Introduction:This manual is for assembly and minimal testing of our laboratory's EPL Microphone Amplifiers. This document will only cover minimal the minimal testing/debugging of the device. Advanced testing, proper operation of the required tools and other related engineering background is beyond the scope of this document. If we did not limit our writing to this the document would grow without bound. For a list of tools and skills we expect you to have for following this document please read the FAQ
The microphone amplifier is powered off of our standard EPL Power Supply. The EPL Power Supply is documented in other places. Since the microphone amplifier might be a considerable distance from the power supply we need C10 and C11. The value of C10 and C11 may be changed if the power cable is really long. C10 and C11 do not really need to be metal film it was just the result of a historical accident. A pair of reasonably sized electrolytic capacitors would do just as well.
The microphone is connected via a CONN2 on the microphone amplifiers PCB. Using a MiniDIN5 for CONN2 is not ideal if the microphone is going to be changed frequently as it is not really meant to be plugged and unplugged very many times. To keep outside noise out the shield on the power cable is connected via the Mini-HEX9 to the chassis of the microphone amplifier and through the Mini-DIN5 to the shield on the microphone cable. This keeps the precious microphone signal in a continuous clean environment.
The microphone amplifier also has a circuit built into it for providing the bias current to the preamplifier inside the microphone. For the FG series microphone used the current EPL DPOAE systems this is approximately 1.5V. The important thing is that the 1.5V be very quiet. VREG1 regulates the voltage down from the +15V supply. VREG1 has it's voltage output set by the combination of R22 and R23. To keep the noise going to the microphone down there is an RC filter composed of C6, R24 and C7.
This just leaves the amplifier itself. There is an input filter that also blocks DC made with C1 and R1. The first amplifier stage sets the gain and second provides output drive. The gain is set by selecting ether R13 or R11 from S1, for approximately 20db or 40db. S1 only needs to be SPDT not DPDT but this will be explained later. To make the amplifier a more stable C31 was added. The second stage of the amplifier, for driving output cabling, is configured for unity gain. R10 is there to protect U2 from an accidental short of the output. C12 and C13 provide simple decoupling. The output is fed to the outside world through a female BNC connector.
The amplifier is built in a metal box. Power is routed via Mini-HEX9 connectors, one for input and another for loopthrough. You may choose to use a different connector. The EPL Engineering Department selected the MiniHEX9 in another era when it was easier to come by. To make wiring the Mini-HEX9's easier there are two connectors CONN1 and CONN4. The most noticeable thing about the printed circuit board is the way the ground was wired. The bias circuit has a ground plane for simplicity but the rest of the layout uses a star topology. A ground plane would add a variety of unwanted capacitance and opportunities for noise to traverse the board.
The negative feedback path connections from S1 to U1 are very sensitive. Ideally S1 would just be a SPST but the switch needs to be PCB mounted and the authors prefer to also have the option of mounting the board to the chassis via the BNC and switch. To make that work a DPDT switch was used because it provides enough height to allow for an extra nut or washers as a spacer to keep its two front solder tabs spaced back from being flat with the edge of the board.
For manufacture of the case a simple template is provided along with the PCB files. It was made for use on the inside of the chassis. If you need the raw dimensions they can be read off the PCB or Gerber files. Originally we built these amplifiers in individual boxes. Later on it was realized that most installations would ultimately need enough electronics to drive two complete DPOAE sources so we have been making these in a single U1 sized box. Finally, like any other electronics devices, proper grounding of the person doing the assembly is important. The author EEF? has broken many amplifiers in his life because of his failure to wear a ground strap with his favorite wool/artificial fiber sweaters.
Testing:For those readers who have done their fair share of basic assembly this paragraph and the one that follows it are not necessary reading. Before everything else take a minute to inspect the board. Check that all the integrated circuits U1, U2 and VREG1 are oriented correctly against their outlines on the silkscreen. Then check the orientation of the two electrolytic capacitors C7 and C6. Capacitors of this size are not always made with stamped vents in the top. You don't want to get hit by the metal can of a capacitor as it shoots off it's rubber stopper base like a bullet. Look for any shorts made by solder bridges or other debris that might have stuck to the board when you were working. If you are testing this board without a chassis around it make sure the area it is on is also clean. This may all seem obvious but you would be surprised how easily these mistakes can happen.
Now take a meter and check that you have not shorted the power rails. It is important to do this test from twice once with the probes on one way and once with them swapped. If you have added a power connector or power supply you need to check that the board has been connected correctly. Check with the meter set the same way check from the – terminal on C7 or C6 to the -15V on the power supply/connector. Then repeat the process from pin 7 on U1 or U2 to +15V on the power supply/connector.
At long last you get to apply power to the board. Test the 1.5V output of the bias circuit by measuring from the CONN2 pin 2 to CONN2 pin 4. If the output voltage is too high or low this usually happens when R22 and/or R23 are not the correct values. It is also important to check that the bias voltage is not shorted via a solder bridge to the microphone's output, verify no voltage from CONN2 pin 2 to CONN2 pin 3. After this is verified the microphone can be connected with less fear that a wiring mistake will destroy it. The magic smoke inside the microphone costs $90 each time you view it. Then connect the boxes output to an oscilloscope and verify that there is an audio signal out.
If you do not get a signal out first verify that the opamps are getting power by checking for ±15V on the power connections of U1 and U2. If that does not fix it work your way back from there to the power supply looking for the break. If that is all fine do the same process for the ground path. After that verify that the gain goes up/down correctly relative to the position of S1. If the gain is wrong or the signal is distorted then you will need to use the oscilloscope to verify there is a signal at CONN2 pin 3. No signal there indicates a bad microphone or microphone cable, solder bridge on one or more pins of CONN2 or U1. Next examine the signal at U1 pin 3 it should look the same but smaller. If the signal there is good move on to U1 pin 6. If you see a signal and it changes correctly as S1's position is changed check U2 pin 6 for the same. If the signal is clipped then you have R31, R13 or R11 wrong. A distorted or noisy signal mean that you C31 is not installed correctly. If you don't see it or it is weak and distorted check that U1 and U2 are not swapped and that C12 and C13 are soldered correctly. If U2 pin 6 looks good but the output is still weak check that the correct value resistor is in R10. This should cover most common problems cause by a single mistake in assembly.
Future Work:While there are a number of improvements we would like to make to the EPL Microphone Amplifier we believe the most useful improvements would be to the sound source itself. The sound source's acoustic cavities, speakers elements and microphone's. We are constantly looking for improved microphones but there are not many mass produced for the consumer market. We limit our selection to this space by default because outside of it the FG microphone meant for use in a hearing aid looks cheap at about $40 apiece in small quantities. Obviously it is unlikely that any new microphone would be compatible in bias voltage to the current one. We plan to just change R22 and R23 when this happens. We also look for better amplifiers particularly for U1. Those reading the notes on the PCB silkscreen and schematic will notice we have changed this part once recently. It will no doubt be changed again in the future. There are other plans under consideration including using a subrack for an enclosure. That other pieces of EPL hardware can be slotted in next to this one. We used to construct more things in this manner. As with everything else at EPL these devices are experimental and are always going through constant improvement.
How to get it?
|No Associated Items Found|