The link I provided earlier in the week regarding using pin 7 for muting; I don't know if you read it, but he did say that there were only two scenarios where it did mute. One was with the LM386 configured for 26db gain and one for 40something db gain. No longer have the article handy so I don't remember what that latter scenario was.
Vince.
toggle quoted message
Show quoted text
On 08/03/2017 02:34 PM, Jerry Gaffke via Groups.Io wrote:
Good question on what the muting action is like.
I tried simulating in LTSpice using a behavioral voltage source (item
"bv" under the "Component" menu) as a variable resistor:
First with that variable resistor tied between pin 7 and ground,
Vcc=12v, pin2=gnd, pin3=0.1v pk-pk, no connection to pins 1 and 8:
We go from no output (just 11.4v dc) at the output pin with 2.8k of
resistance, to full output at 4.8k,
Between those two points the audio is distorted (flat-topping when it
tries to go above 11.4v).
With a cap between pins 1 and 8, the gain is increased so distortion
will continue well beyond the 4.8k value.
Moving that variable resistor from pin 1 to pin 7, output appears
somewhat distorted at the low end, but not just too ugly.
Minimum output at 0 ohms is between 1.62v and 1.92v, so a minimum
voltage gain of 3.0
At 2k, is between 0.8 and 2.5v, saw an almost linear increase in gain
between 0 and 2k ohms.
By 10k, we're between 2.7v and 4.8v, so maximum gain, from here on out
to 100k it is just a smooth shift in dc offset at the output.
So shorting pin 7 to ground sort of works as a volume control, but
somewhat distorted and you cannot completely silence the output.
Driving pin 7 from Vcc attenuates, but creates very distorted audio.
Better to drive pin 7 toward ground.
Shorting pin 1 to Vcc is a bad idea.
Driving pin 1 towards ground distorts the audio for intermediate values
of R, but can completely shut off the audio to the output pin when pin 1
is grounded.
Many web sources recommend driving pin 7 to ground to "mute". Often
ideal, as you still get low level audio coming through to monitor what's
going on.
But to kill the rx pop to the headphones on the Bitx40 we probably want
to short pin 1 to ground, as we need to completely silence the output.
I've left my LTSpice model with the variable 0 to 100k resistor tied to
pin 1 up in /g/BITX20/files/KE7ER/lm386jg02.asc
Just download and install LTSpice to a Windows box (or under Wine on
Linux), read in my model, and press the "simulate" button
to see a plot of output voltage vs resistance from zero to 100k ohms.
(OK, you need to click on the output pin with the mouse to get it to
show up in the plot.)
Can use the scissors icon to cut the wire to pin 1 in the schematic,
then use the pencil icon to wire the variable resistance to pin 7, and
punch simulate again.
Total harmonic distortion shown in the bottom right of p7 of the NJM386
datasheet looks a little bit better than LM386, but not by much.
Hard to get both low current consumption and low distortion in the same
part.
Jerry, KE7ER
On Thu, Aug 3, 2017 at 06:47 am, KC8WBK wrote:
When the LM386 is muted, is it a sudden, digital switch or is there
a range of voltages that attenuate and eventually mute the audio?
--
Michigan VHF Corp.
