Just checking Chinese Win 10 Funny Character Problem. Last evening I found both Chrome and Firebox have intermittent problem. I thought this morning I might need to switch to Rpi4B Chrome. But now I don't find this problem. I will come back and check later. No need to reply. Cheers.

When polishing my long tunnel diode question in CD, @circuit brought out a couple of inspiring ideas, one of which is "Lancaster Elegant Simplicity". This part was deleted, so I need to google to recover: tinaja.com/glib/elesimp.pdf. I remember I skimmed the idea was impressed by the samples, including the "Schmitt Trigger (HC14)" oscillator.

I did agree the simplicity was elegant, so I said I always recommend system integration newbies to use the Occam Razor principle, and mentioned that I would use it in my proposed tunnel diode plot the I-V curve experiment.

Lancaster gave an example which I think using the hystersis feature, which I think might be useful when exploring the tunnel diode oscillator, because I found that tunnel diode curve, after jumping over the negative resistance, can go back to part of the neg res region which we cannot walk by before jumping over. But I am not sure if it is useful. Anyway, the Schmidtt Oscillator:

Actually I was thinking of using the Schmidt oscillator's cap's output as the Vt voltage source to power the tunnel diode, to replace the triangular wave sig gen when poltting the I-V curve.

Lancaster also points out that his elegant simplicity is also related to ideas such as Occam's Razor, Schumacher’s "do more with less" and Buckminster Fuller’s "Appropriate technology". I also pointed out the KISS principle which I learned when doing computer programming.

Then I pointed out it is not necessary (or overkill) to use an instrumentation grade op amp, because when exploring the tunnel diode missing neg reg region, we only need to find out the cause and answer my long question, not to do any precise experiment. Perhaps my question title is leading, but I do point out that my crux of the mater question is the root cause of the jumping over and no measurement.

As I said earlier in this chat, I already found the cause in GE's 1962 tunnel diode manual. So the next step of the answer is verify that my guess is correct.

Ockham’s Razor principle of the most fundamentally direct explanation often ending up the most correct. Or Einstein’s "Always seek out the simplest possible solution – but none simpler".

So I think PCF8591 for tunnel diode I-V curve plotting is "As simple as possible", "As cheap as possible", "As small a possible", and "As newbie frienly as possible"

Now I have started developing the python program to generate the Vt for plotting the Tunnel Diode TD. So it is a long project, perhaps over 3 months, with a 2,000_ line program, perhaps also a web site for remote lock down lab from home experiment.

So the plotting the TD curve is just a very first step to test the water. If TD is OK, I might try other diodes. It is basically a non profit making, open source (hardware and software) CoLearn educational at community college, technician, hobbyist, maker, Raspberry Pi / STM32/ESP32 based level project.

So the project is too big to fit a Q&A site. But our strategy is breaking the project into smaller, correlated pieces. I know the ideal Q&A for EE　might be very short. For example, like a short answer, "Which document should a hobbyist read to research into Tunnel Diode?", Then an appropriately short answer is "Check out Generic Electric's TD User Guide 1962, with a link to their archived documents.

Ａｈ　ｓｕｐｐｅｒ　ｔｉｍｅ，　ｓｅｅ　ｙｏｕ　ｌａｔｅｒ．　Ｊｕｓｔ　ｎｏｗ　Ｉ　ｓｕｒｐｒｉｓｉｎｉｎｇｌｙ　ｆｏｕｎｄ　ｔｈａｔ　Ｉｆ　Ｉ　ｔｒｙ　ｔｏ　ｔｒｙ　Ｃａｐｉｔａｌ　Ｅ　ｂｙ　ｐｒｅｓｓｉｎｇ　ｔｈｅ　Ｓｈｉｆｔ　ｋｅｙ　ａｎｄ　Ｅ，　ｔｈｅｎ　ｔｈｅ　ｆｕｎｎｙ　ｃｈａｒａｃｔｅｒｓ　ｊｕｍｐ　ｏｕｔ．　Ｎｅｅｄ　ｔｏ　ｅｘｐｌｏｒｅｓ　ｌａｔｅｒ．　Ｉｎｄｅｅｄ　ａｎｎｏｙｉｎｇ，　Ｃｈｅｅｒｓ．

For your purposes, the version with 2 op-amps is more suitable.

Ha, I am too weak in op amp to understand both the Schmitt and the uA741 x 2 function generator. As usual, I will try to cheat and find an function gen IC to do the job. But I am not too sure to find, or actually the same ICL8xxx module.

Anyway, I have already started programming the PCF8591's DAC part, which is more flexible because I can generate any analog voltage level at any point in time. But only 50% make I can make it.

As a contingency plan, and also make thing scalable and concurrent, I am also considering the digital potentiometer (I guess it is more precise and looks like a pro, while PCF8591 looks like a toy. Again, it might take a long while to read the datasheet.

Anyway, my ambitious plan is to write a python program to control two ADC's and plot two tunnel diode I-V curves at the same time, and also display two curves in the scope at the same. Perhaps I am dreaming too early. Anyway, bed time, see you tomorrow.

And I did consider using NE555 timer module to generate the triangular wave. I remember NE555 has one ore two op amp comparators inside the chip, so lazy me will not consider the op amp way. I think I am some how op amp allergic. :)

TL Fong, I hope you already have realized that the IV curve does not depend on time. So, you can measure it by any kind of input signal (triangle, ramp, sine, etc.) Also the frequency of the signal does not matter ... and as I already wrote, you can even use a DS signal (for example, produced by a potentiometer) if the scope is digital with memory. Since scopes in my laboratory are analog type, I use the simplest AC signal generator - the mains voltage reduced by a step down transformer.

Of course, the most professional solution is to generate the signal through a microcontroller and DAC as you have decided to do. I realized this idea in 1986 through a personal computer (MICROLAB system).

This was an AD interface implemented by 4 12-bit DAC 1200 (National semiconductor) and an ADC with 4-input analog multiplexer - see the block diagram and the construction of the laboratory setup.

Here is an example of measuring an IV curve of a Zener diode. I had put a lot of such links under your question but Olin deleted them.

Regarding the forum... The problem is not short questions... neither short answers... neither long discussions... These are just forms of knowledge transfer. It is not the form that matters, but the content. The main problem is the envy that arises in those who know but do not understand circuits. It makes them interrupt such valuable discussions under questions and answers with meaningless remarks such as, "This is a site for questions and answers, not for discussions," and more...