How to use an oscilloscope

[bgolash]

Hi

Just picked up an oscilloscope on ebay. I know nothing about these devices.
I`m interested in checking my encoder and step and direction lines for noise.
I`m wondering what settings to use etc.
This is what I have : Tektronics 442 Dual trace 35 MHz

Regards Barry

[Al_The_Man]

If you understand the basic principal, then it should make it easy to learn.
The sweep (X) or the trace from left to right is TIME, and the verticle (Y) is AMPLITUDE, this can be a + or - figure with ref to the base line, We won't mention the Z at this point.
For example if you wanted to display one cycle of the AC from a low voltage secondary for e.g. the trace time for 1 cycle is 1/60 = 16ms. you would set the TIME trace for say 2ms/div, the amplitude or input volts/division to suit the amplitude of the signal, if the signal was 10v and you set the display for 1v/div then it would display one cycle of the sine wave 10 divs high.
Al.

[One of Many]

Here is a basic tutorial on the function.

Using an O-Scope

One of the first things you should do is isolate the ground conductor from the power cord. I use one of the old adapters from 2 prong to 3 prong rubber block style. Some signals do not reference ground, so in effect are above earth potential where your probe grounding clip will attach in the circuit. If the Scope is grounded, expect the worst. Most circuits and tests are common ground, but those that are not will be loaded(read that as direct shorted to ground) if that is where you need to connect your reference.

DC

[gar]

070502-1432 EST USA

bgolash:

You need a basic understanding on how a electrostatic cathode ray oscilloscope works.

Most of these scopes use a single electron beam focused on the phosopher screen. The beam position on the screen is controlled by a pair of electrostatic horizontal deflection plates. And a second pair for vertical deflection. Deflection is proportional to the voltage difference between the pair of plates. Horizontal and vertical deflection are independent of each other. Spot intensity is controlled by the voltage on the grid electrode.

It takes a large voltage difference on the deflection plates to position the beam at the edge of the screen. Thus, amplifiers are required between the signal you want to observe and the deflection plates in most applications. It should be noted that most TV sets use magnetic deflection instead of electrostatic deflection.

The X-axis is generally associated with the horizontal deflection, and Y with vertical.

Most modren scopes will have amplifiers with DC and AC coupling.

Usually your X-axis is fed from an internal horizontal sweep generator. This is a device that generates a saw tooth waveform. A saw tooth waveform has a voltage that changes directly proportional to time, and then quickly returns to its starting point and repeats in the free running mode. There is also a single shot mode where only one sweep occurs per trigger pulse. The usual mode of operation is for the sweep to start at the left side of the screen and travel to the right side with increasing time.

There is a selector switch on the X-axis that selects the sweep speed or an external signal. Al has already described setting the time base.

Usually you do not use a time base generator in the vertical direction, and therefore there is only an amplifier to which you apply the signal you want to observe with respect to time if the X-axis is set in a sweep mode.

To observe your pulse signals start with the time base in free running mode, internal sync from channel A (or whatever it is called). Start with positive slope, AC coupling, and the trigger point at mid position. Maybe 1 ms/cm for the sweep rate. You should see one or two horizontal lines. If not select channel A, set its gain to 10 V/cm, and Channel A vertical position to mid position. Now adjust the intensity knob for a not very bright line. If only a bright spot, then there is a horizontal sweep problem. Do not run the scope with the spot in one position.

The vertical position knob for the vertical channel of interest should move the horizontal line up and down.

Next connect Channel A input to the 1 V calibration signal. You should see a square wave signal. Adjust your sync threshold to obtain sync. You may also need to adjust the time base.

See if you can get this far.

To see some actual waveforms go to my web site www.beta-a2.com and the CAT-5E waveforms page. Here time is from left to right and some different time base values. The first picture is in dual trace mode, and shows that it takes 200 nS for the pulse to travel 150 ft. This is about 142,000 miles/sec and the speed of light is 186,000. So the velocity of pulse propogation is about 76% of the speed of light and this is a reasonable value to expect.

.

[paulC]

Originally Posted by One of Many Here is a basic tutorial on the function. Using an O-Scope One of the first things you should do is isolate the ground conductor from the power cord. I use one of the old adapters from 2 prong to 3 prong rubber block style. Some signals do not reference ground, so in effect are above earth potential where your probe grounding clip will attach in the circuit. If the Scope is grounded, expect the worst. Most circuits and tests are common ground, but those that are not will be loaded(read that as direct shorted to ground) if that is where you need to connect your reference. DC

While there are times isolating the scope from earth is required, I would not do this as the norm and would only do in controlled conditions. It could prove to be quite dangerous when you connect the ground of the probe to a voltage source. Blown fuses or power supplies are preferable to death.
Paul

[gar]

070502-1656 EST USA

I agree with Paul.

Instead of ungrounding the scope. Use two probes and take the difference of the two channels. May require a slight adjustment of the gain of one input.

.

[Al_The_Man]

Originally Posted by paulC While there are times isolating the scope from earth is required, I would not do this as the norm and would only do in controlled conditions. It could prove to be quite dangerous when you connect the ground of the probe to a voltage source. Blown fuses or power supplies are preferable to death. Paul

I agree, the best way is to test if you have a 'Hot' chassis and use a volt meter from the common point or chassis that you intend attaching the probe to and earth ground, test with DC range and AC range to be sure.
To prevent stray voltages to cause indication on the meter, put a low resistance load from the Common to earth ground when you measure, such as a 120/240v lamp.
If you are working on a hot chassis then extra care has to be taken, in this case it is best to connect a mains isolation transformer.
Al.

[One of Many]

Right you are Paul,

I knew I should have put a disclaimer on that post. I guess I should have stated that as something to be aware of if using the scope in this manner.

I only use this for low voltage circuits that do not survive when signals are floating above ground potential. Where the only fuses "at risk" are components in the circuit.

An O-scope is not often needed for checking basic DC or AC control circuits unless timing is a critical issue. A simple logic probe would work for checking if encoder pulses are present. Unless you really need to get deeper into the wave shape. I have seen encoder outputs fall into the null zone of 2.6-3.3 volts at higher speeds.

Sorry for side stepping the safety side of that suggestion.

DC

[Al_The_Man]

On a side note, one of the things missing now is the Timebase sawtooth output, this was really usefull for 'ringing' an inductance like a transformer or coil etc for shorted turns, the resultant waveform frequency would either decay gradually in a linear fashion (good) , or decay rapidly exponentially (bad).
Al.