I was studying this snippet from a battery driven Stereo Headphone driver.
It is powered by 8 batteries of 1.5V each, configured to provide +/- 6V with a ground in the middle.
To balance the current draw, the designer cleverly decided to use one LM386 between ground and -6V (The blue one) and he smartly used a cap C209 at the non inverting input.
(https://i.postimg.cc/P568WcMm/LM386-Amp.png) (https://postimg.cc/5XC01sT0)
Question ONE:
In this configuration, should the Zobel Network/Boucherot cell go
A) from Output of Blue LM386 to -6V
or
B) from Output of Blue LM386 to ground
or
C) It does not really matter since both configurations are electronically equivalent
In option B, the Zobel will be in parallel to speaker load, which sounds very reasonable
But the points in support of option (A) and (C) are that V- is supposed to have a low AC impedance to ground any case.
QUESTION TWO:
Currently (Unsurprisingly) both channels of the headphones are connected via DC blocking caps to ground
Further to to the same point re Zobel, can it equally validly be said that even the loudspeaker could have been connected from blue LM 386 via output cap to V- because V- is a low AC impedance point with reference to ground
(except for the foolhardiness of running live wire to speaker that could short to chassis and the OCD of symmetry. Ie if there is no fear of a short circuit, after the DC blocking cap, could the speaker from blue LM386 have been connected to either + rail or - Rail or ground and all three configurations are exactly electrically equivalent ?)
QUESTION THREE:
Is there no need for the equivalent of power bypass C210 for the V- rail ?
rockman x100?
Zobel network is always set in parallel with load (speaker + & -).. :icon_wink:
Quote from: Vivek on March 17, 2022, 07:07:53 AM...cleverly decided to use one LM386 between ground and -6V....
Is that clever?
The LM386 will run a very very long time on 4 cells. Two will still run a long time. Keep it simple and put the second set of 4 cells aside for later.
The preamp before the LM386s is a bipolar design running on +6/-6V.
Hence the need for the LM386 stage to balance the consumption between the two sets of batteries
isn't the 386 totally capable of running at 12v? that would be way less convoluted.
cheers
I think you need to be careful of how loud a headphone amplifier might get for the sake of someone's hearing. Probably easier to lower supply volts than design a limiter.
Quote from: anotherjim on March 18, 2022, 05:40:44 PM
I think you need to be careful of how loud a headphone amplifier might get for the sake of someone's hearing. Probably easier to lower supply volts than design a limiter.
its powered by 8 batteries, just replace with only 4 of em for 6v, and place 2 amps in parallel.
also, headphone impedance can vary from als low as 32ohm to more than 1k ohm, so its kind of arbitrary anyways.
cheers