Voltage amplifier circuit for high frequency (1-5MHz)

Thread Starter

chirasya

Joined Dec 28, 2022
7
Hello,

I am trying to build an non-inverting voltage amplifier to convert 1V to 5V DC signal in the frequency range 1-5 MHz. I have built circuts with two opams,

1. AD84AAN
https://au.mouser.com/ProductDetail/Analog-Devices/AD844AN?qs=/tpEQrCGXCyAbMeiKtSvmg==
2. TL972IPE4
https://au.mouser.com/ProductDetail...XUi1CsIDJ3Q==&countryCode=AU&currencyCode=AUD

Attached are the designs for the circuits. Both circuits operate at low frequency (less than 1MHz), but the signal gets crappy at higher frequency (above 1MHz).
Are there any ways to get a good signal at 1-5MHz frequency ranges with these opams?
 

Attachments

LesJones

Joined Jan 8, 2017
4,174
Even with 100 Mhz gain bandwidth product you will only amplify the first 2 terms of the Fourier analysis of a 5 Mhz square wave. ( sin (x) + sin (3x)/3 )

Les.
 

ronsimpson

Joined Oct 7, 2019
2,989
What must be the approximate value of capacitor?
I don't know how far the wires are to you power supply. I would use 0,01uF and 1uF in parallel from supply to ground on each supply. But 0.1uF will work also. The data sheet shows 0.22uF. Not critical the value.

1mhz sine wave is simple but you want 1mhz square wave which has harmonics above 1mhz.

1) You need an amplifier much faster than you think.
The data sheet does not show information on a gain of +5. If you have a typical op-amp rated tor 50mhz then; the gain is down 3db at 10mhz with a gain of 5. So a 10mhz square wave will look like a sine wave at 10mhz because the harmonics are gone.

2) You need fast "volts/nano second" out put speed.
The AD844 has 2V/nS slew rate. The output cannot move any faster than that. At 5V output it will take 2.5nS to move the 5V you want.

Are you using a bread board? They are not good at high frequencies.
1672323584799.png
 

Thread Starter

chirasya

Joined Dec 28, 2022
7
I don't know how far the wires are to you power supply. I would use 0,01uF and 1uF in parallel from supply to ground on each supply. But 0.1uF will work also. The data sheet shows 0.22uF. Not critical the value.

1mhz sine wave is simple but you want 1mhz square wave which has harmonics above 1mhz.

1) You need an amplifier much faster than you think.
The data sheet does not show information on a gain of +5. If you have a typical op-amp rated tor 50mhz then; the gain is down 3db at 10mhz with a gain of 5. So a 10mhz square wave will look like a sine wave at 10mhz because the harmonics are gone.

2) You need fast "volts/nano second" out put speed.
The AD844 has 2V/nS slew rate. The output cannot move any faster than that. At 5V output it will take 2.5nS to move the 5V you want.

Are you using a bread board? They are not good at high frequencies.
View attachment 284138
Thanks for the detailed reply. I am currently using breadboard to verify if the desired values can be obtained. I will design PCB circuit once the circuit is sorted out. Can opamp LM6171 able to meet my requirements of amplifying 1V to 5VDC square wave at 1-5MHz frequency. I am planning to procure that.
 

Thread Starter

chirasya

Joined Dec 28, 2022
7
I have to drive an acousto optic modulator (AOM) with square wave of 5V. The signal to drive will be generated from FPGA which produces 1V square wave. Therefore, I must amplify signal from FPGA to drive AOM. I can't use function generator, as the FPGA signal generated has pseudo random pattern.
 

ronsimpson

Joined Oct 7, 2019
2,989
I think we are going down the wrong road.

I just picked a part at random. "Logic Level Translator"
This part passes signal either direction. I should have picked a part that only works one direction.
Set Vcca to 1.15V and Vccy to 5V. It will take the 1.0V and make 5V. This part has 6nS delay and 2nS edges.
This type of part is commonly used to get from 3.3V to 5V or 1.8V to 3.3V.
1672327966452.png
Here is a example of a part that only moves data in one direction. 1.15V to 5V. It is fast. Simple to use.
1672328399014.png
There is no "gain". It simply translates digital data from one voltage range to another.
Good luck, ask questions, I need to run to work for a while.
RonSimpson
 

Thread Starter

chirasya

Joined Dec 28, 2022
7
I think we are going down the wrong road.

I just picked a part at random. "Logic Level Translator"
This part passes signal either direction. I should have picked a part that only works one direction.
Set Vcca to 1.15V and Vccy to 5V. It will take the 1.0V and make 5V. This part has 6nS delay and 2nS edges.
This type of part is commonly used to get from 3.3V to 5V or 1.8V to 3.3V.
View attachment 284141
Here is a example of a part that only moves data in one direction. 1.15V to 5V. It is fast. Simple to use.
View attachment 284142
There is no "gain". It simply translates digital data from one voltage range to another.
Good luck, ask questions, I need to run to work for a while.
RonSimpson
Thanks, Ron. I will place the order and let you know if it works.
 

Thread Starter

chirasya

Joined Dec 28, 2022
7
Thanks, Ron. I will place the order and let you know if it works.
Ron,

Can this amplify different voltages? Example, if the input voltage is 0.8V, can it amplify to its equivalent? I will be varying signal from 0.1 to 1 V, this signal must be amplified to the value of 1 to 5V.
 

MrSalts

Joined Apr 2, 2020
2,767
Ron,

Can this amplify different voltages? Example, if the input voltage is 0.8V, can it amplify to its equivalent? I will be varying signal from 0.1 to 1 V, this signal must be amplified to the value of 1 to 5V.
No, it is a digital part that has two different supply voltages. So it is more of a level translator than an amplifier. For example, if your microcontroller needs 3.3v inputs but you are connecting to a device that puts out 0 (low) snd 5v(high) signals, they will be translated into 0 or 3.3v signals for your microcontroller. One supply voltage would be connected to 3.3v and one to 5v. It is not proportional.
 

ronsimpson

Joined Oct 7, 2019
2,989
I agree with Mr Salts. The input will switch at about 1/2 supply. Or about .55V for a 1.1V supply.
Data sheet said (/txu0101.pdf) it will work down to 1.08V supply. I think it will work a little lower if you need. It will get a little slower.
The output will go from supply to supply (0 to 5V) There is no middle ground.

74AXP1T34 Input supply 0.7V to 2.75 and the output supply 1.2V to 5.5V. This might be a better part because you can run the input at 1V or 0.9V if your signal is not as strong as you want. I use the SOT353 package because it can be hand soldered but the other two packages should be machine soldered.
 

MrChips

Joined Oct 2, 2009
30,712
You can use an analog comparator such as LM311 and set the input threshold voltage to whatever you wish.
The output voltage will be whatever supply voltage you chose.
 

ronsimpson

Joined Oct 7, 2019
2,989
set the input threshold voltage to whatever you wish
( LM311) The input voltage does not work when it is too close to the negative supply. So you probably need to have a -5V and +5V supply. The input common mode voltage range is 0.5V from the supply so it will struggle with a 0 to 1V signal when the supply is 0 & 5V.
Also the LM311 will add 100nS delay. Output fall time is 25nS and the output rise time is 100nS with 500 ohm pull up resistor.
If you want to go down this road choose a newer faster 5V part. Also many voltage comparators are 'open collector" output. Choose a part that pulls both up and down. Also choose a part that works well at voltages at "0V". There are parts that work well at -0.5V.
 

LesJones

Joined Jan 8, 2017
4,174
In post #11 you say " I have to drive an acousto optic modulator (AOM) with square wave of 5V. " But in post #15 you say " Can this amplify different voltages? Example, if the input voltage is 0.8V, can it amplify to its equivalent? I will be varying signal from 0.1 to 1 V, this signal must be amplified to the value of 1 to 5V. " which to me implies that the drive is NOT a constant 5 volt amplitude square wave but a square wave with a varying amplitude of between 0 and 5 volts. Can you clarify which of these statements is what you require.
I have no idea what a an acousto optic modulator is so I can only go by what you tell us of it's drive reqirements.

Les.
 
Top