so doesn't any one knows why add FB in this PEARL II?
Also, a quick question. I saw the article it said second stage is "same phase
'" so the whole pearl II is phase reverse... But I saw the shematics, the
differential amp only get one signal to one side and use feedback as out side
of input... (feedback happens here)... So the 2nd stage should be reserver phase and the whole system should like Pearl, is phase correct.
Also, a quick question. I saw the article it said second stage is "same phase
'" so the whole pearl II is phase reverse... But I saw the shematics, the
differential amp only get one signal to one side and use feedback as out side
of input... (feedback happens here)... So the 2nd stage should be reserver phase and the whole system should like Pearl, is phase correct.
Nope
1st stage inverting
2nd stage definitely non inverting
It goes 180° (inverted) at the out of Q4 (drain) and goes 180° again in Q2 ... so we're back ...Nope
1st stage inverting
2nd stage definitely non inverting
Oh. I thought Q2 is the 3rd stage... and that is correct. Q2 & Q4 together
is none-inverting. So the whole system is inverted phase.
Q2 is not considered as a "gain stage" in the extent it has no real voltage gain, the diff. pair does this job, and the wole thing with feedback is considered as the "second" stage
One could also add a single stage jfet K170 style at the end of 1st stage, with drain output before the RIAA net. So for polarity would be "on phase" again, but at the cost of an additional gain stage. See the AlephOno MC section as a reference. But I think Wayne's advice to invert polarity at the cartridge leads make more sense than additional gain stage
Wayne, could we know what is the open loop gain / bandwidth of the second stage ? Thanks
nAr
Last edited:
Q2 is not considered as a "gain stage" in the extent it has no real voltage gain and The diff. pair does this job, and the wole thing with feedback is considered as the "second" stage
One could also add a single stage jfet K170 style at the end of 1st stage, with drain output before the RIAA net. So for polarity would be "on phase" again, but at the cost of an additional gain stage. See the AlephOno MC section as a reference. But I think Wayne's advice to invert polarity at the cartridge leads make more sense than additional gain stage
Wayne, could we know what is the open loop gain / bandwidth of the second stage ? Thanks
nAr
Agree. Switch the cartridge is more reasonable. I also hope to know the open loop gain too. OR if we short the R14, that means NO FEEDBACK. How high the gain will be and how much worst of the performance?
You have to try it for yourself
albeit not sure for R14 short 
The open loop gain of the 2nd stage might be of 60 or 80 dB ... closed loop gain value is theoretically (R16+R14/R14) non inverting, so x101 or slightly greater than 40 dB ... If you want lower distortion, better reduce gain. But it means either 1°) use more feedback in order to do so, which you'll probably disagree 2°) degenerate the diff. pair by source resistors to lower open loop 3°) use lower transconductance devices for the diff. pair.
nAr
You have to try it for yourself
The open loop gain of the 2nd stage might be of 60 or 80 dB ... closed loop gain value is theoretically (R16+R14/R14) non inverting, so x101 or slightly greater than 40 dB ... If you want lower distortion, better reduce gain. But it means either 1°) use more feedback in order to do so, which you'll probably disagree 2°) degenerate the diff. pair by source resistors to lower open loop 3°) use lower transconductance devices for the diff. pair.
nAr
I think (2) and (3) are all good ideas. Maybe I just connect Q5's gate to Q4's and make it parallel, it will reduce the noise. Also, add 20R in each source.
And connect Q2's Gate with G5's Drain too. So it not differential (SE) but without feedback. Kill R14,!16, C9 all together...
How do you think?
-> Try & report !
I'll have to do a breadboard research soon on this ... gotten some 2SK363 so higher gm so I could just have 2 jfets at the first stage only instead of 4 ... I think I need only 30 dB of gain at the 2nd stage (MM). So maybe degeneration to keep same amount of feedback, not adding some ...Watch out the output impedance too
Brds,
nAr
Last edited:
it depends on how it is applied and to what extent.
Small update
Today, I did the following mods :
1°) Went for 0,4 uF instead of 0,2 uF for the coupling caps between the 2 gain stages
2°) As I only have high output cartridges for now ( about 3mV to 7,2 mV ) I reduced temporarily reduced the gain of the 2nd stage by having 2*100k resistors in place of R16, for a 50k value instead of 100k original. Following the formula of the non-inverting gain stage, gain equals now to (50+1)/1 so 51 times gain instead of 101 times, so about 34 dB gain or so -
Have to run the burn in again as I soldered/ de-soldered inputs and output wires + new components burn-in. Gain is now more "volume control friendly" and I experience for now overall more transparent sound presentation, although tonal balance is about the same, sound seems "cleaner" (as expected)
Brds,
nAr
Today, I did the following mods :
1°) Went for 0,4 uF instead of 0,2 uF for the coupling caps between the 2 gain stages
2°) As I only have high output cartridges for now ( about 3mV to 7,2 mV ) I reduced temporarily reduced the gain of the 2nd stage by having 2*100k resistors in place of R16, for a 50k value instead of 100k original. Following the formula of the non-inverting gain stage, gain equals now to (50+1)/1 so 51 times gain instead of 101 times, so about 34 dB gain or so -
Have to run the burn in again as I soldered/ de-soldered inputs and output wires + new components burn-in. Gain is now more "volume control friendly" and I experience for now overall more transparent sound presentation, although tonal balance is about the same, sound seems "cleaner" (as expected)
Brds,
nAr
Last edited:
All credit to eclectic2k for helping me with layout & routing
You are welcome of course, but I did not do very much
-- though sometimes the little things are important, of course nAr did all the work! I just helped review for errors, of which there weren't many.
Thanks for all the good info and your thoughts on the sound. I've put off building mine while I concentrate on other things, but will get to it in time.
Parts help???
I would like to build the Pearl II, however, I could use a little help in the parts selection:
1. First of all I have priced out the boards with the stock caps and resistors. What do you suggest that I upgrade for lowest noise (which resistors and caps, and what to brand)?
2. Are these the correct parts for U1 (Mouser 512-LM7924CT) and U2 (Digi-Key LM7824CTFS-ND) ? Also what heat sinks for use with these voltage regulators using the Pearl II boards from Pass?
3. I did not see a spec for C22, will this work – (Digi-Key 338-1042-ND)?
4. Power Supply – Please point me to a good transformer.
Thanks, that should get me started,
Alan
I would like to build the Pearl II, however, I could use a little help in the parts selection:
1. First of all I have priced out the boards with the stock caps and resistors. What do you suggest that I upgrade for lowest noise (which resistors and caps, and what to brand)?
2. Are these the correct parts for U1 (Mouser 512-LM7924CT) and U2 (Digi-Key LM7824CTFS-ND) ? Also what heat sinks for use with these voltage regulators using the Pearl II boards from Pass?
3. I did not see a spec for C22, will this work – (Digi-Key 338-1042-ND)?
4. Power Supply – Please point me to a good transformer.
Thanks, that should get me started,
Alan
C22 I used mouser part 505-fkp2100/100/2.5 wima brand 2.5% tolerance yours should work as well
7824 from mouser part 513-njm7824fa because not stocked at digikey last time I looked
7924 from digikey part njm7924fa-nd because not stocked at mouser last time I looked
I bought heat sinks from digikey part ea-t220-51e-nd
I too am waiting on power recommendations
7824 from mouser part 513-njm7824fa because not stocked at digikey last time I looked
7924 from digikey part njm7924fa-nd because not stocked at mouser last time I looked
I bought heat sinks from digikey part ea-t220-51e-nd
I too am waiting on power recommendations
Last edited:
For the caps, most critical for sound and precise values are the ones seen in the RIAA network. Ideally buy some MKP 1% 5,08 space leads. For the electrolytics, I chose a mix of Panasonic High temp low ESR and Mundorf Mlytic but the Panasonic high temp. can be used conveniently
Not so many energy is lost in heat so good brand 78XX / 79XX and regular sinks will do the job
3. suitable but to my ears sound is little less intersting than with the MKP 1%
4. Any good square trafo or RCore or even toroid can do the job. Put at least 1A models (decent regarding circuit draw) or go for 100 or 300VA if you're paranoid ( my Rcore 300VA sounds fine
Regards,
nAr
4. Any good square trafo or RCore or even toroid can do the job. Put at least 1A models (decent regarding circuit draw) or go for 100 or 300VA if you're paranoid ( my Rcore 300VA sounds fine
Regards,
nAr
Hi, nAr
Could you roughly tell us how much current consumed in each channel of Pearl II in your board? (100mA?)
Thanks,
All is said post #531
If you use BL jfets at the 1st stage maybe max will be 50mA per board
Regards,
nAr