# Increasing the output current to a resistor network using transistors

#### Nabil Miri

Joined Jul 20, 2022
44
Hello all,
I am working on a circuit that senses any voltage difference between Rw and the OpAmpSource and upon that more current is fed into the resistors netwrok so that voltage around Rwrk remains equals to the OpAmpSource. So, hwo can I increase the amount of current that can supplied by the BJT to the resistors? (the OpAmpSource is an AC circuit that has a max amptitude of 5V. NPN works in the positive voltage presence and PNP in the negative.)

#### Alec_t

Joined Sep 17, 2013
14,212
Welcome to AAC!
So, how can I increase the amount of current that can supplied by the BJT to the resistors?
Increase the supply voltages to the output stage and/or decrease the resistor values.

#### Nabil Miri

Joined Jul 20, 2022
44
Increase the supply voltages to the output stage and/or decrease the resistor values.
The idea is that I can't control the value of the resistors. They are a model of electrodes. So their range differs. For example, if the Rw = 1 and Rc = 10, the voltage around should be 21.6 V but this circuit can not supply it as the limiting factor is the voltage supply as you said.
I am trying to see if anyone knows another configuration or component that could be used to increase the possible current.

#### Papabravo

Joined Feb 24, 2006
20,993
The idea is that I can't control the value of the resistors. They are a model of electrodes. So their range differs. For example, if the Rw = 1 and Rc = 10, the voltage around should be 21.6 V but this circuit can not supply it as the limiting factor is the voltage supply as you said.
I am trying to see if anyone knows another configuration or component that could be used to increase the possible current.
Take a step back and review the implications of Ohm's Law. In summary:
1. There are three quantities, current, voltage, and resistance.
2. If two of them are known, e.g. voltage and resistance, the third cannot be controlled independently. It must assume a fixed and CONSTANT value.
What you want to do is quite simply impossible. Changing the bias on the transistors changes the effective resistance which will lower the current, but there is nothing you can do with that circuit to increase the current.

#### Nabil Miri

Joined Jul 20, 2022
44
Understood. I have another question related to the BJT.
When I use an MJE182 NPN BJT, the resistors get a current of 1.19 A. But when I use a BD179 NPN BJT, the current decreases to 855 mA.
So, what is the parameter that makes this difference?
BD179:

MJE182:

Last edited:

#### Ian0

Joined Aug 7, 2020
9,493
Hfe.
Both have a wide spread of values, it is likely that your BD179 has a hfe towards the bottom if its range, and theMJE182 has a value towards the top of the range. The base current is limited by the op-amp.

The art of designing a good circuit is making it work with any value of hfe within the range quoted in the datasheet.

#### Papabravo

Joined Feb 24, 2006
20,993
The base current into the NPN transistor is essentially the same in both cases so the difference between the two is the current gain. The MJE182 has more gain at this current level than the BD179. Do the datasheets confirm or contradict this observation? Remember that any given device will exhibit a current gain within some range. For this reason, it is almost always a mistake to design a circuit that depends on a particular value of a poorly controlled device parameter.

MJE182 gain @1.5A & 1.0C Vce minimum 12, no upper limit
BD179 gain @1A, & Vce(sat)=0.2V @ Ic=1.0A typical 100

So we appear to have a contradiction, however the difference in Vce(sat) is dramatic and this may also contribute to the disparity.

Last edited:

#### Nabil Miri

Joined Jul 20, 2022
44
The base current into the NPN transistor is essentially the same in both cases so the difference between the two is the current gain. The MJE182 has more gain at this current level than the BD179. Do the datasheets confirm or contradict this observation? Remember that any given device will exhibit a current gain within some range. For this reason, it is almost always a mistake to design a circuit that depends on a particular value of a poorly controlled device parameter.
In the MJE182 there is only this info about the hfe:
and I found this in the BD179:

How can I depict what is the cause using these values?

#### Papabravo

Joined Feb 24, 2006
20,993
In the MJE182 there is only this info about the hfe:
View attachment 273059and I found this in the BD179:
View attachment 273061View attachment 273063
How can I depict what is the cause using these values?
The data for the MJE182 is a MINIMUM, with no maximum specified.
For the BD179 they specify a gain of 15 as a minimum with a Vce(sat) of 2V.
EDIT:
The actual gain for the MJE182 is 1.19/20.587m = 57.85, which is certainly above the minimum
The actual gain for the BD179 is 855m/20.591m ≈ 41.57 which is again above the minimum

So, says me, everything is within expectations.

#### crutschow

Joined Mar 14, 2008
34,030
As noted, the rule is: Never design a circuit that depends upon a specific value of transistor current gain.

#### Nabil Miri

Joined Jul 20, 2022
44
The data for the MJE182 is a MINIMUM, with no maximum specified.
For the BD179 they specify a gain of 15 as a minimum with a Vce(sat) of 2V.
EDIT:
The actual gain for the MJE182 is 1.19/20.587m = 57.85, which is certainly above the minimum
The actual gain for the BD179 is 855m/20.591m ≈ 41.57 which is again above the minimum

So, says me, everything is within expectations.
So, if I wanted a BJT that could amplify the current more than the MJE182 for example, should I search for a transistor that can has a higher value of Hfe?
Moreover, they wrote the minimum but is there a way to calculate the current Hfe?

#### Nabil Miri

Joined Jul 20, 2022
44
As noted, the rule is: Never design a circuit that depends upon a specific value of transistor current gain.
Can you please guide me to how can we make a design that works for a certain range?

#### crutschow

Joined Mar 14, 2008
34,030
Can you please guide me to how can we make a design that works for a certain range?
You design it so that changes in the base current with load do not significantly affect the circuit operation.
If the base current is to high for this, then you can go to a Darlington or Sziklai pair to reduce that current.

#### Nabil Miri

Joined Jul 20, 2022
44
You design it so that changes in the base current with load do not significantly affect the circuit operation.
If the base current is to high for this, then you can go to a Darlington or Sziklai pair to reduce that current.
Understood your point. If I know the range of the base current, how can I know or calculate the value of Hfe? (Is this the second step to do after knowing the base current range?)

#### Ian0

Joined Aug 7, 2020
9,493
You only need to know the minimum. If you design for the minimum, you know that any transistor that is not as bad as the worst case will require less base current and the design will work.

#### Papabravo

Joined Feb 24, 2006
20,993
You must consider the current gain of an individual device as if it were a normally distributed random variable with a given mean and standard deviation. You cannot depend on any given device to have a particular value of gain even if it comes from the same wafer. What you want is quite simply NOT possible. You test each and every device that comes through incoming inspection and put them in separate bins, but that is the road to economic ruin. Get your head out of the dark place it is in and adopt a different design approach.

#### Ian0

Joined Aug 7, 2020
9,493
It's worth spending that extra fraction of a penny and buying a Hfe-graded device, such as BC847B. The B indicates that the gain will be between 200 and 450, ungraded ones could be between 100 and 800.
You also might have to downgrade your estimation of gain if you product is going to be used anywhere particularly cold, as the gain is specified at 25°C.
Also, whilst the BC847 is rated at 100mA, the Hfe curve starts to turn south at about 70mA. If your requirement is for more than 70mA, buy a different transistor.

#### Nabil Miri

Joined Jul 20, 2022
44
It's worth spending that extra fraction of a penny and buying a Hfe-graded device, such as BC847B. The B indicates that the gain will be between 200 and 450, ungraded ones could be between 100 and 800.
You also might have to downgrade your estimation of gain if you product is going to be used anywhere particularly cold, as the gain is specified at 25°C.
Also, whilst the BC847 is rated at 100mA, the Hfe curve starts to turn south at about 70mA. If your requirement is for more than 70mA, buy a different transistor.
I need one that can handle at least 1.5 A.

#### Nabil Miri

Joined Jul 20, 2022
44
You must consider the current gain of an individual device as if it were a normally distributed random variable with a given mean and standard deviation. You cannot depend on any given device to have a particular value of gain even if it comes from the same wafer. What you want is quite simply NOT possible. You test each and every device that comes through incoming inspection and put them in separate bins, but that is the road to economic ruin. Get your head out of the dark place it is in and adopt a different design approach.
Ok. So, after I know the gain, is there a specific graph in the datasheet that denotes the gain value for a certain amount of voltage?

#### Papabravo

Joined Feb 24, 2006
20,993
Ok. So, after I know the gain, is there a specific graph in the datasheet that denotes the gain value for a certain amount of voltage?
Not that I am aware of.