tools to measure inductance but don't believe it is necessary.

Since my post, I powered the unit and monitored the base and collector with an oscilloscope while "pinging" one tine of the fork.  I got a very nice sinusoidal envelope and it died away.  Then I removed power and did the same thing.  Nothing.  This leads me to believe that the gain of the transistor (beta/hfe) is too low.  Just a theory at this point.

[ManSkirtBrew]

Of the three 218s I've serviced, the one that behaved like you describe was weak magnets on the fork. If you replaced the fork, probably not your issue.

Two of the other non-runners were a 218 and 219, and both sets of resistors and caps were out of spec. Soldering in new ones fixed the problem in both cases.

Not as scientific, but my anecdotal results so far. I'm a computer programmer, not an EE

[LittleWatchShop]

Just now, ManSkirtBrew said:

Of the three 218s I've serviced, the one that behaved like you describe was weak magnets on the fork. If you replaced the fork, probably not your issue.

Two of the other non-runners were a 218 and 219, and both sets of resistors and caps were out of spec. Soldering in new ones fixed the problem in both cases.

Not as scientific, but my anecdotal results so far. I'm a computer programmer, not an EE

Very interesting.  I would posit that increasing the magnetism is roughly equivalent to increasing the transistor gain, so in that sense, your results are the same as mine.  Indeed mine is a 218 (sorry I forgot to mention earlier).  Similarly, the resistor can have the effect of a low gain transistor (not in the same way).  Transistor beta is a function of collector current, so a resistor going higher lowers the quiescent collector current and thus reduces gain.  But, I would think the resistor would have to be way out of spec...dunno.

I am impressed with your soldering skills.  I have the same iron as you, but perhaps a different tip.

I did a little research on small transistor packages and so far all I have found is SOT23 type.  I managed to solder one in (another case) but it was crowded.

Another thing I did was to fool around with some tiny compasses and several forks to see if I could measure a difference in their magnetism.  Sketchy results.

[JohnR725]

9 hours ago, LittleWatchShop said:

I could measure a difference in their magnetism. 

I believe they make newfangled electronic devices that could measure magnetic fields but they're probably expensive. In the absence of that you can make your own if you had a linear Hall effect sensor. The last time I needed to understand what a magnet was doing I just got myself a linear Hall effect sensor and made by own magnetic linear meter it worked quite nicely.

[LittleWatchShop]

Occam's Razor strikes again.

It appears that the problem was purely mechanical. 

The hack lever in some instances can ride up the fulcrum pin.  When it does, it can hit the fork ever so slightly.  The train bridge will hold it down somewhat, but I don't believe it is always snug.

Also, I have observed different washers.  Sometimes two gold, sometimes gold and silver.  I always to put it back the way it came.  In this instance, I think the wrong washer was in place...the interior one.  So all this worked together to stop oscillation.

 

[HectorLooi]

Good job.

Sooo.... when are you going to start on a 214?

I would like to hear about your experience. And count the number of expletives you had to use. 

 

[LittleWatchShop]

16 minutes ago, HectorLooi said:

Good job.

Sooo.... when are you going to start on a 214?

I would like to hear about your experience. And count the number of expletives you had to use. 

 

I still have a bunch of these 218s to finish. Then I will read the book, "Cuss Words for Watchmakers"

[HectorLooi]

1 hour ago, LittleWatchShop said:

I still have a bunch of these 218s to finish. Then I will read the book, "Cuss Words for Watchmakers"

Is that the International version or Texan version?

Many years back I met a group of Texan oilmen who were staying in the same hotel. And we seemed to bump into them everywhere we went. Didn't understand half the things they said, it was like a foreign dialect. My New Yorker friend who was with me didn't fair much better.