Am 12.10.2012 17:05, schrieb Harold Hallikainen:
> I have not been following this thread closely, but here are a few comment=
s
> on the latest schematic ( https://www.circuitlab.com/editor/#4275j4 ):
>
> 1. Because the load is "ground" referenced, you're going to get quite a
> bit of current through the voltage divider formed by R17, R18, C7, and C8=
..
According to analysis the highest amount of current will be 95.74 =B5A. I=20
consider this low. Please elaborate what you are talking about.
> Since you have a DC blocking capacitor in the output (C5), there's nothin=
g
> preventing you from running the op amp single supply. To do this, I'd
> ground the negative side of the battery, then connect the junction of R17
> and R18 to the non-inverting input of the op amp. C7 and C8 can then be
> substantially reduced in size. In fact, I'd probably just do a 100nF from
> the non-inverting to ground and 100nF from the positive supply to ground.
> The inverting input will be biased to 1/2 supply by the feedback resistor=
..
> A single capacitor from the non-inverting input to the input resistors
> (R24 and R27) can provide DC blocking for the two ground referenced
> inputs.
I did the changes. But it is better if you do them because that avoids=20
mistakes on the communication part. Current schematic:=20
https://www.circuitlab.com/circuit/ywsp6g/update4/

>
> 2. The net named combined will ideally have no AC voltage. It will be the
> output voltage of the circuit divided by the open loop gain of the op amp=
,
> which is normally quite high.
Changed.
>
> 3. Is there a reason to use two resistors (R12 and R29) instead of just o=
ne?
Yes. I usually use the resistors I have instead of a in-between value as=20
I intend to build this circuitry. Resistors I have are in Ohms:=20
1,10,47,100,150,220,330,470,1000,1500,2200,4700,10000,22000,33000,47000,100=
000,220000,470000,1000000,2000000=20

>
> 4. The load resistance here is quite low (less than 10 ohms). The
> frequency response on the low end will be limited to 1/(2*pi*R*C) where R
> is the load and C is C5.
Intended frequency range is 10 Hz to 20 kHz.
> 5. The low load resistance will probably cause current limiting of the op
> amp, clipping the output voltage. The datasheet shows a current limit in
> the 20 to 40mA area, so the peak output voltage would be in the 200 to
> 400mV area before the amplifier current limits. There are amplifiers with
> higher current limits. There are also amplifiers that are designed to
> drive speakers that have higher output limits.
Requirements specified in new schematic.
>   It's also possible to add a
> buffer at the output. This could be a complementary emitter follower with
> the feedback coming from the two emitters. Another clever current boost
> circuit I saw years ago used a PNP transistor on "top", and an NPN
> transistor on the "bottom". The top transistor had its collector going to
> the output, the emitter to positive supply, and the base the the op amp
> positive power pin. There was also a resistor between emitter and base. A=
s
> the op amp drew current through its positive power pin, the current
> through the emitter-base resistor of the top transistor would eventually
> reach Vbe of the top transistor, causing current to go through the top
> transistor, pulling up the load. A similar circuit handles pulling down
> the load.
Include it in the circuit, I can't see your picture.
>
> 6. One way to get rid of C5, it cost, board space, and frequency response
> limits, is to use a "bridge tied load." A pair of amplifiers are built,
> one outputting the inverse of the other. With no AC input, the DC outputs
> are the same. The load is tied ("bridged") between the two outputs. As AC
> is applied, one output goes up, the other goes down, putting double the
> voltage across the load.
Include it in the circuit, I can't see your picture.
>
> 7. 9V batteries have very low capacity. It won't drive this for very long=
..
What does "not very long" mean to you? What alternatives do you suggest?

Thank you for your time. I really appreciate it.

Alex
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.