Operation Amplifiers for high frequency application

[pooyan]

Hi,

We would like to connect an amplifier to produce the appropriate alternative current for a high frequency actuating system.
Who here knows where I can find an opamp with the following charachtristics:

1) Non-inverting
2) Amplifies an alternating input current 3 to 5 times
3) Works in the range of 100-200 khz
4) Is under 500 dollars

Please send me the search link as well or direct me if there is a database that let do parametric search on OP Amps.

Thanks.

 

[dlgoff]

Welcome to PF pooyan.
You may find this selection guide useful. http://www.national.com/analog/amplifiers/precision_op_amps"
And this "www.national.com/an/OA/OA-11.pdf"[/URL] seems good.

 

[pooyan]

Thanks!

 

[pooyan]

Welcome to PF pooyan.
You may find this selection guide useful. http://www.national.com/analog/amplifiers/precision_op_amps"
And this "www.national.com/an/OA/OA-11.pdf"[/URL] seems good.[/QUOTE]

Hi Ddlgoff,

Thanks for the links.

Using the first link, I have not yet found which OP AMP can magnify the input voltage 3-5 times (this is the gain in units V/V).

As for the 2nd link, while the knowledge might help get familiar with Op Amp and RF apps, I don't think it specifically helps my problem that is just finding an op amp with the specifications mentioned in the first thread.

 

[Carl Pugh]

Stronglly recommend you get the book
IC Op-Amp Cookbook by
Walter G. Jung

Most OA are inverting or non inverting depending on how they are connected.

What do you mean by Amplifies AC current.
Will a transformer work?
Do you want a voltage that is porportional to a current.
Do you want a current in another circuit that is 3 to 5 times larger?

Use OA whose gain bandwidth is at least 10 times larger. 10X200kHz=2 MHz.
Where does 100MHz in later post come from?

Many OA meet preceding requirements and are under $5.00 each.
LF411CN is not the optimum OA, but should work.

 

[pooyan]

Strongly recommend you get the book
IC Op-Amp Cookbook by
Walter G. Jung

I have one PDF book on OA, and may be looking at it soon but it is really long and most of the time I am not sure what I need to be looking at.

Most OA are inverting or non inverting depending on how they are connected.

so my design is going to be non-inverting

What do you mean by Amplifies AC current.

the input is AC 10-20 Vp-p at 100-200 khz, we would like to amplify to obtain an output AC which is 30-50 Vp-p with the same frequency as input

Will a transformer work?

I am not too sure, but transformer I thought is for very high voltage apps, not an opamp that will feed to a MEMS device

Do you want a voltage that is proportional to a current.

no, Vp-p has to just get larger 3-5 times

Do you want a current in another circuit that is 3 to 5 times larger?

no, the output voltage be 3-5 times larger i.e. for 10-20 Vp-p get 30-50 Vp-p

Use OA whose gain bandwidth is at least 10 times larger. 10X200kHz=2 MHz.

so for 100-200 khz I need to look at 1-2 Mhz? I do not think that is correct.

Where does 100MHz in later post come from?

sorry, that was an error, I removed it. I mean the OA should work with 100-200 khz signals

Many OA meet preceding requirements and are under $5.00 each.

awesome

LF411CN is not the optimum OA,

what is the output voltage I can get from this OA, at 100-200 khz input. It is not clear to me on the data sheet, some other data sheets had Max Output Voltage versus frequency but should work

please see the blue writings

 

[Carl Pugh]

Gain bandwidth of 10 times maximum operating frequency if just a rule of thumb. Actual gain bandwidth depends on how accurate the output signal has to be. 10 is used for low accuracy amplification.

Google "high voltage" "operational amplifier"

The OPA454 costs $6.10 in quantities of 1 and may do what you require. All the specifications for the OPA454 should be checked againist your requirements.

TO DESIGN AND BUILD AN AMPLIFIER FOR THIS APPLICATION WILL BE EXTREMELY DIFFICULT FOR SOMEONE WHO IS NOT FAMILIAR WITH OPERATIONAL AMPLIFIERS.

A LF411ACN has a maximum power supply voltage of +/-22 Volt. So maximum output voltage is about +/-20 volt. It will not work in this application

A transformer may be the easiest/cheapest approach to this project.

Good Luck
Carl

 

[pooyan]

Gain bandwidth of 10 times maximum operating frequency if just a rule of thumb. Actual gain bandwidth depends on how accurate the output signal has to be. 10 is used for low accuracy amplification.

Google "high voltage" "operational amplifier"

The OPA454 costs $6.10 in quantities of 1 and may do what you require. All the specifications for the OPA454 should be checked againist your requirements.

TO DESIGN AND BUILD AN AMPLIFIER FOR THIS APPLICATION WILL BE EXTREMELY DIFFICULT FOR SOMEONE WHO IS NOT FAMILIAR WITH OPERATIONAL AMPLIFIERS.

A LF411ACN has a maximum power supply voltage of +/-22 Volt. So maximum output voltage is about +/-20 volt. It will not work in this application

A transformer may be the easiest/cheapest approach to this project.

Good Luck
Carl

Thanks. I can already test with OPA454, and so far had configured it in a way to use 2 resistors R1 and R2, where R2/R1 = 2. Just trying to find out what the exact relationship between input and output will be. The schedmatic is a simple non-inverting configuration with one resistor (R1) between ground and inverting (-) input. The other resistor (R2) is between R1 and output pin. Input is connected to (+) of op amp. The supply values to pin 7 and 4 are 15 and -5 respectively. According to the datasheets and some friends the Vout = Vin (R2/R1), I guess using simpple KCL and assuming the input current to opamp os zero, I can arrive at such conclusion.

In any case, the GBP apparently talks about open loop gain and I am not sure how that relates to overall gain either. I am to read parts of that cook book today.

Thanks for any other feedback.

 

[Carl Pugh]

If you require 50 volt peak to peak output, your absolute minimum supply voltage is 50 volt DC. It would be better if supply voltage was higher than 50 V, maybe 60 V.

Your present supply voltage is 15+5=20 volt DC.
With 20 V supply voltage, the absolute maximum output voltage from OPA454 is 20 volt peak to peak with 15 volt peak to peak being more likely.

Good Luck
Carl

 

[pooyan]

If you require 50 volt peak to peak output, your absolute minimum supply voltage is 50 volt DC. It would be better if supply voltage was higher than 50 V, maybe 60 V.

Your present supply voltage is 15+5=20 volt DC.
With 20 V supply voltage, the absolute maximum output voltage from OPA454 is 20 volt peak to peak with 15 volt peak to peak being more likely.

Good Luck
Carl

Thanks Carl, I really appreciate it.
You are absolutely right. Just to make sure, by 50 volts you mean that I could use +30 & -20
volts since sometimes the power supplies have limits on negative and positive amount.

Also, I have been reading from that Cookbook you mentioned and found out about this various gain terms such as open loop, close loop and etc. When we talk about gain band width product the higher it is, the better for high voltage high frequency applicaitons.

But what I don't understand is that the Open Loop gain (I assume it means op amp by itself and no resistors) is shown vs frequency and then closed loop gain is read from this graph. Am I being wrong? I can find you the page number if you'd like.

Finally, while the book is very thorough, I find that the designs and configurations might be too much for this analysis, especially since I was told about this Op Amp and configuration working at 100 khz:
http://ece.uwaterloo.ca/~lab318/modules/AMPX10.GIF

It seems to be limited by a +-22 volt supply though, so maybe I can get a max of 44 volts in this configuration.

 

[Carl Pugh]

Have used the LF357 and it's a good OA.

The LF357 can be used at a gain of 5. At a gain of 3 the LF357 will probably be unstable.

The LF357B is rated for an absolute maximum supply voltage of +/-22V

Using a reasonable safety factor, operate at +/-15V.

From data sheet, Vo is +/-12 volt guaranteed at 10kohm load.

So the LF357B is guaranteed to output 24 volt peak to peak at 200 KHz. In typical circuit it will output maximum of 26 volt peak to peak.
Maximum output voltage will be less at higher frequencies.

Open loop gain is just that, open loop gain.
Open loop gain is used to calculate actual gain in feedback circuits.
Using your circuit for a gain of 10 amplifier, the actual circuit gain can be calculated at different frequencies.

Good Luck
Carl

 

[pooyan]

Have used the LF357 and it's a good OA.

The LF357 can be used at a gain of 5. At a gain of 3 the LF357 will probably be unstable.

The LF357B is rated for an absolute maximum supply voltage of +/-22V

Using a reasonable safety factor, operate at +/-15V.

From data sheet, Vo is +/-12 volt guaranteed at 10kohm load.

So the LF357B is guaranteed to output 24 volt peak to peak at 200 KHz. In typical circuit it will output maximum of 26 volt peak to peak.
Maximum output voltage will be less at higher frequencies.

Open loop gain is just that, open loop gain.
Open loop gain is used to calculate actual gain in feedback circuits.
Using your circuit for a gain of 10 amplifier, the actual circuit gain can be calculated at different frequencies.

Good Luck
Carl

Thanks. but all this sound like I am not able to get 30-50 volts at 100-200 khz and only 15-20 volts. Is there anything else I should be looking at?

 

[Phrak]

Do you need 50 volts peak, or 50 volts peak to peak?

To get 50 volts peak the usual op amps will accommodate the frequency you have in mind, but not the voltage. The op amp will need to be followed by a transistor pair.

What is your current requirement; how much current will your load demand? The load is the circuitry that the amplifier will drive.

 

[vk6kro]

There are high powered audio amplifier chips that are really high powered opamps.

One I found on a quick search was this one:

http://www.national.com/mpf/LM/LM3876.html#Parametrics

It has a Gain bandwidth of 8 MHz and it can deliver 56 watts into 8 ohms. This is 21 volts RMS or about 60 volts peak to peak.
It uses a power supply of plus and minus 35 volts.

Futurlec have them for $6. You would need a heatsink.

But that is just an example. There seems to be heaps of such chips out there.

 

[Phrak]

There are high powered audio amplifier chips that are really high powered opamps.

One I found on a quick search was this one:

http://www.national.com/mpf/LM/LM3876.html#Parametrics

It has a Gain bandwidth of 8 MHz and it can deliver 56 watts into 8 ohms. This is 21 volts RMS or about 60 volts peak to peak.
It uses a power supply of plus and minus 35 volts.

Futurlec have them for $6. You would need a heatsink.

But that is just an example. There seems to be heaps of such chips out there.

I guess I should have known. Where there's a demand, there is a supply.

The highest rated supply voltage I could find by National is the LM3875. The max. supply voltage is 84 volts. They don't design to provide the peak to peak output swing at given load, but with two diodes drops per leg plus current source, it should manage as much as 80 volts peak to peak under small loads.

 

[pooyan]

I guess I should have known. Where there's a demand, there is a supply.

The highest rated supply voltage I could find by National is the LM3875. The max. supply voltage is 84 volts. They don't design to provide the peak to peak output swing at given load, but with two diodes drops per leg plus current source, it should manage as much as 80 volts peak to peak under small loads.

Sounds great! I need 60 - 100 p-p. Today, we managed to get 60 p-p using OPA445 (see above posts) because I changed the power supply to have +/- 31 volts. We were only limited by function generator amplitude to test it becausei t wouldn't go above 10 volts.
The only thing though is that apparently due to slew rate the output looks like a triangular wave and not sin wave as the input. I was told to use a low pass filter to drop the unwanted harmonics or get an opamp with a different slew rate.

This is so exciting really, there is so much to analyze, so little time :D

 

[pooyan]

There are high powered audio amplifier chips that are really high powered opamps.

One I found on a quick search was this one:

http://www.national.com/mpf/LM/LM3876.html#Parametrics

It has a Gain bandwidth of 8 MHz and it can deliver 56 watts into 8 ohms. This is 21 volts RMS or about 60 volts peak to peak.
It uses a power supply of plus and minus 35 volts.

Futurlec have them for $6. You would need a heatsink.

But that is just an example. There seems to be heaps of such chips out there.

Thanks that is great. I will check the slew rate and the frequency of operation.

 

[pooyan]

Ok, I learned some good things about Op Amps but I require :


An Op Amp whose Output swing (Vp-p) is between 60 to 100 volts @ at 100 plus khz. Ideally it would be 60 to 100 volts peak t peak at 200 khz.

How can I search for this?