4

I have found in Dieter Nuhrmann's book an interesting approach to amplifier circuit and decided to test it.

Unfortunately the Multisim simulation does not work as expected. The signal is not amplified at the output.

What's wrong or what I'm doing wrong?

Screenshot of Multisim Photo of the circuit from an old book

try-catch-finally
  • 1,254
  • 3
  • 19
  • 37
Jerzy Przezdziecki
  • 657
  • 3
  • 12
  • A proper schematic would be nice, rather than squinting at that screen shot. What do you get? What do you expect to get? – Puffafish Feb 06 '17 at 10:42
  • 1
    Have you tried turning down the input signal? The example has 0.2Vpp input and you seem to have 2Vpp input. – pjc50 Feb 06 '17 at 10:43
  • 1
    Get rid of the 2k7 and make the 330R feedback resistor more like 10 k and see what happens. You should also try it with a cmos gate – Andy aka Feb 06 '17 at 10:44
  • yes, i have tweaked since 0.1 Vp - 1 V - does not react. – Jerzy Przezdziecki Feb 06 '17 at 10:44
  • 10
    The circuit seems to rely on the analog behavior of the 7400, which is, to my knowledge, not specified anywhere. I doubt multisim is trying to be accurate on this aspect. – dim Feb 06 '17 at 10:45
  • Yes you are right. It's rely on analog behaviour. Maybe MS can't handle it? Anyway the scope screen provided by the author shows that it works with TTL... – Jerzy Przezdziecki Feb 06 '17 at 10:50
  • 2
    I advise to build one in real life. Simulators only simulate... –  Feb 06 '17 at 12:57
  • Are you using 330 or 12k for feedback? 330 is never going to work. – WhatRoughBeast Feb 06 '17 at 17:58

3 Answers3

3

The simulator is not modeling the NAND gate as a linear device, but just its digital characteristics. Also, the resistor bias from the gate input to ground may be touchy, but you're not going to be able to simulate this behavior. If you really want to try this circuit, you're going to have to breadboard it.

Paul Elliott
  • 990
  • 4
  • 7
2

I've built amplifiers out of 7400 NANDs. Excellent bandwidth. Need to bias the input node into linear range, because of the internal silicon resistor pullup. If you use 74LS, the internal resistor is 40,000 ohms (value from memory), and you need a resistor >> 2.7K.

For CMOS, you don't need any pulldown to GND.

Beware these circuits are easy to oscillate. GND inductance is a bother.

analogsystemsrf
  • 33,703
  • 2
  • 18
  • 46
-2

A few possibilities.

  1. The circuit ISS wrong.
  2. The circuit used the wrong parts.
  3. The circuit used the wrong part values.
  4. The simulator is wrong.
  5. The circuit is wired wrong in the simulator.
  6. The simulator had the wrong mods.
  7. The simulator had the wrong parts. ....

It is not hard to rule out each of the possibilities. The circuit requires an unbuffered gate, and. Part values several magnitudes higher than what you have.

dannyf
  • 4,272
  • 1
  • 7
  • 9
  • 2
    I'm sorry, dut I don't consider this answer useful. You don't explain what is the most likely reason and why. You don't even exclude the hypothesis that are obviously not the reason (e.g. 5). You don't give alternatives, nor indicate that there are no alternatives. You also seem to state (but the sentence isn't even correct, so I'm not sure) that he used part values order of magnitude different, but he uses the same values as the original circuit, so I don't understand. – dim Feb 06 '17 at 13:41