DC biasing, headroom, clipping in MOSFET amplifiers

Thread Starter

ericknabe

Joined Mar 3, 2020
11
Okay so my question relates to biasing and threshold voltage in a MOSFET amplifier. So in an amplifier the clipping occurs when the signal hits the power rails according to all the reading I’ve done. That’s how much voltage swing you supposedly have before clipping. So if you have an 18 volt supply you should have +/- 18 volts of headroom. But clipping also occurs if the sine wave goes outside of the saturation region of the MOSFET. If the threshold voltage is say 5V, and you bias is halfway between the threshold and the supply, then you don’t have +/- 18 volts of available swing before clipping, you have much less. DC biasing the transistor obviously turns it on so the 0V AC is superimposed on a suitable voltage to get it over the threshold, but then your voltage is just closer to the supply rail and you have less headroom in the other direction, right? I mean, unless I’m totally wrong about how headroom relates to the threshold voltage (and please god tell me I am,) then why would this not be brought up in any readings on transistor biasing and amplifier design?
 

Audioguru again

Joined Oct 21, 2019
1,488
There are many ways to use Mosfets at the output of an audio amplifier.
You did not show which way you are asking about so I found the schematic of an amplifier that uses complementary Mosfet followers. It wastes a lot of unused voltage:
 

Attachments

Thread Starter

ericknabe

Joined Mar 3, 2020
11
There are many ways to use Mosfets at the output of an audio amplifier.
You did not show which way you are asking about so I found the schematic of an amplifier that uses complementary Mosfet followers. It wastes a lot of unused voltage:
I am using the MOSFET in a common source configuration. I’m trying to understand if a higher threshold voltage limits the maximum voltage swing. From what I understand it seems like it would yet none of the sources I’m reading mention it as a consideration
 

Audioguru again

Joined Oct 21, 2019
1,488
Again, I do not know what you are talking about because you forgot to attach your schematic.
Maybe you have only one Mosfet as a class-A heater or maybe you have two common source Mosfets that have no control for their biasing.
 

Thread Starter

ericknabe

Joined Mar 3, 2020
11
Again, I do not know what you are talking about because you forgot to attach your schematic.
Maybe you have only one Mosfet as a class-A heater or maybe you have two common source
Again, I do not know what you are talking about because you forgot to attach your schematic.
Maybe you have only one Mosfet as a class-A heater or maybe you have two common source Mosfets that have no control for their biasing.
theres no schematic. I am asking about the general behavior of MOSFETS
 

Audioguru again

Joined Oct 21, 2019
1,488
I do not know which "general" Mosfet schematic you are talking about because they are all different.
In Google I found ONE Mosfet in a common source circuit. Its output can swing the entire supply voltage so the headroom is determined by the input signal level, the gain and the supply voltage.
 

Attachments

Thread Starter

ericknabe

Joined Mar 3, 2020
11
I do not know which "general" Mosfet schematic you are talking about because they are all different.
In Google I found ONE Mosfet in a common source circuit. Its output can swing the entire supply voltage so the headroom is determined by the input signal level, the gain and the supply voltage.
So if I’m understanding correctly, the maximum voltage swing in that schematic will be the same regardless of the threshold voltage of the MOSFET? And that amount of swing is the supply voltage?
 

Audioguru again

Joined Oct 21, 2019
1,488
Yes, but you must bias it at the correct voltage needed by the Mosfet you buy, for maximum symmetrical output voltage swing, and each Mosfet will need a different bias voltage, even if they have the same part number.

The threshold voltage of a Mosfet is when its current is very low, almost turned off, that you do not want here.
 

Thread Starter

ericknabe

Joined Mar 3, 2020
11
I am making a solid state recreation of the first two stages of the bright inputs of this preamplifier. I have lowered the supply to 20V, replaces the triodes with transistors or MOSFETS and biased the transistors using a voltage divider with the bottom leg being 1M to prevent signal loss. Yet Mosfets deliver more overdrive than BJTs even when biased properly. Could this be due to their higher input impedance bleeding less signal to ground?
 

Attachments

Audioguru again

Joined Oct 21, 2019
1,488
Overdrive? 1959 vacuum tubes guitar amplifier (61 years old!)?
Oh. You are making an extremely distorted (clipping like crazy) guitar amplifier using Mosfets that are very different to vacuum tubes and bipolar transistors.

1) When the grid and cathode of a vacuum tube have equal voltages then the vacuum tube conducts as much current as it can.
2) When the base and cathode of a transistor have equal voltages then it conducts no current.
3) When the gate and source of a Mosfet have equal voltages then it also conducts no current.

Of course a vacuum tube and Mosfet have no input current therefore the input impedance is very high. A bipolar transistor has some bias current causing a much lower input impedance. A high output impedance of one stage is loaded down and has its gain reduced by the low impedance of a following stage.
If you plan the design of an amplifier correctly so that the output of each stage is loaded correctly to work properly then you can have as much gain as you need.
 

Thread Starter

ericknabe

Joined Mar 3, 2020
11
Overdrive? 1959 vacuum tubes guitar amplifier (61 years old!)?
Oh. You are making an extremely distorted (clipping like crazy) guitar amplifier using Mosfets that are very different to vacuum tubes and bipolar transistors.

1) When the grid and cathode of a vacuum tube have equal voltages then the vacuum tube conducts as much current as it can.
2) When the base and cathode of a transistor have equal voltages then it conducts no current.
3) When the gate and source of a Mosfet have equal voltages then it also conducts no current.

Of course a vacuum tube and Mosfet have no input current therefore the input impedance is very high. A bipolar transistor has some bias current causing a much lower input impedance. A high output impedance of one stage is loaded down and has its gain reduced by the low impedance of a following stage.
If you plan the design of an amplifier correctly so that the output of each stage is loaded correctly to work properly then you can have as much gain as you need.
So the bjt low impedance causes loading which attenuates the signal, but in a high impedance input there’s minimal loading and therefore minimal signal loss?
thank you for explaining that to me. I thought I’d never figure it out. I’m a total novice so thanks for helping me out
 

Audioguru again

Joined Oct 21, 2019
1,488
Billions of audio amplifiers use bipolar transistors. Opamps also use bipolar transistors.
The magnetic pickup of a guitar is designed to have a load that is usually 1 million ohms like a vacuum tube or a Jfet or opamp with Jfet inputs. A bipolar transistor or opamp with bipolar inputs might have an input impedance of only 20k ohms.

Speakers used with electric guitars usually do not produce high audio frequencies that are usually produced by a tweeter and I think guitar players and listeners are deaf to the high frequencies anyway, so the 1M load on the magnetic pickup produces a mid-high frequency boost shown in this graph:
 

Attachments

Top