Oscillator basics

[Ryan]

symbol (the lamp, or its replacement, a 200k resistor and a
capacitor).

I'm confused.

Do you mean that the substitute for the lamp is a 200K resistor along
with a capacitor?

I thought from your earlier post that it might be replaced by a resistor
1/4th that of the feedback resistor.

The + input gets DC voltage only from ground through its 200k
resistor, since its 200k resistor to the output has a capacitor in
series that blocks DC.

Ok, I follow so far.

So the + input has to average the whatever
voltage you are calling ground.

It is about the same as "ground" because there is no current, is this
correct? There would be voltage drop across a resistor if there is
current draw, but assuming a high impedance in the op-amp input, there
is no draw, and therefore no voltage drop?

The - input averages the same
voltage, as long as the output can influence it through the 750 ohm
resistor. If the - input was some other voltage, the difference
between it and the + input would be amplified by the opamp and the

In other words, the - and + inputs need to be the same to begin with,
otherwise at rest, before it starts oscillating, it would amplify the
difference, which would cause it to amplify the difference more, like
the screeching feedback of a microphone on a P.A. system?

output voltage would change, changing the - input voltage in the
direction that moves it toward whatever is on the + input. If the
oscillator does not start, then the - input and output should all be
at the ground voltage, to match the + input.

The difference of 0, amplified is 0.

I hope you haven't blown the opamp up by having the power applied
incorrectly, earlier. If you open circuit the 1K lamp replacement,
both inputs and the output should sit at ground voltage. This is a
pretty fair check that the opamp is functional, since the only way the
opamp could know where ground voltage is through the 200k resistor
connected to the + input, and it has to be working to watch that
voltage without changing it.

I don't follow.

By this, you mean I should remove the 1K pot? If so, wouldn't the
difference between the inputs be null since the - is not part of the
circuit? In other words, it wouldn't have enough parts to be activated?

So during my test, I should read 0 volts flatline between circuit ground
and the inputs/output.

~

Is the lamp what starts this oscillating? Am I correct in thinking that
it starts the device in motion because at first when the lamp is
powering on, it is "consuming" the voltage, therefore the - input is
lower than the + input. Once it starts, the resistor/capacitor network
keeps it going?

 

[John Popelish]

Ryan wrote:

symbol (the lamp, or its replacement, a 200k resistor and a
capacitor).

I'm confused.

Do you mean that the substitute for the lamp is a 200K resistor along
with a capacitor?

There are three components with ground symbols on them. They are
listed above.

I thought from your earlier post that it might be replaced by a resistor
1/4th that of the feedback resistor.

Yes. That is the lamp replacement.

The + input gets DC voltage only from ground through its 200k
resistor, since its 200k resistor to the output has a capacitor in
series that blocks DC.

Ok, I follow so far.

So the + input has to average the whatever
voltage you are calling ground.

It is about the same as "ground" because there is no current, is this
correct?

Yes. The + and - inputs, ideally draw no current.

There would be voltage drop across a resistor if there is
current draw, but assuming a high impedance in the op-amp input, there
is no draw, and therefore no voltage drop?

You got it.

The - input averages the same
voltage, as long as the output can influence it through the 750 ohm
resistor. If the - input was some other voltage, the difference
between it and the + input would be amplified by the opamp and the

In other words, the - and + inputs need to be the same to begin with,
otherwise at rest, before it starts oscillating, it would amplify the
difference, which would cause it to amplify the difference more, like
the screeching feedback of a microphone on a P.A. system?

Not quite. If there was a difference, the negative feedback would
pull the - input closer to the voltage on the + input.

output voltage would change, changing the - input voltage in the
direction that moves it toward whatever is on the + input. If the
oscillator does not start, then the - input and output should all be
at the ground voltage, to match the + input.

The difference of 0, amplified is 0.

And if the amplifier gain is infinite, that is the only difference
that will sustain any non saturated output voltage.

I hope you haven't blown the opamp up by having the power applied
incorrectly, earlier. If you open circuit the 1K lamp replacement,
both inputs and the output should sit at ground voltage. This is a
pretty fair check that the opamp is functional, since the only way the
opamp could know where ground voltage is through the 200k resistor
connected to the + input, and it has to be working to watch that
voltage without changing it.

I don't follow.

By this, you mean I should remove the 1K pot? If so, wouldn't the
difference between the inputs be null since the - is not part of the
circuit? In other words, it wouldn't have enough parts to be activated?

If you pull the 1k pot out and connect nothing in its place, the opamp
has maximum negative feedback (way more than is needed to prevent
oscillation). So all the amplifier has to do is make the - input
match the + input. If it can do that it is probably still good.

So during my test, I should read 0 volts flatline between circuit ground
and the inputs/output.

That is the test.

Is the lamp what starts this oscillating? Am I correct in thinking that
it starts the device in motion because at first when the lamp is
powering on, it is "consuming" the voltage, therefore the - input is
lower than the + input. Once it starts, the resistor/capacitor network
keeps it going?

The lamp forms a voltage divider for the output voltage so that the
signal measured by the - input is only a fraction of the output
voltage. When the filament is cold, this reduction in the negative
feed back is great enough to let the positive feedback (through the RC
network ) dominate, and oscillations start echoing around the positive
feedback loop. Once the amplitude reaches some magnitude, the current
through the negative feedback divider gets high enough to warm up the
filament, and its resistance goes up enough to increase the fraction
of the output voltage that appears at the - input. When the negative
feedback exactly equals the positive feedback, there is exactly
nothing left to amplify, so the echo stops growing and just repeats at
a constant amplitude. In this state both inputs see essentially the
exact same fraction of the output wave, so this is all the bigger the
wave can get. If it somehow got any bigger the lamp would heat up
even more, and the resistance would go up more, and the negative
feedback voltage divider would supply an even larger fraction of the
output wave to the - input, making it large than the wave reaching the
+ input. This would cause the output to add a negative version of the
wave to the wave, canceling part of it and reducing the output
amplitude. This way, the variable resistance of the lamp stabilizes
the amplitude of the oscillation.

--
John Popelish

 

[~^Johnny^~]

On Fri, 18 Jul 2003 01:52:06 GMT, Ryan <quakeserver149@yahoo.com>
wrote:

I'm trying to repair my first high power car audio amplifier, and then
hopefully two or three more.

They should be outlawed.

Reasons:

a) They pollute (noise pollution)
b) They cause undue distraction for those both inside and outside
the vehicle
c) They hamper hearing of outside events from within the vehicle
d) My windows rattle to the beat of mariachi polka, when the mexicans
drive by my house at 3 AM in their cars and SUVs with those
ridiculous one-note-bass "cannons" or "subwoofers" and 600 watt
amplifiers.


They should be outlawed. Period.
--
-john


~~~~~~~~
"The first step in intelligent tinkering is to
save all the parts." - Aldo Leopold
~~~~~~~~

 

[~^Johnny^~]

On Fri, 18 Jul 2003 03:43:33 GMT, Ryan <quakeserver149@yahoo.com>
wrote:

knowing a whole slew of stuff.

....like diagonal clipping
(what you'll get with 'slew rate limiting')

<g,r,d>
--
-john


~~~~~~~~
"The first step in intelligent tinkering is to
save all the parts." - Aldo Leopold
~~~~~~~~