OP-AMP HEADPHONE AMP SUGGESTIONS NEEDED

Status
Not open for further replies.
Jump to Latest
H

horizonrays

Member
Joined 2010
#1
Last edited:
#2
All these are basically the same, bypolar 15v power supplies, and opamps with their outputs buffered.

Ill ask a few questions.
Do you have a power supply allready?
Do you want to bread board only or will you by a pcb?
What are your headphones? Big diff in requirements for 250ohm units, 32 ohm units, etc.
Tubes, no tubes?

Id check out out the popular pcbs.

The Mini³ Portable Stereo Headphone Amplifier
Ventus
How to Build the CMoy Pocket Amplifier
PIMETA v2 Headphone Amplifier
DIYForums.org
Pete Millett's DIY Audio pages


Then there are the bread boarders:
DIY IRF610 MOSFET Class-A Headphone Amplifier Project
NP-100v12: DIY 12AU7 (ECC82) Tube / IRF510 MOSFET Headphone Amplifier
GWado - DIY CMoy / Grado RA1 Headphone Amplifier
 
#4
night , i use 32 ohms Philips headphone , yew i have power supply , -/- 15v first i will do it on breadboard , then i will etch the pbc , and yes it's just solid state , no tubes , just op-amp and transistors , well you , as far as i understand both headwirez project is same , just different transistors , but rod elitot's design is bit different thanks for the resources , i am looking at them , in the meanwhile , may be some one will be able to tell me exact difference between both schematic .
 
#5
nightanole ,
i was having trouble understanding the circuits how they work , now you posted , yet another more circuit , :headbash:

someone explain how the transistor are working in these circuits ,

the only circuit i understood is headwize circuit , someone explain me the working of ebay circuit and also the rod eliot's circuit .. :yawn:
 
Last edited:
#6
Ebay and rods are almost the same. They are just standard opamp setups that have a buffered output. I think with headphones you dont need a buffer, but some people want them. I just dont see the point of making an amp with 10x more power then what the cans can take. Id rather have a single high current opamp.

So like i said. Standard opamp setup with a +-15v (really +-6v would be enough to blow your cans)

signal comes in, Gets rid of dc offset with a cap in the audio path. Then goes to the opamp for a voltage gain. Then the opamp "plays" into a several K load of the output buffer. The output buffer then takes the load off the opamp and drives the headphones.

Personally I would try a cmoy first. it does everything the other amps do, its just not as complex. Take away the output buffers and lower the power supply voltages alittle and you have a cmoy.


If you want to try something different. Here is another buffered circuit that is pretty simple and ment for your cans.

How to build the Apheared 47 Headphone amplifier for Grado Headphones
 
#7
nightanole said:
Ebay and rods are almost the same. They are just standard opamp setups that have a buffered output. I think with headphones you dont need a buffer, but some people want them. I just dont see the point of making an amp with 10x more power then what the cans can take. Id rather have a single high current opamp.

So like i said. Standard opamp setup with a +-15v (really +-6v would be enough to blow your cans)

signal comes in, Gets rid of dc offset with a cap in the audio path. Then goes to the opamp for a voltage gain. Then the opamp "plays" into a several K load of the output buffer. The output buffer then takes the load off the opamp and drives the headphones.

Personally I would try a cmoy first. it does everything the other amps do, its just not as complex. Take away the output buffers and lower the power supply voltages alittle and you have a cmoy.


If you want to try something different. Here is another buffered circuit that is pretty simple and ment for your cans.

How to build the Apheared 47 Headphone amplifier for Grado Headphones
Click to expand...


actually i want to know what happens in those circuits , after opamp ,
since op-amp amplifies the signal , and send it to transistor as most opamp cannot provide enough current to driver the transistor ,

but in the circuits i posted , two different approaches has been used , ib both circuits , transistors are given signal in two different way from opam, and actually i want to know the difference between them , :headbash:
 
#8
They are not fed diff, they are just drawn diff. Rods and earle's are almost the same. The only diff is how they are drawn.

Rod cap coupled the outputs of the opamp, and dc coupled the outputs of the buffer.
Earle dc coupled the outputs of the opamp, and cap coupled the outputs of the buffer.

Redraw both circuits without caps in the audio paths, and no power supplies. Both circuits will be the same other then rod's bias diodes, and earle's output resistor.

I dont like either one of them because they have caps in the audio path after the signal has been amplified. I like to have just 1 cap in my whole audio path, in between my source and amp to prevent dc offset. Earle decided that a electrolytic the size of a battery would be good for an output for some reason.

I can blow up any set of cans 64 ohm or lower with my unbuffered amb mini3. it puts out 300mw into 32 ohms with 1 opamp for both channels. Why dont you axe their power supply. use your own design, and use their schematic and parts selection for the opamp and gain resistor values?
 
Last edited:
#10
i dont like caps anywhere except maybe filters in feedback, but then i drive all my cans balanced so offset as long as its symmetrical is almost meaningless. i disagree about not needing a buffer, i use very low impedance headphones, so presenting very low output z is key. there are some nice new chips that will do it on their own if implemented well though, so i guess you are right, just depends on components used.

like this iv stage with gain i'm using in my portable dac/headamp, using +/-13vdc from lifepo4 battery pack to power these 2 opa1632dgn, with powerPAD it will output enough current without buffer.

still a very simple circuit, i went a bit crazy with parts selection, but the only cap in the signal path in the whole build is those relcap RTE polystyrenes to filter hf junk; the others are for decoupling the supply. there is another set of resistors underneath the vishay mkp1837 box caps, so there is multiple feedback, one for gain and one for filter. i omitted the output filter to keep damping high as possible

this is not my desig, its an early ackodac iv stage that i have modified to suit my purposes; these SuSy chips sound very well indeed and while the smd zfoils are not as fiddly as i thought, i sure needed a stiff drink before hand
 

Attachments

  • portabuff IV-headamp w smdz.jpg
    portabuff IV-headamp w smdz.jpg
    230.8 KB · Views: 559
  • Zfoil SMD ASMP identical lorez.jpg
    Zfoil SMD ASMP identical lorez.jpg
    135.9 KB · Views: 541
  • TSMP naked smd Z foils on IV closeup.jpg
    TSMP naked smd Z foils on IV closeup.jpg
    323.2 KB · Views: 531
Last edited:
#12
instead of the voltage being measured/referenced against ground '0v' the second wire that would be signal ground is actively driven in the opposite polarity by a matching inverted amplifier, so the headphone cable needs to have 4 wires, 1 for each phase per channel. so often the only reference to ground is in the power supply. then at the driver any errors that exist in both phases/wires are cancelled out (to the degree that the common mode rejection ratio (CMRR) of the circuit/device permits) when they are added together. that can be distortion, noise, interference etc. purists object that also some good stuff like 2nd harmonic distortion, which is pleasant to the ear, is also cancelled out and makes it sound sterile, i disagree.

so as a result you end up with a transmission line that has double the voltage swing (only if the power supply is doubled to drive the second phase), very low distortion and very good immunity to noise. i find that low level detail is preserved to a higher degree and that bass has more slam, an overall cleaner sound to these ears. then you also have the pass SuSy designs but you can search the pass forum for that, this is what is called a balanced single ended amplifier and that is the type of circuit that the opa1632 opamp in the board i posted uses.

as a result, with my portable headphone rig, not only does it sound brilliant, but i'm never bothered by that horrible cell phone interference when someone takes a call or sends a text, its cancelled out

hope that helps
 
Last edited:
#13
horizonrays said:
actually i want to know what happens in those circuits , after opamp ,
since op-amp amplifies the signal , and send it to transistor as most opamp cannot provide enough current to driver the transistor ,

but in the circuits i posted , two different approaches has been used , ib both circuits , transistors are given signal in two different way from opam, and actually i want to know the difference between them , :headbash:
Click to expand...

OK...I'll take a stab at answering your question- Ther are more experienced people out there as I'm new to all this, but this is what I see-

All the schematics you posted are class A/B output.
So all use the NPN/PNP complimentary resistors.

The main difference I see is the amount of capacitors and filtering by the "added" circuitry of the more complex looking designs.

The other important difference I see is the amount of diodes...and one uses an LED (light emitting diode- which I add the def just in case).

The diodes cause current and voltage to continually keep the transistors active so you never have one fully off. This reduces crossover dstortion.
One uses 2 diodes, one uses one diode the other uses an LED. I have had success using all combos. So play around to find out what you prefer...Class B can sound suprisingly good if sensitive high speed transistors are used. Class B is when the transistor closes completely until enough voltage opens it...this ommits the need for the diodes and the two bases of the BJT's are connetced with nothing in between.

I have done a circuit for larger 2 way speakers- 10"woofer x 1"tweeter- 90db. I have driven this speaker directly from op amps. LF353 and TL082 were able to drive the speakers louder than you would like sitttin' with your ear to 1" from the speaker. So smaller transistors will only help and larger power transistors are not necessary- something that can be heatsinked will work fine...unless you do pure class B...then less heat is generated.

Just get a decent hfe and audio grade. Darligntons are ok...though I prefer straight BJT- richer smoother sound- at least I think. I've had the best sound with TIP41/42. MJ200/ 210 give decent sound and are smaller. I've driven smaller speakers with 2n3904 and 3906.

Another great sounding combo is NTE11/12- even in class B. They are high current and have a max ic of 5amps, but a low power dissapation.
(I know NTE is hated- but I haven't found the cross reference) Class B will give off very little heat. Just do a straight wire connecting the 2 bases of the BJT's.

I also keep my opamp designs very low gain. 1K - 1K....5k -5k on the feedback resistor/ gain adjusting resistor. Using +/-12 with a 2Vrms input results in 10V pk/pk. So I only need to amplify current This leaves me with 8.5V pk/pk after going throu the output BJT stage using .7ohm resistors on the emitters. This max volume results in zero clipping.

With an 8 ohm speaker I get 6.6V rms, 818mA rms wich translates into 5.39W rms. So I need to be sure everything can handle peak to peak ratings of 17v, 2.1amps. Then everything will be safe at full volume and will result in zero clipping. Class B will result in very little heat.

One thing that is important is the resistors on the on the emitters. This will help prevent thermal runaway should the resistors overheat. Though the risk is low at such low power, the last thing you want is to destroy your cans...especially when wearing them.
 
Last edited:
Status
Not open for further replies.
Top Bottom