Amplifier with very low input- design search

• posted

Hello guys,

I am building an extremely low input (voltage) amplifier, with about .2 volt spikes and also -0.2 volt spikes. I just need to detect if there is a voltage, either positive or negative then turn on a led when it is detected. My questions are as followes,

- do I need negative supply ( in the case i use an opamp, i guess I will need +/- supply)

- Why can I use at the input to invert the voltage? (is it possible with that low voltage)

- I will be driving this with with hopefully a 3v battery

Regards,

ken

• posted

probably not

try capacitive coupling into a 2 x comparators or a window detector comparator arrangement.

Maybe stick a monostable on the output to keep the LED on long enough for you to see.

• posted

Would be a lot easier to help if you'd explain what you're trying to do. Are these spikes a picosecond wide or an hour? Fast rise? slow rise? What if you get two or five or a dozen? Do you care whether it was - or + voltage spike? If it's only +0.19V is it not considered a spike? Accuracy required? How long does the led stay on? forever? how reset? What's the source impedance of the signal you're measuring? Has the Source got some DC voltage on it?

Something that works for the general case is likely very hard. What you need may be trivial...if we knew what that was.

Defining the problem is the most important part of finding a solution.

• posted

If you capacitively couple the input, you may be able to get away without a negative supply. You would bias up the output side of the capacitor to a higher voltage, suitable for the "untriggered" voltage of the detector circuit.

One way to do this would be to detect the positive- and negative-going pulses separately. For the positive side, you'd run the signal through a small capacitor, and bias the output of the cap up to about .6 volts using a pair of high-impedance resistors to V+ and V-. This point would then go through a small resistor to the base of an NPN transistor, emitter grounded, collector tied to V+ through a few K-ohms and feeding the output. Normally the transistor would be off; when a positive-going pulse comes in, the 200 millivolt signal would pull the base up above .7 volts and turn the transistor on, pulling the collector down and making a nice strong negative-going pulse at the output.

Repeat the whole thing, upside down, with a PNP transistor, to get a positive-going pulse (up to +3 volts) for a positive-going pulse at the input.

You could then use the two (amplified) pulse outputs to trigger some sort of latch or other circuit to turn on an LED for as long as you want it. The LMC555 chip (CMOS version of the venerable LM555) will run down to 1.5 volts.

• posted

HI,

its actually the output of a walki talkie. Instead of the earphone output, I hooked up to wires to output the signal on a board and amplify it. When t he button is push and release (without talking with the sending radio) the receiving walkie talkie, there is noise coming in, sometime negative , pos itive or both. Even that is good enough for me to catch the signal and ampl ify it. But its not hign voltage 100 to 300 mv.

Ken

• posted

.

, I hooked up to wires to output the signal on a board and amplify it. When the button is push and release (without talking with the sending radio) th e receiving walkie talkie, there is noise coming in, sometime negative , p ositive or both. Even that is good enough for me to catch the signal and am plify it. But its not hign voltage 100 to 300 mv.

Prebias a transistor at 0.55v Vbe. Capacitively couple the audio into it.

NT

• posted

ut

I
e

gh

do.

.

ut, I hooked up to wires to output the signal on a board and amplify it. Wh en the button is push and release (without talking with the sending radio) the receiving walkie talkie, there is noise coming in, sometime negative , positive or both. Even that is good enough for me to catch the signal and amplify it. But its not hign voltage 100 to 300 mv.

HI NT,

I'm not quiet sure I understand what you mean by 'Prebias a transistor at 0 .55v Vbe' can you elaborate on this? thank you,

Ken

• posted

.

, I hooked up to wires to output the signal on a board and amplify it. When the button is push and release (without talking with the sending radio) th e receiving walkie talkie, there is noise coming in, sometime negative , p ositive or both. Even that is good enough for me to catch the signal and am plify it. But its not hign voltage 100 to 300 mv.

What are you making?

• posted

ut

I
e

gh

do.

.

ut, I hooked up to wires to output the signal on a board and amplify it. Wh en the button is push and release (without talking with the sending radio) the receiving walkie talkie, there is noise coming in, sometime negative , positive or both. Even that is good enough for me to catch the signal and amplify it. But its not hign voltage 100 to 300 mv.

Its kind of a light alarm for the depth people. exept on a wireless radio.

Ken

• posted

bout

if

ed

s I

ble

or

ough

o do.

e?

on.

tput, I hooked up to wires to output the signal on a board and amplify it. When the button is push and release (without talking with the sending radio ) the receiving walkie talkie, there is noise coming in, sometime negative , positive or both. Even that is good enough for me to catch the signal an d amplify it. But its not hign voltage 100 to 300 mv.

t.

0.55v Vbe' can you elaborate on this?

. +----------+------o + . | | . R1 LED . | | .o-||---+--------(_) . | | . R2 | . | | . +----------+-----o 0v

NT

• posted

At the risk of repeating myself...the hardest part of finding a solution is defining the problem. You're vague and people have to make a lot of assumptions. So, here go some assumptions...

If you mean 3.00V battery, you may have enough volts to light a LED. If it's two NiMh batteries that start out at 2.4V and are expected to work at full discharge 2.0V, you may not have enough volts.

Still unclear whether you want to latch the led on or make the person stare at it all the time??? Latching multiplies the amount of stuff you need.

What walkie talkies??? Many FRS radios have a privacy code that transmits a sub-audible tone with the voice. If they don't do a very good job of filtering that, you may get enough out the headphone jack to detect that. Set it for the highest sub-audible frequency. Some have a call or alert button that puts out a BIG audio signal that you could detect. With that, you could put a voltage doubler on the output and drive the LED directly.

Some FRS radios have an alert indicator that tells you when someone called.

The second hardest part of a typical DIY project is packaging. Starting with something already packaged and modifying to suit solves that problem.

There are many devices for the hard of hearing. You may be able to start with something like this and modify it:

If you don't need long range, EBAY is full of wireless doorbells that could likely be modified to do what you want.

The devil is in the details.

• posted

Too tired to draw ASCII or fire up PSpice, but in words...

First NPN, emitter at ground

Resistor from Vcc to base of first NPN, setting 1mA

Resistor from base of first NPN to collector of first NPN, 100 Ohms, thus collector voltage of first NPN is at Vbe-0.1V ~ 0.5V

Second NPN, emitter to ground, base to collector of first NPN.

Resistor from Vcc to collector of second NPN, say nominally chosen such that IF second NPN was fully on, IC=1mA

However base of second NPN is at 0.5V, it's off, or very low conduction, so collector voltage is essentially at +Vcc

Add coupling capacitor from signal to base of second NPN.

When signal positive peak exceeds 0.1V, second NPN conducts and collector voltage falls.

Solution for negative peak is left as an exercise for the student ;-)

(I'll provide that tomorrow after you scratch head for awhile... it's wine time in the old west >:-} ...Jim Thompson

```--
| James E.Thompson                                 |    mens     |
• posted

More complex than it need be, OP only wants to detect audio

Uhh, I left out the LEd resistor :)

NT

• posted

Did the OP say he was running it off a car battery with all that bias current?

• posted

o.

ut, I hooked up to wires to output the signal on a board and amplify it. Wh en the button is push and release (without talking with the sending radio) the receiving walkie talkie, there is noise coming in, sometime negative , positive or both. Even that is good enough for me to catch the signal and amplify it. But its not hign voltage 100 to 300 mv.

d.

2CX5WGK

tem43d5f8a1d5

Hi,

I believe I will re explain my project better, it seems it is unclear... I am using two uniden GMR3740 (radio). This come with a headset. My device is hooked up in the headset audio output. My device is only an opamp, a sc r, few resistors, a led and a small battery for powering those components. Wherever some else tries to communicate with me with the other radio, the l ed on my device lights up. Then I unhook my device to be able to talk. It d oes work, but only at 75% of the time.

Ken

• posted

Repeating one more time...there are features in your radios that can be used to advantage. You have a fixed environment you want to work in.

I see three quick possibilities...

ONE Set the subcode on the transmitting end to the highest frequency. Code 38 is 250.3 Hz. If they don't do a good job of filtering the low frequency codes, you may be able to see it at the headphone output of the receiver and use that to trigger your led.

Sub Code Each of the channels 1-22 may have any one of the codes, OFF, or 1-121 selected. Code oF(OFF) indicates no Sub code selected and your radio can receive a signal regardless of the code settings of the transmitting radio.

1) Press MENU/ to enter Menumode. The subcode indicator flashes. 2) Press CH ?/ CH ?key to increase or decrease the code number displayed. You can also select oF(OFF) at this stage. 3) Press MENU/ to exit Menu mode.

dos Looks like call tone might be used to trigger your led.

Call Tone Your radio is equipped with 10 selectable call tones that are transmitted when CALL/LOCKis pressed. To select a call tone:

1) Press MENU/ repeatedly until the CALL icon and Call Tonestart to blink on the display. 2) Press CH ?/ CH ?to move through the available call tones. Each tone will be heard through the speaker but will not be transmitted. Cycle through these tones and stay on the tone you want to select. 3) Exit Call Tone to set the selected tone.

Press transmit and release to get the rogerbeep to trigger the led. Press transmit again if you want to talk. Rogerbeep should be plenty high amplitude to trigger your SCR directly...as is call tone.

three RoGeRBeep Roger Beep is a BEEP that is sent to notify the end of transmission (both PTT and VOX transmission). Roger Beep can be heard through the speaker when both Roger Beep and Key Beep are set to ON. When Roger Beep is set to ON and Key Beep is set to OFF, Roger Beep will not be heard from the speaker but it will be transmitted to your party. When Roger Beep is set to OFF, Roger Beep will neither be heard nor transmitted. To change Roger Beep setting: Press and hold VOL ?while turning the radio on and off. If Roger Beep is ON, it will be turned off; if it is OFF, it will be turned on.

You'll have to try it, but I'd expect that the last two are high enough amplitude that you can use a voltage doubler to light the led directly...but it won't latch.

• posted

Some chips can handle small negative inputs, eg: LM339 however you have nothing egative to compare the negative signal to. so a closer look at the problem is probably a better starting point.

```--
umop apisdn```
• posted

Are you permitted to transmit data on the band they operate on ?

```--
umop apisdn```
• posted

Maybe something like this...

You could also make this up with a dual comparator. ...Jim Thompson

```--
| James E.Thompson                                 |    mens     |