Design an amplifier with wide voltage range ;)

Jump to Latest
Javier Boada

Javier Boada

Member
Joined 2023
#1
Hello diyaudio friends, I hope you are very well, I would like to ask a question, I am looking for a Transistorized amplifier that is capable of working in a wide range of voltages, let's say from 25 +/- VDC to 80 +/- VDC, I know that there are parameters that are very difficult to adjust automatically for that wide voltage range, such as bias, etc. but they can be adjusted manually, I'm a bit of a novice in audio amplifier design, but I've been working on an amplifier that can meet the needs I'm looking for, I just don't have anything concrete yet, and I would like some help, guidance or advice from the most experienced, if you could help me with some information I would greatly appreciate it 😀
 
#3
An amplifier can easily be designed to run over that range, with the only adjustment being the output stage bias. You start with the design being suitable for the full voltage, and use current source biasing for the input pair and VAS to make their operating points mostly supply voltage independent. You’d probably want a triple EF output stage with multiple pairs, so that a modest current can be used in the VAS (say, 6 mA). The extra output pairs could be deleted at lower voltage, but retain the triple EF. Maybe Unnecessary, but it won’t hurt anything either.
 
#5
A amp with current sources for both LTP and VAS should do from 15V to 90V rails.
That is with 2Q CCS's. Led or 2X diode CCS's might require "tweaks". I usually design with a trimmer
to facilitate a wide range of supplies.
The output bias will typically be determined by the VAS current , not the rail voltage.
The downside would be you would have to de-rate for the highest expected voltage.

Also , a amp with cascoded stages will allow you to use premium low voltage semi's with high voltages.
Otherwise , you would have to use less ideal high voltage components throughout.
OS
 
#6
I might cascode the input stage and darlington the VAS. You might not miss 3 or 4 volts that a cascode VAS costs you at 80 volt rails, but on only 25 volt supply you may want those volts back.

Even if output stage bias can be made independent of supply, it still needs to be adjustable - unless you’re on an IC process.
 
#7
wg_ski said:
I might cascode the input stage and darlington the VAS. You might not miss 3 or 4 volts that a cascode VAS costs you at 80 volt rails, but on only 25 volt supply you may want those volts back.
Click to expand...
But for a 25V supply , you would use (only) those "premium" low voltage components (BCxxx).
VAS is different , TO-126 devices are both high voltage and quite precision. I cascode VAS's for clipping
behavior (and speed).
I would design totally different for 25V. 45-70V , I get "fancy"😉 ...
OS
 
#8
25-35V amps bring back memories of my first HI-FI , a Kenwood KA-305. TO-220 60V outputs , all premium Japanese 2SC/A
simple LIN amp topology. It ran my Radio Shack speakers well with my SL-D2 turntable. Amp only had 7 devices per amp.
Bootstrapped LIN .... .08% 20-20K. Yikes , only 25W/CH !!
OS
 
#9
rayma said:
What is the purpose of this amplifier?
Why not have two or three amplifiers, each optimized for its range?
Click to expand...
the purpose of the amp would be to accommodate a wide range of voltages without having to change components or PCB layout, and yes having multiple amps running at different voltages is a good solution, but I would like to produce a single PCB. model that meets my requirements. needs, since where I live it is very complicated and expensive to get transformers or voltage sources similar to each other, so I have to make do with what I have.
 
#10
:cop:
English only please.

Javier Boada said:
el propósito del amplificador sería ajustarse a una amplia gama de voltajes sin tener que cambiar componentes o el diseño de la PCB, y sí, tener múltiples amplificadores que funcionen a diferentes voltajes es una buena solución, pero me gustaría producir una sola PCB. modelo que cumpla con mis requisitos. necesidades, ya que donde vivo es muy complicado y costoso conseguir transformadores o fuentes de voltaje similares entre sí, entonces tengo que conformarme con lo que tengo
Click to expand...

the purpose of the amplifier would be to adjust to a wide range of voltages without having to change components or PCB design, and yes, having multiple amplifiers running at different voltages is a good solution, but I would like to produce a single PCB. Model that meets my requirements. Needs, since where I live it is very complicated and expensive to get transformers or voltage sources similar to each other, so I have to settle for what I have.
Click to expand...

Further non English posts will not be approved.
 
#12
ostripper said:
A amp with current sources for both LTP and VAS should do from 15V to 90V rails.
That is with 2Q CCS's. Led or 2X diode CCS's might require "tweaks". I usually design with a trimmer
to facilitate a wide range of supplies.
The output bias will typically be determined by the VAS current , not the rail voltage.
The downside would be you would have to de-rate for the highest expected voltage.

Also , a amp with cascoded stages will allow you to use premium low voltage semi's with high voltages.
Otherwise , you would have to use less ideal high voltage components throughout.
OS
Click to expand...
I understand your point very well, thanks for the comment
 
#15
There are two main factors to consider:
  1. SOA of all transistors. Either use high voltage / power transistors and / or use cascoding.
  2. Bias drift with supply voltage. Amplifiers are typically biased using CCS and most types of CCS vary output current with supply voltage, thus are not very constant. In order to have stable bias independent of supply voltage, you need CCS that output constant current no matter what the supply voltage is. Best performers in this field are "ring of two" CCS.
 
You must log in or register to reply here.
Top Bottom