opamp rf noise/background meter

[KludgeIt]

I am attempting to put together a project which is more of a toy/coffee-table-curiosity. I have a rather large meter from some govt surplus which has been in my basement forever, 2.4 ma. I would like to construct a meter which just responds to the background rf, or any rf, and is just entertaining, and "mysterious" to those with little or no exposure to such things. However, I am wondering if I can't actually find a use for it, like finding noise sources and applying filters, etc. ... who knows?

I have had to do a crash course in gEDA to get a schematic I could present, which might give someone enough informationl to help me, and provide enlightenment on my ignorance of such things as I am attempting here ... I usually just stick to software, and binary firmware.

So, with no more explanation to waste your time, I will attempt to post it as an attachment. My main questions are, what would it take to actually make my kludged circuit work, and there is a note on the schematic of other points which the device needs, and I do not know how to implement properly ... thanks in advance!

[KludgeIt]

Typically there is a capacitor on the output of the diode across the load resistor of the detected RF input:
https://en.wikipedia.org/wiki/Envelope_detector

I didn't look at the opamp data sheets but watch DC errors due to input offset voltages and input offset currents / input bias currents.  There's quite a lot of gain shown in the chain which is, in some ways good since you may well get XX microvolt level signals out of the detector, but which can also be bad if you have maybe millivolts of offset voltages or whatever depending on your opamps and input current related potentials from the input resistances.  And of course a strong signal could give you millivolts of output.  But operating in the log detector weak signal region of the diode is probably good so that the dynamic range needed at the output is minimized, you'll just have to deal with weak detected signals at the low end of the range.

I see there's a LC coupling circuit on the antenna.  You can probably tweak that for frequency selectivity or diode matching if you like.  See old HP app notes on the topic of such detectors.

Yes. As I pointed ouit, this is just being done as a "toy", or coffee table "conversation piece." So, the main point is just that the meter needle moves about enough to be interesting, and let people contemplate what they are seeing, and be amused, so errors are highly acceptable, indeed, they might even be an asset.

The biasing is something I am concerned with, as since this circuit has ~62db gain, even slight bias errors could peg the meter at the end of all of that 1000x gain! And, I was specifically asking for any thoughts on a bias network which would allow these bias errors to be controlled, and the meter made easy to zero when the antenna was disconnected, to set a base reference.

... and, of course, I was also hoping for some thoughts on adding a gain control in the circuit, where it should be added, and how it should be implemented in the hardware.

And, after looking at the specs on my general coverage communications receiver, I seen its' sensitivity was < 1uv, so I wanted to be able to meter microvolt levels. And, I tend to favor keeping the front end "wide open" in amplification, so those very low microvolt levels are reflected in the chaos of the wide band noise and background the meter was reproducing.

And, I dropped any thoughts of using a cap on the output from the diode, with the thought that I did not want any "smoothing/averaging" of the signal, and am thinking the rough nature of that signal will only make the movements of the meter more interesting.

Thanks for taking a look, and your comments. If you have any further thoughts or ideas on the above, I would appreciate your reflecting on them here.

[Kleinstein]

The low µv sensitivity of receivers is for a small bandwidth - your circuit is large bandwidth and thus will see higher levels, unless you go to very remote areas. A little more adjustment to gain might be needed, e.g. a switch to chose a factor of 10.

There is no need to spread the gain over so many OPs. As the output BW is low a single OP is good enough.
More OP only increase current consumption and chances for unwanted oscillations. The single OP version often also include offset trimming pins - so you could use these.

A little smoothing cap after the diode is a good idea, to keep the RF away from the OP(s).

[KludgeIt]

The low µv sensitivity of receivers is for a small bandwidth - your circuit is large bandwidth and thus will see higher levels, unless you go to very remote areas. A little more adjustment to gain might be needed, e.g. a switch to chose a factor of 10.

There is no need to spread the gain over so many OPs. As the output BW is low a single OP is good enough.
More OP only increase current consumption and chances for unwanted oscillations. The single OP version often also include offset trimming pins - so you could use these.

A little smoothing cap after the diode is a good idea, to keep the RF away from the OP(s).

Since this will be sitting on a coffee table, with a small antenna, and one constructed to simply add to the "mysterious factor" of the contraption, I see the added gain as a plus, but I am looking for ideas in how to implement a suitable gain control to manage it. Having gain which is not needed, but can simply be dialed up when needed appeals to me. I live in a semi-rural area, and escape much of the rf pollution of the city, so will scale back the amplification factor if it is found to be a problem. But, I would prefer to simply have a gain control which could deal with the problem nicely. And, since this is not a real measurement device, demanding any real accuracy, the noise introduced by the added stages will probably go largely unnoticed.

My junk box determined what I used, and the TLC274s were left over from another project of my sons, and I am anxious to be rid of one of them here, plus they boast extremely high input impedance, and will load the input ever so slightly -- the best in my limited junkbox (either this, or a LM324.) And, since it is a quad, I simply used them all, saving me from having to tie their pins to ground, to stop any unwanted oscillations. But, I am looking for ideas on how to implement a suitable external bias control.

Since the device will be operated from lithium rechargeable batteries, recharged by solar from light coming though a window in the afternoon, and the small solar panel simply adding to the devices "mysterious factor", and having line power at its disposal for sunless days, the power consumption of the extra few milliamps is of no real concern here -- its main power source being 2 x 18650 lithium batteries, it should be able to run for many days before there is ever a danger of running short on power, and the battery protection circuit kicks in to prevent a over-discharge which would ruin the batteries, given that the sun is not shining, and grid has gone down.

The cap will be experimented with, to see its affects, and whether they are desirable, or not, at the breadboard stage. No sense in having the rf bleed hamper the amp factor, if that is found to be a problem, correct, your caution will be implemented.

[vindoline]

Here are a few simple projects you could check out:

http://www.vk2zay.net/article/9
http://www.techlib.com/area_50/Readers/Karen/misc.htm#Electrometer