Doubts about the "8 Million Gain!" circuit from Talking Electronics

[davethecipo]

Hi,

I built the "8 million gain" circuit http://www.talkingelectronics.com/projects/200TrCcts/200TrCcts.html#8 from TE. It works even if I use an alligator clip cable instead of the copper plate (I don't have the latter). I'm not entirely sure about how it works, so I'd like to ask a few questions:

1) From what I understand, the metal strip has some capacitance to ground. This strip works as one "plate" of the capacitor, while the other "plate" would be the wire connecting the last emitter to the battery negative terminal. AC current flows into this capacitor and gets amplified by the cascade of transistors. Is this correct?
2) Are the resistor values critical? I'm talking about the first two (the last one is obviously limiting the current through the LED). I tried lowering them and the light from the LED looks the same to the naked eye. From what I understand lower resistor values will consume more current but the final result is the same because the last transistor is always saturated. Am I wrong?

Thanks for any help

[rstofer]

Hi,

I built the "8 million gain" circuit http://www.talkingelectronics.com/projects/200TrCcts/200TrCcts.html#8 from TE. It works even if I use an alligator clip cable instead of the copper plate (I don't have the latter). I'm not entirely sure about how it works, so I'd like to ask a few questions:

1) From what I understand, the metal strip has some capacitance to ground. This strip works as one "plate" of the capacitor, while the other "plate" would be the wire connecting the last emitter to the battery negative terminal. AC current flows into this capacitor and gets amplified by the cascade of transistors. Is this correct?

Thanks to the electric fields all around us, our body works like a voltage source (relative to some concept of ground) and the voltage is high enough to eventually provide enough base current to the first stage.

2) Are the resistor values critical? I'm talking about the first two (the last one is obviously limiting the current through the LED). I tried lowering them and the light from the LED looks the same to the naked eye. From what I understand lower resistor values will consume more current but the final result is the same because the last transistor is always saturated. Am I wrong?

If you assume each transistor has a gain of 200, you can kind of work out the math and, frankly, it doesn't work as well as it could.  The first stage will have a collector current, max, of 6V/1M of 6 uA.  To saturate the transistor will take as little as 6 uA divided by 200 or  30 nA of base current.  Now that 6 uA goes into the base of the second stage and, again, there is a x200 multplier so the collector current could be 1.2 mA.  But, wait, the 100k resistor will only allow 60 uA to flow into the base of the third stage.  Again, we have a x200 so we could have as much as 12 mA of collector current and this just might be enough to get decent brightness out of the LED.  All numbers rounded, hacked and bashed.

If I were experimenting, I would leave the 1M resistor alone, reduce the 100k to 47k to get more base current to the 3rd stage.

Since I am too lazy to do the actual calculations, I would put the circuit in LTSpice and see how it works out.  It's a lot easier to experiment by banging on a keyboard than it is to solder bits and pieces.

[davethecipo]

Thanks for the explanation, it's very clear 

[Zero999]

One thing to note is that cirucit does not have a voltage gain of 8 million!

It has high current gain, probably not 8 million, since the Hfe drops at very low currents, but not far off.

The voltage gain is actually slightly less than a single transistor amplifier and needs 3×V_BE drops to turn on. Without doing any calculations, the gain is probably a couple of hundred.

What this circuit does have is a very high input impedance, although there are better methods to achieve this, even with BJTs.

[davethecipo]

@Hero999 thanks for the info. Could you share some link about the better ways?

[Zero999]

The easiest way I can think of is to use a MOSFET with a zener diode to protect its gate from high voltages.

The tiny leakage current through the zener diode should be enough to turn it off. If it's too sensitive, connect the input to 0V via a high value resistor: 1M to 10M.

[MagicSmoker]

@Hero999 thanks for the info. Could you share some link about the better ways?

A similar circuit that is a bit more linear replaces the MOSFET with a JFET (Junction-FET). Look up "JFET electrometer" for examples.

[Zero999]

@Hero999 thanks for the info. Could you share some link about the better ways?

A similar circuit that is a bit more linear replaces the MOSFET with a JFET (Junction-FET). Look up "JFET electrometer" for examples.

Yes, a JFET will work but bear in mind it's a depletion mode device, so the LED will remain on, until a negative electric field is detected.

[yada]

@Hero999 thanks for the info. Could you share some link about the better ways?

A similar circuit that is a bit more linear replaces the MOSFET with a JFET (Junction-FET). Look up "JFET electrometer" for examples.

Yes, a JFET will work but bear in mind it's a depletion mode device, so the LED will remain on, until a negative electric field is detected.

I have some RF jfets and want to actually use one. Do you have an example circuit of how this would work?

I built this circuit with a wire instead of a copper plate and it works differently then I thought. Putting a 8cm lead it keeps the LED off but attaching a 20cm alligator clip to it makes it so sensitive that when you touch/ go near the insulator of the alligator clip it turns on. But some times you cant get it to turn on by touching even the metal of the lead, even with the additional wire.

Is there a way to make your multimeter detect microvolts with this circuit? Or is it just too unreliable to get a real reading; ie it will show nothing or the amplification is so high that when its "on" the voltage is goes out of range. In other words is it so sensitive that its hard to get the voltage inside of the narrow window where its sensitive?

[Zero999]

@Hero999 thanks for the info. Could you share some link about the better ways?

A similar circuit that is a bit more linear replaces the MOSFET with a JFET (Junction-FET). Look up "JFET electrometer" for examples.

Yes, a JFET will work but bear in mind it's a depletion mode device, so the LED will remain on, until a negative electric field is detected.

I have some RF jfets and want to actually use one. Do you have an example circuit of how this would work?

I built this circuit with a wire instead of a copper plate and it works differently then I thought. Putting a 8cm lead it keeps the LED off but attaching a 20cm alligator clip to it makes it so sensitive that when you touch/ go near the insulator of the alligator clip it turns on. But some times you cant get it to turn on by touching even the metal of the lead, even with the additional wire.

Is there a way to make your multimeter detect microvolts with this circuit? Or is it just too unreliable to get a real reading; ie it will show nothing or the amplification is so high that when its "on" the voltage is goes out of range. In other words is it so sensitive that its hard to get the voltage inside of the narrow window where its sensitive?

Which circuit did you build?

[Richard Crowley]

Is there a way to make your multimeter detect microvolts with this circuit?

https://en.wikipedia.org/wiki/Electrometer

[yada]

[yada]

Is there a way to make your multimeter detect microvolts with this circuit?

https://en.wikipedia.org/wiki/Electrometer

That's explains how I would build this:


But I'm not building that device (yes I understand they both measure electricity but I'm not interested in building a historical replica of a hundred year old device), I'm using transistors from that website and a modern day DMM, also that link doesn't answer my question, which was can you use this circuit with a multimeter or is the useable window of sensitivity too narrow?. See this link:
https://en.wikipedia.org/wiki/Question

Those generic answers in a beginners forum are not helpful and kind of irritating when you can't figure something out. That is like asking "How do I change the alternator in a 2012 Honda accord v6?" And some one gives you a course catalogue to the local vocational college. Yes technically that might help them get their car fixed but realistically its not, which anyone with common sense would know. Isn't that the definition of "trolling"?

[Richard Crowley]

The point was that the formal name for an instrument that measures electric fields is an "Electrometer". There are modern versions (a couple of which Dave has demonstrated on his videos.)

For example:
http://www.w-tech.fr/en/component/virtuemart/consumable/digital-electrometer-?Itemid=0

Another thread discussing electrometers and measuring static charge:
https://www.eevblog.com/forum/projects/measuring-static-charge-with-an-electrometer-19267/

[yada]

The point was that the formal name for an instrument that measures electric fields is an "Electrometer". There are modern versions (a couple of which Dave has demonstrated on his videos.)

For example:
http://www.w-tech.fr/en/component/virtuemart/consumable/digital-electrometer-?Itemid=0

Another thread discussing electrometers and measuring static charge:
https://www.eevblog.com/forum/projects/measuring-static-charge-with-an-electrometer-19267/

Thank you thats much more helpful.