PG 100 PULSE GENERATOR

The purpose of the PG 100 is to provide a pulse with a relatively fast risetime that, with an oscilloscope, can make quick time-domain measurements of cables and other components around the lab.  Measurements of this kind will have a tolerance of about 5%.  While more accurate measurements can be made with other instruments, this is an inexpensive highly adaptable method of getting answers quickly.  Operation of the device is very simple: turn the switch on, and there’s the pulse.

The PG 100 is in a small box measuring 4.1” long, 1.5” high, and 2.6” deep and is powered by an internal 9 volt battery.  The output connector and power switch protrude through the top of the box.  There is only one output from the unit: an 8-volt pulse through 50 ohms, with a 1.5 ns risetime, pulse width of 380 ns and repetition rate of about 200 KHz.  These parameters will vary between units approximately ± 10%.

For a complete set of specifications for the PG 100, see here.

With the use of an oscilloscope, the PG 100 can be used to make a number of measurements.  These include several characteristics of cables, capacitance, inductance and resistance.  The PG-100 can also be used to verify the scope risetime, trigger capability, and as a source of trigger for other devices.

The following pictures show the pulse generated by the PG 100.  The risetime shown averages 1.44 ns.  It was captured using an oscilloscope with 1 GHz bandwidth, or 0.35 ns risetime.  When the scope’s risetime is removed using the square-root-of-the-sum-of-the-squares method, the PG 100 risetime is 1.4 ns.  Using the value in Fig. 1,

Or in our case,

Fig. 1.  Leading edge of the PG 100 pulse as measured with an oscilloscope with 1 GHz bandwidth, 0.35 ns risetime.  With the scope’s risetime extracted, the PG 100 risetime is calculated to be 1.40 ns.

The width of the pulse is shown in Fig. 2, reading 376 ns as measured by the two vertical cursor bars.

Fig. 2.  The pulse width shown is about 376 ns.

Fig. 3.  The two vertical cursor bars in the picture show that the period of the waveform is 5.2 microseconds or repetition rate of about 190 KHz.

Inexpensive TDR

The PG 100 and the oscilloscope are connected as shown in the diagram in Fig. 4 to make an inexpensive though slow-speed Time-Domain Reflectometer (TDR).  The scope should have a 50 ohm input and the cables between the PG 100 and the “T” and the oscilloscope should be as short as possible.  The “T” for the pictures following was connected directly to the PG 100 and the oscilloscope.  The scope should have a risetime of no more than 3.5 nanosecond.  To get the most resolution from this poor man’s TDR, the scope’s risetime should be significantly faster than the PG 100’s risetime.   The risetime of the 1 GHz oscilloscope used for Fig. 1 is 0.35 nanoseconds and is more than adequate.

Fig. 4.  This shows how to connect the PG 100 with an oscilloscope, a BNC “T” connector and a cable leading to the device under test to make a simple Time Domain Reflectometer, a TDR.  The coax represents the device being tested.

Measuring Cable Length

The picture in Fig. 5 shows the waveform from this TDR setup as used to measure the length of a coaxial cable.  The rising edge on the left represents the step from the PG 100 as it reaches the oscilloscope and the cable.  We wait almost 5 horizontal divisions for the reflection from the end of the cable.  The 5 divisions, nearly 100 ns, represent the time for the pulse edge to travel to the end of the cable and be reflected back to the oscilloscope.  The length of the cable is half of that time, slightly less than 50 ns.  The reflected step is positive, showing that the cable is open at its end.  If the cable was shorted, the reflected step would have been negative, and if the cable had been terminated in its own impedance, there would be no reflection.

The waveform was strategically positioned to not show additional reflections caused by the pulse being reflected by the parallel combination of the PG 100 and the oscilloscope input impedance, 25 ohms, out to the end of the cable and back to the oscilloscope again.

This is a very useful measurement technique for finding faults in cables, coax or twisted-pair or whatever.

Fig. 5.  This is the TDR waveform.  The rising edge near the center of the display represents the pulse seen by the oscilloscope.  The second vertical step of smaller amplitude represents the pulse having reached the end of the cable and reflected back to the oscilloscope.

Measure Cable Impedance

Now let’s determine the impedance of the coaxial cable.  While there are several ways of doing that, the easiest one is to connect a small variable resistor across the open end of the cable and adjust it so there is no reflection.  Disconnect the variable resistor and measure it with your DVM.  Since you adjusted it so there was no reflection, the pot is the same resistance as the impedance of the coax.

Another method is to use the height of the reflected waveform.  To make this method easy to use, remove the 50-ohm termination from the oscilloscope.  This will cause a reflection from the unterminated transmission line from the “T” to the scope, but we can ignore it.  Connect a piece of known impedance coax to the other port of the “T” connector.  Measure the height of the first edge and the height of the reflection, i.e., from the top of the first step to the top of the reflection.  Calculate the Reflection Coeficient ρ which is the reflection amplitude divided by the incident step amplitude.   The polarity of ρ is important.  It is positive if the reflection is more positive than the first level and negative if it is below.  The impedance of the cable¹ is:

                                   Z = 50 (1 + ρ)
                                               (1 - ρ)

As an example, I connected a 5-foot cable to the “T” and put an unterminated 2X attenuator to the other end and observed the waveform shown in Fig. 6.  I measured the height of the first step at 4.0 volts.  The reflected step was 1.0 volts.  I computed ρ as 1.0/4.0 = 0.25.  Then

                                    Z = 50 (1 + 0.25) = 83.3 ohms.
                                               (1 – 0.25)

A quick check with an ohmmeter gave 84 ohms looking in one end of the unterminated X2 attenuator.  Not bad.

Fig. 6.  Using the reflection coefficient to determine an unknown impedance.   The spike on the leading edge of the step is due to the reflection of the short section of coax between the PG 100 and the oscilloscope with high impedance input.  In this case, the coax is really the length of a BNC-to-BNC adapter, part of the “T”  and the signal path inside the scope.

Determine Cable Dielectric Constant

Another thing we can learn about a piece of coaxial cable is what kind of dielectric it has.  We need to know the mechanical length of a piece of cable and its electrical length (delay time).  The velocity factor² is given by:

Velocity factor = Time for light to travel the equivalent cable length in air
                                 Delay time of cable with unknown dielectric

The velocity of light is 300 million meters per second, or more appropriately for our needs, one nanosecond per 30 cm of cable.

A table of common velocity factors² is shown below.

So, if we had a coax cable 6 feet or about 183 cm long, and it had a solid polyethylene dielectric, we would expect its electrical length to be:

Measure Capacitance

Now let’s measure capacitance.  I had a very pretty little red-bodied capacitor with color-coded dots around the edge, and I’m sure there’s a way to decode the value from them, but here’s another way to find the approximate value of the capacitance.

Remove the coax connected to the “T”, and replace it with the capacitor.  Depending on the size of the capacitor, you should see a display like the one shown in Fig. 7 below.

Fig. 7.  This shows the RC charging curve for a capacitor that replaced the coax in our poor man’s TDR.  The curve doesn’t quite get to the full amplitude of the pulse, but without the capacitor connected, the full height can be measured.

Measure the time between the beginning of the pulse and 63% of the pulse amplitude like the vertical cursors positioned.  This is the RC time constant.  Time Constant, t = RC where R is the resistance in the circuit which is 25 ohms.  Capacitance, C then equals t/R.  In the example, the time constant is 100 ns, and the capacitance, C = t/R or 100ns/25 ohms = 4 nf.  The resistance, 25 ohms, is made of the 50 ohms inside the pulse generator in parallel with the 50 ohms of the scope input.  It’s a pretty little capacitor.

If the RC curve doesn’t go to the full height of the pulse, you can add resistance across it until it does.  Then the R in the equation is the parallel combination of the generator, the scope input and the parallel resistor you added.

Alternatively, if the capacitor is small enough for the RC curve to go to full height in time that is close to the risetime of the pulse, you must add a resistor in series between the generator and the capacitor to achieve good accuracy.  You must also use a high-impedance probe to get accurate measurements.  Connect the probe across the capacitor.

Measuring small inductors like those found around some switching regulators is also possible, but the curve is inverted from that of the capacitor.

There are lots of other measurements that can be done with a pulse generator and an oscilloscope.  These examples are meant to be the beginning of a learning exercise for those who don’t have experience with time-domain measurements.  We’re interested in learning about other examples you might have experienced with an oscilloscope and pulse generator.

^1. Reference Data for Radio Engineers, fourth edition, pp562.
^2.  The ARRL Handbook for Radio Amateurs, 1999 Table 19.1.

Neither picosecond ATE Inc. nor its owners or assigns are to be held liable for any statement or any misunderstanding of the information contained in this document.

Information: picosecondate info