How to control Peak LED sensitivity in this circuit?

Hello,
I got a fully working VU meter module from Tascam 112 tape deck.

Analog meters sensitivity can be controlled indenpendently by potentiometers R103/R203 and this settings does not affect the peak leds.
My question is how to change the peak leds sensitivity? For now they lights up when signal on input is about 0.480V. I want to be able to adjust it so they will light up on any other level, for example +3dB (1.094V).
My first shot was to manipulate with R107 and making it a bigger value. This negatively affects whole circuit sensitivity, which is not acceptable 🙁
Is there an easy way to do that? I assume the resistors R108 and R109 determines the input voltage for Q101 transistor (2SC6203F).

Here's the complete schematic diagram of the circuit.
Any help is appreciated 🙂
4593615600_1669985934.jpg
 
You may either increase R109 or decrease R108. Computing is easy: the thereshold level would be R109/(R108+R109) × [voltage_at_pin_8_of_U101].

You may try a preset at R108 wired as variable resistor (wiper to one end) or rheostat of, say, 5K and try at eye the point you want.
 
So I enterd the default values from schematic:
1670177222316.png

It says 0.787V but it's not true. I've measured the voltage when the peak diode lights up and it starts to light when 0.48V is on the input of the opamp, to be precise Pin 1 of the J003 connector.
 
Your voltage divider app is working properly.

There are several possible reasons for the performance discrepancy. (A good first test would confirm that the voltage at pin 2 of the opamp is about 0.787VDC in your equipment.) When you report threshold is 0.48V, what sort of instrument are you using to measure? A DVM on AC volts? A scope? Other? Please advise details of your measurement method.

Another complicating detail is circuit speed. C103 and R110 (predominantly) slow response speed, so peak values of the signal won't persist long enough for the LED to light. If you have a scope, study U101 pin1 re input signal for insight. Let us know if this doesn't make sense to you.
 
  • Like
Reactions: Kudlatydm
Agree with @BSST. You are measuring mean or RMS values wirh a DMM. But circuitry takes "attention" to positive peaks only. Speech and music waveforms aren't symmetrical, thus you can also see that reversing the polarity of the source will also change the level of the rudimentary indicator.
 
For measurement I'm using a digital multimeter on AC volts which can correctly read voltage from 10Hz to 50kHz.
Also I've connected the oscilloscope to output to confirm that my measurements are not fake. So when the signal from generator is set to 1kHz with 0,775V output the DVM correctly reads the voltage, also the frequency is showed correctly on it.
On oscilloscope I got the same values on Vrms which is 0.77V where Vmax and Vmin are 1.09V and -1.09V frequency is also 1kHz.
 
But your circuit looks to peak voltage, not RMS one.
The opamp is open loop. So its gain is quite large. When the applied signal's peak voltage is bigger than the fixed value setting with R108 and R109, its output goes to a top supply voltage. Thus diode 104 conducts and make base of T101 positive saturating it and closing led path to ground. It is a very simple device.
Any signal 25mV over the level of resistance divider will light the led no matter its waveform.
 
  • Like
Reactions: Kudlatydm
Output voltage on pin1 of U101 is 9.5V when 0.775V is on its input. After D104 we got about 2V.
I've messed around with resistance of R109 and by making its value higher I can control when the Peak Led wil light up. So this is sorted out.
I got question about C103, R110 and R111. This looks like a lowpass "T" type filter. What is the delay when the led should light up? 1 second? Or more?
 
Osvaldo's description is good. Output before and after D104 is best viewed with a scope; Opamp output will show slew-rate dynamics. The scope should show diode conduction following opamp on peaks (about 0.6V lower). Scope will show stored voltage on C104 when D104 stops conducting. The C104,R110 time constant is in the order of 1ms, so it will respond quickly visually re human perception.

From your description, it seems like the circuit is behaving as it should.