Видео

Not every college or university teaches Op Amps in depth. It depends on the instructors and program objectives. Some feel in this digital age that the focus needs to be on software platforms like microprocessors. At the college I taught at, we covered both areas. Teaching Op Amps wasn't just about learning how the devices worked and how to use them, it was also about learning how to solve circuits across a wide range of applications using math and electrical circuit theory. In my Op-Amp courses, we covered everything from the differential amplifier itself through to instrumentation, filters, PID, oscillators, all while exploiting Kirchhoff's laws, Calculus, complex numbers, and Laplace Transforms. These skills are timeless, and foster higher understanding in students about all circuit types. When I was first learning electronics in high-school and later in university, I stumbled upon a book written by Albert Malvino entitled Electronic Principles. He is one of the most intuitive writers I have encountered in a college level text book. Since then, the book has undergone several revisions and I was delighted to find that the book was still available for purchase many years later and recommended it to our electronics program as a student text. I believe it is now on its 9th edition; our students used the 8th edition five years ago. If you can get your hands on this book, it may help you. It covers semiconductor theory, diodes, including zeners, numerous transistor types, numerous thyristor types (SCRs, Triacs), IGBTs, Op-Amps, Amplifiers, Filters, Oscillators, and Power Supplies.

​@@ProfessorV.Thank you for your reply. Would you happen to have courses available for auditing? Thank you for the suggested reading, I'll find the book. If Analog circuitry is my passion, where would you recommend I go? It is a shame that Analog isn't taught as much anymore. I have never been able to digest an answer I didn't fully understand the implications to. So I simply need to understand the fundamentals before I can move onto more complex ideas. I have just completed a circuit which required 8 op amps to execute my desired function. Originally 9, but I managed to simplify it so I wouldn't have to buy more IC sockets just yet ;) It worked as designed. That was a good feeling. It took me 3 days to assemble it. I'm gonna have to start designing the pcbs and ordering them at this point, that was way too much soldering for me ;) What are you working on right now, if you don't mind me asking?

Thank you for your warm comments, much appreciated. The primary reason for using SCRs (Thyristors) instead of diodes is because SCRs provide variable voltage to the DC load. Diodes conduct as soon as they are forward biased, SCRs begin conduction only when they are forward biased AND have been gated. By delaying when the gate signal is applied, you can delay the output which reduces the average output voltage. SCR controlled rectifiers were essential for controlling DC motor armature voltages, field exciters, and plating rectifiers, all applications requiring variable DC voltage. With a diode rectifier, you have to vary the AC voltage to vary the DC voltage which either requires a mechanical tap changer on a transformer or large motorized variac (transformer with a continuously variable wiper tap like a potentiometer). Alternatively you can vary AC voltage with back to back SCRs but then why not just use SCRs as depicted in this video which would be more efficient?

@@ProfessorV. No need to thank :). I give positive feedback to those who really deserve it. But let me thank you for the advice and fast reply. I didn't get everything tbh.. but I got about half :D.. Thyristors are not so common to me, but I guess I understand how it works. The thing is in my case that I have 50kVA distribution trafo, which is YZN 11 configuration, it's ballasted and connected with big capacitor banks (so the input of trafo is acting like LC circuit with the ballasts and cap banks) The trafo is connected to step up way.(BTW, only 1 phase connected to input all the time every time) The maximum output voltage what I have got out of it by testing different connections, were about 70-75kV. (This is based on how far the arc jumped through air, meaning the "ignition spark gap") At conditions -15°C it jumped about 7-8 cm. The output current was a lot smaller, compared what it's with "normal kinda setup" And that is 20-25kV with 0,75-1,0 amps.. So quite hard to find thyristors with quick search online that can handle such high voltages.. I have couple of *made in china* stacked diodes, which i havn't tried out yet. They promise that one diode stack can handle 30kV and 2A.. I x-rayed one of em at work... And, big suprise I am not convinced :D but we'll see.. My idea is to take my experiments further and further. If I'd get DC output from it, there would be a lot bigger selection of caps to the HV end then and if it would be high frequency also there could be nice chance to build a damn big tesla coil. :D But I guess i need 3 phase input to get that 6pulse rectifier work atleast somehow with the 3 HV phases...?

@@mcsaatana1614 For really high voltage applications, SCRs are problematic because you also have to use HV isolation for the gate drivers, which have to be floated from ground and each other. It is also essential that all series SCRs be simultaneously gated to ensure that no one SCR absorbs all the voltage. It is far safer and more practical in that case to use back to back SCRs to supply the step up transformer's lower voltage primary and control its AC input with high voltage diode stacks on the secondary. The diode stacks would generally employ special parallel resistor / capacitor circuits to maintain equal voltage drop across all series connected diodes to avoid any one diode exceeding its PIV rating (i.e. we want the voltage to be balanced across all series connected devices). The SCRs are only needed if you require variable high voltage DC output. High voltage precipitators used this technique with SCRs controlling the transformer primary voltage for air purification in industrial settings.

@@ProfessorV. Thanks for answering and trying to solve this with me. And yes, with that high voltages the proper isolation of the HV bits is hard.. I didn't rly know that it would work, that way. That i connect SCR's with the LV end. I thought that it wouldn't magnetize the trafo's coils right.. But I believe you, cause clearly you know a lot more about these things than me. So I gotta ask you, that what SCRs you recommend me to use? Are there a lot of differences between em, if two different SCRs were same voltage and current rating, but from different manufacturer? Thanks! :)

Ah, Reliance. I've serviced a number of different Reliance DC drives over the years, particularly in the paper mills. My primary expertise was on Westinghouse DC drives and electronic systems, as well as GE, but we serviced all makes and manufacturers back in the day when you could get schematics for those drives and most electronics back then were built on discrete electronics and therefore serviceable. Today, everything is proprietary with embedded software controls, so only the OEM can repair the electronics, if they repair them at all. Many systems today are so integrated that no one repairs them, the drive itself has become the smallest serviceable component meaning the only solution is replacement. I'm not a fan of the current environment.

Wow, you have a lot of experience. I imagine how hard it must be to synchronize all power elements with the correct trigger sequency. Thank you again.@@ProfessorV.

Each device has a unique line to line voltage that corresponds to its acceptable firing range. For example, device Q1 (positive phase A) can be triggered anywhere over a 180 degree window that corresponds to phase voltage Vac's positive half cycle. If you convert the Vac sinusoidal voltage to a square wave that is high when Vac is positive, and zero when Vac is negative, then you have a logic signal that gives you the range over which Q1 may be fired on. Once fired, each device remains on for 120 degrees at which point the next SCR in the sequence will take over. In practice, we use phase margins to restrict alpha (firing angle) to remain within say 5 degrees to 175 degrees to compensate for measurement error in our zero cross detector. For forward/reversing DC Drives, larger phase margins that adjust with motor amps are employed to avoid firing both reverse and forward bridges at the same time or suffer a shoot through fault if inductive current holds the last conducting SCR on too long before its negative partner turns on generating a short across the motor armature during regen. For my students at college, we built a simple SCR controller (single phase) using a PIC18F4520 on the PICDEM 2 Demo Board that detected the AC line voltage zero cross through an OPTO isolator. Timer zero was configured to count from 0 to 255 over each half cycle of the AC mains voltage and preloaded with an alpha value set by a potentiometer through an analog input on the PIC. The SCRs were gated through pulse transformers when Timer zero rolled over (as determined by examining its interrupt flag bit T0IF). This method uses a simple ramp generator but for commercial motor drives, a cosine lookup table is employed for more linear control of motor voltage since the average DC output voltage is really a function of the cosine of the firing angle, and not the angle itself (Vavg = 1.35 VLL cos(alpha) for a three-phase rectifier).

@@ProfessorV. un gusto tu conversacion: tengo hecho un variador con pic 18f4520 el cual utilizo sus tres entradas de interrupcion que a su vez dispararan su correspondiente temporizado y asi variar los angulos de disparos aca en argentina tenemos red de 3x380+neutro cada sincronismo lo hice con un opto entre fase y neutro, si bien el equipo funciona ok pero tiene muchos componentes, detectar un solo cruce por cero y desfasar en el tiempo los disparos es mucho mas simple, la unica duda que tengo ya que el patron de secuencias de disparo ya lo hice y lo vi con el osciloscopio seria : una vez detectado el cruce por cero de una fase deberia generar un delay y luego iniciar toda la secuencia correspondiente? de acuerdo al diagrama de tiempos expuesto genere disparos utilizando los timer 1 y 2 del pic(en este caso 16f883) y obtuve las tres señales correspondientes desfasadas 180° entre sus semiciclos asi como 120° entre fases, para poder variar la velocidad del motor deberia "reducir el angulo de conduccion "a menos de 120° o simplemente correrlo en el tiempo con respecto a su cruce por el cero? desde ya gracias por tu tiempo y tu respuesta

For many applications, adding the derivative to attain a full PID control also risks adding instability (oscillations). In your application, derivative action may be beneficial and speed response. PI controllers tend to be better behaved however for things like motor speed control but I have seen the derivative used on tachometer feedback to reduce overshoot. For an OP-AMP circuit, adding a capacitor in series with the input resistor effectively introduces derivative action.

I video record myself with a green screen, while my computer video captures everything I'm writing on a Huion Drawing pad along with voice audio using Camtasia software. On my drawing pad, I used Microsoft Whiteboard to write and draw in while Camtasia recorded the session in the background. I then layer my green screen footage of myself together with the drawing pad footage in Camtasia. You also have to sync the audio files of both video clips so they correctly time together. It takes some time to do this. I use other software like Blender, Davinci Resolve, and Hitfilm Pro for more sophisticated 3D and visual effects. Some professors write on a piece of glass lit from the camera side in a darkened room with black sheets behind them. They then flip (mirror) the video later in post editing so that the viewer can read what they wrote on the glass in the correct orientation since the professor speaking is standing behind the glass relative to the camera. This produces a different effect and requires less editing in post, but my method gives me more control and allows me to have drawings and page notes prepared in advance in MS Whiteboard that I can then markup while recording.

@@ProfessorV. Thank you very much for sharing your knowledge.

Once turned on an SCR remains on until its current falls below a minimum current level. This occurs naturally when the AC sinusoidal voltage is too low to support enough load current for the SCR to remain latched (natural commutation) or because another SCR is triggered on which imposes a reverse voltage across the first SCR that quickly drives its current below the minimum threshold (forced commutation). On a three phase system, and assuming that the load is sufficiently inductive to keep current flowing continuously, once turned on, each SCR will remain forward biased and conducting for at least 120 degrees before the next SCR is gated. This also assumes that the firing angle remains constant. If the firing angle varies within one AC cycle of the line frequency, the conduction time of any given SCR can be shorter or longer than 120 degrees, depending on when it is gated relative to the other devices. If the firing angle is constant, then all devices will conduct for 120 degrees for symmetry (all devices perfectly share conduction time over the cycle). The three SCRs Q1, Q3, Q5 alternately deliver positive current to the load over the cycle, so each of those three is conducting for 360 degrees /3 = 120 degrees. SCRs Q2, Q4, and Q6, similarly switch one from the other to share negative return current over the cycle, again for 120 degrees conduction time each. The switching times of the positive leg SCRs will be phase shifted 60 degrees relative to the gating times of the negative devices, but individually, all SCRs will conduct for one third of the AC cycle (under balanced conditions when alpha is constant) because three SCRs are used to deliver positive current to the load, and three are used to return the negative current back to the source. At any given instant, only two devices are on at a time, one in the positive leg and one in the negative leg. As a footnote, if the load is only resistive or very light (high impedance, low current), then it is possible for the load voltage to go negative at some point in the cycle, at which point the SCRs would naturally commutate off as current drops below the minimum threshold. Under this condition we say that load current is "discontinuous" (appears to be pulsating on and off at 360Hz for a full wave bridge of six SCRs) and we would then note that SCR conduction is now less than 120 degrees. Hope this helps.

Assuming you are firing all SCRs at 65 degrees relative to their 0 degree reference points, let's consider the case for Q1 whose anode is connected to phase A of the power supply. At the moment of 65 deg., phase A is the most positive and phase C is the most negative (relative to neutral, not shown). Therefore, Q1 will turn on and diode D2 will return the load current to the three phase supply on phase C. However, at angle 180 degrees, the positive phase B SCR (Q3) is still off and won't be turned on for another 5 degrees (its 65 degree point). At 180 degrees (60 degrees relative to Q3), phase A is becoming more negative than phase C so current will naturally commutate off of diode D2 over to diode D4 on the negative load side. But at this point, SCR Q1 and its opposite diode D4 are now the only two devices conducting load current, which provides no path back to the AC supply, so what's happening? Well essentially, for this last 5 degrees, inductive load current is simply free-wheeling its decaying energy, which is no longer fed by the supply, through the Q1/D4 pair and voltage across the load will sit flat at the volt drop of the two devices (about -2V). As soon as SCR Q3 is fired on, load voltage will immediately step up to the Vba line to line voltage, resupplying load current, and the process continues. So to answer your question, relative to SCR Q1, D2 conducts from 65 to 180 degrees, then D4 takes over from 180 to 185 degrees for a total of 120 degrees conduction time on Q1. The same logic applies to the remaining two SCRs. This refers to the semiconverter of course, and not the full wave bridge made of six SCRs.

What do you mean by "RLE". Can you be more specific?

Thanks Steven, Always happy to hear someone derived some benefit from my work.