Common collector load driver

[rich]

Hello all,
I need to to drive a 300mA resistive load with an variable voltage
of 0-30V. Right now I have a NPN transistor in the common collector
configuration.

The collector is tied to the supply voltage which can be 10-30V. The
base is driven from 0V up to the supply voltage.

This works just fine up to a supply voltage of about 20V. I can vary
the base voltage from zero to that and get a nice output without too
much heat.
Above this however I get way to much heat, even at small loads.

The transistor is in a sot-223 package with a pcb copper pour area as
a heatsink.

What is a better way to construct a voltage follower that will
dissipate less heat? Perhaps switching to a p-channel mosfet?

Thanks

Rich

 

[[email protected]]

Hello all,
I need to to drive a 300mA resistive load with an variable voltage
of 0-30V. Right now I have a NPN transistor in the common collector
configuration.

The collector is tied to the supply voltage which can be 10-30V. The
base is driven from 0V up to the supply voltage.

This works just fine up to a supply voltage of about 20V. I can vary
the base voltage from zero to that and get a nice output without too
much heat.
Above this however I get way to much heat, even at small loads.

The transistor is in a sot-223 package with a pcb copper pour area as
a heatsink.

What is a better way to construct a voltage follower that will
dissipate less heat? Perhaps switching to a p-channel mosfet?

Thanks

Rich

Rich,

Max power will be about 9 watts at 30 volts. What is the number of
the transistor you are using and the config. one side of the resistive
load is connected to the collector and the other side to the 30 volt
supply, the emitter is grounded or tie to the 30 volt power supply
return. The control voltage is connected to the base thru a base
resistor. Is this your config?

Den

 

[Mike Silva]

Hello all,
   I need to to drive a 300mA resistive load with an variable voltage
of 0-30V.  Right now I have a NPN transistor in the common collector
configuration.

The collector is tied to the supply voltage which can be 10-30V.  The
base is driven from 0V up to the supply voltage.

This works just fine up to a supply voltage of about 20V.  I can vary
the base voltage from zero to that and get a nice output without too
much heat.
Above this however I get way to much heat, even at small loads.

The transistor is in a sot-223 package with a pcb copper pour area as
a heatsink.

What is a better way to construct a voltage follower that will
dissipate less heat?  Perhaps switching to a p-channel mosfet?

Worst case, if you have to drop 300mA at up to 30V (30V supply,
desired output of 0V (OK, your load is not likely to be drawing 300mA
at 0V, but the principle doesn't change)), you have to dissipate 9W in
your transistor - no way around that. A mosfet will have to dissipate
the same 9W under the same conditions. If you can reduce your max
supply voltage or load current, or if you can drive your load in a
pulsed manner rather than an analog manner, you can reduce the power
you need to dissipate.

BTW, I'm reading your load as a non-reactive load that draws 300mA
regardless of voltage. If that's not correct (and I kind of doubt it
is), then your maximum dissipated power amount will change based on
the load V-I curve.

Mike

 

[mkr5000]

If you have a 2N4401 laying around you may want to try it.

I know they're good for 700ma but I'm just throwing that out.

Or beef up to a MJE3055 or something?

 

[Spehro Pefhany]

Hello all,
I need to to drive a 300mA resistive load with an variable voltage
of 0-30V. Right now I have a NPN transistor in the common collector
configuration.

The collector is tied to the supply voltage which can be 10-30V. The
base is driven from 0V up to the supply voltage.

This works just fine up to a supply voltage of about 20V. I can vary
the base voltage from zero to that and get a nice output without too
much heat.
Above this however I get way to much heat, even at small loads.

The transistor is in a sot-223 package with a pcb copper pour area as
a heatsink.

What is a better way to construct a voltage follower that will
dissipate less heat? Perhaps switching to a p-channel mosfet?

Thanks

Rich

The power will always be Iout * (Input_voltage - Output_voltage) for
a linear regulator. That's the same whether you use a MOSFET, a BJT or
a rheostat operated by a servo.

If I take the worst-case interpretation of your requirements (that it
might be required to supply 300mA into a 0-ohm load with a 30V supply)
then you need to dissipate 9W, which will require a substantial
heatsink.

You might consider using a switching regulator, but I can'tt think of
a simple single chip solution off the top of my head. Check out Linear
Technology's line, there may be something you can use if cost isn't a
factor.

Best regards,
Spehro Pefhany

 

[[email protected]]

The power will always be Iout * (Input_voltage - Output_voltage) for
a linear regulator. That's the same whether you use a MOSFET, a BJT or
a rheostat operated by a servo.

If I take the worst-case interpretation of your requirements (that it
might be required to supply 300mA into a 0-ohm load with a 30V supply)
then you need to dissipate 9W, which will require a substantial
heatsink.

You might consider using a switching regulator, but I can'tt think of
a simple single chip solution off the top of my head. Check out Linear
Technology's line, there may be something you can use if cost isn't a
factor.

Best regards,
Spehro Pefhany

Rich,

You can use a positive 28VDC reg chip such as a LM350TG 3 A,
Adjustable Output 1.2 to 33 V, Voltage Regulator TO-220 case. Instead
of connecting the gnd lead to gnd you would connect it to your
controlling voltage. This will vary the output as you want. The
bottom line is with a 300 ma load at 30 volts you will be generating 9
watts, thats the way it is. The output will be what ever you set it
for even if the load may vary.

Good luck,

Dennis

 

[Spehro Pefhany]

Rich,

You can use a positive 28VDC reg chip such as a LM350TG 3 A,
Adjustable Output 1.2 to 33 V, Voltage Regulator TO-220 case. Instead
of connecting the gnd lead to gnd you would connect it to your
controlling voltage. This will vary the output as you want.

Well, except it won't go below 1.25V (typical) without a negative
input.

OTOH, it will be short-circuit proof, SOA-protected, and
over-temperature protected (sort of).

The
bottom line is with a 300 ma load at 30 volts you will be generating 9
watts, thats the way it is. The output will be what ever you set it
for even if the load may vary.

Yup. Unless you have a switching regulator.

Good luck,

Dennis

Best regards,
Spehro Pefhany

 

[rickman]

As long as you use a linear pass transistor you'll find your circuit
very inefficient at output voltages lower than the supply voltage, and
the power lost to the inefficiency will get burnt up in the pass transistor.

About the only way that you can get away from that situation is to use a
switching amplifier, which is _much_ more involved to design, build and
test.

Assuming a mild-mannered load, the power in your drive transistor will
be (supply V - output V)(output current). For the extreme case where
you're driving 300mA at 0V, with a supply voltage of 30V, that's 9 watts
you've got to find a home for before it burns out your transistor.

Really, your only two choices involve either biting the bullet and using
a pass transistor and heat sink that can handle the power dissipation,
or biting the bullet and using a switching amplifier with all of its
design complexity.

Complexity? Switchers are pretty tame these days. Pick one of many
available at your current rating, add an inductor, a few caps and a
resistor divider. The only part that is not required with an LDO is
the inductor! The manufacturer does all the work of design for you.
All you have to do is follow their cookbook.

 

[Jamie]

Hello all,
I need to to drive a 300mA resistive load with an variable voltage
of 0-30V. Right now I have a NPN transistor in the common collector
configuration.

The collector is tied to the supply voltage which can be 10-30V. The
base is driven from 0V up to the supply voltage.

This works just fine up to a supply voltage of about 20V. I can vary
the base voltage from zero to that and get a nice output without too
much heat.
Above this however I get way to much heat, even at small loads.

The transistor is in a sot-223 package with a pcb copper pour area as
a heatsink.

What is a better way to construct a voltage follower that will
dissipate less heat? Perhaps switching to a p-channel mosfet?

Thanks

Rich

that's because you're creating to much resistance in the transistor keep
the voltage down on low levels when loaded.. This translates to heat.

When using lower input voltages, the transistor is at a lower
resistant value and thus, less heat.

A Buck regulator will correct for this error.


http://webpages.charter.net/jamie_5"

 

[[email protected]]

that's because you're creating to much resistance in the transistor keep
the voltage down on low levels when loaded.. This translates to heat.

When using lower input voltages, the transistor is at a lower
resistant value and thus, less heat.

A Buck regulator will correct for this error.

http://webpages.charter.net/jamie_5"

Rich,

Can you give us a little more info as what the load is and what your
driving it with. It seems to me the ans. are getting a little complex
for all we know it might be just changing the brightness of a lamp.

Dennis