Standard non-inverting opamp with a cap across + and - terminals...... Why?

We have that issue at the hopping frequency (200Hz) with one of the ROHS replacement audio amps. The 2.4GHz signal overloads something (looks like the driver output of all things) and we can hear the 200Hz buzz. DOesn't happen with the non-ROHS (older) part.

A pal of mine was making an AM detector with one of those ALD zero-bias FETs. I think it was even a "crystal" radio, made better. I wonder if it worked.

James

On a sunny day (Wed, 08 Apr 2009 22:56:40 GMT) it happened James Arthur wrote in :

Yea, 1GHz.. for me RF starts a bit earlier. I have experimented recently with that idea of a cap between + - input on a LM324 with 150 v 27 MHz RF next to it... It does make some difference, but it its much better to decouple both inputs to ground or to be more precise add a lowpass before the input. Just looked how I solved that, yes 2 caps from the inputs to ground.

On a sunny day (Wed, 08 Apr 2009 17:42:59 -0700) it happened Joerg wrote in :

light

Laws have changed, these things are made locally, every farmer has one :-) The 'radio controle dienst' only comes looking when somebody complains. Last time I was in one of those CB meetings, years ago actually, all sorts of heavy duty stuff was to be had. Thse days even AM and SSB are allowed, AM and SSB causing much more audible effects in a=udio equipmwent then FM.

You are close.

through

In the tube day, when I used my PE1/100 linear, it melted the germanium output transistors in my tape recorder, the speaker leads picked up the RF..

:-)

grounded.

That's just a "grid detector".

John

ive

tching

al.

Bode

ting an

nd

ecial.

I have gone nearly up to a 0.1 with a 10nF. The opamp in question had a gain bandwidth of about 4KHz IIRC. It was one of the super low power sort. The resulting circuit rung quite a bit but appeared to repeat well.

special.

Digging out rusty old memory:

On a tube op-amp with a output voltage swing of +/- 200V, the input section was a long tailed pair with about 6V of grid to cathode voltage. The circuit would go nonlinear long before the grids rectify because the one side of the pair would end up with all the current. The gm was about 0.01 and the nominal plate current was 5mA. So the input went nonlinear at about 1V.

ftp://jjlarkin.lmi.net/Philbricks.jpg

I have a schematic somewhere among the litter of files...

John

an

special.

Yep. Back in the early '70's I was called into Interface, Inc, at Scottsdale Airpark. Nasty noise pick-up. Bypassed a summing node to ground... instant cure ;-)

You _can_ overdo it and cause oscillation.

...Jim Thompson

ut

so

BE

an

has

uit

ounded.

There's no abrupt change in detection efficiency at 26mV, but above

-- Bill Sloman, Nijmegen

matching

an

special.

Looks like a good rule of thumb is to make the added pole equal to, or less than, the GBW of the OpAmp.

...Jim Thompson

This is a myth that's unfortunately very much propagated by the many articles and books that show AM detection by describing an ideal diode, showing a full sine wave, and then showing a half-wave rectified sine wave, suggesting that this is what you actually would see if you probed around a traditional radio. :-(

(Well, OK, with a strong enough signal source you might see it, but when your antenna is typically providing you with a handful of uV, you sure don't -- plenty of the original signal is still present, of course.)

matching

an

special.

The sledgehammer testing method is to place a cap 10 times larger there for a test. When still not oscillating, the smaller cap should be fine. Usually ...

Probably some commission in Brussels has outlawed anything below 26mV because it causes global warming ...

You still have some of these? Amazing. They look like new.

grounded.

It's basic math. Diode current is exponential on voltage. An exponential can be approximated by a power series. Only even-order terms rectify. The dominant even-order exponent is 2. So diodes act like square-law detectors at low levels. And a millivolt of RF is plenty enough to cause trouble.

I'm surprised I have to explain stuff this simple.

But you said...

"My point is that small RF currents won't get rectified; if they are big enough to get rectified they are going to show up as a more- than-26mV voltage excursion between the base connection to the input transistors."

so now you're weaseling.

John

It actually was over on s.e.basics but I am sure John knows this stuff:

In his case one would have to make R as small as practical without turning the phase margin into a white-knuckle affair.

In my cases most of them just can't achieve the desired noise performance.

grounded.

So? You expect some other kind of response from a weasel ?:-)

I'm always pleased to note that I'm the highest standard for Slowman's disdain, but please don't feed the jerk. Let him die that most unpleasant of deaths... alone ;-)

...Jim Thompson