SMSL Sap II too loud

[thekarma]

Hi, so shorted the pins and cant tell any difference.

I assumed the OPA2604 is used for gain and the TPA6120a is used as a buffer... could it be the other way round?

Thanks

 

[Mooly]

No, the TPA6120 is a dedicated headphone driver IC.

Assuming the configuration is how you suggest... and that is shown in the diagrams I posted earlier... then I think you have to look at changing the input attenuator around, which is no bad thing tbh.

 

[thekarma]

So I did try reducing the 30k (R19) to 5k... little to no change. Put it back to 30. Then changed the 1k (R20) to 15k... again no noticeable change.What sort of values should I be looking at?

I really appreciate all the help. This is all new to me.

 

[Mooly]

No problem 🙂

Look back at the first diagram in post #18 and first of all re-check that the basic configuration seems correct. In particular that the audio input from the sockets is applied via the cap and series resistor to pins 3 and 5

Assuming its as in the diagram, can you tell me what the marked values are of the resistors corresponding to Rin, Attenuate, Rf and Rg in that diagram.

 

[thekarma]

Assuming it is the same.

Rin = 1k
Attenuate = 30k
Rg = Rf = 10k

 

[Mooly]

So overall then, those values give a gain (as near enough as doesn't matter) of 2. If its way to loud then you need to alter the two input resistors.

Something to bear in mind is the effect of the 1uf coupling cap. That currently gives a -3db point of around 5 Hz (how low in frequency it can work to) so anything we do needs to preserve that.

If you made Rin say 22k then that gives a lot of freedom to select an appropriate value of 'Attenuate'.

For example 22k and 8k2 would give a gain of around 0.5. If that was still to much you could drop the 8k2 further.

 

[thekarma]

I'm going to need some more resistors! If I set Rin to 30K as I have plenty what sort of attenuate value would I need for 0.5 gain? Thanks

 

[Mooly]

30k is fine for Rin. Attenuate would need to be 10k to give an overall gain of 0.5

 

[thekarma]

Cool I'll give that a try! Thanks

 

[Mooly]

Sounds like a plan 🙂 You can make the 'attenuate' resistor as high or as low as you want.

If by chance this doesn't have the desired effect, then the circuit is going to be different to how we imagine it... and that would need a re-examine of the layout and a careful redrawing of it. As it stands, its a classic configuration though, so

 

[thekarma]

Ok, progress! So I can get to about 10o'clock before it gets too loud now vs 9o'clock before. I did put the RF 10k back (instead of just a short). So maybe put the 1k back in place of the 10k RF?

Is there a formula I can use for the Rin/Attenuate? 🙂

 

[thekarma]

Sweet! Thank you. It gets too loud at 2 o'clock!... I'll probably leave it there. Just a thought, are there any preferred good ratios/values that would produce the best SQ. The values that I have are just the resistors/combinations I had.

I'm running Rf 1k, Rg 10k, Rin 30, Attenuate 7.5...

Thank you so much, is there no rep button on this forum?

 

[Mooly]

There's no magic formula simply because all operating conditions are different. Your headphones could be very sensitive (produce a lot of sound for little voltage) and also the source component could be feeding a lot of voltage into the amp.

So its trial and error I'm afraid. The values you quote are giving a gain of around 0.22

Thank you so much, is there no rep button on this forum?

Nope 🙂 but perfectly acceptable are cheques made payable to ..... 😉 We enjoy doing what we do.

 

[hajosattila]

Green board?? 😱