diyAudio (
-   Solid State (
-   -   NoGNFB integrated amp with Ge transistors at output (

VladimirK 23rd September 2012 12:07 PM

NoGNFB integrated amp with Ge transistors at output
4 Attachment(s)
Listening ofthis NoGNFB integrated amp with germanium transistor output has confirmed an idea that Ge transistors are obviously superior to other types of transistors at the output of NoGNFB designs. Output impedance with one transistor (without paralleling) came about 0.08 ohms. Bandwidth is greater than 1 MHz. Even in comparison with the well-sounding bass from good GNFB amps, I have distinguished some new tiny bass details. Interestingly, the h21 vs Ic curve for 1T910A Ge transistor shows only growing behavior even in the range of collector currents 10 ... 20A.
The quiescent current of 1.25 A was chosen, since I was afraid of burning out the Ge.
But, the main heat sinks are only a bit warm, I guess the idling current can be increased up to 2A.
Voltage gain of the input differential stage, for the Idss of KP903 and resistor values indicated, is near 6.5. It is quite enough for me, usually I listen with the voltage gain range 2 ... 5.
Power supply is built with two R-Core transformers, every has 24V-3A and 10V-1A windings. Output voltages are further smoothed by "cap multipliers". Power supply socket is placed at the side panel, in order to avoid the neighborhood of power and input wiring.
The whole schematics is assembled point-to-point. The input pot is DACT 20 k.
Overall, the sound is excellent. As if the high-quality tube SE amp has been magically added by bass, and this is the bass that creates a new reference for me.

Mooly 23rd September 2012 12:13 PM

Nice :) I can imagine it does sound excellent.

VladimirK 23rd September 2012 05:12 PM

1 Attachment(s)
I was surprised with how does current transfer ratio of 1T910A Ge transistors look like (only increasing behavior up to 20A)

Mooly 23rd September 2012 05:30 PM

Interesting although I'm not familiar with the device/s you mention.

My first ever amp used Ge outputs (AD149).

VladimirK 23rd September 2012 05:58 PM


Originally Posted by Mooly (
Interesting although I'm not familiar with the device/s you mention.

My first ever amp used Ge outputs (AD149).

I would not use Ge transistors, if they are slow, and if schematics would suffer from their thermal instability. These drawbacks are not relevant to the present schematics.
Also, I guess, the base region resistance of 1T910A is unusually small, since I have got output impedance modulus around 80 mOhms. Passband -1dB is more than 1MHz.
Not bad for simplest No GNFB schematics.
If somebody mentioned difference in bass between lateral MOSFETs and BJTs at output, that one could guess further improvement with Ge compared to Si BJTs.

Mooly 23rd September 2012 06:13 PM


Originally Posted by VladimirK (
Not bad for simplest No GNFB schematics.

Indeed :) It's not bad at all, and your quoted output impedance seems very good.

It is a really nice design... well done :)

jean-paul 23rd September 2012 06:31 PM

Güzel !

destroyer X 23rd September 2012 08:08 PM

One of the best i had used AD161/AD162
2 Attachment(s)
A very long time ago... was a Brazilian kit made by Philips/Miniwatt factory.



jean-paul 23rd September 2012 08:50 PM

How did it sound ?

VladimirK 24th September 2012 04:33 AM


Originally Posted by destroyer X (
A very long time ago... was a Brazilian kit made by Philips/Miniwatt factory.



Hello, Carlos
Please, pay attention to the differences between that kit and this schematics.
Some of them, are:
1) this thread schematics does not experience power supply current modulation under signal conditions, PS does not stand in the signal path
2) it has No GNFB (Ge output transistors are not good for GNFB schematics, they are relatively slow)
3) it has No input cap
4) output cap is made with all possible care
5) passbands of the two designs I guess differ drastically

As for sound estimates, I have very good tube 300B SE amp, and it gives very good reference for mids and highs quality, I can easily estimate sound differences between traditional GNFB schematics and special schematics, designed for extreme high quality.

Comparing the 1T910A power Ge transistor (p-n-p) with AD162 (p-n-p) we can see:
AD162 Transistor Datasheet. Parameters and Characteristics.
Material of transistor: Ge
Polarity: pnp
Maximum collector power dissipation (Pc): 6W (35W for 1T910A)
Maximum collector-base voltage (Ucb): 32V (33V for 1T910A)
Maximum collector-emitter voltage (Uce): 20V (32V for 1T910A)
Maximum emitter-base voltage (Ueb): 10V
Maximum collector current (Ic max): 1A (10A average current for 1T910A, 20A pulse current)
Maximum junction temperature (Tj): 80°C (358K = 85C for 1T910A)
Transition frequency (ft): 1MHz (30MHz for 1T910A)
Collector capacitance (Cc), Pf: 150
Forward current transfer ratio (hFE), min/max: 50MIN (100...250 for 1T910A)
Manufacturer of AD162 transistor: PHILIPS
Package of AD162 transistor: TO37
Application: Medium Power, General Purpose

I guess, intrinsic resistance of the base region of 1T910A is some orders of magnitude less than that of AD162.
1T910A is really a peak of Ge technology. The output impedance 0,08 Ohms with 1T910A is not emulated by GNFB, it comes from intrinsic properties of the transistor, connected as emitter follower.

All times are GMT. The time now is 01:09 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2