Author Topic: Rectifier and Peak detector  (Read 2010 times)

0 Members and 1 Guest are viewing this topic.

Offline akis

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: gb
Rectifier and Peak detector
« on: May 25, 2012, 05:23:55 am »
I would like to construct a simple rectifier and peak detector to allow me to measure floating voltages up to +/- 100V @ 500 KHz

The requirements are as follows:
1. ability to measure the peak of  signals up to +/- 100 V AC
2. up to 500 KHz = DC not required
3. ability to work well for small signals, eg +/- 1 V and large signals eg +/- 100V.
4. Ability to measure floating voltages, ie no earth connection on the probes
5. Powered by a battery, so a virtual ground would be required

My digital multimeters work very well, but cannot handle "high" frequencies (anything over 10-20K)
The oscilloscope works well, but is quite indirect (ie you have to align the curve and then try to count divisions and sub-divisions etc), plus it is grounded and makes life very difficult
An DSO would ordinarily have the value displayed on the screen, but I have not bought one yet.

Initially I constructed a simple peak detector, ie a diode charging a cap and a resistor. With values of 100nF, 1M and for diode 1N4148. This works remarkably well but
(a) it loads the component under test
(b) the diode voltage drop error is very pronounced at small voltages, for example pct-wise, 600mV matter a lot on a 3V signal
(c) the choice of diode depends on small signal accuracy vs large signal capability (eg I need to use a 1N400x when the voltage rises).

Trying to improve on the above, I constructed an op-amp half-wave rectifier which will eliminate the diode voltage drop error and also provide high input impedance, for example using maybe a JFET op-amp.

To solve the problem of high input voltages, I will have a switch with a fixed resistor potentiometer to do 1X, 10X, 100X, presenting a 1M resistance to the component under test (hopefully). On the diagram it shows as 10X.

However the choice of the charging cap and discharge resistor is problematic, as is performance at higher frequencies (up to 500 KHz). For higher frequencies we want a small cap and large resistor (as shown), and for low frequencies we would want a larger cap.

Is there a way to modify the circuit to be useable in both low and high frequencies?

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo