simple headphone amp

comment pleas, I know there're few mistake..... :roll:

http://img491.imageshack.us/img491/3283/headphoneamp4xm.jpg

Looks pretty much o.k. to me, although I just got up :green:

Except you are shorting out your signal to U2a---you need an R between the noninverting input and the half-supply.

The absolute polarity is inverting, which some people care about (the so-called Wood effect), and if you wanted to avoid this you could make the first op amp noninverting.

I would a.c.-couple into the volume control as well as into U2a to avoid scratchies due to offsets. And then you will need to a.c. couple into the input since it is sitting at half-supply. If your sources are always floating and have no large d.c. offset in their outputs, you could tie your input source common to the half supply, but probably better just to put a cap in series with the input.

I would isolate the noise filtering of the zener rather than just have the big cap across---zeners are quite low Z so the filtering action of the 100uF will stop at a relatively high frequency. A naked zener with then 1k to a cap to common at the input of IC5 pin 3 would be effective.

Mik,

Actually, you not need this "half voltage reference" IC5 buffert, because in this design with only two op-amps, it is ok with a passive divider done with two equal resistor around 10k.
And you get rid of one op-amp and the zener diode that can generate noise, as Bcarco say.

1. Take away IC5 and replace zener D1 to a 10k resistor and change also R1 to 10k. (C1 placing is right)
Now you shall have the junction between C1, R1 and the new "D1" 10k resistor direct connected to "IC3ba" and "U2a" pin 3.

2. Next thing, fix the "shorting out" fault, that also Bcarso talk about.
Take away and move the R205, 100k resistor so you insted connect it in series with the "halv voltage ref. rail, that goes to "U2a pin 3

3. You are finish, and have a more simple and working design.

BTW, If yo not have use for a inverted (virtuell earth) input as a sum-amp or similar, you can leave out "IC3ba", because it have no gain, when the R2, 47k resistor are in series with the input signal.
Connect the input signal direct to the R17, 10k volume potentiometer.
(in this case you also be rid of the "phase reversed" signal from input to output

--Bo

thank you Bcarso, and Bò Hansén, I've corrected the schema, but i'm not sure to have grab the concept of the "wood effect" bcarso mentions..
wath the trade off to use an invertig source coupled with a noniverted one ? I meen in sopund..
BTW now first Op Amp it an noninvertig one ,
Yes Bo you right, I don't really need ic5 , I've short it out , I need a buffer for isolation of the network because the signal come from the main amp and go to the control room amp and vu meter amp, so I need to avoid the interaction between the volume controls of the different stage,
( pleas take a look at the second schema to get a complet vision of the situation)
mabe a voltage follower is more apropriate for this porpouse, it should ?
i'm not sure yet of the positions of the electrolittincs between the 2 ics
othe question is obuot the input impedance of IC3b because the signal got very low source directly out fron other buffers, do I've to put some R in the input ?

haead phone amp corrected:
http://img325.imageshack.us/img325/7357/headphoneamp1pn.jpg

control room amp corrected
http://img325.imageshack.us/img325/4229/controllroom4lb.jpg

thank you very mutch :thumb:
Mik.

Mik,

I have only take a quick look on the first schematic, and there are still some misundrestanding.

1. R205 must be conected on the +/positive side of C204, next to the pin 3 on "U2a"

2. When you use pin 3 as a input on "IC3ba" you must even here have a 100k resistor in series with the "half voltage ref. rail" (same as "U2a")

3. If you not need any gain in "IC3ba" you can take away C4, C10, R2, R10 and link between pin 2 and 1, and you have a "unity gain" input buffer.

4. You can also leave out R15, (link) and change R204 to some hundred ohms. (good to have this also on the first op-amp input, helping to reduce RF/radio distrubance)

--Bo

thank you for the patience Bo
1) ok, now I've understund the concept of the voltage reference. it need to be the same for all the IC, and isolated one from each other. I've dimensioned the first ic as voltage follower.

2) the vistrual ground created by the voltage reference is a ground ? I meen R205 will "shunt" in it ?


3) R205 was there to limit voltage going in the amp in conjunction with R204. if I've to lower this value because the amp go into distortion , do I will have to lower R2 also?


4) How may importance Has got the Value of R2 /R205, should it be say 22K or 10 K

thank you BO

Yes, now it is very near ok, only some small things.

1. Move R204 to between the potentiometer wiper and C204

2. Do C8 bigger, approx. 1 to 10uF. (same as C204)

(maybe also C7 and C203 are a bit stingy in the value, and for headphone drive, the C6 need to be around 470uF, and C9 not need)

Check also the quiescent current trough the output transistors, if you not have them on a heatsink, you shall have approx. 10 to 20 mA through the transistors.
You can adjust this in two way.

Cut up the wire to the Q1 collector leg, and insert a "DVM-meter" in dc mA mode.

1. Change D4 to a resistor, between 10 to 100 ohm, to get 10 to 20 mA.

2. Use D3 and D4 as orginal, and instead change the R210 and R211 equal between 1 to 5 ohm, so you get 10 to 20 mA.

If the transistors are month on a heatsink you can have more quiesent current, maybe up to 100 mA. (the amp will working in very good "small signal" class-A)

--Bo

corrected one. thank you Bo

I'd add a 1 nF between R15/C8 to chassis for RFI protection--otherwise looks fine now.

Samuel

done, Samuel , maby you can aswer to my questin about the voltage reference:
in a big sistem with many ICs the V reference got to the same for all ICs ?
M.

other quastin:
if the signal cames directly from the secondary of an output trasformer, is C8 needed ?
M.

In a big sistem with many ICs the V reference got to the same for all ICs?

Click to expand...

Not sure I understand your question correctly--do you ask if it is OK to use the reference voltage set by R1/R3 for several opamps? If so, I'd say yes--but for big systems (e.g. a console above 8 channels) it is IMO wise to use dual supplies anyway.

If the signal came directly from the secondary of an output transformer, is C8 needed?

Click to expand...

Yes--but I see what you are thinking. Note that you have an unbalanced input. To make an output transformer working you need to ground the unused side of the secondary. This forces the noninverting input of the input IC to zero volt, which is bad news for one half of your signal.

Samuel

Thank Samuel for quick repply.

for real the signal cams from an ICs ( the INA2134 ) that macke the balanced siglal from Main Amp unbalance, the signal then it goes to the control room amp to the headphone amp and to the VU meter amp in a spilt way. all the amp have a single anded PSU +24V



About the V reference I Meen: It cams from the voltage divide R1/R3 and then through R2 100K to pin 3 in the first IC, the same for the second, in this particulary case we have 2 ics biased by the same value.
If I would give the same reference at all the sistem , do I have to use the same value of 100K , or this value is just an arbitrary one and I should use, say 10K ??

This portion ( the Headphone amp ) is part of an entire sistem, and I think imporant to understend this concept :?
M.

If I were you I would take advantage of your "extra" op amp and make a differential input stage out of it.

[quote author="bcarso"]If I were you I would take advantage of your "extra" op amp and make a differential input stage out of it.[/quote]

but I do not need, siglals there are unbalanced... :roll: you probably mean to take the signals directly from the main amp, that in this case is a Neumann V475 amp card..... this is the first time I'm try to plan a circuit as complicated as this ( yea, for you is just not ) , and with your help, I've lerned how to maka a passive summ. ( thank you steve and NYD )
Now I've to put all tougether because, and i need to make a box with a control room, mono stereo option, head phone amp , and a pair of VU Meter.
but I'm still obsessed by the reference voltage......

thank you all for the help
:thumb:

Balanced ins though might help you with noise rejection if the headphone amp is any significant distance away and you are sharing power supply rails and grounds.

thank you Bcarso :thumb:
Mik.

How about I make it easy for you?

http://focus.ti.com/docs/prod/folders/print/tpa302.html

[quote author="Svart"]How about I make it easy for you?

http://focus.ti.com/docs/prod/folders/print/tpa302.html[/quote]

:shock: ah!ah!ah!ah! :green:
and now tell me is it worth it ?

I built a rack of headphone amps out of the TPA302 ICs and they sound just fine for monitoring. they aren't for critical listening, but you probably shouldn't be mixing through headphones anyway!

:thumb: