LM386-N1 based amp for Sony NW-HD1 Network Walkman

[quantumtangles]

Schematic and printed circuit board layout I put together for a little LM386-N1 based mono amp.

Improvement suggestions most welcome.

Breadboard photos tomorrow

 

[davenn]

it would help if you had an input connector

as it stands neither the schematic nor the PCB layout show where the input is

also C3, 470uF doesnt need to be that high, 10uF is the common value
you can instead use the 470uF in the output to the speaker C4, instead of the 1000uF that you have in there at present


Dave

 

[quantumtangles]

it would help if you had an input connector

as it stands neither the schematic nor the PCB layout show where the input is

also C3, 470uF doesnt need to be that high, 10uF is the common value
you can instead use the 470uF in the output to the speaker C4, instead of the 1000uF that you have in there at present


Dave

Sorry. If you look at the printed circuit board, the non-inverting input is connected to the upper terminal of the non-electrolytic (ceramic) capacitor marked as C2. This is a 103.

This terminal of C2 (103) is itself connected to Pin 3 of the IC.

The inverting input on the other hand is directly attached to Pin 2 of the IC. Pin 2 is also connected to Pin 4 which is ground.

I tried lower values for the gain cap and they did not work. Terrible distortion, but when I whacked in a 470uF (not much maths going on this end) it worked perfectly

I should also point out the amp is designed to work without a pre-amp directly from the output of a Sony Network Walkman model NW-HD1. Trial and error design using the signal from that device.

Will amend shortly

 

[davenn]

Sorry. If you look at the printed circuit board, the non-inverting input is connected to the upper terminal of the non-electrolytic (ceramic) capacitor marked as C2. This is a 103.

This terminal of C2 (103) is itself connected to Pin 3 of the IC.
The inverting input on the other hand is directly attached to Pin 2 of the IC. Pin 2 is also connected to Pin 4 which is ground.

you are missing the point ... you show the power connections and speaker connections on the PCB layout but NOT the input connections

also with C2 to ground in the input, its not really doing anything. you should be making C2 ~ 1uF electro and your input singnal goes through that cap and then into pin3

I tried lower values for the gain cap and they did not work. Terrible distortion, but when I whacked in a 470uF (not much maths going on this end) it worked perfectly

which sorta indicates a problem somewhere
datasheets are a great thing to follow, if it doesnt work per the datasheet, then you must be doing something wrong

without anything between pins 4 and 8 the gain is fixed internally at 20
you are probably overdriving the input, you should put a trimpot in there to have it as a level control

I should also point out the amp is designed to work using without a pre-amp directly from the output of a Sony Network Walkman model NW-HD1. Trial and error design using the signal from that device.
Will amend shortly

yes, it only need quite low input to drive it fully

Dave

 

[quantumtangles]

I did not mean input connections were shown. I was just explaining where omitted input connections should have been marked Fully agree when you say I did not mark input connections. Just explaining where I thought they should be

Your further comments are interesting. Looking forward to trying them out tomorrow. Really helpful. Thanks for the 'input'.




More later from the 'ready, fire, aim' school of electronic engineering

 

[davenn]

have fun, look forward to seeing the results

 

[quantumtangles]

Here is an amended schematic, amended printed circuit board and fuzzy breadboard photos (the camera on my iphone was great but I took it apart with relish...obviously).

The sound is superb (more by accident than design). Note that I connected the useless 104 cap and attendant resistor not to pin 3 as Dave correctly pointed out, but to pin 8 (Gain).

This was because inappropriate components elsewhere caused distortion when connected to pin 3.

The upshot, for reasons buried in the datasheet, is that the crispness of higher frequency sound improves audibly when the previously useless cap and resistor are connected to pin 8 (where the negative pole of the 470uF electro cap sits, its positive pin just going across the IC to pin 1 (this cap straddles both gain terminals).

The bill of parts makes for a very inexpensive amp for this device.

1 x 0.01uF ceramic capacitor (103)
1 x 0.1uF ceramic capacitor (104)
1 x 10 Ohm 0.25w resistor
1 x 1000uF Electrolytic capacitor
2 x 2-pin terminal block
1 x 470uF Electrolytic capacitor
1 x LM386N1 integrated circuit
1 x 8 Ohm speaker

Just realized pin 3 works great with a higher value resistor. Hand capacitance can cause the distortion so changing resistors when it is working can be misleading. Also works great with higher value resistor in pin 3. The resistor vlalue is determining gain. As the device already has quite a powerful audio output, the resistor value cannot too low as that overamplifies it and causes distortion. Thanks Dave. Would appreciate your views about the pin 3 pin 8 situation, which seems to behave as though it is a filter. Sounds good now.

In any event, I hope the circuit is of interest. I had a lot of fun making it and appreciate your comments

 

[quantumtangles]

I also changed the battery in the Sony Network Walkman ND-HD1.

Be careful when changing the battery. The micro-electronics make this device very delicate. After removing two small screws at the side of the device, the cover slides upwards but the screen remains connected to the main device so be careful not to break the connection.

The problem I had was I dislodged the power connector soldered to the logic board. You should disconnect the old battery by pulling it OUT OF the white connector. Don't pull the white connector from the logic board as I did. I used too much force.

Looking at the soldering, it would have required me to solder two connections, less than 0.4mm apart with pinpoint accuracy. I was also worried about causing thermal damage to components with my soldering iron.

What to do? The solution (to tearing off the main power connector) is to put it back where it was, as accurately as you can, and then turn the machine on to see if it is working.

Fortunately there was some power in the new battery and the machine came back to life.

Knowing the connection was good, but wanting to avoid any soldering, I hot glued the old connector in the precise position I knew was working. Two tiny blobs of hot glue fixed the broken connector in place, and also secured the new battery in position. Wont be changing that battery again any time soon.

Carefully close the case to avoid snapping the overhanging battery wires and it should work fine, but this can be tricky so try not to be as clumsy as me

 

[quantumtangles]

Hot glue for the battery fix seemed to work, but the device stopped working after being charged. Glue moved as it cooled around the white power socket.

Conclusion...imminent microsurgery on 0.4mm terminals.

Arranged two magnifying glasses, one on top of the other to get a good focal length. Desk lamp for visibility. Countless times tried to get the white terminal block in place but it was just too small to solder. Connections seemed smaller than 0.4mm.

Plan B. Removed two pin connector at the end of the battery. No going back now. Exposed ends of the wires (red and black). Red wire has to go to the tiny terminal towards the side of the device. Black wire the terminal towards the center.

The soldering tip looked really massive under the magnifying glasses...because it was really massive. Solder everywhere. Not working. Used more flux.

Eventually, pre-soldered the red wire and then it stuck to the tiny positive terminal. Black wire proved much more difficult because the two terminals were side by side on the logic board and kept shorting. Until that is I thought about it. As a ground connection, the black wire could be soldered anywhere at all on the large metal battery plate (the metal plate the battery sits on).

Both terminals were soldered, and then covered with insulation tape to keep them steady when putting the device back together.

Two hours and three solder burns later, it works.

 

[davenn]

there's some errors in there that you didnt make in your first schematic

1) ... the R1 10 Ohm and the C1 0.1uF in series are supposed to be between audio out and ground NOT power rail and GND

2) ... You still have the 0.01uF between the input and GND ... get rid if it and instead have the 1uF in series between input and pin 3 input

3) ... you have the negative of the battery going to the speaker ONLY instead of there and all the other negative ( GND) points eg Pin 4
It aint gonna work the way you have it

Dave

 

[quantumtangles]

Thanks Dave. I will amend it.

 

[davenn]

Thanks Dave. I will amend it.

thats cool, you will get there eventually
just keep chugging away at it.

I'm more than happy to help people like you who are at least giving things a good go
unlike some who wont even help themselves

Dave