Tuesday, January 8, 2013

But Why Did Swamping Resistors Cure JBOT Oscillations?

OK,  so as I mentioned in SolderSmoke 148 (apprently with screeching in the background -- sorry) I managed to tame a bad JBOT oscillation problem by placing 680 ohm resistors across the primaries of the broadband transfomers in the JBOT.  Now I'm wondering how/why this worked, and what we could do to make this kludge unnecessary.  

Am I correct in suspecting that at the problematic frequency of 250 kHz, the one or more of the tranformers are self resonant, and that the resistors just bring down the Q of the -- in effect -- tuned circuit?  This lowering of Q makes feedback and oscillation more difficult, right?  

Here is the background.   Read from the bottom up.   


There is an old saying in Spanish:  "No hay mal que por bien no venga."   More or less this is the same idea as: "Every dark cloud has a silver lining."  Well, the dark cloud was my techno-agony with the parasitic oscillations (see my plea for h elp from yesterday).  The silver lining was the e-mail from Edgardo, LU1AR, in Buenos Aires that helped me get rid of them.    Edgardo advised putting a resistor across the primaries of the JBOT amplifier stages.  This is an old "lower the Q" trick, the idea being that lowering the Q might help prevent the amp from self-oscillating.     I used 680 ohm resistors.   First I put one across the primary of Q1.  No joy. Then Q2.  No luck.  Then I put one right across the primary of that big output transformer.  That did it!  The parasitics disappeared.   And I still get a nice 4 watts of output.  Thanks Edgardo.  I hope to make a contact with this rig today. 

The real silver lining in this story comes, however, in the form of Edgardo's blog site.  Wow, what an inspiring example of Argentinian Knack.  Radios, telescopes, auto-giros.  This guy is also into homebrew DSB rigs.  Wonderful stuff.  Check it out.  (Google will translate it for you, if necessary, but even if you don't read Spanish, the pictures tell most of the story.) 
I put this URL blog up on the SolderSmoke blog.  
Thanks to all who sent advice and encouragement.  Merry Christmas! 

--- On Sun, 12/23/12, lu1ar  wrote:

From: lu1ar
Subject: [BITX20] Re: JBOT taking off at 250 kHz
To: BITX20@yahoogroups.com
Date: Sunday, December 23, 2012, 9:56 PM


First. You must secure the RF decoupling of +B line. Use a bunch of small capacitors in paralell 1n and 100 nF with 22 uF electrolitic.
Second: Use 1Kohm resistor in paralell with primary of the transformers. Begin with the driver, then the 1st amplifier and then the output transformer.
Let us know results.
Regards.
Edgardo Maffia LU1AR
Buenos Aires - Argentina

--- In BITX20@yahoogroups.com, "iam74@..." wrote:
>


> --- In BITX20@yahoogroups.com, Bill Meara wrote:
> >
> > I'm building my FOURTH JBOT and this one is giving me more trouble than all the others combined.   I seek the help of the group. 
> > Here are the details of my tale of woe:
> > 20 meter rig.  Double Sideband.   The JBOT is fed by a simple two-diode, trifilar toroid singly balanced modulator.  NO DIPLEXER TO SPEAK OF.  At the output of the balanced modulator there is a 1000 uH choke to ground and a .001 uF cap to the input of the JBOT.  AF amp is an LM-386.  VXO is a very simple MPF-102 one stage ceramic resonator VXO with no buffer stage.  7 element low pass filter (three coils, four caps). 
> > All the transformers are wound on FT-37-43 cores.  T3 is four such cores stacked 2X2.
> >  
> > The arrangement works beautifully into a 50 ohm dummy load.  But as soon as I connect it to an antenna (a dipole out in the yard, fed with 50 ohm coax) the output signal gets ugly.  
> > Looking at it on the scope, instead of the nice figure eight pattern (similar to the two tone test pattern of SSB) I get ugly fuzzy outcroppings from the peaks.  Looking at the signal more closely I can see that in addition to the 14.2 Mhz signal, there is another LF oscillation at around 250 kHz. 
> > I've been struggling with this.  I can't get rid of the LF oscillations. The leads are all short and the inputs are far from the outputs.   I've beefed up shielding, and decoupling.  I've sacrificed a chicken to Papa Legba. Nothing seems to help.  
> > I THINK the feedback/oscillation is taking place in the JBOT itself -- not through the
> > audio amp or the balanced modulator or the VXO.
> > I watch the RF and the AF inputs to the balanced modulator to see if there is any difference between the stable situation (with the 50 ohm dummy load) and the unstable situation (with the antenna),  I don't see any differences. 
> > I even put an antenna tuner between the final and the antenna anb made sure that the antenna looks like 50 ohms non reactive.  This seems to help a bit, but the ugly instability is still there. 
> > Help me! 
> > 73  Bill N2CQR    
> >


Our book: "SolderSmoke -- Global Adventures in Wireless Electronics" http://soldersmoke.com/book.htm Our coffee mugs, T-Shirts, bumper stickers: http://www.cafepress.com/SolderSmoke Our Book Store: http://astore.amazon.com/contracross-20

No comments:

Post a Comment