The rated values are C2 15nF, C2Y 100pF, but I don’t have a meter to measure them. I will include some more options in my next Mouser order to tune by ear. You also mentioned that changing the loading resistance may be good. The recording was done at 220 Ohms, I will play with this as well after doing a better job on the cartridge setup.
Most MKP1837 (1%) 15n nominal for C2 that I'd measured with a reliable LCR on Kelvin clips are nearly spot on (10-30pF lower usually but with enough samples right on the nose also) then there are the odd few which are veering off occasionally to higher like 15.1nF. Same for NOS 1830 ERO film & foil (1%) if not ever higher than 15n at all.
You could have a mild upwards tilt of 0.2dB across the whole curve in the FSP when using 15.1nF actual total, measured including C2Y, while 15.3nF-15.4nF bringing ruler flat but 15.2nF seems to be a good spot of subjective neutrality given the 2nd harmonic THD nature of this non loop feedback staged design.
Anyways, those are quantitatively small tilts but influential enough, bcs spectrum wide, for you can carefully use them to advantage in tuning against a darker or brighter than neutral front end. Thus the multi-pitch C2Y dedicated area on the pcb is a useful tool.
Test value caps there can be put in and out easily without disassembly because the pad holes are metalized so touch soldering C2Ys on top will electrically continue to the bottom layer as well.
You could have a mild upwards tilt of 0.2dB across the whole curve in the FSP when using 15.1nF actual total, measured including C2Y, while 15.3nF-15.4nF bringing ruler flat but 15.2nF seems to be a good spot of subjective neutrality given the 2nd harmonic THD nature of this non loop feedback staged design.
Anyways, those are quantitatively small tilts but influential enough, bcs spectrum wide, for you can carefully use them to advantage in tuning against a darker or brighter than neutral front end. Thus the multi-pitch C2Y dedicated area on the pcb is a useful tool.
Test value caps there can be put in and out easily without disassembly because the pad holes are metalized so touch soldering C2Ys on top will electrically continue to the bottom layer as well.
Hello Salas,
I added the second pair of 369's on Q2 position along with the R3 but the result was rather disappointing ; no sound out of the build ; the TP1-TP2 voltage was ~7.6V and tweaking VR2x did absolutely no effect. I checked the jfets with a general purpose tester; they are fine; i checked the position of the jfet it is fine.
After this failed attempt i just extracted Q2 and the sound was back. Rebiased TP1-2 to 3.6V warm up and now i wonder what I did wrong during adding the second input stage jfet. I am aware that the gain should not increase ; and my only guess is that there is not enough reserve of current from the shunt to support the consumption added by the second jfet thus keeping both closed.
Forgot to mention I have the 56DB build with R13 2.2K and R2 5.1R
What can i do to adjust and accommodate Q2 ?
I added the second pair of 369's on Q2 position along with the R3 but the result was rather disappointing ; no sound out of the build ; the TP1-TP2 voltage was ~7.6V and tweaking VR2x did absolutely no effect. I checked the jfets with a general purpose tester; they are fine; i checked the position of the jfet it is fine.
After this failed attempt i just extracted Q2 and the sound was back. Rebiased TP1-2 to 3.6V warm up and now i wonder what I did wrong during adding the second input stage jfet. I am aware that the gain should not increase ; and my only guess is that there is not enough reserve of current from the shunt to support the consumption added by the second jfet thus keeping both closed.
Forgot to mention I have the 56DB build with R13 2.2K and R2 5.1R
What can i do to adjust and accommodate Q2 ?
I wait the current compilation to finish and i will try.
The reason is simpler , i wish lower the noise floor(white noise) now that i found a combination with absolutely no hum.
And the main provider of white noise seems to be the input stage. according to the Apx i used a few months back.
Its not a meaningful plan in this case because what gained in input stage noise spec by paralleling yet another 2SK369 is meant to counter some higher output noise for the extra voltage gain that will occur. As the transconductance also doubles, the gain goes 6dB up. Higher voltage gain not only amplifies the cartridge more but the background noise as well. In other words what we aim for in this design by doubling the input JFETs is comparable hiss results for both MC and LMC sensitivities.
What is the coil resistance of your cartridge and what is its nominal output? What is the IDSS of your 2SK369s? Maybe we can work out a meaningful way for values that will keep the gain as needed while using both 2SK369 with hopefully the least possible noise contribution from the need go higher value source resistors.
What is the coil resistance of your cartridge and what is its nominal output? What is the IDSS of your 2SK369s? Maybe we can work out a meaningful way for values that will keep the gain as needed while using both 2SK369 with hopefully the least possible noise contribution from the need go higher value source resistors.
coil resistance: 500R
coil inductance 450mH
output 2.7mV
369's Idss 12.4mA
I chose the load that gives the lowest impact on the midrange and highs:47K
And listening to it, makes me very happy.
I think I am still comfortable with the input capacitance increase of the second 369;
I attempted to keep the gain in the same zone switching to R13 3.3K 2W and R2&3 46R going for roughly 52db but black background.
Typically for 300mV RMS output at 5cm/sec velocity (-10dBV standard hi-fi level) you would only need 41dB gain with your cartridge. Due to its high coil resistance the R2//R3 parallel total value will not start contributing noise before 170R which will never need to be reached in your case anyway. That's fortunate because it kicks one sneaky noise source out of the equation.
If your system voltage gain seems needing the 52dB then the compromise is mainly in raised THD levels and lowering of headroom to clip for ticks and pops. Maybe you could sacrifice even more phono stage gain and still have good max volume in your system? That will lower the hiss accordingly even more. As you raise the R2&R3 values towards that goal the input stage capacitance also lowers because of the higher local feedback by degeneration they create. R13 may need further adjustment for proper TP1-TP2 reading (i.e. input stage DC centering) as you go down that route.
P.S. The 500R coil resistance cart has 2.8693 nV/√Hz source noise at 25C. Or 0.4056 uV for 20kHz bandwidth. That's 76.465dB dynamic range for 2.7mV output.
If your system voltage gain seems needing the 52dB then the compromise is mainly in raised THD levels and lowering of headroom to clip for ticks and pops. Maybe you could sacrifice even more phono stage gain and still have good max volume in your system? That will lower the hiss accordingly even more. As you raise the R2&R3 values towards that goal the input stage capacitance also lowers because of the higher local feedback by degeneration they create. R13 may need further adjustment for proper TP1-TP2 reading (i.e. input stage DC centering) as you go down that route.
P.S. The 500R coil resistance cart has 2.8693 nV/√Hz source noise at 25C. Or 0.4056 uV for 20kHz bandwidth. That's 76.465dB dynamic range for 2.7mV output.
Thank you for the great insights and calculations; I reached a desired behavior using a 3.3K 2W R13 and 46R for R2-3.the noise floor is black @ 1cm near the very unforgiving high-mid range speaker. It becomes audible when volume is way too high for anyone to enjoy any vinyl recording.
Thank you so much for guiding trough!
I still have a question: Is Q3 providing any voltage gain? Being open with tp1-2 @ 3.6V it provides a more linear pathway than a resistor?
Thank you so much for guiding trough!
I still have a question: Is Q3 providing any voltage gain? Being open with tp1-2 @ 3.6V it provides a more linear pathway than a resistor?
Is R2,3 playing a feedback role (What kind of feedback)? Or just limits the swing of the transistor?
See about emitter degeneration in BJTs, source degeneration is the analogous thing with JFETs. Local feedback is the term I used because its debatable:
"While this is often described as "negative feedback", as it reduces gain, raises input impedance, and reduces distortion, it predates the invention of negative feedback and does not reduce output impedance or increase bandwidth, as true negative feedback would do"
Common emitter - Wikipedia
"While this is often described as "negative feedback", as it reduces gain, raises input impedance, and reduces distortion, it predates the invention of negative feedback and does not reduce output impedance or increase bandwidth, as true negative feedback would do"
Common emitter - Wikipedia
Hi Salas
Would you care to explain how you got to 300mv output knowing only the cart specs ?
Thank you Salas
Now I understand.... 300mV is consumer level spec.
V2 300mV
V1 2.5mV
L = 20 × log (voltage ratio V2 / V1) in dB = 41.6dB
Regulator voltage
I’ve built up one of Teabag’s boards aiming for ~43 db.
I am only able to get 30vdc after the shuntreg on my Folded phono board with VR2X all the way open.
Thus the lowest reading I can get between TP2 and TP2 is 4.99V.
TP1 to ground is 7.93 for both channels.
I originally used the LEDs from the minikit, but these measured low, so I replaced them with a string of all green LEDs measuring 1.98Vf each for D2X-D4X. The result is that I am still getting only 30Vdc after the reg. And 4.99V between TP1 and TP2 with both VR1 and VR2X wide open.
Any thoughts? Is this just fine?
It certainly sounds pretty good. Will it sound better if I manage to raise the rail voltage?
Thanks for your help!
I’ve built up one of Teabag’s boards aiming for ~43 db.
I am only able to get 30vdc after the shuntreg on my Folded phono board with VR2X all the way open.
Thus the lowest reading I can get between TP2 and TP2 is 4.99V.
TP1 to ground is 7.93 for both channels.
I originally used the LEDs from the minikit, but these measured low, so I replaced them with a string of all green LEDs measuring 1.98Vf each for D2X-D4X. The result is that I am still getting only 30Vdc after the reg. And 4.99V between TP1 and TP2 with both VR1 and VR2X wide open.
Any thoughts? Is this just fine?
It certainly sounds pretty good. Will it sound better if I manage to raise the rail voltage?
Thanks for your help!
This sounds as if VR2X is capped; tweaking R3x by a Kohm or so would give the coarse headroom needed to bring TP1-2 in ballpark.
Last edited: