2
\$\begingroup\$

Electret amplifier LTSPICE I am trying to build an amplifier to allow a cheap electret boom microphone to be used with an aircraft radio, but am unsure how to increase the gain to a usable level.

I derived the attached design from experimentally trying this Circuit to make an electret microphone simulate a carbon microphone and Andy AKA's electret microphone amplifier design, neither of which had sufficient output when tested in the aircraft, and experimenting with schematic changes and component values in LTSPICE.

The attached design works though the output is lower than required when using a normal voice, but when speaking very loudly the output level matches the output level of commercially available aviation headsets and the signal is still very clear and intelligible. This tells me this amplifier can easily drive the aircraft radio's input but lacks sufficient gain to be driven by this microphone.

My previous attempt used a single 2N3904 (measured hFE = 110) and had the same problem of low output. Replacing the 2N3904 with an MPSA13 darlington (hFE = 5000) had no noticeable effect on the amplifier gain. I don't know enough about this stuff to understand why. How can I increase the amplifier gain without adding another transistor stage before the MPSA13? Cost isn't a factor but component count is as this amplifier will need to be made much smaller, and it seems absurd to require another stage when the MPSA13's gain is already so high. What am I missing?

I don't have access to an oscilloscope to measure actual signal levels and can only test each design in the aircraft, so there's a lot of guesswork involved. Any help will be greatly appreciated.

@Olin Lathrop - I simulated your amplifier design in LTSPICE (see image) Amplifier as suggested by Olin Lathrop

\$\endgroup\$
8
  • \$\begingroup\$ Replacing the 2N3904 with an MPSA13 That's because hFE is the current gain which has almost no influence on the small signal gain of a transistor. To understand that you need to follow a course on transistor circuit analysis. My advice: get a different microphone/headset/amplifier because modding an existing one is nearly impossible when you don't fully understand transistor amplifiers. And learning how they work will take you even more time. The extra difficulty in an aircraft is the high noise levels of the engine, some kind of AGC is probably needed. \$\endgroup\$
    – Bimpelrekkie
    Commented Feb 3, 2018 at 22:24
  • 1
    \$\begingroup\$ Thank you for pointing out that hFE has little influence on small signal gain, that explains why the MPSA13 didn't help. I'm stuck with the microphone and radio, so modifying the amplifier design is my only option. The noise level is actually low because it's a jet aircraft so AGC isn't needed, and the microphone is a dual port noise cancelling type. \$\endgroup\$
    – Adam
    Commented Feb 3, 2018 at 22:28
  • \$\begingroup\$ Small size doesn't require discrete components (individual transistors.) Integrated amplifiers are available in packages smaller than a single 2N3904. You have two things you need to do. One is to amplify the microphone, and the other is to drive the low impedance microphone input of the radio. I think you problems are in the amplifier section. A simple opamp microphone preamplifier can be made to provide enough gain and a high input impedance. That can then drive another amplifier with a low output impedance \$\endgroup\$
    – JRE
    Commented Feb 3, 2018 at 22:30
  • \$\begingroup\$ Thank you JRE, however the problem is that the aircraft radio is intended to only supply a very small current (maybe 0.5mA?) to power the FET amplifier in an electret microphone, and a separate battery to power the amplifier is not an option. The amplifier needs to simulate a carbon microphone as that is the standard aircraft radio input (even in 2018!). The attached design has been demonstrated to do this, but lacks gain. My problem is how to increase the gain. \$\endgroup\$
    – Adam
    Commented Feb 3, 2018 at 22:34
  • \$\begingroup\$ A better amplifier doesn't need to use more power. The odds are very good that an opamp preamplifier will use less current than your single transistor amplifier. \$\endgroup\$
    – JRE
    Commented Feb 3, 2018 at 22:42

1 Answer 1

Reset to default
3
\$\begingroup\$

Here is something to try:

This circuit will load Vout with a current proportional to the microphone sound. It's not totally clear from your description that is what the input to the existing system wants, so this might not work. This is assuming the existing system powers Vout, and reacts to the current signal drawn from it.

R5 and C1 filter the AC from the input voltage to make a reasonably steady average DC voltage. This voltage is then used to run the electret mic, with R6 being the load. The mic signal is then AC coupled thru C2 into the amplifier section.

The amplifier is similar to what you already have, but with a more predictable gain and operating point. The emitter of Q2 should be held at around 4 V when there is no sound. Vout minus this 4 V causes a current thru R1, which is the output signal. That current will be perturbed by the signal from the microphone.

R3 and R4 bias the amplifier stage by providing DC feedback. This DC feedback is high enough impedance that it can be perturbed by the microcphone signal.

Its not clear how much current change the existing system is expecting for a normal amplitude. You'll just have to experiment. Hopefully this circuit provides either enough or too much gain. To reduce the gain, increase R2.

R1 should be a "½ W" resistor if you keep the current value. Again, it's not clear what current range the existing system is expecting.

\$\endgroup\$
5
  • \$\begingroup\$ Thank you Olin. I'll give this a try but first I'll have to obtain some 2N4403 or similar, all I have available right now are 2N3904 and MPSA13 (I know, I know, great junkbox). The radio input expects around a 0.5-1V input signal from a carbon microphone of around 150 ohms... these are very approximate specs since the actual 'standard' dates back to WWII bombers. I think the 200 ohm resistor R1 might limit this current too much. \$\endgroup\$
    – Adam
    Commented Feb 3, 2018 at 23:29
  • \$\begingroup\$ @Adam: Pretty much any jellybean NPN and PNP will work for Q1 and Q2, respsectively. If I remember right, a 2N3904 is a NPN, and therefore should work for Q1. Maybe the radio wants to see a lower impedance load, but it's safer to start with higher impedance. There are some good reasons not to draw too much current. If the 200 Ohms works, I'd stick with it. Certainly try that first. \$\endgroup\$
    – Olin Lathrop
    Commented Feb 3, 2018 at 23:33
  • \$\begingroup\$ Will your suggested amplifier have significantly greater gain than the design I attached? It's undoubtedly a better design but just as a practical matter (parts on hand vs mail order) as a short term solution I'm looking for ways to increase my amplifier's overall gain (preferably without adding a gain stage) since it's already been demonstrated to drive the radio input adequately when using a loud enough voice into the microphone. \$\endgroup\$
    – Adam
    Commented Feb 3, 2018 at 23:39
  • \$\begingroup\$ @Adam: It's not clear what the gain of your existing design really is. \$\endgroup\$
    – Olin Lathrop
    Commented Feb 4, 2018 at 12:57
  • \$\begingroup\$ I simulated your amplifier design in LTSPICE (see image) and found the output to be very low, well below the 0.5-1.0V p-p expected by the aircraft radio. I tried removing the 10R resistor and/or decreasing the 200R resistor to 20R and both increased the output to around 1V p-p, but the gain is about the same as the amplifier I posted first and the two other designs referenced in my original post. I still need to find a way to increase overall gain about 5-10x to accomodate the low microphone output (I'm simulating 10mV but think it might be closer to 1mV in practice). \$\endgroup\$
    – Adam
    Commented Feb 4, 2018 at 17:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.