Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


praezis last won the day on September 29 2018

praezis had the most liked content!

About praezis

  • Rank
    WRT Addict

Profile Information

  • Location

Recent Profile Visitors

5,678 profile views
  1. praezis

    Timing and Lift Angle!!

    @John your picture shows even more: If you try to poise at >220 deg, you will make the error worse, as the effect of poise error is inverted there. @diamondslayer There is no danger at all using 52 lift angle with poising: if the real L.A. is e.g. 42 (pocket watch, chronograph etc), you will poise at a lower amplitude than expected. This just amplifies the visible rate differences in vertcal posirions! Frank
  2. praezis

    Loose roller on new balance staff

    If you need another opinion: hole closng with a domed punch (no hole) works. Just apply a light tap. Another method: a light tap in the middle exactly like tightening loose cannon pinions . Both only if its really tolerances, not a wrong, wobbling roller. Frank NB. I would not apply above methods on a hard steel roller.
  3. Soft soldering is the way this work is done on enamel dials. But: You have to heat the whole dial equally, usually with an alcohol lamp from below. Then there is no danger for the dial. Don't use brass wire for the post but copper wire. Frank
  4. Round mineral glasses are available in all sizes for little money. Thickness starts at 0.8mm afaik. They are glued in or pressed using a gasket. Window must have 90 deg edges. A too thin back can be an issue, test before cutting! Frank
  5. praezis

    Alloy vs Steel

    There is little danger to get a Nivaflex spring, as their price is a multiple of the stainless ones, so wholesalers will order the cheaper springs. Frank
  6. praezis

    D. I. Y. Watch Timing Machine.

    Hi, in my workshop I service mainly vintage chronographs. To adjust the beat error, I always have to rotate the collet. To see the pallet position and in which direction to rotate, I use to let down the mainspring, look, wind again. This means, after every check there is a different state of winding and amplitude. Beat error will show different [ms] values for the same error angle then, but beat error [deg] always shows the same value for the same error. That is why I like [deg] display for the beat error. Frank
  7. praezis

    D. I. Y. Watch Timing Machine.

    Hi Guido, here is your file running on PCTM. Amplitude varies 228...235 degrees. Frank
  8. praezis

    Sourcing a replacement mainspring

    Apart from the best reply of Khronos: exact length is least important. Frank
  9. praezis

    D. I. Y. Watch Timing Machine.

    Hi, yes, searching can be cumbersome. Microchip makes a few, e.g. mcp6002 or mcp6402. But their max allowed supply voltage is much lower than of lm358 . You will have to study the datasheets for details. Frank
  10. praezis

    D. I. Y. Watch Timing Machine.

    Guido, beware, the input resistance of that circuit (1 k) is too low for a piezo disc! With a bit electronic knowledge much can be adapted at this standard design. E.g. plug in a more modern OP replacing that grandfather of IC and it will work at 2V and lower. Frank
  11. When slotted balances are out of poise, you should also consider bent balance halves. Your slots look as if they are of different width. Frank
  12. praezis

    Help finding lathe collets

    The 40g is used in Germany. g for "Gang" means tpi. What is the diameter of the lathe bed, 25 mm? Frank
  13. praezis

    Tap and Die Kit

    Bergeon sold 4.7 mm punches with a set of collets. There are also collet holding arbors for Seitz jewelling tools, I have one in the big Seitz set. Frank
  14. praezis

    WOS mic & watch holder.

    Hi, rather late, I know. But just today I read this interesting thread again and looked at the sound file "Ingersol-trenton.wav" from post #1. From the speaker I could already hear considerable echo. Curious, I ran the file in PCTM (after conversion to mono). But there was a nearly perfect graph line and waveform, I could not see any evidence of suspected echo: Ok - filter is on by default. Switched filter off: Now the original sound appears. What a difference, the tic noise can hardly be recognized. Apparent is a wave at about 600 Hz, which probably comes from heavy resonance in the pickup device. And some 50Hz hum is visible, too. Frank