From: info@turneraudio.com.au   
      
   the.audio....@gmail.com Oct 21     
   Other recipients:    
   Good article. One thing that isn't mentioned, though, is the calculation for   
   f: = 1/(2π·R·C·√6) (although the nomogram helps). >...unlike so many   
   modern descriptions of anything technical, the old guys could   
   On Tuesday, September 10, 2013 1:36:27 AM UTC-4, patrick-turner wrote:    
   > There is an artical posted here about tubed phase shift oscilators....    
      
   Good article. One thing that isn't mentioned, though, is the calculation for    
   f: = 1/(2π·R·C·√6) (although the nomogram helps).    
      
   >...unlike so many modern descriptions of anything technical, the old guys   
   could    
      
   >say a lot, and not leave out very much. I think my father's generation were    
      
   >better communicators than many of today's website creators...    
      
   I concur.  I think that people (authors AND readers) were a good deal more    
   patient back then than they are now.  Today, we absorb (or disseminate) just    
   enough information to IMPLEMENT, but not enough to truly UNDERSTAND.  Hence,    
   the "sound bite" method of teaching circuit theory.  Even in educational    
   settings, truly expository explanations (such as that article) are rare.    
      
   The irony that exists between the early and modern ages of electronics is that    
   today, we have (particularly passive) parts that are quite a bit better than    
   what was available "back in the day", but what's lacking is the workmanship,    
   attention to detail, and the "make it as good as humanly possible regardless   
   of    
   the effort involved" attitude that prevailed during that era.    
      
   It's no secret that the best equipment available is being produced on garage    
   workbenches and kitchen tables – DIY'ers combining the best of today's    
   components with the diligence and perfectionism of yesteryears' craftsman.    
      
   > Has anyone else built a decent phase shift oscillator with wide range of F?    
      
   I guess it would depend on what's considered a "wide range of F".  I've built   
   an op-amp-based PSO that goes to just under 1MHz (from about 1KHz).  But the   
   tube PSO's I've made were all LFO's for guitar amp vibrato (1-10Hz).    
      
   """"""""""""""""""""""""""   
      
   I didn't get anywhere with trying to make a wide band phase shift oscillator   
   as dedicated bit of test gear with an accurate single dial for many decade   
   ranges of F.   
      
   But I have now got my tubed wien bridge oscillator sorted and free of any   
   instablity and which has open loop gain at 1kHz of about 2,000. Tubes are 6BX6   
   for V1, V2 gain, then White follower with 2 x EL86. NFB is to V1 6BX6 cathode   
   from network using 4 x    
   12V-50mA "grain of wheat" lamps.   
      
   The high gain at 1kHz is reduced above 10kHz and below 100Hz by gain shelving   
   R&C networks. I have nearly completed the schematic drawings for my next   
   website page on wien bridge ocillators.   
   The unit gives 5 decade ranges of 20Hz-200Hz, and up to 200kHz - 2MHz. These   
   ranges require fixed R between 2M8 to 253r, and a quad of 3 gang tuning C   
   which gives about 2,500pF to 250pF. The changing of F by string dialcord and 4   
   wheels, one for each cap    
   gives very stabgle Vo with no bounce that one gets with some pots. The sixth F   
   range obtained by turning the rotary rage switch gives 1Hz to 20Hz, using a   
   dual gang 25k log pot + 1k25 and fixed C = 6u6 approx.    
      
   I have yet to instal a rotary switch for the output Vo range levels,   
   7V,2.2V,0.70.22V,0.07V, and 0.022V if I have a suitable 6 pos switch in junk   
   bin. Following this switch will be a 2k2 log pot for adjusting level, and then   
   a solid state complementary    
   darlington pair emitter follower buffer to maintain Rout at about 60r, from   
   class A bjts working from +/- 15Vdc, and idle current 200mA. I had thought of   
   using a tube CF but when output load is say 100r, the bjts give far less gain   
   loss.    
      
   All R&C values are trimmed for wanted F range. There are two dials, one for   
   the 5 tuning gang operated ranges and another for the LF pot range. I have to   
   calibrate cardboard templates of the F outcomes, then scan into PC and tidy it   
   all up in MS Paint.    
   Then I print 3 copies of each about the right size by trial and error, OK,   
   because the exact size is not critical for a dial with rotating pointer mving   
   around a fixed centre point. Then the paper copies are soaked in varnish and   
   laid on the front panel    
   and the text shows good contrast when you have 3 layers all soaked in varnish.   
   I could have illuminated dials, but that is not necessary.   
      
   In my retirement, I can afford to fiddle around making gear if I want.   
   Retirement is a wonderful change from my 20 year stint as volanteer amp worker   
   fixing every other man's old junk for net wages less than the dole, currently   
   $220 a week in Oz, while    
   the mean gross wage in Oz is now about $1,280, giving over $800 net after tax.   
   I live quite well on old age pension of $350. I owe the world SFA. What I   
   found sickening was that all customers I ever had never ever made sure I got a   
   socially just wage for    
   which they would be prepared to work for themselves. I could not be friendly   
   to any, because if I did, they'd just abuse the friendship by expecting me to   
   fix gear for zilch. Nobody actually mixes business with pleasure ethically.    
   But I digress.   
      
   Patrick Turner.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)