In sci.electronics.design, "a.n" <recycle@no-spam> wrote:
>> I'm gonna go way out on a limb here and say NO.
>>
>> You're asking for a lot of sensitivity. My cheapo RS
>> SLM bottoms out at 50 dB.
>> If you want any kind of dynamic range, the switching
>> will dominate the circuitry anyway.
>> Why only 2-3 transistors? quad op-amps will take up
>> less space, use less power, have more gain.
>> Why no other semiconductor devices? You'll probably
>> want at least one diode in the meter rectifier.
>>
>> Suggest that <3 semiconductors is not the best metric
>> for choosing a design.
>
>You may be right, and I suspect you are, but the reason I ask
>the question is that I have such a circuit, here it is:
>http://www.geo cities.com/jacklearner/101_0185.jpg
First off, the LED's might be better driven with National
Semiconductor's LED Bargraph driver chip, though that only gives fixed
ratios between segments, and you might want something different (which
would probably take a restitor divider string and a comparator for
each LED).
Next, here's an interesting link:
http://www.gti.net/wallin/audio/rsmeter/33-2050/33-2050.html
This thing only uses a single transistor for the front end, then it
uses an op-amp after the scale switching to get enough level to drive
the meter.
>In practice it is almost useless, but can you explain why it is so
>bad?
I think the other poster gave a reasonable circuit description.