Can You Do Component Testing On Digital Oscilloscopes?

[nmoore]

I am in the market to getting an oscilloscope, I have see that the older analog cathode ray oscilloscopes (cro) are able to do component testing of different circuit elements (please see below). My question is do digital oscilloscopes have this ability?  This feature seems good for electronic repairs as you can test components directly on a soldered PCB. I am new to oscilloscopes and trying to find a beginner oscilloscope that has this feature. Any assistance and recommendation is greatly appreciated. Thank you.

[4thDoctorWhoFan]

Most O-scopes whether analog or digital have XY mode which is what you need to get results like in your pictures.

[artag]

The XY mode can be pretty awful on digital scopes though.
 This may be because there's no obvious digitiser rate and no retrace period.
I've tried to use a scopetrex (rebuilt vectrex vector-plotted games machine) and even on a decent Agilent it looked poor.
 

[alm]

This was a feature that was part of many Hameg analog scopes, and quite possibly other ones, too. You could replace it with any dual channel or more analog or digital scope that supports XY mode (which is pretty much all the serious ones) and what is often called an octopus component tester or curve tracer, which you could easily build and probably buy. See for example this video:

[coromonadalix]

 tested many zenner diodes and transistors like this ...   we could saw the leakage ...

[wasedadoc]

As noted above this requires X-Y mode which is one thing that an analogue CRT scope can do much better than most digital scopes.

See https://www.eevblog.com/forum/testgear/shopping-for-used-scope-to-use-as-curve-tracer/msg4391185/#msg4391185 for an example of what I achieved with a Rigol DS1054Z and a few components configured as an 'Octopus' type tester.

As for testing components in circuit, you need to be disabused of that belief.

[tautech]

I am in the market to getting an oscilloscope, I have see that the older analog cathode ray oscilloscopes (cro) are able to do component testing of different circuit elements (please see below). My question is do digital oscilloscopes have this ability?  This feature seems good for electronic repairs as you can test components directly on a soldered PCB. I am new to oscilloscopes and trying to find a beginner oscilloscope that has this feature. Any assistance and recommendation is greatly appreciated. Thank you.

Welcome to the forum.

At this time no DSO makers that I'm aware of offer a proper curve tracer of which the old stuff offered some very high stimulus levels. The display part is simple and any scope with an XY feature can display the curves.

Member mawyatt opened a couple of topics not long back that are worthy of a read:
https://www.eevblog.com/forum/testgear/fooln-around-with-dso-awg/
https://www.eevblog.com/forum/projects/dso-awg-based-curve-tracer/

[mawyatt]

Another link that folks may find interesting.

https://www.eevblog.com/forum/testgear/capacitive-impedance-plots-with-sds2104x-plus-bode-function/msg4335745/#msg4335745

Best,

[MarkL]

Here's an app note from Agilent that describes some component testing methods that can be used with any oscilloscope with either a built-in or external function generator:

  https://www.eeweb.com/wp-content/uploads/articles-app-notes-files-oscilloscope-with-integrated-waveform-generator-1339692370.pdf

[james_s]

The XY mode can be pretty awful on digital scopes though.
 This may be because there's no obvious digitiser rate and no retrace period.
I've tried to use a scopetrex (rebuilt vectrex vector-plotted games machine) and even on a decent Agilent it looked poor.

That's been my experience as well, however I suspect that plotting semiconductor curves is far less demanding of the XY capabilities than drawing complex vector graphics.

[Darkover]

> My question is do digital oscilloscopes have this ability?

Yes, there are a few from Hameg/RS (HMO2022 for example)

in case you don't have it, you can build it yourself. I did it:

https://www.mikrocontroller.net/topic/528893#new

Discussion is in German, but schematic and picture is readable. 

Olaf

[Wallace Gasiewicz]

I have a BK  transistor curve tester, It works very well on my analog scopes. You do not need a high end analog  scope for a curve tracer to work. I think any analog scope will work. I even had one that just had X and Y inputs and NO time base. That should work also.

When I put it on my Agilent 54820, the result is barely readable. You can see the traces but they are in a mess of other stuff I don't know if this is due to "aliasing'.  The retracing is horrible to look at.
ff you are used to curve tracers you can pick out the curves of interest, but i find them unusable.

However the 54820 does make a very nice XY trace using sine or square inputs. Nice circle with proper sine input, pretty colors too. So maybe testing things that would generate a stable repeatable scope input would be displayed better. I did read the article posted by Mark, that article deals with tests that can be done WITHOUT using XY mode. Nice post.
Transistor curve tracers produce an output that includes about 5 different traces. So there are two different changing signals going into the scope. And then the scale jumps ranges Maybe if I were to modify the curve tracer so I would only see one curve at a time the 54820 would display it in a readable pattern. I did post my efforts on the same thread that wasedadoc mentioned.If I had the room I would just set up the curve tracer with one of my old scopes and use it as a dedicated curve terser.

[Darkover]

> It works very well on my analog scopes.

I don't think you need an analog scope. My RTB ist very digital.


Olaf

[wasedadoc]

When I put it on my Agilent 54820, the result is barely readable. You can see the traces but they are in a mess of other stuff I don't know if this is due to "aliasing'.  The retracing is horrible to look at.

Absolutely nothing to do with aliasing.

[Wallace Gasiewicz]

Olaf:Your trace is not quite applicable to the traces you get with a curve tracerThe inputs to the scope are very different.
My comments were to the compatibility of a curve tracer to a digital scope. Other devices can be measured nicely, as you have demonstrated.

[FrancisM]

This is what you can expect from a well known brand on this board, pretty usable.
I wanted to show the effect of using high voltage transistors in a Wilson current mirror or cascoded at a very low Vce.

Francis

[themadhippy]

Another diy option
https://sound-au.com/project119.htm

[Darkover]

> My comments were to the compatibility of a curve tracer to a digital scope.

I would say a curve tracer and a component tester (subject) ist something complete different isn't?

Olaf

[Swainster]

I think that the OP is talking about analog signature analysis rather than what is typically understood by the term curve tracing (at least in the sense of accurately characterising a particular component). ASG can be done on a populated board, but then it tests the characteristics between 2 circuit nodes rather than a specific component. Of course, this characteristic may well be dominated by a particular component, but this isn't always the case. The main thing is that by comparing the signature between a good board and bad one (or from a reference document), it helps to locate a fault in a circuit.

[bdunham7]

That's been my experience as well, however I suspect that plotting semiconductor curves is far less demanding of the XY capabilities than drawing complex vector graphics.

You don't need Z-axis control for one thing, and that's probably the most important.  The only remaining issue would be getting your acquisition period right and I think very fast and short might be good enough for a 60Hz or 1kHz based component tester.

[bdunham7]

I would say a curve tracer and a component tester (subject) ist something complete different isn't?

A curve tracer can operate on two-leaded devices like diodes and will be essentially the same thing as a component tester, but typically with greater current/voltage options available.

The device typically known as a curve tracer will also have the ability to scan a third lead of a component with steps of current and voltage while measuring the current vs. voltage on the other two leads.

I like the design of the device you posted.  Would I be stepping on anyone's toes if I laid out a PCB for that design and posted it here?  If I ever get the time and motivation I might just do that.

[Shock]

For testing components, a component tester on old an analog oscilloscope is the last tool I would get. For it to be effective with in circuit testing you need familiarity with the circuit, or a spare known good board or have previously recorded results.

In my opinion the far more practical tools are:
Multimeter (essential for quick live and in circuit testing)
LCR meter (more of a precision tool with inductance as well)
ESR meter (measures electrolytic caps for a specific common failure)
Transistor/Multi component tester (does a bit of everything)
Then a lab power supply, digital oscilloscope and if you need one a function generator.

Don't cheap out on the oscilloscope if you are taking up electronics for a while, the modern entry level models (digital) are quite advanced and better value for money. You can purchase a component tester circuit for your digital scope or make one if you are curious, fairly inexpensive or use a function generator to achieve similar result.

[Stray Electron]

  I agree with Shock's last paragraph about not cheaping out on a scope for troubleshooting but I disagree with his statement that a CT is the last thing that you need.  You can learn a LOT about component behavior by using a CT.  My advice to buy a good digital scope and not worry using it for CT'ing. Then buy a cheap analog scope and build an octopus CT circuit.  All of the scopes that I have seen that have built in CTs are 20MHz or less and they're not adequate for troubleshooting modern electronics.  I recently passed up a B&K scope with built in CT for $15 in a surplus store (a B&K 2120 IIRC). It was only 10 or 20 MHz and only tested two terminal devices (such as diodes) and not three terminal devices such as transistors.

   I also agree with Wallace's statement that it's next to impossible to understand the hash that digital scopes make of X-Y curves. 

   I was lucky enough to find one of the little B&K Component Testers (a curve tracer) that about 6 or 7 inches wide and about 3 inches tall and that uses an about 3" CRT and I use it all of the time, even though I have two Tek 575s and a Tek 576.