EXPERIMENTAL
Design of a sound to light effects Circuit.
DESIGN:
First experiment and gather data considering the biasing of the mosfet.
" * " = arbitrarily chosen value.
--------------------------------------------------------

*VCC = 8v.
*VGQ1 min. = 2v.
*VGQ1 max. = 6v.
*VSQ1 off = 1v.
VDQ1 off = 8v.
*VDQ1 on = 6v.
*RDQ1 = 68 ohms
IRDQ1 on = calculates to be 30 mA.
RS1Q1 off... calc.to be 33 ohms.
RS2Q1 off... calc. to be 220 ohms.
*RG1Q1 = 470 ohms
RG2Q1 calc. to be 160 ohms.
*RL1 = 100 ohms.

Now some thing to consider is that when Q1 is conducting RDQ1 causes a heavy positive voltage due to the small resistance across Q1 so VSQ1 rises from 2v. to around 3v. because of the parrallel resistance between RS2Q1 and RDQ1. But since VGQ1 is at 6v. the mosfet is able to continue conducting properly.
This is breadboarded and it measures out to around 6v. at the gate, VGQ1.


Mosfet stage
--------------------------------



Now that is the voltage needed to turn the output ON. (LED ON)
Now to turn off the output (LED OFF) the VGQ1 needs to be around 2v.
Calculations are as follows.
With 2v. VGQ1 then 6v. is dropped across RG2Q1, which makes a total current of 37.5mA.
Now 2v. will drop across RG1Q1 which gives a current of 4,25mA. Therefor the remainder current will flow through Q2 which is 33.25mA. (IEQ2)
*VBQ2 = 1.5v.
VEQ2 = 0.8v.
So REQ2 calc. to be 24 ohms.
*RB1Q2 = 240 ohms ( 10 times REQ2.)
RB2Q2 calc. to be 1K ohms.


Q2 stage
-------------------



Now with the VBQ2 established at 1.5v. then the VCQ3 is 1.5v. so make VBQ3 lower for proper biasing at around 1v. That makes VEQ3 around 0.3v.
REQ3 calc. to be (VEQ3 / IEQ2) = 9.1 ohms.
Now split the REQ2 into 2 resistors with a 9.1 ohms going to Q3's emitter, and (24 - 9.1) = ~15 ohms to go to Q2's emitter.
--------------------------------------------------------------------------------------------------------------------------------
Q3 stage
----------------


INPUT STAGE
--------------------------------




Now I need to implement a PNP transistor for the inputs of the next stages, because the signal output is a negative going signal.
*RCQ4 = 1K ohms, this is to tie the base of Q3 to ground so it's not floating and keeping any eratic RF noises away from base of Q3.
*VBQ3 on = 2v.
So IRCQ4 on = 2mA.
and I'll make around 2.5v. dropped across Q4 .....*VCEQ4.
REQ4 calc. to be around 2.7K ohms.


Now that is the configuration of the succeeding stages after the first input stage.
Now I breadboarded the input stage plus 3 more of the above stages taking the output of each stage into the base input of the PNP (Q4) transistor, then I hooked up my audio freq. generator and varied the amplitude, as I varied the amplitude the LED's came on one at a time, but they came on to fast one after another.
When I increased the freq. to 500KHz, then as I varied the amplitude the LED's came on one at a time slower the way I want them to.
So I concluded that the Mosfets are picking up voltage spikes as input signal varies, thereby charging the gate capacitance causing the output to go high and stay high longer, before shutting off.
To remedy this I placed for starters a 0.1uF, capacitor across RG1Q1, from gate to ground, to act as a speedup capacitor to discharge the gate capacitance quickly, and experimentally it worked much better, Now at a much lower freq. on the audio generator (2khz) the LED's came on slower one at a time as the volt. amplitude was increased.
So I need to experiment to find the best value of speedup capacitors, would work to discharge the gate capacitance quickly to ground. Looks like the 0.1uF works good with an actual audio input.
--------------------------------------------------------------------------------------------------------------------------------
COMPLETED succesive stages.
------------------------------------------------


I quickly put together a class A amp and biased the collector voltage to be lower than the base voltage of Q3 of the input stage so that the circuit is not prematurely triggered until a high enough signal comes in.

I hooked up my portable CD player with 100mV. output, and a spkr, to this circuit and it works nicely.

All together it works real well, all 4 LED's come on one after another in sequence in accordance to the amplitude of the signal voltage, since there are no filter circuits for this experimental circuit, then all audio freqencies modulate the system.
-------------------------------------------------------------------------
Example of hookup from input stage to second stage....ect...
---------------------------------------------------------------------------------

 

Why are your resistor values so low? They make very high currents in the circuit.
Why don't you drive the VU meter LEDs with an LM3915 IC like every other VU meter circuit?

 

Why are your resistor values so low? They make very high currents in the circuit.
Why don't you drive the VU meter LEDs with an LM3915 IC like every other VU meter circuit?

The reason for the low value resistors, is because I'm not using battery power, but plan on building a linear supply.

My hobby goal with electronics is to design everything, as possible using only discrete components, I feel that is the best way for me to really understand electronic circuit design at the core.
I feel that designing from scratch using discreet components, gives me a better understanding of how and why circuits work, the way they do.

Also this is experimental, (breadboarded) when I build the real thing, printed circuit, then I will design it with 3 of these and RC filters (low and high) pass for low mid and high freq.
The LED's will be 3 different colors also, and a input amplifier,

Then I'll post it in the projects section.

 

There is no sense in using very low value resistors that waste a lot of power.
For example the Mosfet gate uses no current so it can be fed with very high value resistors that do not waste a lot of power.

Oh. You are using a breadboard that has such long wires that it picks up a lot of interference unless the resistor values are very low.

That is why I never use a breadboard.

 

There is no sense in using very low value resistors that waste a lot of power.
For example the Mosfet gate uses no current so it can be fed with very high value resistors that do not waste a lot of power.

Oh. You are using a breadboard that has such long wires that it picks up a lot of interference unless the resistor values are very low.

That is why I never use a breadboard.

I agree but when I was using very high resistors to conserve power for my H- bridges (mosfets) every one was telling me that the resistors were to high, and the gate capacitance in conjunction with these high resistors, will cause the mosfet to stay on in the linear region overheating them,
so that's why the low resistor values to use as pull up and pull down for the gate capacitance.

My breadboard has very few wires, on it, I keep all the components real close to eachother side by side and stretch the resistor leads from component to component, as possible.

It's late now but I'll post a picture of my breadboard work sometime tomorrow.

 

I laugh at photos of breadboard messes with hundreds of wires all over the place.

A Mosfet that drives a very high current motor with high frequency PWM needs its high capacitance gate driven from a low impedance.
Here you are driving low current LEDs slowly so the driving impedance can be much higher.

 

AG - Well I'm thankful for breadboards, why should I waste a whole PCB if I just want to experiment with one thing? I do, however, use a circuit board when I'm building a important project that I will continuously use.

Hobbyist - My compliments on your design, it's great to see people experimenting and trying to understand the core of electronic circuitry! Most people are apprehensive when It comes to seeing such a elaborate schematic and also all of the math that's involved. Where did you learn all this transistor stuff? In accordance to your name, I'm assuming you're not an engineer and you just do electronics as a hobby? What kind of references did you come across to obtain all the information that you have?

 

My breadboards use to look like a spaghetti mess,
but I bought the wire kits, and now I use wires like printed circuit tracks, I keep them as neat as possible flat against the board if I can, and try to keep color coding to so as to trace my circuit paths easier.

I'll look into trying to bring the resistances higher, but I must take into account, that the current output from each mosfet, is feeding the base input of the succeeding stage, and there needs to be enough current to transfer this signal to the 4 stages.
A REdesign may need a higher supply, I'll experiment with this.


Electronerd:

Thankyou.

I used to belong to the electronics book club, and I bought as many transistor books I could find,
then I took 2 courses in NRI (Basic Electronics) and (Electronic Circuit Design), then lastly 1 course from CIE (Electronic Engineering).

But the courses I took showed how to design transistor circuits from a FORMULA type introduction.
But there was really no design techniques that showed LOGICAL approach to circuit design.


But lately I just started looking at designing Basic transistor circuits from a systematic Logical approach. I'm learning how to design these circuits by first, having a thorough understanding of what I want the circuit to perform, then learning required information about components I'm using,

and then taking one step at a time, designing using a calculator, and a good understanding of "OHMS" law, and current and voltage laws, basic theorems, and then applying that with aq good scientific calculator, to designing one stage at a time.

Before when I would make a circuit it used to be MOSTLY trial and error until it worked somewhat.

But NOT anymore, I want to calculate ahead of time what volytage and current should be at a certain place in my circuit, and design using the value of bias resistors required, and if the results measured are way off my calculations, then I'm not satisfied with the design being trhat way and so I'll take the time to investigate why it didn't work out according to my calculations.

When I'm fully satisfied why it works out the way it does, then I'll redesign it using that new knowledge gained from that.

I want to be able to give an answer for every component used in my design, if asked, with the logic, and math needed to confirm the component used, also I want to be able to give what the voltage is at key points in my circuit and why those voltages are present.

To me that is a successful design when it works purposefully and not by some fluke of component tolerances.
I designed this whole circuit using nothing more than ohms law and some voltage, and current divider equations. and basic transistor theory.

I've learned how to get from the FORMULA approach to the LOGICAL approach in my circuit design experiance.

To me that's the real enjoyment of designing circuits from scratch.....

 

I laugh at photos of breadboard messes with hundreds of wires all over the place.

.

Here is the circuit breadboarded for experimental purposes.

 

Your breadboard circuits look like a mess of wires all over the place to me. They have thousands of intermittent connections.

I use Veroboard (stripboard) where each connection is soldered for reliability. The strips make half the wiring of a pcb and the parts and a few short jumper wires make the other half. It is easy to use a solder-sucker to replace a part if it is needed. The circuit is compact so it works at high frequencies where a breadboard cannot. It also does not pickup interference like a breadboard circuit.

 

There is no sense in using very low value resistors that waste a lot of power.
For example the Mosfet gate uses no current so it can be fed with very high value resistors that do not waste a lot of power.

Oh. You are using a breadboard that has such long wires that it picks up a lot of interference unless the resistor values are very low.

That is why I never use a breadboard.

Thankyou,
For the constructive criticism, it made me aware of an improvement I could make to this circuit, with the knowledge of using a capacitor from ground to gate to take care of the gate capasitance charge, I was able now to raise all the bias resistors from the gate through to the voltage peak detector, to conserve more power.

There is a significant change in the current consumption showing on my supply current meter.

Here is the first improvement made to this design.



Thanks again...

 

The Mosfet switches on and off slowly because its gate to ground and drain to gate capacitances are high.
ADDING ANOTHER CAPACITOR FROM THE GATE TO GROUND MAKES IT WORSE!.

 

The outputs come on too fast even when there is a small glitch, once the mosfets turn on there gates are charging through the pull up resistors, I remedied this by placing the capacitor from ground to gate so as to ground the gate and dischrage the internal capacitance quickly, and this improved circuit performance imensly.

The performance I designed this for is that the outputs (LED's) do NOT appear to come on at the same time, but that there is enough delay to give it a sequential display of the LED's rising from the bottom to the top.

Without this capacitor the mosfets stay in conduction to long with respect to the input audio modulating signal.

I built this and see this difference with and without this capacitor.


It is doing nothing more than keeping the gate at ground potential when needed during eratic glitches throughout the stages.

And it is small enough to be overcome with a substantial valid inputto the stage.

This is my understanding thus far concerning gate capacitance in a mosfet, as I'm slowley learning this NEW component in my circuit designs.

Thanks again..