EMRFD Message Archive 13042

Message Date From Subject
13042 2016-08-08 09:12:47 K5ESS Question RE EMRFD Fig. 9.82 - 10 pF across feedback resistors

I’m laying out a PCB (mostly SMD) for a phasing SSB exciter based on EMRFD Figs. 9.82 and 9.83.  My question has to do with the function of the 10 pF capacitors in parallel with the 10K feedback resistors in the all-pass circuits.  I’ve found no discussion regarding them and they are absent in the circuits for the mini R2 (Fig 9.62) and the many previous articles on similar projects.  I’d like to eliminate them if possible.  I analyzed the circuits in LTSpice and saw a reduction in gain well above the audio range (around 3 MHz as I recall) .  Were these added to reduce the possibility of a high frequency oscillation ?  Any comments on the pros/cons of eliminating them would be appreciated.

Mike, K5ESS

 

P.S.  I thought I posted this online but it never showed up.  I apologize if it now shows up twice.

13043 2016-08-08 12:14:02 iq_rx Re: Question RE EMRFD Fig. 9.82 - 10 pF across feedback resistors
The 10pF chip capacitors across the 10k feedback resistors in the all-pass networks reduce gain well above audio frequencies.  They were originally added to cure a high frequency oscillation experienced with some particularly hot varieties of op-amp included in experiments during the writing of EMRFD.  I've never had an oscillation with NE5532 op amps in that application, but your results may vary.

I suggest you select an op-amp type, by a batch of new ones so they are likely built on a recent IC process, design your first-pass experimental circuit board without the 10p capacitors, and then use a high impedance oscilloscope probe to confirm that everything is stable.  If not, you can add 10p chip capacitors to reduce high frequency gain, experimenting to find the offending stage or stages.

Enjoy the layout experiments.  I expect that you'll find the layout an equal engineering challenge to the schematic, and no simulator can yet handle all the interactions: electromagnetic; milliohms of common trace resistance; current crowding; thousands of transistors in op-amp circuits etc. in finite time.  For a circuit of the complexity of an SSB exciter, I generally plan on 4 revisions of the circuit board from first pass through production boards.  Then after several hundred have been built, I often do a further bit of tweaking.  Until I have a prototype working on the bench and I can make reliable measurements, I don't take the layout details too seriously, because during the prototype measurement phase I always discover a few things I wasn't aware I didn't know.

Best Regards,

Rick KK7B
13044 2016-08-08 15:19:42 K5ESS Re: Question RE EMRFD Fig. 9.82 - 10 pF across feedback resistors

Thanks Rick.  I suspected that they were to fix an oscillation problem.

Mike, K5ESS