Sgheadphones DIY A47 pocket amp kit

[heady 0]

jason, why can't use the TLE instead of the opamp, won't it be cheaper? Just a thought.

[huangyong 0]

heady, use opamp as rail splitter is cheaper solution than TLE.

btw, my experience with NJM4556 is not good when function as rail splitter, it tend to drift few mV to one side. NE5532 is good for this application.

[heady 0]

huangyong, good to know that opamps cheaper than TLE. The TLE is very fragile to reversed polarity of power, burnt two due to accidental connection. Thanks.

[bpribadi 0]

The main reason is to stabilize the ground, and some op-amp has higher output current than the TLE. Inside the TLE is also resistor voltage divider and an op-amp, so same thing.

[huangyong 0]

can try discrete rail splitter too. make it like 8 PIN dip configuration.

Edited December 7, 2004 by huangyong

[Firefox 0]

yes jason sorry didn't see your sms today

no work today, playing online game whole afternoon

 

i don't quite understand the new board...if the iG & oG is already connected together, what is the third op-amp for?

can the new board work with just 2 op-amps? left & right channel op-amps only, the first version can't work this way

It's used as a rail-splitter, not an isolating ground channel.

 

Resistor dividers will easily drift because of the load.

[heady 0]

Huangyong, is there any advantage of using the discrete rail splitter from Sijosae as compared to using a resistor/opamp setup? (besides the fact that it looks real cool )

Edited December 7, 2004 by heady

[tropicalrips 0]

just finish soldering the new version 2.0 of this amp

 

first off, my opa2132 does not work on the board...strange...will need jason to explain why, there's just lots of distortion with the 2132, i think the chip is osilating badly on the board.

 

i use 4.7k for the rail spilter resistor, not sure if thats the cause.

 

also tried LM6172, used them on left/right + ground, this chip works but heats up...after about 5 mins of play, i remove them...fearing i might send them to opamp heaven

 

next up, opa2228, agian this was the surprise opamp for this board, though it also gets warm, it was not as bad as the LM6172.

 

my final configuration for tonight, opa2228 for left/right + JRC4556 on ground, the 4556 also gets quite warm, but i hav place a piece of ram-sink on it to help it stay cool

 

first impressions of the new board : the sound is an improvement over the first board, the 2228 sounded fast, has good punch, this is without any modification, i also did not use my Alcaps on the input.

but the board don't seem to ba as stable as the first one...i did not hav the 2134 to try out, but shouldn't the 2132 be almost same as 2134? anyway i like how the amp sounds with the 2228, its better on this baord then on the first one.

 

will still need to confirm with jason on the resistor value that i used, maybe i got it wrong

 

starting with R4 & R8, i used 100k,100k - 1.5k,1.5k - 43r,43r

my gain should be more then 6 i think...need to sleep npw

will post pics tomorrow...

[jtfoo 0]

Maybe some output resistors might help? Paralleling opamps without output resistors is not a good idea. It's akin to stacking opamps.

[bpribadi 0]

Tropical, I can pass you my version for your carefull comparison

 

Hopefully the new board sounds better, even though I think the ground channel circuit will easily oscillating and will cause instability

Edited December 8, 2004 by bpribadi

[huangyong 0]

Huangyong, is there any advantage of using the discrete rail splitter from Sijosae as compared to using a resistor/opamp setup? (besides the fact that it looks real cool )

the fact is that it's real cheap. cheaper than some good opamps; TLE, and TLE+buf634.

The discrete buffer's performance is claimed to be as good as TLE + Buffer in term of current output capability, as i rememebr it's around 200mA (not too sure, can somebody confrim this).

Compared to the discrete rail splitter, the TLE + buf634 can output 250mA. but the price difference is huge, around S$2 for discrete rail splitter and $13++ for TLE + buf634.

 

[heady 0]

Huangyong, thanks, understood. Electronics is not my strong point, I only like to listen to music.

[tropicalrips 0]

 

as promised, here's a pic of the new board, the sound has improved, my only concern now is heat from the op-amps, there also seems to be noise being picked up by the board or cables...

[jasonhanjk 0]

Ahh yes. Just came back from cinema.

National treasures is a good show. DO NOT MISS!!!

 

Ok, here's the resistor for this board that I used.

 

R1, R5 = 10

R2, R6 = 91k

R3, R7 = 33k

R4, R8 = 33k

R9, R10 = 10k

 

The gain is (1+ R2/R3)= 4

If you got it wrong I still can spare you one more board. I feel uneasy if this proto don't work and I call for production. It will be a very costly move.

 

For the ground channel, a few configuration.

1) OPA2132BP, the 2 output of this opamp in this single IC is short together. The chip will get warm but it gives the highest output current theoritically. Problem is it may oscillate and hear heavy distortion. It will get very very hot. Try option 3 if it doesn't work.

 

2) OPA134, this single opamp IC can be use with this board by bending the IC leg. Cut away pin 1, 5 and 8. Bend pin 6 to pin 7 location and pin 7 to pin 8 location.

 

3) Adding a resistor between the output pins with opa2134. Under the board, you can see thick traces short between 2 & 7. Cut them and join a 10 ohm resistor underneath. Use this config if option 1 causing distortion and heat up.

 

The measurement I done lately is the improvement in the THD, Total Harmonic Disturbance. Revision 1 has THD of 0.03 but this new one is less than 0.01.

Better then my mp3 player. Problem is that cable is picking up a lot of noise so it cause other measurement to go bad a bit. I am not so worry cos the results are still in the good region.

[skyskraper 0]

sounds like u guys still have a lot to work out with this. if you guys are getting oscillation etc then it'll be easy for kit builders to get it!