op amp buffer

Thread Starter

Ilovecircuits

Joined Oct 15, 2009
5
I've built a oscillator, using a 4Mhz, crystal. I want to amplify the 4 Mhz sine wave coming out of the crystal. I used a op amp buffer. Connecting the inverting input to the ouput, and the non-inverting input, to the output of my oscillator. To no avail will it amplify the signal, the output of the buffer looks nothing like my signal. Any suggestions for amplification of such a signal, and yes I'm using a high speed op amp.
 

hgmjr

Joined Jan 28, 2005
9,029
It does sound like your opamp is unable to respond to the 4MHz signal. Not surprising really. Commonly available opamps are not typically capable of handling frequencies in the 4 MHz range.

hgmjr
 

ELECTRONERD

Joined May 26, 2009
1,146
Buffers are more for impedance matching, and so I doubt they would amplify your signal much. Is impedance a primary issue in your project? Otherwise, you might consider using a different format design.
 

Thread Starter

Ilovecircuits

Joined Oct 15, 2009
5
I just want to amplify the signal of the oscillator. I've attached the schematics of what I'm using. If I don't use an omp -amp since ELECTRONERD said a different design might be better ideas?

But how come my op amp doesn't have the same signal output, is the signal generated from the crystal too weak?



The top circuit is my oscillator

The middle circuit is the voltage follower.

The bottom circuit is the voltage regulator coming into the oscillator +12V

I'm using a 2N222A npn transistor. The voltage regulator uses the LM317 chip. The crystal is just a standard 4 Mhz crystal
 
Last edited:

beenthere

Joined Apr 20, 2004
15,808
What is the op amp? It is arranged as a follower, so even if it can run at 4 MHz, it will not amplify the signal.
 

Audioguru

Joined Dec 20, 2007
11,251
The opamp is not powered.
It is not DC biased properly.
Since it is a follower with no voltage gain then its output will be exactly the same as its input if it is powered and biased properly.
 

Thread Starter

Ilovecircuits

Joined Oct 15, 2009
5
I guess I didn't make it apparent in my diagram, I am hooking up my op amp to +15 and 15 voltage sources.

Shouldn't this give me atleast unity gain?
 

russ_hensel

Joined Jan 11, 2009
825
A video op am should have high frequency response, but normally fairly low input impedance. Can easily turn into ossc. But do some research it may do it.

Since the transistor works why not add another transistor stage.
 

Thread Starter

Ilovecircuits

Joined Oct 15, 2009
5
I've used the LM318 and MC33074 op amps.

Do I have to connect pins 1,8 and 5, or can I ignore those on the MC33074.
I do have two supplies also, for +15, and -15 biasing.
 

Audioguru

Joined Dec 20, 2007
11,251
The datasheet for the MC33074 shows that its output swing at 4MHz ia almost nothing.
At 4MHz its gain is less than 2.
 

steveb

Joined Jul 3, 2008
2,431
Look at bandwidth spec's - http://www.national.com/mpf/LM/LM318.html

http://www.onsemi.com/PowerSolutions/product.do?id=MC33074

15 MHz & 4.5 MHz. You need an op amp with the ability to give gain (how much) at 40 MHz, plus it need to be one for dual supplies, unlike the MC33704.

Something like the AD829 might work out - http://www.chipcatalog.com/Analog/AD829.htm
This is good general advice, but there is still a mystery here. Both opamps should give some response at 4 MHz when used as a unity gain buffer. And the MC33704 should still operate with dual power supplies, even though it is a single supply opamp.

I think audioguru was on the right track previously. He mentioned the DC biasing. You can't drive the input of the opamp with capacitive coupling. There needs to be a DC path into noniverting terminal on the OPAMP.
 

beenthere

Joined Apr 20, 2004
15,808
Agreed about the response - I misread it as 40 MHz. A resistor to ground can give a 0 volt reference on the noninverting pin - the question is still how much gain?
 

steveb

Joined Jul 3, 2008
2,431
what if I change the npn transistor im using a 2n2222, is that not puting out enough power?
Are you still using capacitive coupling from the output of the oscillator to the input of the opamp. If so, you need to provide a DC path on the input of the opamp noninverting terminal. "beenthere" mentioned using a resistor to ground which should work if the value is large enough to not load down the output of the oscillator.
 
Top