amplifying DC signal - BJT?

Discussion in 'Homework Help' started by edgetrigger, Jun 29, 2011.

  1. edgetrigger

    Thread Starter Member

    Dec 19, 2010
    I want to use a BJT for amplifying DC dignal, say a signal of 10mv originating from a sensor like PT100, here both the sensor and the transitor is powered by a single dc supply. A coupling capacitor wouldn't help as i want to amplify steady state dc signals.

    It is very easy to amplify dc siganls using op-amps but i am not able to figure out how to do it with a single transistor.

    I graduated with a decent score in engineering today i feel like surrendering my degree back to the university. I am not in a postion to simply crack this task.:mad:
  2. steveb

    Senior Member

    Jul 3, 2008
    Don't feel bad. It's not an easy task. Actually, unless you specify constraints on what defines a good measurement, it's impossible. This example is one of many reasons why OPAMPS were invented. Why do you need to do it with one transistor? Why wouldn't you use the OPAMP?
  3. Adjuster

    Well-Known Member

    Dec 26, 2010
    You should simply be thankful that good inexpensive IC amplifiers are available these days - they make life a lot simpler.

    It is of course possible to make discrete transistor DC amplifiers, and before suitable ICs were available this had to be done. That said, the kind of simple unbalanced amplifier that is satisfactory for AC signals would have poor temperature drift if used for DC. Normally some sort of differential (long-tailed) pair amplifier would be used to reduce drift, unfortunately, even with selected matched transistors, the balance is unlikely to be as good as it could be in an integrated amplifier.
  4. edgetrigger

    Thread Starter Member

    Dec 19, 2010

    No i am not against using opamps, humble 741s to classic burr-browns have come handy in my designs. I just thought amplifying dc is a simple task using a transistor and ventured, but to my dismay i found out that it is a daunting task.

    I thought that i have reached a position wherein i could build mixed signal circuits without refering to text book/google. But hard to digest that is not the case today. But even after thinking for 24 hours i am not able to figure out the basic logic to handle dc signals. That is what bugging me!
    Last edited: Jun 30, 2011

    Senior Member

    Jun 29, 2010
    I think you can use Two transistor to pull up the supply, in common collector first one than common emmiter....!!
    as C.C. will have high input impedance then C.E. will gave high current gain....!!
  6. WellGrounded


    Jun 19, 2011
    Which specific PT100 model are you looking at and what is the temperature range you are testing? What out put ratio do you want e.g 1mv =1 Volt change?

  7. steveb

    Senior Member

    Jul 3, 2008
    A carpenter would not let the fact that he can't build a house with one board and one nail, bug him. :p

    But, I understand your point of view. It's all part of the learning process, and in engineering/design we learn over a lifetime, not just in school. To do a real DC measurement with high quality and with good temperature performance, one transistor won't cut it. However, if you want to sense a change that can be converted to a course estimate of the DC voltage, it is possible. This is a good mental exercise for learning, but not an issue for practical design.
    Last edited: Jun 30, 2011
  8. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    Experience begins where you start...

    Personally I wouldn't give up on op amps just yet, but do toss those 741's and anything with "classic" in it's description. The latest generation of very high quality op amps will run off very small voltages (3 volts?) and a single ended supply.

    A simple (and traditional) non-inverting op amp buffer could give you all the gain you wish and not swamp out the tiny 10mV signal you have with offset and drift voltages.

    They are cheap and available, perhaps the only drawback is being very recent parts they tend to only come in surface mount packages.
  9. hobbyist

    Distinguished Member

    Aug 10, 2008

    Are you asking, "how can this be done using only ONE transistor" ?


    Are you asking " how can this be done using transistors ONLY" ?

    Please me more clarifying.
  10. hgmjr


    Jan 28, 2005
    You have not yet defined what accuracy you are looking for in the final circuit. Do you have a specific performance requirement for your circuit or are you simply experimenting as means of developing more familiarity with transistors and temp sensors.

  11. hobbyist

    Distinguished Member

    Aug 10, 2008

    This is a circuit I built a few months back, to experiment with light tracking.
    It works real well.


    This is an example of DC amplification, however as can be seen, there needs to be more than one transitor.

    The reason being, to bias a transistor stage you need to have the proper resistors, to keep the transitor stable, from temp. and param. drifts.

    To introduce a DC signa of low valuel to that same stage, will only cause the transistor to react according to the biasing applied to it, that's why the transistor is biased using resistors at it's emitter and base terminals, to keep it as stable as possible, during fluctuations in it's bias voltages, due to temp. drift, and parameter changes.

    So it would take a larger DC value to cause the transistor to amplify it, accordingly.
    That's why there needs to be more transistors introduced, so as to couple the small cahnges and to amplify them through succesive stages.

    I'll try to explain why One transistor is not enough to be a DC amplifier, for very small signals.
    By looking at my circuit.
    Q1 and Q2 are biased heavily into there conducting region. So as to apply the same voltage at the emitters and bases of both Q5 and Q6. With the same voltage applied at the emitter and base of the transistors, the transistors are NON conducting.

    Since only around 0.7 vbe. is needed to turn these transistors on, then a small change of voltage at the emitter of Q1, will be enough to cause Q5, or Q6 to conduct heavily.

    Q1 and Q2 are biased heavily, to keep a lot of temp. drift, and param. changes under some controle. But a large change in DC bias at the base of Q1 will cause a significant change in its emitter voltage, that when this change is compared to the voltage at the emitter of Q2, there will be enough Vbe, at either the Q5 or Q6 transistor to cause one of them to turn on, heavily.

    However it does not take much voltage change at the emitter of Q1, just enough Vbe to be applied to either Q5 or Q6.

    Now with this in mind, a change in resitance of either R10 or R11, will have a small change in DC value at the base of Q3 or Q4, but this change will be somewhat amplified as it is applied at the base of Q1.

    The change in resistance of R10 or R11, is small, it is too small to have any great effect upon Q1, so this small change of either R10 or R11, will have to go through a stage of amplification, hence that is the reason for both Q3 and Q1, OR Q4 and Q1, depending on which sensor is activated.

    So the very small change in either R10 or R11, is to small to effect Q1, but it is enough to cause a small chenge in voltage output of Q3 or Q4, due to the biasing voltages, however Q3 and Q4, do not have enough current capacity due to there biasing, to send this signal to the output, that's why the signal needs to go through 2 more stages, of amplification, the next amp stage is the emitter output of Q1, referenced to the emitter output of Q2.

    But still not enough voltage output to be useful, so the next stage of amplification, is the base emitter junctions of either Q5 or Q6, depending on sensor inputs.
    It is at the collector of Q5, or Q6, that there is significant amount of current flow to produce an useful output voltage.

    So you can see why a DC amp. needs more than one transistor to be useful.
    Now I am NOT an experianced electronics professional, EVERY one who answered you posts ARE.

    This is up there ally, there teritory,
    I on the other hand am ONLY a HOBBYIST in this field.....
    I have NO professional experiance, so everything I wrote is up for scrutiny, by those who are more knowledgable in this field.

    I am NOT trying to stir anything up here guys, I just think that maybe I could help this individual understand a little bit about why you need more than ONE transistor for a DC amp.

    It seems like everyone is saying to use opamps, but not answering his question about using a single transistor. And the detailed reason why one transistor is not feasible.
    So I'm doing just that.

    And ,,,Yes, I did build this amp and it works great, with my transistorized H-bridge and a hacked servo motor.
    Also I am fully aware that it is not the best design for the least power consumption.
    However it was a quick design for a proof of concept idea, I had for a light tracking circuit.
    And I wanted to see if it would work.

    Have a great day....
    Last edited: Jul 6, 2011
    RRITESH KAKKAR likes this.
  12. edgetrigger

    Thread Starter Member

    Dec 19, 2010
    Thank You Hobbyist, thank you very much for your circuit and your explanation. I'm planning to build your ckt this weekend and try it out.

    I wanted to build a single transistor dc amplifier for amplifying dc signals, the catch is a common power supply is used for the amplifier and the transducer ( thermistor, pt 100, thermocouple etc) generating the dc signal.

    If i split the supply to +Vcc and -Vcc and use -Vcc to pull the emitter low then easily this arrangement can work, but then it will be a dual supply ckt.

    Stability, drift, gain variation is not my concern.
  13. hobbyist

    Distinguished Member

    Aug 10, 2008

    Your welcome. I'm glad I was able to help out,

    However, the circuit values I came up with for my design, was with the 2 LDR's in the circuit, everything was calculated using those parameters, it may not work properly using the same values with your input device.

    Depending on your input device will depend on how to design your circuit.
    You may be able to use part or all the circuit configuration, but I doubt it would work properly just using the values I used.

    If you could post the input device values then someone could help you get a circuit underway.
  14. SgtWookie


    Jul 17, 2007
    Well, with a 10mV input signal, you'll be hard-pressed to amplify it with a single transistor; you'd need to get it above ~620mV just to start turning the transistor on.
  15. hobbyist

    Distinguished Member

    Aug 10, 2008
    A 10mV. signal, is very small for discrete transistors to operate at in detecting and amplifying, it takes a lot of trasnsitors in different configurations to make up a good amplifier for that low of value.

    My circuit will work fine for input signals in the 100ths of mA.'s, but never in the 10's of mA.'s.

    The reason being, that small of signal, can never be detected in line of the temp. drifts, and transistor parameters, B, Vbe change, all kinds of things like that.

    My voltmeter readings change in the 10's to 30 mA. readings just by holding the probes at the base of the transistor, without any input signals.

    That's the reason for very unique designs like opamps and things when it comes to extremely small signals to work with.

    Hope this helps in clarifying the potential problems with using discrete transistors in a small signal DC amp.