10KHz signal amplification

I want to amplify a 10KHz square wave signal. What Op-Amp and configuration will be be best?

Input signal: 0-1V at 10KHz
Output required: 0 - 4V at 10KHz without any shape loss of signals. Please suggest me what Op-Amp I should use?

I tried with LM358, LM741, TL072, TL084, LM386D but none of the working fine.

Hi,

a square wave has infinitely high overtone frequencies.
Therefore you need a very fast OPAMP if you really want to "amplify" this signal.

If you just want a clean square wave with variable amplitude, then I recommend to use a variable DC voltage and analog switches.

Klaus

I've to read an analog voltage coming serially as a square wave with variable amplitude at 10KHz. As the signal is small, and some signals are too small so I was planning to amplify it a little bit. So that I can measure it easily. Do you have any suggestion?

Hi,

If I understand correct then there are analog voltages from several channels serialized (multiplexed). After each 100us the analog channel is changed and thus the analog voltage changes.

Like:
SYNC | Analog_voltage_0 | Analog_voltage_1| Analog_voltage_2| Analog_voltage_3 ... and so on ...
(usually you should provide such information)

Now my questions are:
* do you want to de-serialze the signals, so that you get n analog outputs for n channels?
* or do you want to convert the signal into digital using an ADC?
* is the signal asynchronous (without clock signal)? then how do you sync it?
* or is there an additional digital sync signal?

***
OPAMP.
As said before a fast OPAMP can amplify the signal. For sure it introduces some delay, maybe ringing, noise, offset...
Therefore you should define your specification, like: expected voltage accuracy, expected settling time, and so on...

Klaus

ok, I have 216 data first and then 168 data next serially. What I'm doing is, reading analog signal using ADC one by one and saving the digital value in a table serially. there is additional signal for synchronization. Once it is saved in the table, its being compared again with previous data. If there is any miss match, a output is generated. Its working fine with what it is but problem is some data are too low to compare in common.

Example:

data patter is as like...

102,240,100,65,50,268,167

then I'm comparing them at level 50. When the circuit is working, data will change to lower value such as...

10,22,10,0,0,26,15 etc (from previous sequence)

But problem is, some data are around 5 to 10 initially and its can not be compared in common. So those are missing from reading. You ma suggest to lower the compare the level to 5/10 but then other data are always triggered.


This is why I'm trying to amplify the signal a little bit so that I can compare it easily.

Hi,

Amplifying with an OPAMP is one method. You may use it.
Do you need for this additions informations?

******
Other solutions:
100us per channel is sufficent time to do multiple AD conversions. Then sum them up or average them. The value becomes more reliable and noise is partely cancelled out. I think this gives better quality than one conversion per channel. Because of cancelling out noise the signal-to-noise ratio becomes better.

In opposite to this: The amplifier method surely amplifies the noise and adds new noise, therefore the signal-to-noise ratio is getting worse.

Klaus

If the signal is "too small" is that not enough information?
Can you not simply deal with that, "binning" everything
below "good enough" range into "not good enough"?

Certainly there will always be a point of "too small" even
with gain in the lineup (which also gains up noise floor,
perhaps yielding no improvement depending on just why
the comparison fails to satisfy).

Slew rate, my friend. Slew rate.

An LM358 has a typical slew rate of 0.3V/us. So it takes 13.3us going up and 13.3us going down (assuming symmetrical SR) for a total of 26.6us total.
That on a waveform whose period is 100us. Of course it will look distorted.

You need to find an opamp with high slew rate.

schmitt trigger said:

Slew rate, my friend. Slew rate.

An LM358 has a typical slew rate of 0.3V/us. So it takes 13.3us going up and 13.3us going down (assuming symmetrical SR) for a total of 26.6us total.
That on a waveform whose period is 100us. Of course it will look distorted.

You need to find an opamp with high slew rate.

Click to expand...

It was a very helpful reply. Ok, I'm trying with high slew rate Op-Amp.

As you already sorted out higher slew rate OPs like TL07x/TL08x, I guess that there may me other unsaid prerequisites like supply voltage range.

TL084 can nicely reproduce a 10 kHz square, but only within its specified input and output voltage range. A "rail-to-rail" OP extends the range, but you still can't exactly reproduce a voltage near zero with single OP supply voltage. In case you used TL084 with appropriate dual supply, there must be a different (also unsaid) problem in your setup (e.g. capacitive load, bad circuit layout).

If you can live with an output somewhat less than 4 V, then this may be an option.
https://www.ti.com/lit/ds/symlink/sn74aup1t34.pdf

I have used these up to 2 MHz, so 10 kHz should be no problem.

then this may be an option

Click to expand...

I presume it isn't. The OP wants to transmit a square wave as analog signal, with continuously variable amplitude.

SN74AUP1T34 is not available in local market. Anyway, I tried with all available one. But distortion is happening. Also there are some other issues involved. Maybe PCB layers are generating a capacitance.

I'm still trying to find what is happening. But using Op-Amps performance has become worse than before. I've to find the actual reason first. So I'm trying what is the problem and why its happening.

Hi,

if you need help you need to give informations.
The above post gives no information, so we can´t help. In the end it is useless for you.

--> Give pictures, timing diagrams, schematics, scope pictures, values, specifications....

Klaus

Any suitable high speed opamp should do the job for you. It sounds like you have some layout/construction problems.