Thanks for the instant response, Artur!

Here, saying current/voltage source, I mean constant current/voltage source... and saying constant current/voltage source, I mean that its current/voltage is constant regarding the varying load (a diode here). Actually, regarding the input quantity, it is a varying current/voltage source.

Cyril - what is a "constant current/voltage source" with a current/voltage that is  constant regarding the varying load? I suppose you mean "current OR voltage, correct? 

Yes, Lutz... I have combined the two definitions into one. I have tried to say the definition of a CCS and CVS in a more interesting and intriguing way exploiting the little confusing fact that the source is both constant and varying:)

'

I can't imagine why such good fellows are so confused over such

a simple matter!  A diode (or junction device in general -- or ANY

electrical device, come to that) has no idea how it is being applied,

whether using a (variable) source of either current or voltage, or in

the case of active devices, a combination of both. But, obviously,

some methods, in selected circumstances, are better than others.

.

Assuming you are speaking of practical testing, it would not be a

good idea, in general, to apply a constant voltage of 2 V to most

small-signal junctions, because the current would read either an

extremely high value (if you had just melted the poor creature) or

a near-zero value, after you had blown it to smithereens.

.

But neither would you be any better off by applying a test current

of 2 A, for roughly the same reasons.

.

And in either (practical) case, most semiconductor junctions will

get very hot even before breaking point is reached; and at much

lower applied voltages you will see the current RISE accordingly,

and perhaps progressively until thermal runaway occurs; while at

much lower applied currents you'll see the junction voltage FALL

as the device heats up, in which case, the dissipated power will

conveniently fall, thus automatically preventing thermal runaway.

.

Thus, as a purely practical matter, your best "guess" would be to

use current sourcing as a general rule, if a guess is needed at all.

.

The best equipment for performing such measurements is either

a Tektronix 575 Curve Tracer or the later Tektronix 576 -- which I

happen to have had a part in designing, long decades ago. These

can still be found on ebay, and should you be so lucky, my advice

is to grab one up as quickly as possible. (I have one of each in my

home lab, and there are constantly in use). HP also made one or

two such instruments and called them 'Semiconductor Analysers'.

.

The key question is ever: "What am I trying to determine?".  In

many cases, theory alone is adequate; in understanding practical

devices, the modeling accuracy in simulation will usually provide

the best general insights. But what happens in practice will often

only be a "one-off" result, having limited value. And you may well

end up with an injured device.

.

But I presume you knew all this before my fingers hit the keyboard.

.

Your ever-practical friend,

Barrie

Dear Barrie Gilbert,

Thanks for the comprehensive and so valuable answer.  As you probably guess, my purpose is educational - in this lab, students to not only monitor diode IV curves, but also to understand the idea of the measuring setup... to form a notion about graphical solution, operating point, load line, diode log/antilog convertors, temperature sensitivity, etc.

From this educational viewpoint, it is interesting to drive a forward biased diode by both current and voltage source to see the differences between the two arrangements, which you are talking about. For example, to show how difficult is to obtain a good resolution (a sufficient number of points) along the vertical part of the IV curve, I conduct such an interesting experiment - replace the diode with a voltage source (battery). They see that, in this case, it is impossible to measure the IV curve of a voltage source by driving it with another voltage source...

My observation is that the humble resistor-type current source is most appropriate for these educational purposes. By changing its voltage and resistance, we can move and incline (rotate) its scanning IV curve to obtain a sufficient number of points along the investigated section of the scanned diode IV curve. In comparison, while the voltage-driven arrangement is bad along the vertical part of the curve, the current-driven arrangement is bad along the initial horizontal part of the curve.

.

Cyril:

.

I see... It is a matter of didactics. In that case, it will be hard

to better the use of a dynamic curve-tracer - either a home-

brew instrument or as stated earlier, a fine, old, and reliable

Tektronix 575 or 576. Either that, or build yourself a pulsing

and calibrated current source for observing the time-varying

forward voltage on a diode,

'

Static stimuli can be very misleading - and they are virtually

bound to produce very limited insights.  They are analogous

to visiting a beautiful and vibrant city like Sofia and admiring

it solely from one's small hotel window looking out on to the

Tsarigradsko shose near the Mladost district - at night-time..

,

Barrie

 '

There are plenty of Tektronix 576's on ebay. Get one today!

http://www.ebay.com/sch/i.html_from=R40&_trksid=p2050601.m570.l1311.R1.TR5.TRC1.A0.H0.Xtektronix+57.TRS0&_nkw=tektronix+576&_sacat=0

Dear Barrie Gilbert,

Thank you for the colorful metaphor which gives the impression that you have ever been in Sofia (and maybe you really have been). It seems, you are not only expert in the field of analog electronics, but also in the field of geography:) Incidentally, "Tsarigradsko Shose" means "A road to the city of the King (Sultan)" (i.e. Istanbul), and "Mladost" means "youth"...

About measuring instruments, I have the promise of the dean of the Faculty of computer systems that the laboratory will be equipped very well - with a sufficient number of digital oscilloscopes and, of course, with the recommended by you curve tracers... For now, I must be content with "home-brew" devices.

So, I imagine the following sequence (scenario) to study the elements by inexperienced students (illustrated it in the next comments below):

1. First, they should measure (at least one time) an IV curve of a diode at the lowest level, in the most primitive way - by using variable sources and multimeters, filling the data in a table, and plotting the curve (manually or by some program as Excel). The measurement will be accurate but tedious and painful for them... and they will not see in real time the movement of the operating point along the curve:

https://drive.google.com/open?id=0B45uRPpHPD9hOHhDTWEyWjJPcVE

Manual investigation of Si diode

2. Then, to alleviate their suffering:), we include an XY digital oscilloscope in parallel to the multimeters (I mean digital because it can store the curve and highlight the operating point). It will draw the curve on the screen, and multimeters will show the exact values of the quantities (unfortunately, TBS 1102 has very limited options for this). Now, students will move the operating point around the curve like the Papert's tortle and, having fun in this way, by the help of these psycho-motor actions, they can "feel" it:

https://drive.google.com/open?id=0B45uRPpHPD9hMTZkbnNJT0pBTWM

Semiautomatic investigation of Si diode by a digital oscilloscope

https://drive.google.com/open?id=0B45uRPpHPD9hcWVqUXNCbW9aczg

Scope and multimeters - big picture

https://drive.google.com/open?id=0B45uRPpHPD9hOXE3SWJ6WTlwMFk

Diode IV curve on the screen

In these static experiments, there are no DC and AC signals; there is only one slowly varying DC signal (that can be boring for students to change). Of course, it is possible to split it into two ingredients (DC and AC) by summing the signals of two potentiometers... but I аm not sure if this makes sense here... maybe, it is better to do this in the next dynamic investigation.

3. After that, we drive the element by an AC signal (ramp or simply 50Hz sine). The whole IV curve will be automatically plotted on the screen of the scope (not storing). Then we add an DC signal (biasing) to set the operating point at the desired place, and decrease the magnitude of the AC signal so that to picture a section around the operating point. By varying the DC signal, students can move, in the Papert's tortle manner, the whole section around the curve.

The advantage of this dynamic investigation is that the image of the curve is automatically refreshed; its disadvantage is that the curve on the screen is static.

4. Finally, we scatter a huge pile of components on the table, and give a professional curve tracer to students to explore them. The advantage of this method is that students are not distracted by the measuring setup and focus on the components themselves.