DOPA’18 – Classic Discrete Preamplifier – Design Guide.
The circuit presented in this article is a non-LTP (Long Tailed Pair) preamp design. The DOPA’18 is a classic transistor preamplifier similar to circuits popular in the 60′s and 70′s. Many famous vintage consoles were based on simple 3 to 5 transistors circuitry. Each block of the signal path in the console was designed for a specific task. The microphone input preamp circuit was different from the “opamp” used in the EQ section. Today same IC operational amplifier can be used in mic pre and in the EQ section.
This article will explain how the DOPA’18 preamp works. Although all calculation are presented and explained, good knowledge about electronics is required. For the purpose of this article and future publications the circuit will be called DOPA’18. This article however is not the design guide for the DOPA’48 discrete operational amplifier, which will be released for use in your own projects and equipment.
The circuit was built and tested, frequency responses and THD+N plots are showed in the article. Anybody who wishes to build and experiment with the preamp should have practical experience with circuit assembly as bad assembly will lead to oscillation or the circuit will not work at all!
The DOPA’18 can be used as microphone preamplifier, general purpose gain stage, filter building block or output transformer driver.
The DOPA’18 is not suitable for EQ application since it is non-inverting amplifier. Inverting version of discrete preamp will be introduced in the following articles. An example of the complete microphone preamp will be the subject of next article.
2.0 Preamplifier Circuit.
Figure 1 shows the test circuit diagram. The preamplifier operates in a non-inverting configuration. DOPA’18 is a Class A circuit with high current low impedance active emitter follower output.
Classic Discrete Microphone Preamplifier – DOPA18 – Bart HRK
2.1 Input section
The input section of the DOPA’18 is built with T1, R1, R4, R7 and C1. C1 and R7 form a hi-pass first order input filter. By changing the value of C1 or R7 the frequency response for LF can be changed. Equation for first order hi-pass filter is shown below:
For components values shown on the diagram Fc=0.16Hz.
It is good practice to use a larger value of C1 than the value from the above calculation. I prefer to use at least 100uF, especially if more stages are used in the signal path of the whole preamplifier or channel strip unit. In practice the value of R7 shouldn’t be increased over 100k. The base current of the bipolar transistor has significant value which will increase the noise level. This could be an issue for mic preamplifier. Values over 100k can also cause DC biasing problems. If any components are changed it should not be forgotten than a bipolar transistor is a current amplifier.
T1 works in common emitter (CE) configuration and introduce 180° phase shift. R4 sets the current for the input stage. Approximately 1mA will ensure stable DC condition for the input stage. Theoretically according to some datasheets lower collector current of the input stage should improve noise performance. Practically good assembly, proper shielding, use of high quality input transformers and use of low noise transistors will have bigger impact on noise performance of the circuit. EIN of -126dBu at +60dB of gain should be considered as a good result.
R1 is biasing the second stage and the voltage across R1 should be close to 0.7V. R1 can be replaced by a trimmer which will help to tune the output voltage of the preamp close to 0V. The output voltage drift could be an issue with non LTP circuits. In reality the output offset voltage can vary between +/-200mV.
Tip: Before fine tuning the output voltage allow the circuit to operate for 10 minutes. This should be enough to stabilise thermal condition of all transistors.
2.2 Calculation for the input stage.
Because the T1 emitter is tied to -12V the base is on a higher potential. In order to simplify the calculation we can assume that the base has potential of +12V when referenced to the emitter. Therefore the emitter voltage is equal to the base voltage minus Vbe of the T1. The emitter current is equal to the emitter voltage divided by R4. The voltage across R1 is the result of the collector current which is almost equal to the emitter current. In reality the emitter current is a sum of collector and base currents. If the voltage on the base changes the emitter current will change as well, which will cause the change the voltage across R1.
All calculations are presented bellow. Anybody who wishes to learn more about how to design discrete preamps should follow all the calculations and check the results with values on the diagram shown on Figure 1.
For a start we can assume that and R1 connected to T1 emitter is on potential.
Where: is voltage on T1 emitter, is the voltage on the base of T1, is the voltage drop between base and emitter (usually: 0,65 – 0,7V).
Where: is current of T1 collector, is voltage on T1 emitter.
Always remember that:
Where: is voltage across R1, is current of T1 collector.
The value of R1=640 was picked to set the output voltage close to 0V. R1 can be replaced with the 1K trimmer and the output DC voltage can be fine-tuned. In reality DC output offset shouldn’t be a problem for audio circuits. Value of R1 between 620R or 680R will do the job.
Tip: If you use electrolytic capacitors to couple stages ensure correct polarity of the cap. For example if DC output voltage holds stable at -100mV the negative lead on the cap should be connected to the output of the circuit. Reversed polarity can cause problems and unstable behaviour of the whole preamp. The use of bipolar electrolytic caps is also a good idea.
2.3 The VAS.
T4 forms the VAS (Voltage Amplifier Stage). In a static condition T4 is conducting a small current. It is biased on the edge of being fully open. When the voltage across R1 changes, the current of T4 is changing as well which will change voltage across R5. Very small change of voltage across R1 will create a big current change in the collector of T4 and therefore a large voltage change across R5. This is why T4 is called the voltage amplifier. More about the VAS can be found in this article.
The collector current of T4 is quite important, but first we have to decide what the function of the preamp is. If circuit will drive an output transformer and output headroom is expected to be high, the T4 current should be increased even up to 3mA, simply because the following output stage will draw more current. The T4 stage will have to provide enough current to drive base of T2.
If we just need a voltage amplifier and we know that the load of the preamp will be more than 1k we can leave at 1mA.
Calculation for VAS:
Where: is voltage of T4 collector, is the preamp output voltage, is the voltage drop between base and emitter of T2 (usually: 0.65 – 0.7V).
Where: is the current of T4 collector, is voltage of T4 collector from previous equation.
The T4 works in CE configuration which means that VAS introduce another 180° phase shift. The DOPA’18 is non-inverting amplifier and is impossible to use it in the inverting configuration.
2.4 The output stage.
The T2 output stage is built with an active emitter follower circuit. T3 works as the current source and the current is set by resistor R2. The Red LED is very useful as the reference voltage source.
The calculation for current source is very simple. Usually Red LED has Vf =1.65V. This applies to cheap low efficiency Red LEDs. Because T3 Vbe=0.65V emitter voltage is equal to approximately to 1V. The math is very simple: 1V divided by 100R gives 10mA of current drawn by the current source.
I believe the calculations are easy to understand.
The LED current is set by R3. If the preamp is used as the gain stage the LED current is sufficient. Only if T3 is replaced by a medium power transistor such as BD139 the LED current should be increased. Medium power transistors have lower Hfe factor and the base current will be higher. This could be necessary if the DOPA’18 is driving a low impedance load such as an audio transformer, therefore higher current of the output buffer would be required. One of the Application Notes for the DOPA’18 will present an output transformer driver circuit. The current of the current source from Figure 1 is 7.29mA because a LED SPICE model used in simulation has different Vf value. The Vf of LED can be checked with standard multimeter.
3.0 Real life circuit.
The schematic diagram of the DOPA’18 which can be used for experiments is shown on Figure 2.
Classic Discrete Microphone Preamplifier – DOPA18 – Bart HRK 1
Compared to the diagram in Figure 1 a few capacitors have been added. C5 and C6 are standard power supply rail decoupling caps. It is a good idea to add an additional two 100uF caps in parallel with C5 and C6 to ensure the power supply provides filtered and stabilised voltages and to remove any crosstalk between other circuits via PSU rails. I would recommend using bench PSUs with split stabilised rails and earthed ground connector for any experiments.
C4 is essential for filtering any AC signals and PSU rail ripples on the input of the current source. The value of C4 can be increased up to 100uF if needed. An additional 100nF ceramic cap can be added in parallel with C4.
The supply voltage can be increased up to +/-15V, however power dissipation limits of the T2 and T3 should be taken under consideration.
R9 improves stability of the preamp. R9=100R is necessary if the DOPA’18 is needed to drive longer cables or complex impedance loads such as filters network or inductors. R9 should be removed if the DOPA’18 is used as the audio transformer driver.
3.1 The feedback loop and gain.
R6 and R4 are creating the feedback loop. The minimum gain of the preamp will be always equal to +6dB. To calculate the voltage gain the standard equation for a non-inverting amplifier can be used.
Because R4 is setting static DC conditions of the circuit we need to add R8 in parallel with R4. Capacitor C3 is necessary to block gain for DC. For R8 in parallel with R4 the equation for gain will change:
With R8=1k the gain of the DOPA’18 will increase to +21dB. The capacitance C3 should be at least 1000uF to avoid a loss in LF for higher gain values. To get +60dB of gain, resistance of R8 will have to be as low as 10R. High gain is possible to achieve with this simple circuit if required. According to SPICE simulation the open loop gain can reach +78dB in the 20kHz bandwidth – this is very good result! However I don’t recommend trying this with the circuit built on the test bench as the preamp will most likely become a radio receiver… 🙂
Capacitor C2 – 10pF connected between the base and the collector of T4 limits the HF response of DOPA’18. I don’t recommend using a capacitor across R6 in order to limit HF response nor to fix the problem of oscillation. If the preamp is oscillating it means at least that assembly is bad or the ground layout has issues, or the supply rails are poorly stabilised. If the circuit is unstable limiting the bandwidth will not help.
Limiting the bandwidth of the feedback loop can result in a higher level of THD especially in the HF region, simply because the feedback loop loses the ability to control the THD for high frequency.
4.0 Performance of the test circuit.
DOPA18 THD vs Freq
Figure 3 shows plots for the THD+N vs Frequency performance. Green plot represents THD for Gain=6dB and the yellow plot for Gain=21dB.
DOPA18 Freq Resp
Figure 4 shows the Frequency Response of the DOPA’18 for Gain=21dB. 0dBr is referenced to 0dBu@1kHz.
All measurements were done with an Audio Precision on the test bench.
Picture 1 shows how many parts are required to build the DOPA’18 preamp.
Classic Discrete Microphone Preamplifier – DOPA18 – Bart HRK 2
The test circuit is quick and easy to build, just perfect for weekend afternoon!
In the next article I will explain how to build microphone preamp with a step-up input transformer.
I hope you will enjoy experimenting with the DOPA’18 circuit and as always feedback is welcome.
If you have any question please email me. It would be great to create the Q&A section bellow the article.