I use AD797 on old VSPS boad as MC preamp.. in power supply after LM317/337 only got 1uF Wima MKS2, works without problems..
Since RIAA network is in feedback does not need any compensation..
Attachments
I sad R2, not R1.. if you know anything about how opamps works, you understand why there.. before that, you supposed to wonder how is it possible that AD797 play worse than OPA27..
Things are very simple, if in some circuit twice better opamp works bad, you must find why.. always must be a reason and explanation.. in this universe, and with these laws of physics, OPA27 can be better than AD797..
Last edited:
I am the first to admit that I know absolutely NOTHING about how Op Amps work. I can follow instructions, and solder to a reasonable standard but give me a box of bits and ask me to make an amplifier, then forget it.
I don't dispute that the AD797 is better than the OPA27. I just posted my findings as previously stated.
I recently rediscovered the initial thread about the VSPS and there are some very interesting suggestions in there by Thorsten
http://www.diyaudio.com/forums/analogue-source/33555-diy-phono-preamp-cheap-simple.html#post388315
Especially the "including the output cap in the feedback loop" and "use x10 resistors and /10 capactitors for the RIAA" are very interesting...
Anyone tried them?
http://www.diyaudio.com/forums/analogue-source/33555-diy-phono-preamp-cheap-simple.html#post388315
Especially the "including the output cap in the feedback loop" and "use x10 resistors and /10 capactitors for the RIAA" are very interesting...
Anyone tried them?
Oh and has anyone tried the pin 8 decompensation thingy on AD797?
~50pF from output (pin6) to pin8?
It is supposed to significantly reduce output impedance and distortion on gains above 10 or frequencies above 30K (could be perceivable on lower ones too?)
We could also use C2 to slightly increase bandwidth even further, but it is supposed to have no significant effect on gains below 100
Page 15 of datasheet
~50pF from output (pin6) to pin8?
It is supposed to significantly reduce output impedance and distortion on gains above 10 or frequencies above 30K (could be perceivable on lower ones too?)
We could also use C2 to slightly increase bandwidth even further, but it is supposed to have no significant effect on gains below 100
Page 15 of datasheet
The datasheet suggests
"Use of this feature improves distortion performance when the closed-loop gain is more than 10 or when frequencies of interest are greater than 30 kHz"
so I guess the first it is within the scope of our thingy?
It is a fairly easy test anyway. I am receiving my stuff today, so I will report back
"Use of this feature improves distortion performance when the closed-loop gain is more than 10 or when frequencies of interest are greater than 30 kHz"
so I guess the first it is within the scope of our thingy?
It is a fairly easy test anyway. I am receiving my stuff today, so I will report back
imd of frequencies slightly above and well above the audio spectrum can fold back into the audio spectrum.
That means your amplifier should not distort signals that are still relatively strong but outside the audio spectrum.
Once you "KNOW" that all signals above 500kHz have been attenuated to 100dB below your wanted signals you can probably then ignore those HF signals.
But the signals below your brickwall filtering will still be affecting the performance of the amplifier and the effects of such should be minimised to preserve the wanted signals.
That means your amplifier should not distort signals that are still relatively strong but outside the audio spectrum.
Once you "KNOW" that all signals above 500kHz have been attenuated to 100dB below your wanted signals you can probably then ignore those HF signals.
But the signals below your brickwall filtering will still be affecting the performance of the amplifier and the effects of such should be minimised to preserve the wanted signals.
you seem unable to understand that frequencies above 30kHz are of interest to an audio amplifier designer.
I looooooooooooooooove AD797 on my VSPS
I got it in there bypassed by 100nF TDK ceramics
Currently it seems a bit trembly but I will give it a few days.
Unfortunately I am a moron and ordered 5, so I cannot try them on the regulators... The LME49710 works very nicely though
And yes it did a big difference replacing the OPA27 there
Tomorrow I will try to get some more locally, along with some 47pF ceramics to try the compensation thingy
I got it in there bypassed by 100nF TDK ceramics
Currently it seems a bit trembly but I will give it a few days.
Unfortunately I am a moron and ordered 5, so I cannot try them on the regulators... The LME49710 works very nicely though
And yes it did a big difference replacing the OPA27 there
Tomorrow I will try to get some more locally, along with some 47pF ceramics to try the compensation thingy
You don't need compensation in VSPS, you have RIAA network..
Compensation cap is necessary at first chip position in phonoclone..
I m not talking about usual feedback compensation for stability.
I am talking about utilizing the special feature in pin 8 to reduce distortion, applicable for gains bigger than 10.
The interesting thing is that it is also suggested that if connected after the following buffer or capacitor, it can reduce its distortion as well
http://www.analog.com/library/analogdialogue/archives/39-05/Web_Ch6_final_I.pdf
Page 6.59
This actually allows us to implement something like Thorsten's suggestion to include C3 in the feedback loop, without actually having to modify the VSPS or PC3 pcb.
We just have to run a 50pF cap from pin8 of the output opamp to the output side of C3. Perhaps we only have to move the 47R after C3 to keep the capacitance isolation?
I am talking about utilizing the special feature in pin 8 to reduce distortion, applicable for gains bigger than 10.
The interesting thing is that it is also suggested that if connected after the following buffer or capacitor, it can reduce its distortion as well
Source:
http://www.analog.com/library/analogdialogue/archives/39-05/Web_Ch6_final_I.pdf
Page 6.59
This actually allows us to implement something like Thorsten's suggestion to include C3 in the feedback loop, without actually having to modify the VSPS or PC3 pcb.
We just have to run a 50pF cap from pin8 of the output opamp to the output side of C3. Perhaps we only have to move the 47R after C3 to keep the capacitance isolation?
Last edited:
I understand what you mean..
But you do not understand me.. capacitor at pin 8 is the only additional compensation, and you have two caps already in the chip.. unnecessary in this case..
OK OK I made a mistake. Its not the first time, and it won't be the last.
The question still stands. Does the quality of the caps matter?
I like using the more stable variety of ceramics like C0G which are less microphonic and more stable with frequency