But still my students have managed to stop them and observe the values of the address and data (read the binary numbers:)

No... they did not pause the signal... they did just the opposite:)

Data on all system buses (Data, Address, and Control) are changing all the time according to processor read/write cycles, governed by system clock. Clock is usually in a MHz range.

We can use LED-s (with some driver) or Logic probe to observe activity in any of line. Intensity in "low" LED and "high" LED will correspond to relative percentage of presence of given logic level. If some line is mostly on one level with very short activity on other level we may miss seeing it in corresponding LED. To catch these situations Logic probes usually have some transition indicator in for of blinking LED.

Classical oscilloscopes rarely will help, because usually signals are not periodic.

Digital oscilloscopes sometimes may help if proper triggering is used and if used in single sweep mode.

Both of them can be used very successfully if we are writing code and provide some programmed trigger signal at proper time.

Logic analyzers will do the job. Even in this case proper triggering is very important to catch the desired moment.

Road to follow will depend on type of information we want to observe.

In 1980's when we wanted to observe details of bus activity we lowered the clock down to DC and gave clock low/high states by two keys.

This was possible if devices could operate with down to DC clocks (static operation), and at that time not all could. There were also asymmetrical cases where one level could be kept indefinitely, but other had some higher bound.

If operation is fully static, than we will be able to observe whole activity in any bus either with LED (with driver) or with voltmeter.

In asymmetrical case we used latches with LEDs for quasi-static state and direct LEDs for static state.

If comparators are used, depending on circuit, it will be possible to detect also High-Z.

If specific event is important then we can make it "static" and show it on LEDs (or measure with voltmeter) by using proper synchronization for gating state - stroboscope like effect.

An excellent explanation, Lavdim! I read with great pleasure this "bag of microcomputer tricks":) How interesting it is to slow down the operation of such a "super fast acting" device ... to make it act step-by-step at the slow speed of human thought ... and so calmly to explore and present it to students...

Dear Cyril,

You always ask didactically interesting questions.I hope you are fine.

Your question is devoted to the possibility of observing the high speed electrical activities on the computer Bus. The signals on the bus are binary digital representing logic 1 and logic 0. The format of the binary digital signal is TTL one, such that the logic one is represented by a pulse with width equal to the bit time while the zero logic zero is represented by zero level.This format is termed unipolar nonreturn zero.

Accordingly,the electrical signals on the bus are streams of pulses. The pulse duration is equal to the bit time. From the point of view of an electronic engineer, the signal on the bus have specific waveform, that is v(t) From the point of view of informatic engineer it is a stream of logic values,ones and zeros.

If you want to observe the waveforms we use oscilloscopes with sufficient bandwidth to avoid the waveform distortion.From such waveforms one can extract the logic values. Lavdim explained the use of the oscilloscope to capture the Bus signals.

To fix the waveform on the screen and make it static to the observer, the time base of the oscilloscope must be synchronized to the signal waveform by either an internal or external triggering.

If you are interested to verify the logic functions of the different logic devices in the system you may be interested only to acquire the logic values of the signals on the bus you use logic analyzers. They display decoded values of the signals on the screen of the analyzer. The rise and fall time detail are eliminated by decoding. and all analyzer signals are synchronized to the system clock.

It is common now to use data acquisition cards that is suitable for observing both analog and digital signals. The signals are sampled with sufficient speed, converted into digital form and stored in the memory of the computer. Then the signal can be displayed and proceed at any time. Even, if they are digital they can be decoded and stored as bit map file. In this case you can compare your test results with the expected

results to verify the functions. These cards are supplied by many vendors with different performance parameters and prices.

That is the computers are used to test other equipment and to test themselves. By the acquisition process you stopped the signal and stored it to be able you see it on your request.

To decrease the system clock to be slow enough to observe it by blinking LED, It is really used in the digital circuits verification in the logic laboratories. It is suitable for demonstrating the logic functions rather than large systems testing.

I hope that i shed some light on this important didactic topic.

Thank you

The moment we are stopping it, we can measure the current flowing across the bus by putting an ammeter in series across it. But this changes the flow of binary information as they are flowing at a particular speed depending on the clock frequency.

As Dr. Abdelhalim said you can connect Logic analyser at the interface pins of either side of bus. Never tried this but I am sure you will work on all the suggestions mentioned earlier and do share the results with us.

Hi, Abdelhalim! Your answers are always so professional, thorough, profound and comprehensive so you do not give a chance to others after you say something more... but I still try:)

All these solutions are excellent (I love most the data acquisition approach - "computers testing themselves":)... and they are the real ones... The only disadvantage when used for educational purposes at the initial stage of teaching microcomputer devices is that they are a little abstract and hide the physical nature of such real things like buses. At this stage, connecting simple Indicators to buses convinces students that the address and data are actually there... So unless these straightforward techniques, it is worth to use some "lateral" tricks (like these proposed by Lavdim at the end of his comment) to show with simple means (an oscilloscope, "LED + resistor", "voltmeter + diode"... or something similar...) what is on the buses...

So the question remains open: How do we read in this "elegantly simple" way (only by a humble LED or voltmeter) the addresses, data and other signals appearing on the buses? Maybe the snapshot of the "frozen" (D7=1) data signal on the oscilloscope screen below can help:)

Best regards, Cyril.

Thank you Cyril for your complements. Only jewelers know the value of real jewels and also real scientists know the value of scientists. Thenk you again for your encouragement.

I agree with your view that the beginners need to make themselves familiar with the signals in the logic devices , subsystems and systems by fetching the signals on the buses of the systems.

One way is to write down simple programs and observe its execution steps by observing the signals on the buses.This can be achieved by slowing down the processes through timing control. In this way one can use LEDS, Voltmeters to display the instantaneous signals on the bus. It is so that the eyes can follow changes up to about 14 per second. One per second rate could be very comfortable to observe the light pulses.

Other way, one can repeat the same instruction many times to elongate the observation time of it and hence can be captured by the eye.As you said there are many tricks to make the verification in a convincing manner.

Good evening

Lavdim and Abdelhalim,

Maybe the picture below (I made it yesterday evening with my students in the lab) will illustrate your remarkable thoughts and give an idea what the trick is:

Lavdim: "If specific event is important then we can make it 'static' and show it on LEDs (or measure with voltmeter) by using proper synchronization for gating state - stroboscope like effect."

Abdelhalim: "Other way, one can repeat the same instruction many times to elongate the observation time of it and hence can be captured by the eye."

Regards, Cyril

Cyril you are a very good teacher. Please can you admit me in your class?

Wish you success.

You are welcome, Abdelhalim! But if two very good teachers in one class will not be too much:)))?

Thank you for your friendly answer.This will be for the benefits of the students.

To catch and observe the operations of even a simple microprocessor, 8 bit data, 16 bit address, read, write requires that you use a logic analyzer. These can be purchased pretty cheaply to connect a PC with USB to the analyzer. The waveforms can then be grouped in busses, address, data, control and displayed in binary hex or other format as desired. It is even possible to debug your students code using such a setup.

To use a scope to capture just one signal at a time even if the same instruction is repeated multiple times is sure to be error prone as it takes different people different amounts of time to move the probes.

An LED on the line will potentially drag down the line and prevent a true logic signal from appearing on the bus. To use this method would be to recreate a logic probe. Probe -> High Impedance OpAmp buffer -> LED. Simple and cheap. http://www.bellnw.com/manufacturer/BK-Precision/DP-21.htm is a low-cost version. Google Logic Probe will return projects for build it yourself versions.

My colleagues, thanks for the clever suggestions... but I still think this clear and simple question deserves the same concrete, clear and simple answer...

So eventually, where (between which two points) do we connect the LED network?

In today's complex digital systems, LED may be useful only as an "assertion" check - e.g.whether the system has reached a certain state, thereby helping the debugger to focus on the portion of the code that needs attention. There is a problem of inserting controllability and observability points in digital circuits that scientists in the field of digital testing have studied. There are measures of controllability and observability such as SCOAP measures. I suppose by adding an LED, you improve the observability; in the same way, one can add a multiplexer or a jumper to improve controllability. Inserting test points can reduce the number of test vectors required to achieve good fault coverage. But the algorithms for adding test points tend to be heuristic in nature. You will find literature on test point insertion online. Hope this helps.

All the wisdom above is true... but still it is a great challenge to find a way to see all these bus signals only by a humble LED or a voltmeter... Let's recall the situation:

Besides the humble indicator (LED or voltmeter), we have a looped program writing/reading data to/from a memory cell. Also, we have an access to the steady circuit voltages (Vcc and ground) and dynamic bus signals (E, R/W, CS, A15-A0, D7-D0, etc.) The "puzzle" was:

Between which of these circuit points do we connect the humble indicator?

Sorry for the persistence ... but I still think that except professional and sophisticated modern solutions we should encourage simple but clever ideas... just to motivate and awake curiosity in our students...

Regards, Cyril

Thank you dear Aparna for your (ever) clever thoughts! We (my students and I) have used a LED network consisting of the very LED and a 1.5 k resistor... We have not used a 7-segment LED but the idea was the same clever trick used in the 7-segment dynamic LED indicator:) Regards, Cyril

http://www.circuit-fantasia.com/ResearchGate/Enthusiastic_students_observe_bus_signals_by_LED....AVI

Hello Cyril,

Here's another idea: supposing your system has some kind of I/O port - a latch that your processor can write to to set its outputs high or low. If you wish to examine the bus state occurring on a particular instruction, you insert an instruction in the code (just before the instruction of interest) that sets the I/O port outputs low, and insert another instruction (just after the instruction of interest) that sets the I/O port outputs high again. LEDs wired with their anodes to the bus lines and cathodes to the I/O port outputs will then light up to show which bits are high on the bus when the instruction of interest is being executed. The obvious limitation of this is on the duty cycle defined by instruction execution time and repetition period of the loop - in some cases, a darkened room may be necessary!

Of course, you can trigger a 'scope from the I/O port, and observe the bus line states on the 'scope, but typically you will only be able to examine the bus one bit at a time, whereas with the LEDs, you can observe all bus lines simultaneously, which creates a more immediate and intuitive demonstration for you students.

Perhaps you will have to reveal the 'trick' you are using in your class!

Best regards,

David

Hi David,

An excellent answer - another "elegant simplicity"! I like it very much (like all the clever answers above)... although it seems there is a little problem... You select the LED during the whole instruction of question (and even during the port instruction) but an instruction consists of several cycles during which the addresses and data on buses change...

Still this is the simple but clever idea widely used in dynamic 7-segment LED indicators from "ancient" times:) We can control a LED from the side of the anode keeping its cathode at the steady ground; conversely, we can control it from the side of the cathode keeping its anode at the steady Vcc; finally, we can control it simultaneously from both the sides (of the anode and cathode) simply realizing an AND function without any additional circuits. Thus the signal applied from the side of the cathode can serve as SELECT (CS of the observed device) in respect to the INPUT bus signal of interest (e.g., A15) applied from the side of the anode. So, every time when CS is active (low), and the observed bus signal (A15) is high, the LED will flash... and if we repeat it quickly enough, we will have the illusion the LED lights up continuously (the Edison's cinematograph idea:)

So, the 'trick' is extremely simple - just connect a LED between these two points (the cathode to CS and the anode to the bus line) and loop a simple program (e.g., a single LDA or STA instruction followed by BRA). It can be any but looped program but "a darkened room may be necessary":) As you suggest, we can also connect multiple diodes to observe all bus lines (hoping that CS can endure the sum of the LED currents:) But wait... How do we observe the very CS signal?:) And what will happen if a (real) program addresses several cells inside the same address space? Maybe we can solve this problem by combaining your "port idea" with this "CS idea" (AND between time and address enabling)?

Perhaps you and all the contributors above (my colleagues and friends) will note some disadvantages... incorrectness... and even some "deception" in my "puzzle":) But this is still a "trick"... and tricks always contain some portion of cheat... but people still watch them with interest... perhaps because, except the raw truth, sometimes we prefer the alluring lie?

Thanks for the responsiveness,

Cyril