View Single Post
  #8   Report Post  
Posted to rec.audio.tubes
patrick-turner patrick-turner is offline
external usenet poster
 
Posts: 119
Default Wien Bridge Oscilator, 1Hz to 1MHz.

Phil mentioned this about my butchering of a crummy low quality budget iBWD141 oscilator...

""The Turneroid abortion has horrible, discontinuous frequency adjustment,
massive transients in the output and very high THD. So it is utter garbage
for audio testing.
As I expected, when you throw a Pearl in front of a swine. ""

The original 141 I was given had a ****ed Compton pot, and no amount of fiddling with adjustments would improve it.

Rather than chuck it out, I tried implememting a home brew op-amp using discrete bjts and j-fet input. Such a voltage amp can amplify signals from 0.1Hz to about 4MHz at 3.6Vrms output without slew THD and have low output Vdc offset and have low Rout, and low noise, all better properties than the original 141 amp.

But anyone should feel free to use some of the op-amps available which are not to be found at Jaycar or wescomponents.com.au and which may produce higher F with wien bridge than 1MHZ.

The output on my oscilator amp has a voltage clamp to prevent Vo swinging wildly during switching F ranges, (and unavoidable in the original 141.) The make before break switches used for switched resistors instead of a damned pot prevent noticeable transients. Any transients are so short lived during normal use that any amplifier being tested will cope well with the transients because they have peak voltage of only +/- 1.4db around normal output voltage of 3.6Vrms. If a tested amp doesn't cope with such mild transients then it is a really awful bloody amp which needs serious re-design.
The Turnerized 141 replaces a similar unit I buit from scratch some years ago but which only goes from 2Hz to 200kHz in 5 ranges, and often I wanted to know what happens between 200kHz and 1MHz. That unit has square waves to 470kHz, and not quite as nicely square as the Schmitt trigger I have made, shown at my page on the revised oscillator.

The overall build quality of the original 141 was Bluddy Crummy. Low budget **** quality, and nowhere near what was then being made in USA and which was of much more expensive, and mainly only purchased by government depts, armed forces, universities, but rarely by private service ppl. The original BWD 141 board is a monument to bean counting and penny pinching cost cutting and parts are crammed onto a board that is too small and very low quality like you'd find in really poor quality kits from Jaycar or Dick Smith. Almost anything any one does can be better.

I also have a 161 which is a function generator with sine, square and triangle wave and done using an early function chip with a single pot for F control. Its not bad, but very imprecise and with high THD and varying square wave symmetry and atrociously low Vo level so had to instal a wide band home brew op-amp and do considerable repairs and alterations to make it as usable as my existing totally home brew gear. I did once have a Topward function gene, with far more sophisticated circuitry, and AM and FM functions up to 2MHz, and far less THD than the BWD but the pot wore out, then another wore out, and the output amp burnt out so badly plus many other chips when a lead touched a tube amp B+ that I binned the bloomin thing. Its power tranny hummed loudly, i'd replaced bits so often it was a mess, and time to say farewell. The use of a tubed cathode follower buffer after the unit Vo was a good idea, but I got lazy, forgot to include it, and POOF, smoke, and no sig.
Later, I found diode voltage clamps are OK most of the time.

I have a pair of HP 333A THD testing contraptions and a spare 4 gang tuning cap, and time permitting, I may build up a better unit than the original HP.
But I also have about 6 triple gang tuning caps from old radios and an ex-ham's collection so I can get 400pF to 4,000pF C variation and with low R get up to 5MHz with wien bridge if I wanted to with suitable SS devices. But for 1Hz, the R must be 40Meg, so fixed C and variable R is better for 1Hz to say 100Hz, and then the tuning C for the rest, so TWO oscillators need to be made.

Phil goes on to spit the dummy, and to insist I am a copulating female sex organ. IMHO, he really should stick to electonics, not delve into incorrect personel assessment, lest everyone begin believing all he says is incorrect.