The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
Subject: Re: Opamp GBW tempco ?
Date: Wed, 18 Sep 2002 22:26:52 +0100
NNTP-Posting-Date: Wed, 18 Sep 2002 21:30:44 +0000 (UTC)
X-Newsreader: Turnpike Integrated Version 5.01 U
In article <email@example.com>, Ian Buckner
>"fred bartoli" wrote in message
>> Opamps for my app may not be as silly as it looks at first.
>> My gain stages are 30dB and 10dB, I have 3x30dB+3x10dB switched on
>> depending on the wanted gain and I'm seeking 0.01dB, 0.01° accuracy
>> full gain range for the IF-detector block. This is about a 30ppm
>> for the the phase accuracy. Not a trivial task for a discrete
>> Also, the IF section has switched gain stages, which means load
>> (yes they can be made small, but compared to required accuracy...)
>> input and output impedances of opamps is definitely a plus in this
>> Also the lock-in detector is sensitive to IF harmonic contents. It
>> rejects even orders harmonics but odd orders harmonic are taken into
>> accounts (1/3 for H3, 1/5 for H5...) and the full scale voltage is
>> high as possible for obvious reasons. That sets the distorsion
>> pretty high for single transistor stages.
>> Also the VNA has to provide gated measurements which means highly
>> operating conditions (IF levels) and asks for the best transfer
>> stability over time.
>> IMHO we are definitely in the modern Opamp domain.
>> For your remark about the 100kHz "which is almost DC", its
>> depends on how accurately you look at this.
>> Just consider the 350MHz (TI's THS4021) opamp I've found, set for a
>> In a non inverter gain block of gain G, an opamp of gain bandwidth
>> give a phase error (when angles are small) of :
>> Alpha = G freq/GBW * 180/Pi
>> This is about half a degree at 100kHz for a 350MHz GBW opamp.
>> Of course I could set the feedback loop to correct that 0.5° error,
>> will not compensate for GBW variations.
>> Taking the derivative versus GBW gives :
>> d_Alpha / d_GBW = -Alpha/GBW
>> which means that for a best case 20% variation of GBW over the
>> range that will give a 20% variation of Alpha, which is 0.1° (per
>> A 100MHZ GBW will give 0.35° per stage.
>> While it's not bad for a standard IF use (radio, SA), for a
>> like my app, *only one* gain stage is at least tenfold off from the
>> all the IF section.
>> In fact, even a single 100MHz GBW follower will just give a
>> 0.01° over the temperature range.
>> Of course I could use active feedback and achieve a near perfect 0
>> error, but for the input stage it's 3db more noise, and for the
>> filters stages the passive tempcos will implies real time phase
>> And one more point in favor of opamps : this is a homebrewing
>project and I
>> don't have access to a precision VNA to make sure the IF perfs are
>> temp. Also I need to optimize the oneshoot-prototype-hardware-cost
>and a few
>> opamps are cheaper than a new PCB.
>> It's definitely time for opamps IF block, and furthermore with real
>You might look at, for example, the OPA687 or the AD8351. These have
>GBW products of several GHz. The open loop phase shift for the OPA687
>is only about 35 deg at 100kHz. There are a number of other
Im successfully using the opa686 at 100kHz from 50r source with gain of
200. low phase shift and low noise as per spec. 40r resistor in series
with o/p tames the oscillation!.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup