EEVblog #652 - Oscilloscope & Function Generator Measurement Trap

[EEVblog]

Dave explains a big trap in high frequency measurement with your oscilloscope.
Based on a viewer request, Dave demonstrates how to incorrectly and then correctly measure the signal output level over frequency of your function generator using your oscilloscope.
Some whiteboard transmission line theory is thrown in as well.

[apelly]

Useful background/primer from w2aew:

How to measure coax velocity factor VF and impedance Z


Transmission Line Terminations for Digital and RF signals - Intro/Tutorial


Edit: Oops. Forgot to say thanks, Dave. Please excuse my manners.

[Co6aka]

Yo Stubby!    Yous shoulda also shorted da enda da stub 'n show'd 'em what appens den -- fuhgeddaboudit!   

[HP-ILnerd]

Dave,

Great video!  It's a nice one particularly since there's a lot for viewers to try at home (assuming they have appropriate kit).
This is actually the sort of thing that lured me back into the hobby from the software side of things.  In software, everything
is so perfect and pristine--i.e, the ones and zeroes don't ring, reflect or otherwise distort.  But that's only because some EE
made sure they were that way in the end product.  At the bottom, the electronics are entertainingly messy, and there is physics
involved, not just logic.

I suppose it's normal practice when debugging something complicated to try your probing setup against known voltages, etc.
to make sure you are seeing real data and not an artifact of your measuring arrangement?

Anyway,   

[w2aew]

Useful background/primer from w2aew:

How to measure coax velocity factor VF and impedance Z


Transmission Line Terminations for Digital and RF signals - Intro/Tutorial


Edit: Oops. Forgot to say thanks, Dave. Please excuse my manners.

Ha, you beat me to it!  Dave's now playing in my RF world...  Here's a few more besides those that you mentioned...

This one is on how to make your own TDR (take advantage of the reflection to measure the coax length):


My initial video on how to measure length and impedance of coax using scope:


This one goes into more detail on 1/4 wavelength transmission line (stub filters):

[martinv]

It seems you can think of this in both frequency and time domain.   On the white board at 16:00 on the portion of the pulse labeled "Out of Phase Cancels",  I tend to think (perhaps abnormally): what if that pulse is the first pulse?  There is no reflection yet, as the pulse has only just reached the oscilloscope input.  The oscilloscope sees the signal of the 50 Ohm signal generator terminated into 2 50 Ohm loads (25 Ohm), one being the terminator and the other being the characteristic impedance of the stub cable.  The level at the scope remains about the same as the signal travels down the open ended cable.  When the first reflection coming back from the open stub cable hits the oscilloscope input, revealing the open stubs termination impedance (infinite) and the voltage jumps up (labeled "IN PHASE" on the white board).   Beyond that you get more back and forth reflections adding and subtracting. 

It's kind of fun to view it on the scope, then model it in LT Spice and see how the results agree, but the real measurements are what I tend to remember and make the concepts sink in.  I also found w2aew's videos on this subject very good and it's always interesting to see other perspectives.

[apelly]

Ha, you beat me to it!

Sorry mate. Just happened to be online. And I really like your videos.

[EEVblog]

It seems you can think of this in both frequency and time domain.

Yes, and ultimately there are many ways to think of and explain this sort of stuff.

[EEVblog]

Great video!  It's a nice one particularly since there's a lot for viewers to try at home (assuming they have appropriate kit).

Yeah, it's a fun and easy way to play with transmission line theory without actually having to build anything.

[KD0RC]

Thanks Dave.  As someone who likes to play with RF, but without an EE degree, I don't usually think of the termination when connecting a pair of BNC connectors with coax.  This kind of tutorial really helps keep me from falling into those famous traps for young players!

[hamster_nz]

At 14:21, the 'bench' on the rising edge is caused by the signal being split, not the reflection. The short stub looks like a parallel resistance until it is 'loaded up'.

If the stub cable was infinitely long, the 'bench' would be infinitely long too, as a portion of the signal's energy sails off down the stub...

Or to look at it another way, it can't be a reflection, as the signal hasn't reached the unterminated end of the stub yet!

[motocoder]

Dave explains a big trap in high frequency measurement with your oscilloscope.
Based on a viewer request, Dave demonstrates how to incorrectly and then correctly measure the signal output level over frequency of your function generator using your oscilloscope.
Some whiteboard transmission line theory is thrown in as well.

Nice video, although I was a bit disappointed that you didn't whip out a Smith Chart.

[BravoV]

Thanks for this video Dave ! 

It will be a good reference for helping "tons" of young players that were facing this exact problem that we've encountered and helped in this forum alone. 

Just to complement this video, you also can use a cheap 50 Ohm terminator + a T connector instead of using feed through terminator as Dave used in the video as usually its harder to find and more expensive.

[miguelvp]

And if the terminator is color coded, then Green is 50 Ohms, Purple is 75 Ohms

[OilsFan]

Thanks Dave good info though I don't usually play up in that Hz range being an audio guy. I have a few side questions related to coax for a noob like me.

1. What exactly is termination and what does a terminator do? Does it simply ground the signal wire to the shield?

2. Where do you get good 50 ohm coax leads? Also can a person make them easily?

[SNGLinks]

I wish my 'scope (DSO-X 2024A) had switchable input Z (50/75/1M) as I do loads of RF and video work.
I have to use T pieces and terms - very messy - easy to use the wrong term.

Using a spectrum analyzer and a tracking generator is a great way to show the effect of a stub.

[HP-ILnerd]

Thanks Dave good info though I don't usually play up in that Hz range being an audio guy. I have a few side questions related to coax for a noob like me.

1. What exactly is termination and what does a terminator do? Does it simply ground the signal wire to the shield?

This is a broader question than you may realize.  Not to step on a question directed at Dave, but this will answer your question pretty effectively:

[jesuscf]

Transmission line reflections could happen at any frequency.  For instance at 50/60Hz.  This is a big issue with power grids where the transmission lines could be hundreds of kilometers long.  Just imagine what would happen in a a 400kV power line if  the voltage doubles...

[w2aew]

I wish my 'scope (DSO-X 2024A) had switchable input Z (50/75/1M) as I do loads of RF and video work.
I have to use T pieces and terms - very messy - easy to use the wrong term.

Using a spectrum analyzer and a tracking generator is a great way to show the effect of a stub.

Here's the video I did on that exact topic:

[w2aew]

Thanks Dave good info though I don't usually play up in that Hz range being an audio guy. I have a few side questions related to coax for a noob like me.

1. What exactly is termination and what does a terminator do? Does it simply ground the signal wire to the shield?

2. Where do you get good 50 ohm coax leads? Also can a person make them easily?

To answer #1, a termination is usually a resistive load connected at the end of a transmission line (from center to ground), whose resistance matches the impedance of the transmission line.  Here's a video I did on the topic of terminations and reflections, from both a digital and RF perspective:

[w2aew]

Thanks Dave good info though I don't usually play up in that Hz range being an audio guy. I have a few side questions related to coax for a noob like me.

1. What exactly is termination and what does a terminator do? Does it simply ground the signal wire to the shield?

This is a broader question than you may realize.  Not to step on a question directed at Dave, but this will answer your question pretty effectively:

That AT&T video is one of my favorites - absolutely brilliant way to visualize wave propagation and reflections on a transmission line.

[rstoer]

This whole topic brings up a question I've had for awhile:
When do you need a feed-thru terminator vs. the end-cap type?

[eendje]

Hi All,

please also add a bit about different cables, and connectors, there is a world of difference.

maybe this wuld be a good read for your young ones as well

http://www.l-com.com/content/Coaxial-Cabling-Tutorial.html

my two cents,

Eendje

[w2aew]

This whole topic brings up a question I've had for awhile:
When do you need a feed-thru terminator vs. the end-cap type?

A feedthrough is used when connecting the end of a transmission line to something that has a high input impedance (scope input, counter input), and it is placed between the end of the coax and the input jack on the instrument/device. It provides a proper termination to the line, with ideally minimal issues from the short "stub" connecting the high-impedance device.

A end-cap type termination is used when you need to terminate the end of a length of transmission line to prevent reflections.  In many applications, this would be called a "dummy load". For example, it may take the place of an antenna so that transmitter tests can be run into a proper load, but not radiate.  It may be used at the end of an old Ethernet coax run.  They may be used to terminate unused outputs of an amplifier, or unused halves of a differential output device.  In a pinch, they can be used along with a T connector at an instrument/device input to substitute for a through terminator when one isn't available.

[rstoer]

This whole topic brings up a question I've had for awhile:
When do you need a feed-thru terminator vs. the end-cap type?

A feedthrough is used when connecting the end of a transmission line to something that has a high input impedance (scope input, counter input), and it is placed between the end of the coax and the input jack on the instrument/device. It provides a proper termination to the line, with ideally minimal issues from the short "stub" connecting the high-impedance device.

A end-cap type termination is used when you need to terminate the end of a length of transmission line to prevent reflections.  In many applications, this would be called a "dummy load". For example, it may take the place of an antenna so that transmitter tests can be run into a proper load, but not radiate.  It may be used at the end of an old Ethernet coax run.  They may be used to terminate unused outputs of an amplifier, or unused halves of a differential output device.  In a pinch, they can be used along with a T connector at an instrument/device input to substitute for a through terminator when one isn't available.

Got it. Thanks for the clear explanation.
BTW: I subscribe your U-Tube channel and always enjoy your videos. I used to work as an electronics technician but switched to software development over 20 years ago. Now as I approach retirement I'm back to electronics as a hobby, trying to understand more about the design process. Videos from you, Dave and others are making this possible. Your efforts are appreciated!!!