Hi EvanF,
First a disclaimer. I have no first hand experience with the AD797; perhaps other members will weigh in. This part is capable of good performance in many applications (I can't speak to its use in the 339A oscillator, others might) but is has a reputation of being susceptible to oscillation.
I would assume the previous compensation should be removed: i.e. C44, C45, C46, C48 and CR7.
Is there any symptom of oscillation? Check with a 'scope at U1's output. Be sure to use a x10 scope probe to minimize capacitance loading the opamp--- it can provoke oscillation. An added precaution would be to look at U1's output at the R29/C46 node (C46 removed). Probing through R29 will further reduce cap loading of the opamp.
Is there evidence of distortion on the monitor output? Would you offer details--- dBc, perhaps a scope picture of the residual at the monitor output?
The AD797 data sheet discusses a distortion reduction capacitor, but perhaps you might avoid using it, atleast initially, until you see low distortion with no oscillation. Then you could introduce the cap to see improved distortion results. In this application, the opamp must be unity-gain stable because the bridged-T network delivers unity feedback at high frequencies.
Good luck, keep us posted.
Steve
First a disclaimer. I have no first hand experience with the AD797; perhaps other members will weigh in. This part is capable of good performance in many applications (I can't speak to its use in the 339A oscillator, others might) but is has a reputation of being susceptible to oscillation.
I would assume the previous compensation should be removed: i.e. C44, C45, C46, C48 and CR7.
Is there any symptom of oscillation? Check with a 'scope at U1's output. Be sure to use a x10 scope probe to minimize capacitance loading the opamp--- it can provoke oscillation. An added precaution would be to look at U1's output at the R29/C46 node (C46 removed). Probing through R29 will further reduce cap loading of the opamp.
Is there evidence of distortion on the monitor output? Would you offer details--- dBc, perhaps a scope picture of the residual at the monitor output?
The AD797 data sheet discusses a distortion reduction capacitor, but perhaps you might avoid using it, atleast initially, until you see low distortion with no oscillation. Then you could introduce the cap to see improved distortion results. In this application, the opamp must be unity-gain stable because the bridged-T network delivers unity feedback at high frequencies.
Good luck, keep us posted.
Steve
What is setting of frequency switches? Roughly 500 Hz displayed if I read picture correctly.
Try shorting U1's non-inverting input (pin 3) to ground. Oscillation should stop, output should show 0V, there should be no high frequency oscillation.
If response is as described, then remove short, try adjusting R30 through it adjustment range to see if amplitude will drop into linear operation. If no change, would you report voltages at TP4, TP5, TP6, TP8. Both DC voltages and any peculiar waveforms.
Thanks!
Try shorting U1's non-inverting input (pin 3) to ground. Oscillation should stop, output should show 0V, there should be no high frequency oscillation.
If response is as described, then remove short, try adjusting R30 through it adjustment range to see if amplitude will drop into linear operation. If no change, would you report voltages at TP4, TP5, TP6, TP8. Both DC voltages and any peculiar waveforms.
Thanks!
what is the part you're looking to replace?
AD797 is not really a "drop-in" replacement for much of anything - it's too special.
btw, just a hint - there's a LONG thread on this board with some REALLY good improvement suggestions for this analyzer. it worth slogging through ...
AD797 is not really a "drop-in" replacement for much of anything - it's too special.
btw, just a hint - there's a LONG thread on this board with some REALLY good improvement suggestions for this analyzer. it worth slogging through ...
Just powered the 339a on again. I now get a .05% THD sine wave on the output. I did read that thread, seems RN Marsh used an AD797 too
I finally put an LT1468 in. Only removed c44, c45, c46, c48, and cr7. about the test the loopback THD+N. I can adjust the amplitude on tp8 just fine at least
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.