temperature compensation in collector common transistor amp
- Thread starter xela_92
- Start date
Help Support GroupDIY Audio Forum:
bcarso
Well-known member
A schematic would be helpful.
ruffrecords
Well-known member
[quote author="xela_92"]Does anyone can help me in order to compensate drifting due to temperature in transistors ?
the schem is a simple collector common ( class a amp using a 2N2222 transistor)... [/quote]
The standard/classic way to do this is to use a pot divider bias network and ensure the current through it is 10 times the required base current (which of course you know from the collector current and the hFE of the chosen transistor).
Ian
the schem is a simple collector common ( class a amp using a 2N2222 transistor)... [/quote]
The standard/classic way to do this is to use a pot divider bias network and ensure the current through it is 10 times the required base current (which of course you know from the collector current and the hFE of the chosen transistor).
Ian
bcarso
Well-known member
[quote author="ruffrecords"][quote author="xela_92"]Does anyone can help me in order to compensate drifting due to temperature in transistors ?
the schem is a simple collector common ( class a amp using a 2N2222 transistor)... [/quote]
The standard/classic way to do this is to use a pot divider bias network and ensure the current through it is 10 times the required base current (which of course you know from the collector current and the hFE of the chosen transistor).
Ian[/quote]
Yes, you are describing a so-called Vbe multiplier, which is often used to stabilize a complementary emitter follower. However, xela_92 speaks of a single Q, suggesting that it's a resistor or I-sink loaded e-follower. Again, we need to see the circuit to go further.
the schem is a simple collector common ( class a amp using a 2N2222 transistor)... [/quote]
The standard/classic way to do this is to use a pot divider bias network and ensure the current through it is 10 times the required base current (which of course you know from the collector current and the hFE of the chosen transistor).
Ian[/quote]
Yes, you are describing a so-called Vbe multiplier, which is often used to stabilize a complementary emitter follower. However, xela_92 speaks of a single Q, suggesting that it's a resistor or I-sink loaded e-follower. Again, we need to see the circuit to go further.
> you are describing a so-called Vbe multiplier
No. I think ruffrecords is describing the basic voltage-divider, applying a "fixed" voltage to the Base. The Emitter "follows", so knowing the voltage and a target current you know an emitter resistor.
The "10X" thumb-rule keeps errors around 10%, which is good enough for any small audio work.
Oh, and you want that emitter voltage to be "large" compared with Vbe (~0.6V) to get "reasonable" temperature stability. For room-temperature work, Vbe may drift 50mV. If the emitter voltage is 500mV normally, a 50mV change is only a 10% change of emitter current, which is good enough for any small audio work.
When you add all the possible errors, or work in extreme temps, you may want a higher voltage across the emitter resistor. Use cheap dirty Silicon (or Germanium!) at high temps, you may even want to increase the current in the base voltage divider.
But these are rarely issues in modern home audio. (I remember when bleeding-edge Ge needed half the supply voltage across the emitter resistor, just to work in studios....)
On the other hand, if you are measuring thermocouples or strain gauges, microVolts matter. Then (as you well know from your telescope days) you get into very different techniques than we normally need in audio.
And since it IS such a broad topic, from Volts to microVolts, I think xela_92 needs to speak up and say what he is doing, guitar-buffering or precision measurements.
No. I think ruffrecords is describing the basic voltage-divider, applying a "fixed" voltage to the Base. The Emitter "follows", so knowing the voltage and a target current you know an emitter resistor.
The "10X" thumb-rule keeps errors around 10%, which is good enough for any small audio work.
Oh, and you want that emitter voltage to be "large" compared with Vbe (~0.6V) to get "reasonable" temperature stability. For room-temperature work, Vbe may drift 50mV. If the emitter voltage is 500mV normally, a 50mV change is only a 10% change of emitter current, which is good enough for any small audio work.
When you add all the possible errors, or work in extreme temps, you may want a higher voltage across the emitter resistor. Use cheap dirty Silicon (or Germanium!) at high temps, you may even want to increase the current in the base voltage divider.
But these are rarely issues in modern home audio. (I remember when bleeding-edge Ge needed half the supply voltage across the emitter resistor, just to work in studios....)
On the other hand, if you are measuring thermocouples or strain gauges, microVolts matter. Then (as you well know from your telescope days) you get into very different techniques than we normally need in audio.
And since it IS such a broad topic, from Volts to microVolts, I think xela_92 needs to speak up and say what he is doing, guitar-buffering or precision measurements.
bcarso
Well-known member
You're right---I see the thrust now. To reiterate, we need more data about what's desired.
