Highest speed/frequency Op Amp

[Higgler]

Hopefully Op Amp users can assist me,
who makes the highest frequency Op Amp out there? I plan to put RF through it, so slew rate per se isn't important, just highest frequency spec. Low power too.

Hoping for frequencies in the GHz band. i need to make negative capacitors and negative inductors.

Thanks,

kch

 

[IRstuff]

http://www.analog.com/en/prod/0,,759_846_ADL5542,00.html

TTFN

FAQ731-376

 

[itsmoked]

Whoa, that's smokin!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jimkirk]

Are you looking for a true operational amplifier? The ADL5542 is a fixed gain RF amp. The AD8003 is a triple current feedback opamp with 1650 MHz @ unity gain, 730 MHz at gain of 2.

http://www.analog.com/en/prod/0,,759_786_AD8003,00.html

The LT6200-10 from Linear Technology is a voltage feedback op amp with 1600 MHz gain bandwidth product (stable at gain >= 10).

http://www.linear.com/pc/productDetail.jsp?navId=H0,C1,C1154,C1009,P1784

 

[Higgler]

Thanks jimkirk,
I am looking for a true op amp, just to make a negative capacitor and negative inductor at high frequency. Higher frequency than a regular op amp works at it seems, want 1.6 GHz operation as an op amp.

Question;
If an op amp starts with super high gain (50 dB+? or is that 100 dB+ ?), and it works because feedback is minimized and it doesn't oscillate, can one use a cascade of 3 RF amps like the one IRStuff suggested to equal to a crude RF Op Amp (then add feedback)?

Sounds like an oscillator ready to happen, but with perfect isolation between all the Rf amp stages and the feedback isolating the output from the input adequately in an isolated can, may be some higher frequency results are possible?


kch

 

[sreid]

I've got to ask; what's a negative capacitor/inductor?

 

[benta]

Perhaps he means a Gyrator?

 

[itsmoked]

Over unity?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jimkirk]

Cascading multiple similar RF amplifiers is going to triple the phase delay at a given frequency which will likely lead to oscillation if you try to use feedback. Also keep in mind that the RF amplifiers will probably have a 50 ohm output impedance, which might limit their use. The output impedance of an op amp will also increase with frequency, so you'll have to deal with that. Depends on exactly what you're trying to do with this.

sreid, imagine you have a 100 pF capacitor and a high input impedance amplifier monitoring the left side with the input and driving the right side with the output.

Set the gain of the amplifier to zero (essentially grounding the right side). Looking into the left side will look like a 100 pF capacitor.

Now set the gain of the amplifier to unity. Now looking into the left side will look like zero pF, because the voltage across never changes.

Now increase the gain the capacitance looking into the left side continues to decrease, going negative.

With a gain of 2 the capacitance will look like -100 pF.

This is a common technique to develop ultra low capacitance input impedances. Stability can be touchy if things drift. Compensating a 10 pF input capacitance with a -9.9 pF capacitance gets you down to 0.1 pF, but a small change in either the negative capacitance or the original capacitance you're trying to compensate leads to instability.

Of course if you use an inverting amplifier, you'll increase the capacitance. A gain of -1 doubles the capacitance. This is the Miller effect in transistors.

You can do similar things with inductors and resistors. It's a neat trick to use to compensate the DC resistance of an inductor to extend low frequency response.

 

[Higgler]

jimkirk,
thanks, do you have a reference book suggestion on high speed opamp designs?

sreid, the purpose of negative C's and L's is to impedance match short antennas which look like a series RC. At low frequencies, some people have made receive antennas, though transmit antennas are more troublesome. Op Amps can effectively make a negative c or l using feedback. Ham Radio folks would like the wideband VSWR matching this can provide for a single short monopole. Typical high frequency matching using this technique is 30 Mhz so far that I've found in the literature.


The stability issue jimkirk mentions is a key parameter to making these op amp matching circuits usable. It takes some significant design experience. If there's anyone out there with "a ton" of experience, maybe we can talk.

thanks for the inputs

kch
khiggins at toyon dort corm