Definition of amp noise

I'll just have to put it in the computer and let it rip.

Yes, I've read a lot on Barkhausen, the avalanche effect. Actually it's fine even if the pure 37Hz sine wave is distorted, to even say a square wave. It's the unpredictable random noise that I'm trying to eliminate. Pure noise is always changing in phase and amplitude. I'm sure the Barkhausen will add some 37Hz noise, but I look at it this way. If the 37Hz signal were not present then Barkhausen noise would be nearly zero in a toroid. If the 37Hz were present then there would be 37Hz Barkhausen noise. Maybe the Barkhausen will help increase my signal, lol.

Wouldn't the differential help eliminate some of the terrestrial and stray EM noise? Albeit I'm going to use a noise cancellation coil, but it might not cancel 100%. Also, the computer and nearby equipment would generate noise. In that case would a differential amp do wonders?

With L's & C's? If the Q is high enough then the 37Hz signal across the L or C would be amplified. Which method is better, LC or transformer? Or perhaps both.

Thanks for the great tips, Paul

Ken, I hope you are right because so far I've been unable to find any differential amplifiers even close to an op amps gain. I'm looking at op amps with about 7 million voltage gain and 0.8 nV/Sqrt(Hz). Lets say I have one coil that's 1-meter radius-- about 20 lbs of copper. Next to this coil is a ferrite loopstick. The total turns on the loopstick will be adjusted so as to cancel out any 37Hz noise in the large coil. Although the large coil will be closer to the 37Hz signal than the loopstick. The large coil and loopstick are properly connected in series so that any EM noise, except the 37Hz signal, will cancel each other out. By whatever method, the 37Hz signal is amplified and sent to a signal meter. In such a case, do you think the differential amp could help reduce enough noise? I'm thinking that the amplifier itself could pick up some terrestrial noise and who knows what else.

Paul

Huh?

Most opamps ARE differential amplifiers! Or can be combined by fours for even better ones.

I was told by a EE that op-amps can be made into a differential but they make poor differential amps. May be you know something that people who design differentials don't know.

Don Lancaster is an EE, as am I. We're sorry to see you confused.

Opamps are in fact high-quality differential amplifiers, with high common-mode rejection and with very high gain at DC, approaching infinity in the ideal case, and dropping at 20dB/decade until the unity-gain frequency is reached, which is typically 5 to 200MHz. These high-performance amplifiers are used inside feedback loops to make various kinds of circuits. Perhaps you are interested in difference or instrumentation amplifiers rather than opamps? These are sometimes called differential amplifiers, which is an incomplete description. Analog devices offers a free book about these, "A Designer's Guide to Instrumentation Amplifiers," that you can use to enhance your knowledge. Order it on this page,

Burr-Brown (under the name TI), NSC, and Linear Technology are other companies that make similar high-performance amplifiers.

But if Don's 170dB estimate (a factor of 316 x 10^6) is correct, you'll certainly not be able to imply cancel out the offending noise.

Correct, it's low-frequency noise is given by its dc resistance. But that won't abrogate the reality of ambient E-M signal noise.

Are you just arguing semantics? Obviously there's plenty of 50/60Hz EM radiation floating around, and I expect that various subharmonics and switching produce plenty of detectable radiation at 37Hz.

In article , wrote: [...]

The earth's field going through the core and any mechanical stress will cause noise. Also a slowly changing field will make noise.

You are going to use a transformer, I assume. The primary connects to the signal and the secondary to the amplifier. The primary is balanced and the secondary need not be.

Tune the transformer if nothing else.

Graham

As are the 76 and 82Hz ELF/SLF frequencies known to be used by the US and Russian navies respectively to communicate with submerged nuclear attack submarines.

Ken,

Thanks very much! It's very clear. Although I am not certain what C3A & C3B do, but you said they're optional. An increased voltage is on C2 because it's in resonance with the coil. The beads block most of higher frequencies. C1A & C1B routes most of the higher frequencies to ground. And of course the shielded T1 ups the 37Hz signal by 10 to the op-amp. I might even get 200:1 from the Sp-4. It's an audio transformer so the designers should have had a goal of minimum noise in mind. If it has too much noise then there's always the nano-crystalline cores, which have extremely low avalanche noise because they're domains are magnitudes smaller than most cores.

I'll probably buy a few LT1028 op-amps today-- Voltage gain of 7 million min., 0.85nV/Sqrt(Hz) @ 1KHz, CMRR=126dB.

Perhaps a Summing Amplifier could be used to combine both the pickup coil and noise cancellation coil. Of course the wires of noise cancellation coil would be switched to they subtract, not combine.

Here's a list of op-amp types:

Thanks for the superb help! Paul

Now that is very interesting. I see it outputs DC, which is great. I have access to the 37Hz signal only by EM, but I also know the phase. It's the amplitude that I need to detect-- it could from zero on up. Here's a great lock-in amplifier site:

Thank Robert! That will make a huge difference. I'm surprised to not find any lock-in amplifier chips! Sure, you can get as complex as one wants, but it really isn't that complex. So I'll use Ken's circuit as a preamp and then feed it to an analog mixer and then to a low pass filter. Do you have any mixers in mind. The first one I came to was MC12002. Not sure if it works at 37Hz since some mixers have built in caps and generally used for VHF. It would be nice to have a built in oscillator, but problem no at such low frequencies. I could use a simple audio oscillator.

Thanks again, Paul

If it is receiving electromagnetic waves through an antenna like device, it IS a RF frequency.

Bizarre as it seems.

Or, to use a much more modern term, synchronous demodulation.

CMOS 4016 ferinstance.

Needless to say, any decent Q at 37Hz resonance would limit the allowable rate of change of the input signal to be well under one Hertz.

Huh?

One quarter of a ten cent CMOS 4016 hung on any opamp to make +1 -1 gain.

There's even an ugly and unsubstantiated rumor that someone once wrote a book on this.

snipped-for-privacy@gmail.com wrote: ...>

...

Paul,

Be aware that the quoted gain of an op-amp is the DC open loop gain and normally the feedback is selected to bring the closed loop gain to very much less - typically lower than 1000 per stage.

The gain will typically start dropping off at a very low frequency in the 1Hz region.

Not to say that this device is not a good choice for your application, it does have extremely good low frequency noise characteristics.

Also you were asking questions about why Differential Amplifiers had much less gain than opamps: This is because usually such an amplifier is actually an opamp with feedback all in the same package to bring the gain to a well controlled figure (usually in the range of 1 to 100) eg TIs INA114 series.

kevin

There's only been around 1,300,000 of them so far.

ASCII-ART

GND ! --- C1A --- T1 = Triad SP-4? ! C3A from--[RF bead]-+---+----!!-- ------ OP-amp input coil ! !. ! --- ).( ---C2 ).( ! ).( -----[RF bead]-+---+----!!-- . ! ! C3B ! ------ GND near op-amp ! SHIELD ! --- C1B --- ! GND

The beads and the C1A and C1B serve to keep the local radio stations out of the system. C2, tunes the receiver coil. C3A and C3B are fairly optional. The tune the transformer. The transformer is a step up by about 10:1.