basic digital troubleshooting

[danny davis]

I know what timing problems look like on an oscilloscope and logic logic analyzer

I just don't know what timing problems does to a circuit and how it reacts from what you guys have seen in your experience with timing problems

Can you give me some insight on what circuits you have experienced with timing issues and problems

 

[(*steve*)]

Timing problems can cause anything from a weird small glitch every blue moon to complete and utter chaotic behaviour.

It's a bit like asking:

I know what oil contamination looks like under a microscope and spectographic analyzer

I just don't know what oil contamination problems do to a car and how it reacts from what you guys have seen in your experience with oil contamination problems

Can you give me some insight on what vehicles you have experienced with oil contamination issues and problems

Oil contamination issues can cause anything from a door that squeaks in cold weather to the complete destruction of the engine.

 

[danny davis]

The Chop mode and Alt Mode on a analog O-scope is ment to display 2 signals at once and to measure the time difference between them

When using chop mode or Alt mode , the frequency has to be the SAME frequency?

What do u guys use the chop mode and Alt mode for?

At work when measuring Logic signals , the square waveforms look like the square are dirty signals and the lines of the transitions and alternation and edges are not straight

Why don't they look like straight lines on the logic waveforms?

When Measuring 20uS pulse waveforms on my oscilloscope when i zoom in and stretch out , it changes a pulse waveform either to a cone shape or adds a rise time and fall time to a square waveform, The tech says its a o-scope resolution problem

Another Problem I'm having on the O-scope is that i have to measure 20uS pulse waveforms but that are at different times in the time line like 20mS, 50mS , 80mS, 100mS, 300mS , 500mS in the time line

How can you measure a 20uS pulse waveform when its 500mS in the timeline?

If you use a storage O-scope and zoom in on the 20uS pulse waveform and set the O-scope to 50uS timebase , it will stretch out the pulse waveform or make it in to a cone shape triangle, so there is no way I can measure the Pulse Width when it does this

How would u measure the Pulse Width when there is a Pulses at 20uS , going down a time line from zero to 500mS?

 

[danny davis]

A rising edge is the transition from low to high. It is also named positive edge
A falling edge is the high to low transition. It is also known as the negative edge
A leading edge is an event that is triggered on the front edge of a pulse
A trailing edge is the opposite of a leading edge. It is triggered on the back edge of a pulse.

I don't understand the difference between Rising Edge Vs Leading Edge, are they different? and is the falling edge different than the trailing edge?

 

[Rleo6965]

This might help.

 

[Harald Kapp]

If you have a continuous clock signal, the definition of leading edge and trailing edge is more or less arbitrary. It is just a custom to mark the low-high transistion as the leading edge since this is the edge where circuits are many clock sensitive

If you have a signal that does not run a continuous clock but is normally idle, followed by one or more clock pulses, the definition of leading and trailing edge also depends on the logic state of the signal when idle.

1) Assume your signal is Low in the idle state. Then a clock pulse starts with a low-high transistion from the idle state (remember: low). In this case the leading edge would be the low-high transistion (imagine standing on the signal line and watching the edge of the clock rolling right up to you: the first sign of the clock is the L-H transistion leading the pulse towards you).

2) Assume your signal is High in the idle state. Then a clock pulse starts with a high-low transistion from the idle state (remember: high). In this case the leading edge would be the high-low transistion (imagine standing on the signal line and watching the edge of the clock rolling right up to you: the first sign of the clock is the H-L transistion leading the pulse towards you).

You should always keep in mind the exact circumstances of the circuit and the signals within the circuit when dealing with such matters. There is no general truth. Rleo's signature states this in a rather precise way.

Harald

 

[(*steve*)]

The leading edge may be a rising edge or a falling edge.

If the signal is normally high and the pulse is a short duration at a low voltage, then the leading edge is a falling edge.

If the signal is normally low and the pulse is a short duration at a high voltage, then the leading edge is a rising edge and the trailing edge is a falling edge.

It all depends of what you define as the start and the end. Whatever way the signal goes at the start, that's the leading edge and the same for the trailing edge.

 

[danny davis]

In Trigger Normal mode. You can get the last pulse of the pulse train
But if something triggers the last pulse than the normal mode won't work, so what do u do?

How do u guys measure Logic pulse signals that are 10uS pulse width that are at different mS times in the timeline

 

[(*steve*)]

If your scope is digital, it *may* be able to show the waveform before the trigger pulse.

Showing data at or before the trigger point, whilst very useful, can be very hard to achieve.

If you have a *very* long roll of coax, you can trigger on the signal, and display on another trace the signal after it goes through the roll of coax. This signal may be delayed long enough to see the stuff you require before the trigger point.

 

[Rleo6965]

Have you tried using "Ext. Trigger Input" of scope to sync waveform in your scope?

 

[Harald Kapp]

A logic analyzer is the tool of choice for this purpose.

 

[danny davis]

Have you tried using "Ext. Trigger Input" of scope to sync waveform in your scope?

No, I have never used that, how do u use that? and what do u do?


Yes, I have to use the Pre-Trigger Delay, but my problem is trying to find the 10us or 20us pulses down the timeline because the Pre trigger delay is like hunting for the pulse waveform because its hard to added up the pre-trigger delay to know you're going to get the waveform down the timeline

example.

The Pulse width is 20uS
I set the O-scope to 5uS
Timeline is at 430mS
What is the Pre Trigger Delay value?

The Pulse with is 400uS
I set the O-scope at 50uS
Timeline is at 620mS
What is the Pre Trigger Delay Value?

Ya, and I have tried just setting the Pre- Trigger delay below the Timeline value at which the pulse waveform should be at and it still is hunting for the right pre trigger delay value

The math to get the right pre trigger delay is? because I'm just guessing

 

[danny davis]

I just don't know how to get the right POST Trigger delay value

Even when I know where the pulse is at in the timeline , I enter the post trigger delay value and on the oscilloscope and the pulse signal is not there , so i have to keep going entering in post trigger delay values that are below values or above values where the pulse should be at until i find it

Example, There should be 10uS pulses in the timeline at 500mS

I enter in 500mS on the post trigger delay on my O-scope but i still don't see the pulse , I have to keep entering in values below 500mS or Above 500mS until i Find the 10uS pulse

How do I know what is the right post trigger delay value?

 

[(*steve*)]

I assume you're wanting t see stuff that happened BEFORE the trigger pulse, right?

If so, the post trigger delay needs to be negative.

Or have we changed onto a different subject again?

 

[danny davis]

yes the trigger delay is Negative

 

[(*steve*)]

You probably want a very small time. a couple of ns should suffice to see the start of the pulse that triggers the display.

If the entire sequence does not fit on the display, it may retrigger again and you'll end up with a muddled looking display.

The normal way to handle that is to set the trigger delay to a positive value , long enough that the signal returned to its quiescent state before triggering is enabled again.

Trigger holdoff is described here.

Note that trigger holdoff and trigger delay are two totally different things...

 

[danny davis]

I use the O-scope in Single mode not Auto mode

The entire 4 channel pulse train signals will display at O-scope time base at 50mS. You will see all the 10uS pulses. The Timeline of the pulses goes as far as 900mS and then the pulses stop.

I have to count how many 10uS pulses there is and measure the time for each pulse and the pulse width.

You can't use positive trigger delay because the 10uS pulses are way after in the Timeline , you have to use negative trigger delay

How can you measure the pulse width and time distance when a 10uS pulse is at 500mS timeline?

You want me to use positive trigger delay and use the trigger holdoff? O-scope set at nS?