06-06-2015, 11:45 AM
PS: "Watts RMS". Tsk Tsk 
Still, 5W isn't far off. The peak voltage might be 6V, depending on the rail, and depending on the losses in the transistors. The EF configuration will drop a minimum of 0.6 to 0.7V, and a bipolar 555 isn't quite "rail to rail".
But of course, the exact power will depend on the duty cycle of the second 555, with a maximum at 50% (which they obviously don't achieve in the default astable configuration). And I suppose that this should be multiplied by the duty cycle of the first so that we can derive a "continuous average power" rating.
OT ramble:
As a philosophical point, it's easy to hung up about the voltage drops in an EF output stage. In a high quality circuit, they are normally higher because, amongst other things, you use Darlington pairs. Many years back, I designed an amplifier for my first car, and I decided that I needed to squeeze every available millivolt from the 13.8V I had available. The circuit used a topology that - looking back - was really very innovative to drive a common-emitter output stage that could get to within half a volt of each rail - it was extremely "voltage-efficient". There were 4 of these, as the output was bridged, so I needed to keep the component count down. But sadly, as a result of that, it did have a major flaw - the quiescent current varied (very!) dramatically with supply voltage! I started making progress with a solution, but then did a "reality check". Exactly how much was my complicated and troublesome circuit gaining me? Let's say I was getting 12.8V pk-pk from mine, while a TDA2005 would get 9.8V (can't remember exact numbers - this was 20+ years back). Basically, it's a 2dB difference. Can we hear a 2dB difference? Probably, depending on the circumstances. But perhaps not... So, I went for the TDA2005 solution. And guess what? There was space in the box for 4 of these chips - each of which was bridged - so I was able to have a four channel amplifier, and that more than compensated for the 2dB loss in each. I made a PCB at work, and it did sterling service until I got a new "head unit" with a 4 channel bridged amp built within. Next time I'm in the attic I must dig it out for old times sake
End ramble. Really must go out now...
(06-06-2015, 08:29 AM)Nick Wrote: This drops the power dissipated in the speaker to about 5W RMS
Still, 5W isn't far off. The peak voltage might be 6V, depending on the rail, and depending on the losses in the transistors. The EF configuration will drop a minimum of 0.6 to 0.7V, and a bipolar 555 isn't quite "rail to rail".
But of course, the exact power will depend on the duty cycle of the second 555, with a maximum at 50% (which they obviously don't achieve in the default astable configuration). And I suppose that this should be multiplied by the duty cycle of the first so that we can derive a "continuous average power" rating.
OT ramble:
As a philosophical point, it's easy to hung up about the voltage drops in an EF output stage. In a high quality circuit, they are normally higher because, amongst other things, you use Darlington pairs. Many years back, I designed an amplifier for my first car, and I decided that I needed to squeeze every available millivolt from the 13.8V I had available. The circuit used a topology that - looking back - was really very innovative to drive a common-emitter output stage that could get to within half a volt of each rail - it was extremely "voltage-efficient". There were 4 of these, as the output was bridged, so I needed to keep the component count down. But sadly, as a result of that, it did have a major flaw - the quiescent current varied (very!) dramatically with supply voltage! I started making progress with a solution, but then did a "reality check". Exactly how much was my complicated and troublesome circuit gaining me? Let's say I was getting 12.8V pk-pk from mine, while a TDA2005 would get 9.8V (can't remember exact numbers - this was 20+ years back). Basically, it's a 2dB difference. Can we hear a 2dB difference? Probably, depending on the circumstances. But perhaps not... So, I went for the TDA2005 solution. And guess what? There was space in the box for 4 of these chips - each of which was bridged - so I was able to have a four channel amplifier, and that more than compensated for the 2dB loss in each. I made a PCB at work, and it did sterling service until I got a new "head unit" with a 4 channel bridged amp built within. Next time I'm in the attic I must dig it out for old times sake
End ramble. Really must go out now...
