0
\$\begingroup\$

My goal is to input 30MHz sine pulses, with a pulse width of 10us and gap of 90us between the pulses. The circuit should convert this pulse train to a TTL signal that goes from 0 to 5 volts.

To achieve this I run the pulse train through a peak/envelope detector, consisting of a diode, a resistor and a capacitor. This will capture the peak of the pulse. These peaks are then recieved by a comparator that outputs 5 volts whenever the input voltage exceeds the 1 volt reference voltage.

Circuit

Circuit

Pulse train

Pulse train

One pulse

One pulse

Green signal is the output from the peak/envelope detector

Green signal is the output from the peak/envelope detector.

When trying to verify this on a breadboard I ran into a problem. The peak/envelope detector does not detect the peak of the input signal but instead outputs the exact same signal as the input (I did not save the oscilloscope picture.) My concern is that the diode is not fast enough to handle signals of this frequency (30MHz.)

When looking at the output of the peak/envelope detector in LTSpice it looks fine:

Output of peak/envelope detector

This (below) is what an input pulse looks like after the diode.

After diode

Does anyone have any thoughts about this?

\$\endgroup\$

2 Answers 2

1
\$\begingroup\$

Your concerns are correct, that the 1N5817 diode has too much capacitance to rectify 30 MHz as it is rated for 1A.

Since diode capacitance reduces with current rating, use a fast low current signal diode with a C rating such as the 1N4148 , \$C_D=4 pF\$ with a bulk resistance ~ 40 ohms max @ 10 mA.

The diode capacitance will need to be less than 2% of the load capacitance to rectify and hold a good signal in the output cap.

The 1N4148 has a \$t_{rr} <= 4ns\$ reverse recovery time

The load R will speed up the response but also attenuate it if low relative to the bulk resistance.

40 Ohms is also the impedance of 40 nF at 100 kHz.

Therefore I recommend 1k//1nF for a load with an RC=T=1us or less depending on your SNR requirements.

For faster RF signals a PIN diode would be used.

\$\endgroup\$
0
\$\begingroup\$

Does anyone have any thoughts about this?

You bet. Your diode output has no load. As a result, with no current through it, it doesn't act as a rectifier.

Put a resistor to ground at the diode output and see what happens. Since this IS RF, the standard load would be something like 50 ohms, depending on what coaxial cable you are using to supply the signal to your circuit.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you! I edited the circuit and the last picture \$\endgroup\$
    – Lucjoh
    May 15 at 11:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.