WOTS Pre-amplifiers

I designed a multi-configurable PCB board for the WOTS pre. It can do all the versions of marigno's wots pre-amplifiers, from the original Wots Tree to the Talia by omitting/adding some components. The board can also use B244/245/Lsk489/other popular NJFET, and all most any popular TO220/TO92 MOSFET/JNFET as follower output.
For testing, I also included Shade feedback network.



  • Picture1.png
    18.4 KB · Views: 100
Last edited:
  • Like
Reactions: 1 user
I know the IXTP01N100D is an expensive transistor, but the cheaper DN2540 didn't work as a CCS in the power supply; perhaps the current we need (1.5mA) is too low. When I changed it to the IXTP01N100D everything worked fine.
I used the IXTP01N100D in the follower and its CCS after reading some @analog_sa 's posts and never used a DN2540 again.
If you want, share your experience with LTSpice using/not using the Sziklai BJT, it is fascinating! The Sziklai solution gives a neutral, perfect sound, but excluding it you get emotion. Health to those who use sunglasses to watch the Monnalisa!
The best dual JFET for a line stage is the LSK489B.
Thank you! Marigno.

This is the first draft design. I really like to get some critical views for the design. The more, the better. Then it will be refined and improved, including the selection of components, layout of PCB.

I also like your double 6922 concept. It is approved that double JFETs can increase gm, further improve THD and reduce Pd strain put on single component.

If checking specs of BF256B and LSK489B carefully, they are very very similar if not almost same, but higher gm. I also may use BF256B as lower JFET(T12) and Lsk489B as higher JFet(T11), since it has a 60V Vgs(break).

The simulation of your WOTS will be done in the near future. I will post them here.
Last edited:
I did some simulation of WOTS pre original circuit. I could not find the model for BF244C and ZVNL120A. So I use BF245C and DN2540 in the circuit. BF245 is identical part as BF244 but its noise figure is not specified. Then I put the original numbers of marigno’s circuit, everything in the sim is very close to those with the original.
Since Idss and Vp of the JFET used in the circuit can be different, Rs and Rd have to be modified. Using B-grade JFET, Rs should be reduced and Rd should be increased.
1. Change Rs to get desired J2 Id
2. Change Rd to get output DC close ½ power supply voltage
3. Change Rs2 to get Is for the follower M2


  • Picture01.png
    14.6 KB · Views: 47
Initial 0.1Vp 1KHz signal to get the gain. Change the input to get THD at 1Vrms and 5Vrms output. See the pictures. Generally it is H2 dominated. Combined with marigno' selection of the JFET, it is may be the main reason for "Warm Sound".

BF244 is used in amplifier for UHF and VFH, and low noise. Its Ciss and Crss are extremely small around 0.x-2PF. these values are comparable with a tube triode. For instance, 12AX7 has a grid to plated capacitance of 1.6PF and grid to cathode/heater 1.7pf. I remembered a Japanese member in this forum mentioned 2SK168 has the most "airy" sound used in preamp. 2SK168 is used to amplify UHF and VFH as well.

BF244C is tough to find now. But BF256 and BF244 belong to a same category. Its ft is even higher than BF244. I could not find its noise figure. But from the noise-frequency curve at few hundred MHz, its noise should comparable to BF244 if not better. BF256B is still available at Digikey and other major distributors. I bought few ten of them few month ago and tested them briefly. Their Idss are in the range 6mA-10mA. By selecting BF256B with large Idss, it could be used as BF244C in the circuit(I think).


  • Picture04.png
    360.3 KB · Views: 24
  • Picture03.png
    406 KB · Views: 21
  • Picture02.png
    95.4 KB · Views: 21
Last edited:
Another concern to use BF244/245 Jfets in the circuit is their min breaking Vgss figure, which is 30VDC. If WOTS pre uses 60VDC power supply, the max Vgs on the top Jfet can reach 50VDC in the extreme case. From Pass's experiences, JFET can work at much higher Vgs than its breaking Vgs. I tested some J113, too. It is breaking Vgs is 35V but there is no problem to go 70V. So I picked up a BF256B randomly and tested its Id-Vgs curve. It is should be safe when Vgs not over 45V.


  • Picture05.png
    8.8 KB · Views: 12
Good job!!!
I left behind all those old obsolete JFETs, even if they are good pieces. Difficult to find, easy to get fakes.
My original "Tre" (BF244B) is in my upstairs set, mainly used by my wife to listen to the TV. I'm not going to touch it anymore, it works and sounds very well.
I'm working with the LSK489B, following your simulations. Today I received Talia's PCBs (motherboard containing the PSU, the delay, and the output capacitor) and SGS' (Shared Gain Stage) which is a tiny PCB containing the gain stage that can be configured with the Sziklai BJT or without it by moving a jumper. This PCB is intended to be installed on Talia's motherboard or Vera's, which is my phono RIAA preamp.
If everything goes in the right way, in a few days I should be able to close the Talia with its new arrangement and proceed to the Vera.
I really like to hear your objective/listen impression of your new Talia version.
I am still swinging between E2 and Talia and have not settled down yet.
From your personal experiences/opinions/favorites, does a amp sound better with minimal 3rd/5th harmonics or with a monotonically decreasing harmonic(how much dB lower should be the high order harmonic?)
We all know (except some self-styled audiophile admirers of the Mona Lisa) that the trend of the harmonic content must be monotonically decreasing, but I don't know how much each harmonic must be at a lower level than the previous one. It would be better to avoid odd harmonics because they are dissonant with the fundamental frequency. I know it can be done, but I don't know how. This is a secret known only to admirers of the Mona Lisa I mentioned earlier. They also know how to avoid the annoying even harmonics. Distortion is not allowed.

I am curious: have these admirers of the Mona Lisa been to see her?

I've finished several "boring post-amps", they all have distortion, so I can't post them.

I enjoyed using old ladies from 90 years ago "VT" with "X5" socket. For us lovers of monotonically decreasing distortion, they sound magnificent. I cannot check their harmonic distortion, but their euphony suggests the distortion pattern is correct.

I also closed a mu-follower a couple of weeks ago. I enjoyed myself that much on this project, so I'm studying a hybrid mu-follower.

The Talia should be ready in a few days. We will see!
I thank you very much for having opened my mind and my eyes.
I just closed the Talia, now I have jumpers that allow me to switch from pure Cascode to Cascode & Sziklay. I used the LSK489B selecting a couple in a bunch of 10 for a matched highest Id.

Here are the results of both configurations obtained by LTSpice.
Gain around 10, or 20dB.

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+3 1.453e+0 1.000e+0 -89.95° 0.00°
2 2.000e+3 8.484e-5 5.840e-5 1.30° 91.25°
3 3.000e+3 5.525e-6 3.803e-6 -90.67° -0.72°
4 4.000e+3 9.114e-8 6.274e-8 127.83° 217.78°
5 5.000e+3 6.450e-8 4.440e-8 87.88° 177.83°
6 6.000e+3 4.328e-8 2.979e-8 87.67° 177.62°
7 7.000e+3 3.268e-8 2.250e-8 92.18° 182.13°
8 8.000e+3 2.351e-8 1.618e-8 89.32° 179.27°
9 9.000e+3 1.699e-8 1.169e-8 89.99° 179.94°
Partial Harmonic Distortion: 0.005853%
Total Harmonic Distortion: 0.004329%

Rsziklai=0Ω (short)
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+3 1.399e+0 1.000e+0 -89.95° 0.00°
2 2.000e+3 2.211e-3 1.581e-3 0.10° 90.05°
3 3.000e+3 1.951e-5 1.395e-5 -90.00° -0.05°
4 4.000e+3 3.430e-7 2.452e-7 -126.91° -36.96°
5 5.000e+3 2.211e-7 1.581e-7 -89.98° -0.03°
6 6.000e+3 1.887e-7 1.349e-7 -88.40° 1.55°
7 7.000e+3 1.659e-7 1.186e-7 -88.95° 1.00°
8 8.000e+3 1.480e-7 1.059e-7 -91.35° -1.40°
9 9.000e+3 1.361e-7 9.731e-8 -90.06° -0.11°
Partial Harmonic Distortion: 0.158115%
Total Harmonic Distortion: 0.158066%

The distortion is very low in both cases. Listening to music is however possible to appreciate some differences between the two configurations. Guess which one I like better? But don't tell anyone.


  • 20240515_162156.jpg
    462.1 KB · Views: 19
  • Like
Reactions: 1 user

It's time to fulfill my promise to post Talia's docs once it is finished.
Here we are.
Talia has an incredible sound and a great soundstage. The quality of the sound could be subjective, google "scarrafone's rule". The soundstage, on the contrary, is objective.
Considering the simplicity of its circuitry and the absence of a feedback loop, I think I have achieved a good result.
The Shared Gain module (used in other projects too) can be configured as a pure Cascode or a Cascode + Sziklai.
LTSpice's results are very encouraging, here they are:
gain 20db, a bit more in Sziklai;
pure Cascode THD: 0.16% with a prevailing H2;
Cascode + Sziklay THD: 0.004% showing a monotonic fall.
All working points in LTSpice are confirmed in the real thing.
The mainboard and the SGS are two PCBs, designed to be symmetrical: one design fits two channels by reversing the PCB upside down as you see in the pic.
PCBs Gerbers are available for free if you don't use them for commercial purposes, send me a PM.

Thanks to:
@wwwtttwww for convincing me to use LTSpice which I have always stayed away from;
my friend Flavio who has always supported me in many projects, teaching me how to listen correctly;
JLCPCB for the high-quality work in making the PCBs.
HiFi2000 for their high-quality cabinets and personalization.


  • Talia Front.jpg
    Talia Front.jpg
    262 KB · Views: 26
  • Talia PCBs.jpg
    Talia PCBs.jpg
    477 KB · Views: 29
  • Talia Rear.jpg
    Talia Rear.jpg
    270.4 KB · Views: 40
  • Talia Schematic.jpg
    Talia Schematic.jpg
    406.2 KB · Views: 41
  • Talia SGS module.jpg
    Talia SGS module.jpg
    373.6 KB · Views: 44
  • Talia SGS Schematic.jpg
    Talia SGS Schematic.jpg
    194.6 KB · Views: 49
  • Talia Top 1.jpg
    Talia Top 1.jpg
    462.1 KB · Views: 42
  • Talia Top 2.jpg
    Talia Top 2.jpg
    412.5 KB · Views: 29
  • Like
  • Thank You
Reactions: 3 users