I am thinking of using 5 or 6 transistors and making an insanely high gain amplifier (not intended for audio use)
for extreme sensitivity
If it works. would it be possible to put an extremely high ohms resistor to bias it (would a million ohms or 500 thousand ohms or 50,000 ohms be enough?) to get a low but clean audio signal out of the amplifier? (only for testing purposes)
using only a low voltage power source... like 9 volts or less.
the absurdly high gain is for testing purposes with my own experiments..
for extreme sensitivity
If it works. would it be possible to put an extremely high ohms resistor to bias it (would a million ohms or 500 thousand ohms or 50,000 ohms be enough?) to get a low but clean audio signal out of the amplifier? (only for testing purposes)
using only a low voltage power source... like 9 volts or less.
the absurdly high gain is for testing purposes with my own experiments..
More parameters are needed, AC coupled, bandwidth? The noise of 1 Meg Ohm "extremely" amplified could be > 9V.
just a simple bias
Runs off of DC battery.. ac-coupled or not i'll try both
doesn't matter what frequency... if it will work from 1hz to 10,000hz that's okay to me.
Runs off of DC battery.. ac-coupled or not i'll try both
doesn't matter what frequency... if it will work from 1hz to 10,000hz that's okay to me.
Last edited:
Not if he doesn't use feedback.
OP are you bothered about distortion? Current drive? How much gain do you want?
OP are you bothered about distortion? Current drive? How much gain do you want?
A very high gain amplifier is quite capable of finding its own feedback mechanisms: grounds, supply rails, circuit loops, stray capacitance.
Yea, but I get the feeling the OP does not really understand what high gain means, or what the implications of various bias schemes are.
In some ways low noise amps at high gain are easier than the same thing at low gain, making say 50dB gain with Ein of say 3nV/root hz is while not exactly easy not massively hard, doing the same thing with only 10dB of gain is a complete pig.
And, yes, there is pretty much always feedback, so you may as well make sure there is lots of it and that it is negative....
Regards, Dan.
In some ways low noise amps at high gain are easier than the same thing at low gain, making say 50dB gain with Ein of say 3nV/root hz is while not exactly easy not massively hard, doing the same thing with only 10dB of gain is a complete pig.
And, yes, there is pretty much always feedback, so you may as well make sure there is lots of it and that it is negative....
Regards, Dan.
it doesnt matter how bad it distorts if it puts out a decent voltage to make a speaker cone move i'll be happy
and if it will turn on at 0.001 of a volt or less I'll be happy
its not an audio amplifier for any clean signals at all.. just extremely high gain.. really really high gain. like higher gain than a 40,000 watt car amplifier in bridged mode but far less output power overall maybe less than a tenth of an amp or so
running off of just a 3 volt or 5 volt power supply... nothing powerful..
it doesnt matter how bad or what the waveshape of the output is... if the output is consistent at least it'll be perfect.
i've actually one time got a amplifier that ran off of 2 triple A batteries it sounded prefectly clear with my computers headphone output hooked to it. not too loud but perfectly listen-to-able a few feet away
and biased weakly (just trying different resistors until it sounded clear and best volume)
and if it will turn on at 0.001 of a volt or less I'll be happy
its not an audio amplifier for any clean signals at all.. just extremely high gain.. really really high gain. like higher gain than a 40,000 watt car amplifier in bridged mode but far less output power overall maybe less than a tenth of an amp or so
running off of just a 3 volt or 5 volt power supply... nothing powerful..
it doesnt matter how bad or what the waveshape of the output is... if the output is consistent at least it'll be perfect.
i've actually one time got a amplifier that ran off of 2 triple A batteries it sounded prefectly clear with my computers headphone output hooked to it. not too loud but perfectly listen-to-able a few feet away
and biased weakly (just trying different resistors until it sounded clear and best volume)
Last edited:
Does it have to be discrete ? Whats wrong with FET input opamp ?
1 Meg ohm input impedance isn't really high in the scheme of things. Use a dual opamp and split the gain.
1 Meg ohm input impedance isn't really high in the scheme of things. Use a dual opamp and split the gain.
I don't know how to do that o-O
I only know how to make a simple (single transistor amplifier)
but i know how to wire more of them together to get more gain..
but I wanted to get lots of gain without having a big output voltage
I have some sensitive transistors... but I need them to be extremely.. extremely sensitive..
the transistors are all the same type.. so far i've gotten 3 of them wired and the gain is massive already.... so I wanted to get enough gain that just touching the positive or negative input wire it would turn on the transistors.. and make the speaker cone move
not a big amount... but just a lot of gain.
I only know how to make a simple (single transistor amplifier)
but i know how to wire more of them together to get more gain..
but I wanted to get lots of gain without having a big output voltage
I have some sensitive transistors... but I need them to be extremely.. extremely sensitive..
the transistors are all the same type.. so far i've gotten 3 of them wired and the gain is massive already.... so I wanted to get enough gain that just touching the positive or negative input wire it would turn on the transistors.. and make the speaker cone move
not a big amount... but just a lot of gain.
Well for 1V out and 1mV (0.001V) input you only need 60dB of gain, and that is nearly trivial, ANY microphone preamp worthy of the name will manage that without breaking a sweat (And will do it with low distortion).
This is not a high gain application, and is nearly trivial with any of the common opamps, something with two stages having 30dB of gain each (7 resistors and a few caps by my count), maybe with a simple totem pole output stage to drive a small speaker will get the job done, for high Z input maybe preceed it with a jfet buffer or such.
What is the application?
Regards, Dan.
This is not a high gain application, and is nearly trivial with any of the common opamps, something with two stages having 30dB of gain each (7 resistors and a few caps by my count), maybe with a simple totem pole output stage to drive a small speaker will get the job done, for high Z input maybe preceed it with a jfet buffer or such.
What is the application?
Regards, Dan.
"There is no such thing as too much feedback"
Bruno Putzeys, Linear Audio Volume 1, April 2011, p112.
Bruno Putzeys, Linear Audio Volume 1, April 2011, p112.
I think the question that needs to be asked is: What kind of gain? Do you want voltage, current, or power gain?
Voltage gain would simply boost the hell out of a voltage signal. A ton of voltage gain with no feedback is going to be all but useless... It's basically going to become a comparator or it will find every source of noise and most likely make you turn it off really fast. Haha. A ton of gain with feedback will have to be carefully controlled to not oscillate. (Just for reference, you can sneeze and get 100dB of OPEN LOOP gain from an op-amp. )
Current gain.. Technically would be something like a BJT, put a little bit of current in, and get a lot more out. Although it can also be construed to mean something that allows you to output a lot more current that the device with which you are driving it.
Power gain, can be a mix of the two above, the basic idea is that it can deliver a lot more power to your load than the device with which you are powering it.
What's you end goal for the amplifier? You mentioned just having something that you can touch and make the speaker cone move. A good depletion MOSFET or a JFET will do this if configured for high gain. Any charge you introduce to the gate or pull from will modulate the channel and cause a voltage swing.
Voltage gain would simply boost the hell out of a voltage signal. A ton of voltage gain with no feedback is going to be all but useless... It's basically going to become a comparator or it will find every source of noise and most likely make you turn it off really fast. Haha. A ton of gain with feedback will have to be carefully controlled to not oscillate. (Just for reference, you can sneeze and get 100dB of OPEN LOOP gain from an op-amp. )
Current gain.. Technically would be something like a BJT, put a little bit of current in, and get a lot more out. Although it can also be construed to mean something that allows you to output a lot more current that the device with which you are driving it.
Power gain, can be a mix of the two above, the basic idea is that it can deliver a lot more power to your load than the device with which you are powering it.
What's you end goal for the amplifier? You mentioned just having something that you can touch and make the speaker cone move. A good depletion MOSFET or a JFET will do this if configured for high gain. Any charge you introduce to the gate or pull from will modulate the channel and cause a voltage swing.
100 dB gain at 10 kHz , this is not trivial or even possible and certainly not with a few transistors.
Perhaps this thread by X-Pro can provide useful ideas.
http://www.diyaudio.com/forums/solid-state/104205-vas-cd-cb-stage.html
Differential input and gain stage, feeding a high input-impedance buffer (to minimize loading of the first gain node and Miller effects), next a folded-cascode additional gain stage, followed by the output stage (preferably also a high input-impedance configuration for maximum gain.
hth
http://www.diyaudio.com/forums/solid-state/104205-vas-cd-cb-stage.html
Differential input and gain stage, feeding a high input-impedance buffer (to minimize loading of the first gain node and Miller effects), next a folded-cascode additional gain stage, followed by the output stage (preferably also a high input-impedance configuration for maximum gain.
hth
I got 120dB of variable gain at 9Mhz on a eurocard easily enough, 4 * AD600 VGA chips with a J310 10dB stage on the front to lower the noise level, based on work done by pa3ake.
100dB of gain at audio is usually pointless (You will have a **FUN** with noise and power line interference), but is not hard to achieve if you really need it (Some sonar applications in the low ultrasonic region are not far off).
Regards, Dan.
100dB of gain at audio is usually pointless (You will have a **FUN** with noise and power line interference), but is not hard to achieve if you really need it (Some sonar applications in the low ultrasonic region are not far off).
Regards, Dan.
- Status
- Not open for further replies.