# Amplification to a constant Vpp

Discussion in 'Power Electronics' started by marcusaurelius, Dec 10, 2017.

1. ### marcusaurelius

8
0
Dec 13, 2016
Consider this like a voltage regulating circuit.

Think of an amplifier where the input is a sinusoidal, perhaps not in the small signal range. And that we want to amplify this signal to a certain Vpp, say for example vpp = 2. To make this a bit clearer, if the input is a sinusoidal with an amplitude of vpp=20 mV, the circuit yields vpp=2 V and if the input is say, of vpp=50mV then again it yields vpp=2 V , perhaps with a bit error.

This is kind of like regulation, except this is done is done to a sinusoidal, of varying amplitude and frequency.

Assuming the variation is not huge, say not from 10 Hz to 10 GHz or 10mV to 100 V, what sort of implementation would you suggest? I am considering to use mosfets and bjts, opamps, and p-n junction diodes, maybe some DC regulators. Is it possible to build what i speak of using such components?

Can't wait to hear your opinion.

2. ### Audioguru

3,389
732
Sep 24, 2016
A signal limiter or compressor keeps the signal level from exceeding a certain amount. A Wien Bridge oscillator has its compressor circuit rectify the output signal and uses the resulting DC to turn on a Jfet attenuator at the input. It is designed to produce very low distortion, not clipping.

3. ### hevans1944Hop - AC8NS

4,615
2,154
Jun 21, 2012
Let's see if I understand what you are requesting... you have a sinusoidal input "signal" that varies in frequency and amplitude, and you want to produce a sinusoidal output "signal" of the same frequency but with a constant peak-to-peak amplitude. In other words, whatever information might have been present as variations of the peak-to-peak amplitude of the sinusoidal input "signal" is discarded as a result of a creating a fixed peak-to-peak sinusoidal output. However, whatever information might have been present as frequency variations of the sinusoidal input "signal" is preserved as frequency variations in the sinusoidal output signal.

Is that about right, or am I missing something here? If my paraphrased description is correct, then you can accomplish this by phase-locking an amplitude-stabilized sinusoidal oscillator (such as the Wien Bridge oscillator mentioned by @Audioguru in post #2) to the variable-frequency input signal.

"Piece of cake" as we say here in the States. But why do you want to do this? What is the application? What are you trying to DO?

Harald Kapp likes this.
4. ### AnalogKid

2,515
718
Jun 10, 2015
Search for audio AGC schematics or audio compressor circuits. There are hundreds out there, most of them variations on the same overall method.

Amplifier stage with a voltage-variable element somewhere in the gain-setting feedback loop or the input

Amplifier output drives a peak detector and lowpass filter

Peak detector drives the voltage-variable resistance.

ak

5. ### Audioguru

3,389
732
Sep 24, 2016
Why does Electro Schematics dot com always make every schematic wrong? The NPN transistor has its collector load wrongly negative instead of positive.

6. ### Harald KappModeratorModerator

11,662
2,697
Nov 17, 2011
Not huge? That's a factor of 10^9 in frequency and 10^4 in amplitude. This is quite a challenge. I doubt that you'll find a simple or even acceptably complex circuit to cover these ranges.

7. ### AnalogKid

2,515
718
Jun 10, 2015
quote:
*not* from...

Harald Kapp likes this.
8. ### hevans1944Hop - AC8NS

4,615
2,154
Jun 21, 2012
Not so fast, Herr Kapp! Ve haf here our newest invention: an hyper-organic fullerine molecule encasing a proprietary high-technology thingamubob that will possibly satisfy the OP's wequest, just as soon as details of the application are posted. Available now, but only in Avogardo's Number quantities at an insane price with preliminary (not guaranteed) specifications.

Harald Kapp likes this.
9. ### Harald KappModeratorModerator

11,662
2,697
Nov 17, 2011
I may need new glasses

O.K., so if really not huge, what are the ranges?

10. ### AnalogKid

2,515
718
Jun 10, 2015
Voltage is stated, freq is not.

ak

11. ### hevans1944Hop - AC8NS

4,615
2,154
Jun 21, 2012
Still waiting for why, what the application is, and what the OP wants to DO... NOT holding my breath though.

12. ### Audioguru

3,389
732
Sep 24, 2016
I hate the effect of an audio compressor on pop music when it causes the background music volume to pump up and down with the beat. Also when a TV news announcer is talking on the street then his voice level disappears when a loud truck drives past due to the audio compressor.

13. ### hevans1944Hop - AC8NS

4,615
2,154
Jun 21, 2012
Yeah! Second that objection! Those cheap skates should look into real-time adaptive audio signal processing with dedicated DSP integrated circuits and programming. This IS the 21st Century! OTOH, I am still waiting for an AI commercial whacker for TV and radio reception... maybe something that would switch the audio over to a favorite music clip instead of the commercial, with automagical fade-out and resume normal programming when the commercial noise has ended.

14. ### davennModerator

13,866
1,958
Sep 5, 2009
I have done a tidy up, deleting a bunch of totally off topic posts

If the OP @marcusaurelius doesn't respond within the next 24 hrs with more info as requested, I will close the thread

hevans1944 likes this.
15. ### marcusaurelius

8
0
Dec 13, 2016
Hello there!

Thank you all for your replies, they provided me with a good starting point.

First off, we say piece of cake here too^^ Although as I am on a ... kind of a diet... I've been trying to avoid cake

I am sorry that I haven't answered already, as this is only a part of a term-project that i am trying to build among finals and what not. I still had a lot to figure out before this issue.

So the application is that, this is something like a microphone driver. And yes, all of the information is contained in the frequency variations of the incoming signal, which is confined to a range of about 1 to 5 kHz. I don't need the data in the amplitude.

Basically, the microphone will be receiving an FM modulated sinusoidal input ; but since the distance will determine the amplitude of the incoming signal (and again, I only need the frequency), I had to find a way around this. It might still be possible to finish the project without this step, but I am also trying to learn how these things work.

I found myself a little confused when I read "phase locking an amplitude stabilized oscillator". What exactly does this mean ? I have seen the Wien Bridge oscillator before and used it once, but how exactly could it be helpful in this case? I don't know what phase locking means either. I have never studied oscillators thus far. Maybe in the near future.

And finally, I am not allowed to use ICs or audio opamps. I am only allowed to use transistors, opamps and regulators.

16. ### AnalogKid

2,515
718
Jun 10, 2015

Microphone *amplifier*, not driver. A microphone is a signal source, so it is received/amplified/processed, not driven.

Does the output need to be a sine wave, or can it be a square wave that represents accurately the frequency of the incoming signal. If all you need is an output squarewave, then the circuit becomes much more simple - a basic comparator with 25 mV of hysteresis.

ak

17. ### Audioguru

3,389
732
Sep 24, 2016
The "frequency modulated sinusoidal input" to the mic is called a warble tone, or music with "vibrato".
What is the signal used for?

18. ### marcusaurelius

8
0
Dec 13, 2016
The sound the microphone receives is a signal whose frequency changes linearly, that is what i meant. Is that called a warble tone ?

Anyways, the application has nothing to do with music. The thing works like a radar. You see, after the sound is received from the mic, it will be passed from a mixer.

The sound the mic receives comes from a speaker. We also have access to the input signal of the speaker. In the mixer, these two signals (the one coming from the mic and the one coming from the speaker) are multiplied. A simple analysis shows that, since the frequency changes linearly, and since the signal that mic receives is only a time shifted version of the input signal to the speaker, then, we get a sinusoid with a constant frequency after the mixer (and some simple filtering). The frequency of the signal at the very end contains data to measure the distance between the speaker and microphone. I hope this was all clear because it sounds a little confusing as I read.

I can provide more details but I avoided doing so as not to bore everyone to death. If you have a time and desire, I can submit actual requests and diagrams. The whole thing is a relatively simple circuit based on FMCW Radar.

19. ### marcusaurelius

8
0
Dec 13, 2016
The reason that I want to keep the amplitude constant is that frequency is the only thing that I need from the mic's signal. This is not necessary, but I am also aiming to learn this well.

20. ### Alec_t

2,969
804
Jul 7, 2015
Then presumably you only need to detect zero crossings of the signal, once the signal has been filtered to remove noise outside the frequency band of interest?