Amplifier for small voice recorder module

Thread Starter

jack33

Joined Dec 31, 2010
42
I have a small voice recorder module based on the ISD1820 chip. It's supposed to drive an 8Ω, 0.5W speaker. It works but the speaker volume is not very high. I thought I could possibly use an LM386 audio amp to increase the volume. I tried the headphone amp circuit by Tracecom shown in this thread:

http://forum.allaboutcircuits.com/showthread.php?t=73725

I connected the speaker + output from the recorder as the + input to the 386 through the resistor and capacitor and the - speaker output to ground. The recorder and 386 are fed by the same 5V power supply. The circuit is on a breadboard.

The circuit does amplify the voice but there is also a very loud buzzing noise.

Is there a way to clean up the background noise?
 

Thread Starter

jack33

Joined Dec 31, 2010
42
Thanks MrChips. I tried the 47μF and also tried a 100μF capacitor across the power supply, but I still get the noise. Is it because it's built on a breadboard?
 

thatoneguy

Joined Feb 19, 2009
6,359
Add both a 22uF (electrolytic) and 0.1uF (ceramic) across the V+ and GND lines on both the recorder IC and LM386, as close to the chips as possible.

The breadboard can be causing the problem if you are using long jumper wires, for on breadboard wiring, you should use/cut solid core wire to length so the wires are all flat on the board. If you use the 12" stranded jumpers with 0.1" pins crimped onto each end, you have a bunch of essentially antennas sticking up. These will pick up many sources of EMI interference, often 60Hz hum from line frequency, nearby equipment, etc.
 

tracecom

Joined Apr 16, 2010
3,944
I have a small voice recorder module based on the ISD1820 chip. It's supposed to drive an 8Ω, 0.5W speaker. It works but the speaker volume is not very high. I thought I could possibly use an LM386 audio amp to increase the volume. I tried the headphone amp circuit by Tracecom shown in this thread:

http://forum.allaboutcircuits.com/showthread.php?t=73725

I connected the speaker + output from the recorder as the + input to the 386 through the resistor and capacitor and the - speaker output to ground. The recorder and 386 are fed by the same 5V power supply. The circuit is on a breadboard.

The circuit does amplify the voice but there is also a very loud buzzing noise.

It's also possible that your input level it too high; maybe you need some attenuation at the input of the LM386 circuit.
Is there a way to clean up the background noise?
I am still using that little amplifier between my MP3 player and a 4 Ω speaker. I use a 9V battery and find that it works well until the battery voltage gets low, in which case there is a noisy output. The circuit can be noisy if the component leads and/or wiring is too long. Try to make your assembly as tight as possible...no antennas.

It's also possible that the input level is too high; try an adjustable attenuator on the input to the 386 amp.
 

Thread Starter

jack33

Joined Dec 31, 2010
42
Thanks to all for the suggestions. I added 22μF caps across the power supplies for the 386 and the recorder. I previously only had 0.1μF caps. I added a 10kΩ pot to adjust the input attenuation and another to adjust the gain (in place of the 1.2kΩ resistor). I also cleaned up the breadboard connections.

With the negativerecorder speaker output connected to ground the background noise was gone, but the voice was distorted. I also could not get much volume gain and sometimes after the message finished there was a constant, periodic clicking noise. Then I tried disconnecting the negative recorder speaker output (left it floating). The positive recorder speaker output was left connected to the 386 input through the pot and 1μF capacitor. Now I get good volume gain, and better voice quality, but there is a bit of background hum but not as bad as before. I don't need great audio quality, but I would like to improve it a bit more, if possible. The best I get is with the attenuating resistor removed from the input and not turning the gain too high.

Would there be any benefit to using separate power supplies for the recorder and the 386? If so, should they have a common ground?

Radiohead, I couldn't see how to get the schematic for the amplifier mentioned in your post.
 

thatoneguy

Joined Feb 19, 2009
6,359
Please post both a schematic with all components you are using, and a clear, bright, in-focus photo of your breadboard with connections.

.PNG is preferred for schematics and photos, there's no loss in .PNG for zoom levels, so it makes it easy to "inspect" various areas.
 

Thread Starter

jack33

Joined Dec 31, 2010
42
Here is a schematic of my amp circuit. I tried a better speaker (a small 8Ω, 0.3W Sony) and got better sound quality. It's not bad now, but I wouldn't mind improving it as much as possible. I need to borrow my daughter's camera, and then I will send a photo of my setup.

Recorder Amp.png
 

Attachments

THE_RB

Joined Feb 11, 2008
5,438
An LM386 works fine from a regulated 5v supply, and will have a nice clean sound, it's just not that loud (with an 8ohm speaker).

Re the "buzzing noises" I think your amp is picking up digital noise from the digital sound module. You can try a low pass RC filter on the input with a cutoff about 8kHz or so.
 

Thread Starter

jack33

Joined Dec 31, 2010
42
An LM386 works fine from a regulated 5v supply, and will have a nice clean sound, it's just not that loud (with an 8ohm speaker).

Re the "buzzing noises" I think your amp is picking up digital noise from the digital sound module. You can try a low pass RC filter on the input with a cutoff about 8kHz or so.
I added the RC filter ( R=200Ω, C=0.1μF). The voice quality is pretty good and loud enough for my puposes. I think the buzz is somewhat diminished but still audible. I tried varying R over the range off 100 - 1200Ω in the filter, but I couldn't detect any difference in the buzz.

My camera is not working. I'll post a picture of the setup as soon as I can borrow another one.
 

THE_RB

Joined Feb 11, 2008
5,438
What frequency is the "buzz"? Is it actual noise, or an aliasing artifact? Does it occur only when the sound is playing? Can you 'scope out if the buzz is present on the audio signal or if it is on the PSU lines?
 

Thread Starter

jack33

Joined Dec 31, 2010
42
What frequency is the "buzz"? Is it actual noise, or an aliasing artifact? Does it occur only when the sound is playing? Can you 'scope out if the buzz is present on the audio signal or if it is on the PSU lines?
Yes, the buzz only occurs when the sound is playing.

The buzz appears to come from the recorder module. I made a blank recording without any voice input or substantial background noise. When I attach the speaker directly to the recorder without the 386 amp, I hear the buzz but at a lower volume. I think the 386 is just amplifying the buzz from the recorder.

I don't have a regular oscilloscope. I have an old Oszifox Penscope which displays on a PC. After the blank recording, I tested the power supply and the speaker outputs from the recorder and 386 amp. I didn't see any noise on the power supply. I didn't see the noise on the recorder output although I could hear it. On the 386 amp output I see a noise signal which varies from about 250 - 500Hz.
 

THE_RB

Joined Feb 11, 2008
5,438
It's probably the sampling noise from the frequency used for the digital recording.

You may not be able to do much about it, but one good tip is to re-record your sound data but at maximum volume, use dynamic compression on the sound too if you can. That will give you better signal/noise ratio.

Otherwise if you show a scope trace of the sound with noise steps visible we might be able to suggest a simple filter of some sort.
 

Thread Starter

jack33

Joined Dec 31, 2010
42
I have some pics of the breadboard setup and the oscilloscope trace of the noise signal. I tried some different sampling rates with the scope and got some different results. The one I attached looks like the noise is more in the 2kHz range.

IMG_0996.jpg

IMG_0998.jpg

IMG_0995.jpg
 

thatoneguy

Joined Feb 19, 2009
6,359
Does your scope have 2 channels? If so, what is does the output trace look like when you have both connected to Recorder board Speaker Out, and put the scope in ADD mode and in subtract (Add invert of B)?

The audio record IC has differential outputs, I'm wondering if the noise would be cancelled using both of them?

When you connected the speaker directly to the IC, did you hear any buzzing?

The circuit shown, you are only using 1/2 of the output signal to feed the amplifier. You could use the differential signal to connect to differential input of the amp to see if noise is removed (provided you don't hear noise from the speaker when driven by the IC).
 

Thread Starter

jack33

Joined Dec 31, 2010
42
Does your scope have 2 channels? If so, what is does the output trace look like when you have both connected to Recorder board Speaker Out, and put the scope in ADD mode and in subtract (Add invert of B)?

The audio record IC has differential outputs, I'm wondering if the noise would be cancelled using both of them?

When you connected the speaker directly to the IC, did you hear any buzzing?

The circuit shown, you are only using 1/2 of the output signal to feed the amplifier. You could use the differential signal to connect to differential input of the amp to see if noise is removed (provided you don't hear noise from the speaker when driven by the IC).
My scope only has one channel.

I did originally try connecting both outputs to the 386, but it didn't work at all.

I do hear the buzzing when I connect the speaker directly to the recorder IC. In that case I do connect both outputs. I get the buzzing only when I am playing back the recording. If I do a blank recording (no direct sound input) I just get the buzzing. If I do a voice recording the buzzing is there in the playback along with the voice.

There must be some problem with the recorder. I will try to examine it more closely to see if I can find some defect. I can't find a specific datasheet for the ISD1820 chip, but I have found one for a related chip (ISD1810) that I think I can use as a guide.
 

thatoneguy

Joined Feb 19, 2009
6,359
Can you easily change the resistor for the sample rate (sizes listed in similar datasheet) to see what effect that has?

The trace you show is a higher frequency than "buzzing", can you get the amplitude and duration between two spots where the output is flat for a bit? That would give a better idea of the noise, and if you can do an FFT on that sample, it may help ID the source better if there's an obvious spike at some fraction of sampling freq.
 
Top