What happens to voltage gain when a bipolar transistor is operated at low voltages (Vce < 1V). In a previous circuit that operated at a quiescent Vce of 0.75V, voltage gain was demonstrably low. This experiment is an attempt to quantify what actually happens to voltage gain and input resistance in a general purpose transistor over the range of Vce = 0.1 to 10V. The results are interesting and informative.
It was desired to keep Ic constant over the entire range, but that is not practical in the test circuit. Also, I wanted to keep the collector load resistor constant because this has a direct relationship with Av. I suppose that I could have varied Vcc for each graph point, but the number of adjustments for each point becomes unreasonable and I desired to use a fixed battery to reduce noise. So to simplify the adjustment procedure and obtain seamless data, I allowed Ic to vary from 5.7 to 3.5mA over the Vce voltage range of 0.1 to 10V. The effect of Ic upon Av over this range I know from experience to be minimal. While Ic varies by a factor of 1:1.6, Vce varies by 100:1
I selected a typical medium gain 2N3904 small-signal transistor with an hFE of 200.
With pot R6 set at zero, the amplifier self biases itself at Vce = 10V by virtue of the base divider resistors that are fed from the transistor collector. As the pot is rotated, current is injected into the base circuit thus reducing the quiescent operating point voltage. At maximum rotation, Vce = 0.1V. Otherwise, the additional resistor (R5) has minimal effect upon small-signal AC voltage gain.
To obtain reasonable data at such low DC voltages without appreciable amplitude distortion, the output AC signal level was purposely kept in the range of about 10 to 20mV. This qualifies it as a small-signal amplifier. This low signal level output and relatively high voltage gain (Av = 500) requires an 80db attenuator at the output of the signal generator.
It was a challenge getting the noise level down to an acceptable level. Because the variable voltage power supply is quite noisy, I opted to use 27VDC from batteries (three 9V batteries in series). Output noise was in the order of 0.5mV – low enough to have minimal effect upon measurements. One detail you experimenters must keep in mind when using a protoboard is that the steel baseplate (shield) must be grounded to the circuit via the corner binding post – otherwise the baseplate acts as an intermediate capacitor coupling ambient noise into the sensitive circuit.
It is easy to determine input resistance by adding a series pot and adjusting it until the output voltage is exactly half. At that point the pot resistance equals the input resistance and may be easily measured via a dmm.
The gain plot indicates relatively constant gain between Vce = 0.35V and 10V, but shows it dropping like a rock below 0.3V. I learned something here because I thought that the transition would occur at a much higher voltage.
The Rin plot shows increasing resistance between Vce = 0.35V and 10V – this is probably tracking the change in Ic over this range – if the current was truly constant, I believe that it would be flat. However, below about 0.3V, the input resistance drops very quickly and actually bottoms out at Vce = 0.15V.
Yes, bipolar transistors can operate at a low Vce, but with degraded performance. The threshold is approx 0.4V or so for the 2N3904 operating at Ic = 5mA. A voltage gain of 10 is possible at Vce = 0.15V, but the input resistance is extermely low (< 100Ω).
Readers are encouraged to duplicate my results. Those that have access to a simulator are encouraged to see if the simulation works at this low Vce level - I have my doubts.
For the future
Single Transistor Amplifier Revisited, Part 5 – effect of load resistance upon voltage gain