Has anyone played with any of the new high-bandwidth op amps in IF
amplifier circuits? I've got an idea I want to experiment with, and
not a whole lot of analog design ability (though I do better with
building blocks than with compensating discretes), and now that
GBw products are cheap above 1 MHz it might look practical to
pursue. I have no standards for comparison since I know nothing
about parameters like noise figures and the like in standard
discrete circuits.

--
All relevant people are pertinent.
All rude people are impertinent.
Therefore, no rude people are relevant.
-- Solomon W. Golomb

clifto wrote:
Has anyone played with any of the new high-bandwidth op amps in IF
amplifier circuits? I've got an idea I want to experiment with, and
not a whole lot of analog design ability (though I do better with
building blocks than with compensating discretes), and now that
GBw products are cheap above 1 MHz it might look practical to
pursue. I have no standards for comparison since I know nothing
about parameters like noise figures and the like in standard
discrete circuits.

GBw products above 1MHz have been around for a long time in op amps, if
you don't mind dropping the bucks then you can even get GHz numbers.
There's very little stopping anyone from using even the old
high-bandwidth op amps in IF's. But there's not much need to.

In IF strips for FM receivers, most op-amps do really bad things when
they hit their limiting amplitudes. Google for "phase reversal" with
"op amp".

"Straight" op-amps are not easily amenable to AGC action, but I've done
some playing around with for example LED/CdS optocouplers as very
linear controllable-loss blocks for AGC action and they work fine up
through the low MHz. Above a few MHz I think capacitance across the CdS
photocell gives too much leakage in typical receiver use, although if
you only want a limited amount of AGC I think you're good to go.

It is MOST interesting to use chips like AD603's in IF's if you want
AGC action. Many many app notes and construction articles on the web
will show you examples of use.

Are you perhaps really talking about RC active filters to replace
crystal or mechanical IF filters? I have always been VERY PESSIMISTIC
about this, usually even few percent tolerance capacitors (not cheap)
cause your passband/ultimate rejection to be way way worse than what
you expect. Different filter topologies have different tradeoffs, but
RC active filters when you want the Q to be 100 and want even mediocre
ultimate rejection are economically unfeasible in every analysis I've
done.

Tim.

"Tim Shoppa" wrote in message
ups.com...
Are you perhaps really talking about RC active filters to replace
crystal or mechanical IF filters? I have always been VERY PESSIMISTIC
about this, usually even few percent tolerance capacitors (not cheap)
cause your passband/ultimate rejection to be way way worse than what
you expect.

I'm not as pessimistic as you, but I share your disappointment to some extent.
However, if a filter doesn't need to be in "active duty" 100% of the time,
having it self-calibrate seems potentially viable, at least for lower order
filters.

Different filter topologies have different tradeoffs, but
RC active filters when you want the Q to be 100 and want even mediocre
ultimate rejection are economically unfeasible in every analysis I've
done.

I'm usually after active filters for the sake of space, and consider Q100
difficult to achieve at HF with any physically small design.

I do enjoy the occasional posts from people whose SPICE simulations show
ultimate rejections in the couple hundred dB ballpark. :-)

---Joel

Tim Shoppa wrote:

clifto wrote:
Has anyone played with any of the new high-bandwidth op amps in IF
amplifier circuits? I've got an idea I want to experiment with, and
not a whole lot of analog design ability (though I do better with
building blocks than with compensating discretes), and now that
GBw products are cheap above 1 MHz it might look practical to
pursue. I have no standards for comparison since I know nothing
about parameters like noise figures and the like in standard
discrete circuits.

GBw products above 1MHz have been around for a long time in op amps, if
you don't mind dropping the bucks then you can even get GHz numbers.
There's very little stopping anyone from using even the old
high-bandwidth op amps in IF's. But there's not much need to.

In IF strips for FM receivers, most op-amps do really bad things when
they hit their limiting amplitudes. Google for "phase reversal" with
"op amp".

"Straight" op-amps are not easily amenable to AGC action, but I've done
some playing around with for example LED/CdS optocouplers as very
linear controllable-loss blocks for AGC action and they work fine up
through the low MHz. Above a few MHz I think capacitance across the CdS
photocell gives too much leakage in typical receiver use, although if
you only want a limited amount of AGC I think you're good to go.

It is MOST interesting to use chips like AD603's in IF's if you want
AGC action. Many many app notes and construction articles on the web
will show you examples of use.

Are you perhaps really talking about RC active filters to replace
crystal or mechanical IF filters? I have always been VERY PESSIMISTIC
about this, usually even few percent tolerance capacitors (not cheap)
cause your passband/ultimate rejection to be way way worse than what
you expect. Different filter topologies have different tradeoffs, but
RC active filters when you want the Q to be 100 and want even mediocre
ultimate rejection are economically unfeasible in every analysis I've
done.

Tim.



The engineers at my last job liked the Mini Circuits ERA and similar
monolithic amps. They also liked to do most of the AGC on the front end
rather than at the IF. This sometimes included an electronic attenuator
to drop the gain by 20 dB to prevent overloading a critical stage. The
customers liked the way they worked, they ordered plenty of them at
$20,000 USD each.
--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

wrote:

On Tue, 18 Apr 2006 17:26:43 GMT, "Michael A. Terrell"
wrote:

Tim Shoppa wrote:

clifto wrote:
Has anyone played with any of the new high-bandwidth op amps in IF
amplifier circuits? I've got an idea I want to experiment with, and
not a whole lot of analog design ability (though I do better with
building blocks than with compensating discretes), and now that
GBw products are cheap above 1 MHz it might look practical to
pursue. I have no standards for comparison since I know nothing
about parameters like noise figures and the like in standard
discrete circuits.

GBw products above 1MHz have been around for a long time in op amps, if
you don't mind dropping the bucks then you can even get GHz numbers.
There's very little stopping anyone from using even the old
high-bandwidth op amps in IF's. But there's not much need to.

In IF strips for FM receivers, most op-amps do really bad things when
they hit their limiting amplitudes. Google for "phase reversal" with
"op amp".

"Straight" op-amps are not easily amenable to AGC action, but I've done
some playing around with for example LED/CdS optocouplers as very
linear controllable-loss blocks for AGC action and they work fine up
through the low MHz. Above a few MHz I think capacitance across the CdS
photocell gives too much leakage in typical receiver use, although if
you only want a limited amount of AGC I think you're good to go.

It is MOST interesting to use chips like AD603's in IF's if you want
AGC action. Many many app notes and construction articles on the web
will show you examples of use.

Are you perhaps really talking about RC active filters to replace
crystal or mechanical IF filters? I have always been VERY PESSIMISTIC
about this, usually even few percent tolerance capacitors (not cheap)
cause your passband/ultimate rejection to be way way worse than what
you expect. Different filter topologies have different tradeoffs, but
RC active filters when you want the Q to be 100 and want even mediocre
ultimate rejection are economically unfeasible in every analysis I've
done.

Tim.



The engineers at my last job liked the Mini Circuits ERA and similar
monolithic amps. They also liked to do most of the AGC on the front end
rather than at the IF. This sometimes included an electronic attenuator
to drop the gain by 20 dB to prevent overloading a critical stage. The
customers liked the way they worked, they ordered plenty of them at
$20,000 USD each.

I tried that once using MAR-6s and -11s for gain and 1n4007s as
pseudo PIN diodes (Pi antenuators) between stages. Produced an
excellent IF for gain, noise and intercepts but it drew near 350ma
at 9V for the IF and AGC. Definately not for the qrp/battery user.

Allison


We weren't in the QRP business. We were building hi rel equipment to
track space probes.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

wrote:

On Wed, 19 Apr 2006 17:42:35 GMT, "Michael A. Terrell"
wrote:

The engineers at my last job liked the Mini Circuits ERA and similar
monolithic amps. They also liked to do most of the AGC on the front end
rather than at the IF. This sometimes included an electronic attenuator
to drop the gain by 20 dB to prevent overloading a critical stage. The
customers liked the way they worked, they ordered plenty of them at
$20,000 USD each.

I tried that once using MAR-6s and -11s for gain and 1n4007s as
pseudo PIN diodes (Pi antenuators) between stages. Produced an
excellent IF for gain, noise and intercepts but it drew near 350ma
at 9V for the IF and AGC. Definately not for the qrp/battery user.

Allison


We weren't in the QRP business. We were building hi rel equipment to
track space probes.

Well that much was obvious. However, despite the higher power needs
it's performance is stellar. I can see no reason why it would not be
very high rel.

The reciever it was used in has a overload threshold and dynamic
range far higher than anything in the ham market. Also the NF is low
giving it an excellent MDS. It was built to work weak signals
effectively without overload because I have a "kilowatt charlie" less
than 300yds away on the same band. Nominal RF comming down
the coax from him is 25-150mw! The whole reciever pulls a whopping
2A (no signal) becuase of all of the high standing currents in RF and
IF processing never mind the required 17DBM LO..

At the other end of the power spectra I've used cascode J310 FETs
to build very good IFs with far lower power needs but still good AGC
and gain.

What's hard is building filters with at least 140db of stopband
attenuation. There nothing like shielding, lots of it. Then testing,
testing, testing.

Allison

We built tubular filters in house, because we couldn't get what we
needed, when we needed it. We built the ground station and a mobile
earth station with a large mobile diesel generator for power, for the
Italian government to track their launches. They had a severe overload
problem, trying to set up the one station near the launch pad.
Engineering was going round and round about a way to reduce the signal
levels prior to the launch, that could be done quickly. My suggestion
was a chain link fence near the pad to block most of the signal, till
the rocket left the pad and was no longer pointing right at the ground
based antennas. The receivers had lots of aluminum between stages to
stop stray signals.


We used a LINEAR AGC system that was a pain to get set up, but the
dual diversity combiner required the linear 0 to 5 volt signals to steer
the combiner.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

Clifto,

Is your goal a project to get hobbyists with AF project background into
RF building with a minimum of giant steps (or something like that)? If
so, the familiarity of op-amp building is probably a good choice.
Introducing ceramic filters and the need for a ground plane can be
enough for hobbyists doubtful about entering a higher frequency range.
Getting something to work will build confidence. You can make more
satisfying results a later step.

Steven

clifto wrote:
Has anyone played with any of the new high-bandwidth op amps in IF
amplifier circuits? I've got an idea I want to experiment with, and
not a whole lot of analog design ability (though I do better with
building blocks than with compensating discretes), and now that
GBw products are cheap above 1 MHz it might look practical to
pursue. I have no standards for comparison since I know nothing
about parameters like noise figures and the like in standard
discrete circuits.

--
All relevant people are pertinent.
All rude people are impertinent.
Therefore, no rude people are relevant.
-- Solomon W. Golomb

wrote:
Clifto,

Is your goal a project to get hobbyists with AF project background into
RF building with a minimum of giant steps (or something like that)?

Actually, I want to play a bit myself, and my analog design "skills"
(if that word applies) have always been weak. I've always been a
*much* better troubleshooter than designer.

--
All relevant people are pertinent.
All rude people are impertinent.
Therefore, no rude people are relevant.
-- Solomon W. Golomb