Exact match. Not showing close matches.
**
PICList
Thread
**

**
'[EE] fast differential in-out amplifiers'
**

2005\12\20@090409
by
Vasile Surducan

I'm looking for a high frequency differential amplifier (> 400MHz BW

at -3dBm) optimized for high input impedance and very large slew rate

(1000 to 5000 V/us) running at 3V or 3.3V single supply.

I found some differential in-out amplifiers (AD8132, LMH6551) but all

are using small input/gain resistors to keep high the bandwidth, and

this is reducing the differential input impedance. Requirement for

differential input impedance is 10K at a differential gain of 5.

I don't want to use any buffers between the input signal and

differential amplifier.

I do not want to design my own differential in-differential out

amplifier using two or three high frequency OA.

If there is any RF master here with experience in high frequency

differential amplifiers, I'm searching for a clue.

cheers,

Vasile

2005\12\21@005727 by Mike Singer

> I'm looking for a high frequency differential amplifier (> 400MHz

> ... Requirement for differential input impedance is 10K

...

> If there is any RF master here with experience in high frequency

> differential amplifiers, I'm searching for a clue.

Vasile,

There is a formula F = 1 / (R * C)

C = 1 / (R * F) = 1 / (1 000 000 000 Hz * 10 000 Ohm) = 0.1 pF

pretty low input capacitance, isn't it?

Mike.

2005\12\21@012433 by Dmitriy Kiryashov

What 10K _differential_input_impedance_ has to do with RC filter ?

Dmitriy.

Mike Singer wrote:

{Quote hidden}

> -

2005\12\21@044958 by Vasile Surducan

with 2pF input impedance, voilla RC filter.

Vasile

On 12/21/05, Dmitriy Kiryashov <spam_OUTvze27bymTakeThisOuTverizon.net> wrote:

{Quote hidden}

2005\12\21@070659 by Dmitriy Kiryashov

Active element ( buffer ? ) in feedback net.

Providing required attenuation and removing

capacitance effect the same time.

Vasile Surducan wrote:

{Quote hidden}

>

2005\12\21@073252 by olin piclist

> There is a formula F = 1 / (R * C)

There is, but if you're looking for the 3dB rolloff frequency then it's

F = 1 / (2 * Pi * R * C)

******************************************************************

Embed Inc, Littleton Massachusetts, (978) 742-9014. #1 PIC

consultant in 2004 program year. http://www.embedinc.com/products

2005\12\21@082007 by Gerhard Fiedler

> There is a formula F = 1 / (R * C)

Actually, it's

f = 1 / (2*pi * R * C)

(R * C) by itself it the time constant of the filter. 1/RC is the "angular

speed", so to speak, so it has to be divided by the angle of a full circle

(2*pi) to get the frequency.

Not that it changes the basic thought a lot... just the numbers a bit.

Gerhard

2005\12\21@131420 by Vasile Surducan

> Just an idea.

>

> Active element ( buffer ? ) in feedback net.

> Providing required attenuation and removing

> capacitance effect the same time.

No, it's easyest to add buffers on the input signal and low down with

the input impedance at about 150 ohm, that's tipical for this

frequency. The better choice is to use a transimpedance amplifier, but

I didn't found any at 3V single supply.

At +/-5V there are plenty differential in /single out or differential

in/differential out

buffers with 5000V/uS or more but power hungry devices or

transimpedance amplifiers with fully differential outputs. But I can't

afford to spend energy for 3/+-5V conversion.

Vasile

{Quote hidden}

> >

On Wed, 21 Dec 2005, Dmitriy Kiryashov wrote:

> Just an idea.

>

> Active element ( buffer ? ) in feedback net.

> Providing required attenuation and removing

> capacitance effect the same time.

Or a 'driven shield' ;-)

Peter

> > There is a formula F = 1 / (R * C)

>

> There is, but if you're looking for the 3dB rolloff frequency then it's

>

> F = 1 / (2 * Pi * R * C)

But Vasya did say nothing about the signal being sinusoidal, If I'm

not mistaken.

Mike

2005\12\22@180739 by Mark Rages

> Olin Lathrop wrote:

> > > There is a formula F = 1 / (R * C)

> >

> > There is, but if you're looking for the 3dB rolloff frequency then it's

> >

> > F = 1 / (2 * Pi * R * C)

>

> But Vasya did say nothing about the signal being sinusoidal, If I'm

> not mistaken.

>

> Mike

Vasile only said 400MHz bandwidth. Joe Fourier added the part about sinusoids.

Regards,

Mark

markrages@gmail

--

You think that it is a secret, but it never has been one.

- fortune cookie

More... (looser matching)

- Last day of these posts

- In 2005
, 2006 only

- Today

- New search...