Compact headphone amp design

[tuck1s]

I'm looking to build a compact, wired, stereo headphone amp design that will be used for live performance monitoring (in-ears on stage, weekend warrior band).

Quality has to be good, with adequate volume levels, but doesn't need to be audiophile quality. I was thinking of taking and simplifying the Nwavguy output design, for example having only one NJM4556.

I'd like to make the unit small, I think it will be powered by +/- 9v or +/-12v DC feed, and built into a small belt-clip box with volume control.

I have 2x XLR balanced outs from the desk, so - optionally - a differential input stage could be a nice-to-have, although we currently use commercial headphone amps that are single-ended input with no significant noise or hum problems.

 

[abraxalito]

Given typical XLR level feeds you probably just need a buffer, not any form of voltage amplification. That is unless your phones are 300ohms+ impedance types. A discrete transistor diamond buffer would be my recommendation, its power draw (quiescent current) wouldn't need to be very high so a single lithium cell could last many days.

 

[OlegSh]

Hi tuck1s,

Do you plan to run it on batteries? agdr has the smallest complete headamp design that I've seen here: http://www.diyaudio.com/forums/head...y-2x-9v-real-ground-headphone-relay-pcbs.html

Depending on your definition of compact I also offer a small headamp PCB using OPA1622. The PCB size is 3x4cm. It can be used with balanced or single ended input. The volume pot and PSU are wired externally. Details are here.

There is also another compact headamp design from AIM65.

Regards,
Oleg

 

[erikb]

Not audiophile quality? Get a 'headphone amp' IC (Headphone Amplifier | Products | TI.com) and power it with a cheap switching power supply. Use a cap multiplier to get rid of high frequency noise and you're done for less than 2 euro's/dollars.

 

[agdr]

Do you plan to run it on batteries? agdr has the smallest complete headamp design that I've seen here: http://www.diyaudio.com/forums/head...y-2x-9v-real-ground-headphone-relay-pcbs.html

Yep you need one of these. 😀

 

[xrk971]

This would work well and sound is fantastic. World class SE Class A sound. Can be driven differentially too and one amp can drive balanced headphone outs.

http://www.diyaudio.com/forums/group-buys/302859-xrk971-pocket-class-headamp-gb.html

You can build it yourself with a $12 PCB from the GB or fully built and tested ones available too.

 

[needtubes]

Can be driven differentially too and one amp can drive balanced headphone outs.

True balanced/differential output requires four channels of amplification - a positive and a negative phase per channel. I thought this amp was just stereo, in which case, you would need two boards to build a balanced version.

 

[xrk971]

I have tested it with balanced inputs driving right and left and right left as 3.5mm output jack to balanced drive headphone. Works very well. Usual test of not having balanced input causes signal cancellation and no music. Single ended drive (one side only) sounds like usual single ended drive amp. With differential (one phase 180 deg out) drive from digital source via bit flip inversion produces a sound that is very dynamic and on crescendos and transients, sounds very crisp and loud - about 12dB louder I would estimate given the opposing signals an double the drive current. So you need two boards for stereo balanced drive of course. Nice thing is that it's still fully Class A SE output into balanced drive with no issuenof unbalanced DC from killing headphones. But amp boards are cheap if you make it yourself.

 

[tuck1s]

Thanks for all your responses - much appreciated. One point of info - I was considering balanced inputs from the desk XLRs, rather than outputs (which will be standard 3.5mm stereo jack).

We have a couple of pairs of Shure SE215 headphones in the band (20 ohms?), as well as SoundMagic E10 (46 ohms) and 1st generation Sennheiser C300 (16 ohms), plus a drummer who insists on using his Apple EarPods (which I think are around 46 ohms).

I wonder if the Super CMOY variants are a better bet than the SE class A, whose article recommends "30 ohms and up".

 

[jcx]

does your drummer "insist" in American Sign Language?

the 1st 2 have some meaningful sound isolation

but your drummer needs hearing loss sense beat into his head - may be too late for the ears





Shure SE215

Sensitivity 107 dB SPL/mW
lmpedance 17 Ω


SoundMagic E10

Impedance: 46Ω
Sensitivity: 100dB at 1kHz/mW


Apple EarPods ( http://www.innerfidelity.com/images/AppleEarPods.pdf )

Volts RMS required to reach 90dB SPL: 0.091 Vrms
Impedance @ 1kHz: 43 Ohms
Power Needed for 90d BSPL 0.19 mW
Broadband Isolation in dB (100Hz to 10kHz): -1 dBr

 

[sgrossklass]

but your drummer needs hearing loss sense beat into his head - may be too late for the ears

ACK.

Just an idea: Get some earmuffs to be worn on top. Problem solved.

Looks like E10s may be using different drivers these days? Mine measure about 17 ohms DC. Isolation can be improved by making custom foam tips.

With 16 ohm loads, supplies as low as +/-6 V have been recommended for the O2 to avoid overheating, not sure whether a single 4556A would be particularly amused.

I would do something like an O2 with the gain stage replaced by a balanced receiver. The XLRs out of a mixer can be assumed to be reasonably hot and required output voltage levels won't be that high, so I'd look at an input gain of -6 to -12 dB. Expected input clipping levels would be about +17 to +23 dBu, respectively.

Still not sure how "+/-12V supply" and "belt clip" mix.

If you wanted you could do something like this on +5V or even a lithium cell (3.7-4.1 V), that does rather limit your parts choices and your SNR though.

 

[PRR]

> powered by +/- 9v or +/-12v DC

Single supply with capacitor coupling is a lot cleaner than more wires for bi-polar power.

Is 18V-24V even needed?

I will assume 120dB peak SPL, though I think this is dangerous.

Sensitivity 107 dB SPL/mW Impedance 17
0.13V for 1mW
0.58Vrms for 120dB

Sensitivity: 100dB at 1kHz/mW
Impedance: 46
0.22V for 1mW
2.2Vrms for 120dB

Power Needed for 90dB SPL 0.19 mW
Impedance @ 1kHz: 43 Ohms
0.09V for 1mW
2.9Vrms for 120dB

Worst-cases: 2V-3V rms for 120dB SPL
Multiply RMS by 3 to get peak-to-peak
Need 6V-9V total supply.

Yes, with output resistors you might need more than one 9V battery.

BUT 120dB SPL is real frikking loud!

I have used such an amp for recording monitoring. I needed to know that any clipping was in the recording, not the monitoring. I actually came close to 120dB maybe 30 seconds in 5 years. I would NOT want to be working near that level weekend after weekend.

I think the max outputs could easily be half the above voltages and still do real harm. (As known from 3V battery Walkman/iPod deafness.)

In the 4.5V-9V area, NJR-made '386 chips are fine reliable low-impedance audio. And it happens they "can" be rigged differential input, though the gain control gets too messy for me to figure today.

 

[tuck1s]

Thanks all for your comments. i first went to in-ears as stage foldbacks / floor wedges are usually pretty nasty. They add to the general sonic mush on stage, are heavy/bulky, and make microphone feedback problems worse. By "stage" I mean "corner of the local bar".

In-ears solved that for us, and give a useful degree of sonic protection against being near to the cymbals and snare of an enthusiastic drummer. I now can finish gigs, remove the in-ears, and have a normal conversation level with no ringing.

I agree our drummer needs something better than Apple earpods. I will remind him again. He plays other gigs with other bands, with open ears which is unbelievably bad for him (he's only a teenager and so adverse effects / tinnitus are going to last a lifetime for him). We're the *least bad* band he plays for in that respect.

I have one of these lying around unused: X5 PCM2706 + CS4344 +Dual TDA1308 USB DAC HiFi mini Headphone Amplifier | eBay

while the USB in and DAC are no use to me in this application, it's a nice, small box, well finished, with a reasonable output stage, and a pot and everything. If nothing else, it's a good donor box for the case, pot and possibly output stage.

I should be fairly easily able to find the tracks going to the TDA1308s, wire a separate analog + 5v DC input on the back panel, and disable the digital stuff. Several comments have suggested that I don't need wide rails for the low impedance cans, to achieve sensible volumes, which makes sense.

To clarify the belt clip idea - we currently use wired headphone amps on the floor, so we're used to not leaping around too much and remembering to unwire them before walking off!