If you short out a few zillion 1Meg resistors, and put them in an insulated box,will they generate enough heat (over time) to be detectable?

You are close to revealing why such a setup won't work.

Hint to #12: consider maximum power transfer from a "source resistor" to a "load resistor". What are the requirements? Draw a schematic and then tell me what happens.

 

Quoting Craig Ferguson: "It's a JOKE!"

 

No...my hint to #12 was to make him realize that Johnson noise is present on all resistors regardless if there is current flowing through them or not. Johnson noise can be modeled as a perfect resistor in series with a voltage source (the voltage source having a particular RMS depending on R, and a uniform power distribution vs frequency).

I think this is rather what I was talking about... I didn't know that Thermal Noise was called Jonhson-Nyquist noise. Nonetheless, I shall try to do a better job band-limiting my amplifier so that I'm not amplifying all this white noise. Can I just add caps in parallel with my feedback resistors to do this, or would it really benefit me to do a bit of research into active filters here?

Alternatively, would D/C coupling be an awful idea here? Could get rid of a lot of this nasty white noise.

Thanks all!

 

Quoting Craig Ferguson: "It's a JOKE!"

Who is Craig Ferguson, why are you quoting him, and what's a joke?

 

DC coupling gets the DC offset out of hand really quick with huge amounts of gain (as Joey said), unless you build a DC feedback path across more than one amplifier and I don't think that would change the noise at all.

You can get the out-of-band noise quieted very well with passive measures.

 

Craig Ferguson is the host of The Late Late Show. He is constantly derisive, insulting, and improper. Then he hollers, "It's a JOKE!" as if that is going to fix everything.

Edit: And I do not suffer from the belief that a drawer full of resistors might overheat. If that was true, the vinyl tile in my shop, being of very high resistance, would have burst into flames a very long time ago.

 

Awesome, thanks. I totally skimped out on the schematic I posted earlier. I'm remembering now that I originally had a problem with my DC offset being a bit out of hand and have added an offset null to the first amplification stage. The offset nulling is simple: Instead of the inverting pin going straight to ground, I have it going to the wiper on a potentiometer that is hooked to the + and - rails. I then would crank the gain up and moved the wiper until my output waveform sat as close to 0 as possible. I realize this isn't a permanent solution since the offset can drift, but it's done wonders for me.

 

Craig Ferguson is the host of The Late Late Show.

That explains it...I neither stay up late nor watch tv...

 

I'm single, I work for myself, and I paid off my mortgage a long time ago. I can play, "nurse the noobies" until 5AM if I want to...and sometimes, I do. Craig Ferguson is a high energy case of silly, which is exactly what I need to counterbalance my tendency to be way too serious.

 

what's a joke?

How could anyone seriously now know he was joking?

Never mind, I get lots of wedgies for my jokes here. A healthy sense of humor is conspicuously invisible.

 

I only wish I could click you a dozen thanks.
Floating a joke here is like pulling teeth.

 

A joke going unnoticed among engineers!? Never!

Thanks for all the help, you guys. I'll likely post more when I'm back to the breadboard. I'm out of town at the moment.

 

Hey Everybody,

So I've come back to my breadboard and my SPICE. I've gone ahead and re-uploaded my schematic. I should note that R10 and R11 are a potentiometer used to cancel my DC offset in the gain stages. My amplifier seems to be working much better, now. It no longer has a point where the gain drops to zero: Decreasing the resistance on the base of the PNP seemed to do the trick, and adding a feedback cap didn't seem necessary.

Now I'm working with the actual receiver for the circuit and need to pin the noise down. People have been telling me that I need to band-limit my gain, but I'm wondering how to do this more specifically. I've used caps in parallel with feedback resistors before, but I was following the instructions of a Godly Engineer and I never got a clear answer as to what they were for: Perhaps this? Can anyone point me to a good resource for figuring out how to band limit my gain?

Also, I think I'm having a problem with my circuit ringing. If I put the receiver near the transmitter for only a moment and then move it far away, my circuit keeps alarming (making sound at the output) even after I've moved it far away. Would this be similarly alleviated by band-limiting my gain circuitry? I'm a little anxious about trying to fix all these problems with band-limiting; Particularly because the "noise" that I hear at the output is near the same frequency (7kHz) as my intended output.

Any ideas you all might have to stop the noise would be helpful. I should mention that I already know the 1M on the input is a noise-source, but couldn't figure out a great way to get rid of it. If I DC couple my input, I get hella DC offset and the circuit tries to blow itself apart (yay for current limited supplies!!) . Thanks for all the help so far, it has been very appreciated.

 

First, your input is 200nf and 1M. That calculates to a low frequency limit of .8Hz. You used the word, "audio". Nobody can hear .8 Hz.

Set the input resistor just low enough that it doesn't load down your input source. Then calculate the input capacitor with F(low) = 1/2PiFC

Then pick a feedback resistor that is connected to an inverting input, R4 for instance.
Use F(hi) = 1/2PiFC to set the high limit.
C= 1/2PiFR
C = 1/2Pi 20KHz 200kohms
C = 40pf

Rinse and repeat.

 

Okay, Thanks. That should've been obvious, I suppose. I've dropped my input resistance to 150 Ohm to put the corner frequency right around 5 kHz. I've also tweaked R4 a bit and added a parallel cap to try and take out some of the high-frequency noise. These have helped to remove a lot of the low-frequency noise, actually, which was quite substantial before: Thank you.

I'm still getting significant ringing, though, when I put the receiver near the transmitter and then remove the receiver. I've measured the frequency of this anomalous ringing and it's around 2.5 kHz. If I put my receiver near the transmitter again, I can pick up the desired 7 kHz, but it is all but drown out by the 2.5 kHz. I'm considering trying to change the last stage to a tight band-pass active filter. Any advice on this front?

Thanks again!

 

To get rid of 2500Hz, Add some more poles on the high pass function. You don't need a buffer amplifier for something that has an impedance of less than 150 ohms, so turn your first amp into something with a gain of 2 and add a capacitor in the feedback path, or design a 2 pole high pass filter as the first stage. Use capacitive coupling between stages to make more high pass filters. put an RC or LC in parallel with R8 tuned to have low impedance at 2500Hz. etc.

You have everything you need to make a "tight bandpass active filter". Those are made out of op amps and resistors and capacitors. You could make tighter frequency points all over this circuit. No need to make a separate tight bandpass filter, you are already making one.

 

I have to confess that I haven't read the entire thread.
Is this about acoustic feedback from a speaker to a microphone?

 

I have to confess that I haven't read the entire thread.
Is this about acoustic feedback from a speaker to a microphone?

Heh. No it's not, but I love that you ask anyway. It ended up being a really long thread. Long story short, I'm trying to get a very high-gain amplifier for audio frequencies while keeping ALL noise to an absolute minimum, as the amplifier will be run with the gain turned up until it "finds" an audio signal.

You have everything you need to make a "tight bandpass active filter". Those are made out of op amps and resistors and capacitors. You could make tighter frequency points all over this circuit. No need to make a separate tight bandpass filter, you are already making one.

Alright; Good advice. I'll work to put more low- and high-pass filters in the circuit, and I might, at your suggestion, put a tuned LC across R8 so that it zeros out the gain at 2.5kHz. Thanks so much for all your help, #12. I may revive this thread again if I find another problem.

I have an alright grasp on the fundamentals, but it's time that I start realizing how to put them all together into a coherent system. Thank you all for being a nice support system throughout that process.

 

Hi again. I've looked at designing a tuned LC to go in parallel with R8 and have a few questions about it. The first thing I noticed was that if I put the resonance at 2.5kHz, then the impedance of the LC would short out a 1kOhm at 7 kHz. I've picked values as shown for my tuned LC. I am concerned how small R7 needs to drop to compensate for the small impedance of the feedback loop. Won't something like this load down U4? I would imagine that U4 will now be trying to output too much current and will not be able to swing sufficiently in voltage...

Am I totally off base here? Any ways around this? I don't think I can really increase the impedance of the feedback loop without increasing the broad-band gain...

 

Have you told us what your required gain range is?