[plug] Altering Gains on Multiple Audio Files

[Denis Brown]

At 09:04 AM 2/04/2009, Jason Posavec wrote:
>Daniel Pittman wrote:
>>What he needs is something that generates a line-level output, which the
>>headphone socket can't deliver.  You might get something that had a
>>headphone input on one side and a line-out on the other...
>>
>>Regards,
>>         Daniel
>>
>So how does an iPod do so well, then? Or does that have a line-out?

$0.02 worth and a chance to wander down memory lane from an [mature | old] 
electronics guy :-)

Devices designed to drive a headphone load have their output stages set up 
primarily for current sourcing I think you'll find, into headphone-like low 
resistance loads.   Typically these have their (output stage) frequency 
responses tailored to the matching headphones so some after-market 'phones 
may not sound quite like the originals due to different acoustic / 
electrical characteristics.

Any adapter leads *should* mimic at least the DC resistance of the 
headphones as a load - anywhere from a few ohms to a hundred or so 
ohms.   The output under these circumstances will be clean but probably low 
amplitude and hence require some form of external gain stage (or a suitably 
sensitive input on a main amplifier.)   Here "low amplitude" will most 
likely be excessive for use directly into a microphone input.   In that 
case a voltage divider would be required before going into the mic input(s)

Because of the tailored frequency response, expect to have to make 
adjustments (tone controls, etc) to compensate for tonal quality.

Line output on the other hand is designed to drive loads of 600 ohms or 
greater and typically do not *require* a comparable load resistance into 
which to operate, in order to get clean output because they are voltage 
sources.   They also typically will export a "flat" frequency response 
characteristic.   For this situation a plain stereo 3.5mm plug to RCA or 
whatever, with no loading resistors, is fine feeding the line input(s) on 
the subsequent device.

In summary my take on this is:
1. headphone drivers in the device -> loading resistors to match the DC 
resistance of the intended headphones -> some sort of gain stage or 
sensitive input.   (Voltage divider if going into mic input(s))
2. line drivers in the device -> should not need loading resistors (or use 
600 ohm if you want to be "good" about it) -> line input on the intended 
A/D card, amplifier, etc.

In my travels I have only ever come across one commercial lead (from the 
Panasonic company some 30 to 40 years ago) that had loading resistors.   On 
the other hand I have built heaps of these, with the "fake headphone" 
resistors wired inside the backshell of the 3.5mm stereo plug.

For completeness, headphone-to-mic circuit...

L----+--------- 4.7 kohm ----+----ML   where L = device Left headphone 
output (repeat circuit for right),
       |                              |                     ML = left mic input
    10 ohms                   100 ohm            G = ground
       |                              |
G---+-----------------------------+----G

The 10 ohm resistor should be changed to the DC resistance of the OEM's 
'phones.
The 4.7 kohm and the 100 ohm resistors are subject to change according to 
the voltage level from the device and the required non-distorting input 
required by the subsequent amplifier / A/D card.   Unless you have a 
turbo-charged iPod or equivalent, the 1/4 watt or 1/2 watt resistors from 
Dick Smith, Jaycar, Radio Spares, Farnell, etc will be fine for the 
fake-headphone load.

*Beware* that DC loads at mic input sockets designed to accept 
electret-style microphones *may* result in damage to the electronics on the 
mic-socket side.    This is because electrets require a bias voltage to 
operate, sourced from the amplifier / A/D card.   Usually no damage, but 
better to cover my backside :-)

In such cases insert a (series-connected) capacitor of <insert favourite 
value here> just before the ML terminal above.    I would start with 0.1 uF 
and if the sound is too "tiny" increase that by an order of magnitude and 
go from there.   Tantalum capacitors will work and ideally will be 
correctly polarised; -ve side to the 100 ohm resistor and +ve side to the 
ML terminal    It would be rare to find a -ve output electret.   (Comments 
from the more up-to-date in this regard, please!)

HTH,
Denis