Quietest FET selection for audio input stage

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Hello gang:

Designing a common source (Vdd 12VDC) amplifier to drive a tube input stage. I need to select a JFET (or mosfet?) that has very low noise. Data sheets don't really show NF's. I am looking at a J (174 or whatever) series or maybe a 2N3819, but I am open to other ideas aside from op amps or BJT's. Any suggestions?
 

sparky 1

Joined Nov 3, 2018
185
The lowest JFET THD I have simulated is 2SK3557-7 under 0.001%
replacing it with 2SK170 fell to 0.08%
however getting to useful amplification at the next stage I can see why the trade off for improved gain
overall performance in completed circuit that some of the experienced audio preamp builders reporting
warm effects of jfet front end. The question then is which second stage can maintain the incredible sound from these jfets.
AG had said that the audiophools believe it is lowest noise highest cost that fall into that trap.
In my setup I can pull the NE5532A and replace it with other op amps the ones that say precision and audio and low noise work best.
The parts used in quality class A with grade AA parts seems like the way to go. With out a lot of audio test equipment it is not easy to be objective.
 
Last edited:

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Thanks folks... I just got reminded to use a LND150 and use the HV supply from the tubes as well. Not sure on the NF yet.
 

sparky 1

Joined Nov 3, 2018
185
JFET or BJT what some people say and how would you decide ?
" Warm sound " I should have used better vocabulary "The tube sounds warmth and richness"
Wikipedia also has less than perfect description of this. it states: higher levels of second-order harmonic distortion in single-ended designs, resulting from the tube interacting with the inductance of the output transformer.
What I should have said is that some feel that the jfet does better job at emulating the tube sound.

I am not a sound expert. I have read that some having talented ears and good hand in musical instruments they can tell the difference. A consideration for which transistor to choose is a personal choice. Because the sound may not be a single instrument and sound has been processed in a certain way so and so. How is making nice sine waves different than doing good audio ?

I feel the audio builder can make a choice. Listening and comparing the preamps.
Listen to what could take another discussion I am sure. The piano, violin, guitar there is no end.
There are different test music so someone else might choose something else.
If I played this I would know which transistor because one preamp would sound better than the other.
So it is a personal choice. For me it is a start. If it sounds like being right there ...
 

Audioguru again

Joined Oct 21, 2019
1,503
Many old tube amplifiers produced odd harmonics that are not musical and even harmonics that are multiples octaves of real frequencies. Some people like "tube sound" because its even harmonics distortion is all that they have heard. When they hear low distortion hifi it sounds odd to them since it has no added harmonic distortion. Rock bands add "fuzz" and "overdrive" I think because they are deaf to the high frequencies of distortion and cannot hear how bad it is.
The amounts and balance of even and odd harmonics cause old tubes amplifiers to sound different.

Since most modern solid state amplifiers have extremely low distortion that cannot be heard then they sound almost the same.

I forgot to say that tubes wear out and their distortion increases with time. Solid state parts do not wear out.
 
Last edited:

Papabravo

Joined Feb 24, 2006
13,741
If you are not one of the people with "golden ears" then it is completely pointless for you to search for parts that will please them. This is all subjective and you will spend a lifetime searching for parts that will not please them. Please do not go down this rabbit hole. It is the road to perdition.
 

Audioguru again

Joined Oct 21, 2019
1,503
Some expensive amplifiers are solid state but have vacuum tubes glowing on top but not in the amplifier circuit.
Other expensive amplifiers are also solid state but have a fuzz circuit to produce tubes distortion and boomy bass caused by the poor damping ratio of a tubes amplifier..
 

Papabravo

Joined Feb 24, 2006
13,741
Some expensive amplifiers are solid state but have vacuum tubes glowing on top but not in the amplifier circuit.
Other expensive amplifiers are also solid state but have a fuzz circuit to produce tubes distortion and boomy bass caused by the poor damping ratio of a tubes amplifier..
That's a laugh riot!
 

Papabravo

Joined Feb 24, 2006
13,741
We got lucky today. Thanks. It is the golden ears. Here is a song for you. 1972 Tom Petty band. It is the wahwah effect that is used to depict the gitty atmosphere. I recall that Tom Petty was taught by Don Feldar. When the Canadian guy begins playing the guitar the effect can be heard.
Sorry, can't hear a damn thing anymore.
 

sparky 1

Joined Nov 3, 2018
185
This is the Philips spice model
.MODEL 2N4391 NJF(VTO=-8.0337 BETA=1.56450m BETATCE=-0.5 LAMBDA=1.00000E-2 RD=8.25115E-1 RS=8.25115E-1 CGS=8.25000p CGD=1.32500E-11 PB=7.72344E-1 IS=6.54620E-16 XTI=3 AF=1 FC=0.5 N=1 NR=2 MFG=PHILIPS)
**********************************************************************************************************************
comparing to op amps, this is a quantum leap ... wow not an lm386 by any means. it is not a whole system
OP AMP selection input transistor type, different types. There is a chart that shows how the various op amps compare
by bandwidth over noise.

INA1650 spectrum chart shows -140dB if you're into audio precision test equipment, probably overkill for most projects.
The headphone amplifiers are really nice the race to high fidelity DAC has some great candidates.
The OPA1612 specs, hard to beat, It is an extremely high performance audio op amp.
OPA1612 ultra quiet has a BJT input, now the board is sort of mid to high level to build and higher level to problem solve.
looking at the evaluation board copper and setup all the parts are probably very close tolerance to match what is inside the chip. Well matched.
I believe most of this requires fair amount of experience. An assembled board might be something to compare with. like the graphs show.
here is your JFET and here is the 1612 but how does it sound what do you not like about it ? can it be objective ?
https://www.ti.com/lit/sg/slot140/slot140.pdf?ts=1590722311050
 
Last edited:

Thread Starter

Dynaman

Joined Jan 17, 2008
94
I thought your use of a Fet is to duplicate the vacuum tube sound of a high input impedance guitar pickup preamp.
Here is one but I do not know how much even-harmonics distortion it produces.
Yes thank you... I think I am going to simplify my circuit and build this exact thing, but use the LND150... which is a high voltage MOSFET, and then just share the B+ vs. using 12VDC.. Not sure on the noise figure yet, but I have some on the way.
 

OBW0549

Joined Mar 2, 2015
3,408
I think I am going to simplify my circuit and build this exact thing, but use the LND150... which is a high voltage MOSFET...
I'd advise caution: because that device is a MOSFET, and because the data sheet doesn't even specify its noise performance, don't be surprised if it proves disappointing in that regard. A good, quiet JFET is really quiet; MOSFETs, not so much (in my experience).
 

Audioguru again

Joined Oct 21, 2019
1,503
Google has an Overdrive preamp project using the obsolete LND150 depletion-mode (like a vacuum tube) high voltage Mosfet that is non-linear and produces distortion like an old pentode vacuum tube. The preamp uses a 234V(!) power supply. An advantage it has over using a vacuum tube is that when you drop the Fet it does not break like a vacuum tube.
 

Attachments

sparky 1

Joined Nov 3, 2018
185
Philips JAN ECG 12AT7WC the effects must be something like the booster and screamer


I think the 12AX7 cake pan with high voltage fet driver must be for audio to HV exciter.
A tesla coil must not be putting out enough streamers. tesla coil modulation with effects at high impulse.
A real 5 stage tesla at 22kHz is too slow for audio so they use 4 Mhz to crackle the air

video shows three different makes of the 12AX7 and the output to large amplifiers for sound.
The third effect crunch is wiped out by the first two.
In this Crunch has reverted to over driven screamer square edge.
 
Last edited:
Top