Problems with ISD4002 audio recorder

[Fearsome]

I'm doing a project with a ISD4002 audio recorder. An electret mic is preamplified and the signal is fed to the ISD4002. The signal out from the ISD4002 is fed to a LM386 amplifier and 8 ohm speaker.

The schematics are here:



Evereything works but I get a lot of low frequency noise, or humming. Here is the output from the ISD4002. The large spike at 1.3 kHz is me whistling.



Now I have the everything on an solderless breadboard (called experimental board?).

When measuring on the supply I get this:



30 mV/Div. A 40 Hz interference that I can't decouple away. The ISD4002 is the source of this interference, that much I've concluded. It propagates to the mic somehow and of course then is recorded. The high frequency stuff superimposed on the low frequency pulses are too high to matter here. ISD4002 has cut off at 3.4 kHz.

So I would appreciate some advice on how to get good sound quality from this device. Anyone been able to get good sound from the ISD4000 series? If so how did you do it? Do I need to put everything on a PCB? Any ideas are welcome!

/Fred

 

[Audioguru]

Each wire on the experimental breadboard picks up mains hum and other interference. Make the circuit on a compact pcb in a grounded metal enclosure.

The wires from the microphone must be very short or use shielded audio cable.

Do you need such a high amount of gain?

 

[Fearsome]

Thanks for your reply.

Yes, following your advice would help. But I'm just curious if it is possible to get decent audio quality on a experimental breadboard. I'm not hoping for perfect here, just something where the sound isn't completely drowned by the low-freq humming.

Anybody succeded?

I will try rearranging and shortening wires and see if that helps.

Regardning the gain (I assume you mean the 17 dB gain of the speaker amp) it's just a note I made of the maximum gain I could use without exceeding the specs of the amp. A gain considerably smaller than that is used.

 

[Audioguru]

The mic preamp has a lot of gain, the ISD4002 has a lot of gain and the LM386 speaker amp also has a lot of gain (26dB, not 17dB).

The total gain is very high at 4000 times. Your circuit feeds the unregulated 9V directly into the input of the mic preamp so it could easily cause motorboating low frequency oscillation.
Add an RC decoupling network at the input of the mic preamp like this:

 

[Fearsome]

Thank you! That worked very nicely! Motorboating gone, only white noise left. Before trying the RC decoupling, I connected the mic directly to the ISD4002 and the motorboating also disappeared. But the signal was too weak and it was almost drowned in the (white) noise. So maybe it is the biasing of the preamp that is the most sensitive, even though I only use a low gain (~x3).

Regarding the overall gain of the design, I wanted it to have some flexibility. It is how hard to know how it sounds before trying it out. I also wanted to amplify the input to the ISD4002 rather than its output, since I thought this would give me better noise performance.

Any ideas on how to reduce white noise?

 

[Audioguru]

The ISD4002 IC is old, simple and cheap. So of course it is noisy and has a limited bandwidth. If you reduce its high frequencies so that speech is hard to understand then its noise also will be reduced.

Have you heard an underwater throat microphone? It produces vowels, groans and growls only. It takes you a few years of training to understand speech from it. Maybe the ISD4002 IC is designed to record a throat mic.

 

[roberto.s]

Hi, this is my first time in a forum, sorry for my intrusion but I have a problem with the same device, ISD4002.

In my case the ISD do not recive any command.....
When I sand a PLAY command, it would sand me an interrupt but it do not appen....
why?

I did a lot of test but anyone work.

I'm using a PIC18F2520 as Master....

For the connection I followed the example on ISD datasheet;

for the PIC I'm using a C18 Compiler and the attached file is the last sample of firmware....

Anyone can Help me????
Please.....

 

[erikodinsvej]

Audioguru write "The ISD4002 IC is old, simple and cheap. So of course it is noisy and has a limited bandwidth. If you reduce its high frequencies so that speech is hard to understand then its noise also will be reduced."

my question is : do you have new devices you can recoment ?
I have used ISD1420 and ISD4002 a coble of times and they work fine
(I only use line in - not mic)
I have also bay a ISD5116 but not yet recieved the device.

On Winbond homepage they say that the ISD4000 series will continue.

- but if you know better chip I will take a look on them.

Best reg. Erik Nielsen

 

[Audioguru]

Sorry.
I do not know of a better recording IC.
A few years ago I found some defective MP3 players and I used parts from a few of them to fix one. It recorded perfectly on a 4GB tiny hard drive.
New MP3 players record perfectly on a memory chip.

 

[sonosun]

Audioguru write "The ISD4002 IC is old, simple and cheap. So of course it is noisy and has a limited bandwidth. If you reduce its high frequencies so that speech is hard to understand then its noise also will be reduced."

my question is : do you have new devices you can recoment ?
I have used ISD1420 and ISD4002 a coble of times and they work fine
(I only use line in - not mic)
I have also bay a ISD5116 but not yet recieved the device.

On Winbond homepage they say that the ISD4000 series will continue.

- but if you know better chip I will take a look on them.

Best reg. Erik Nielsen

how do you record with line in,without mic

 

[Audioguru]

how do you record with line in, without mic

You attenuate the line level signal to a mic level signal with two resistors that make a voltage divider.