When I was a kid tinkering with electronics I "discovered" the concept of 4 diodes arranged such that no matter how I hooked up a battery to a project I would always get the polarity right and not destroy anything. I though I had the million dollar idea and wondered why no one hadn't thought of it already...
Keep doing that! Once you "discover" something, the first you want to do is research. You never know, but you might hit a jackpot with something someday, and be sure to document it along the way.
If you basically loose half of your battery voltage with the forward voltage drop of the diode, it just does not make economic sense. A simple [+] sign indicating the polarity and 2 brain cells to figure out how to install the battery correctly solve that problem for most applications just fine.
I "discovered" the bridge rectifier as a kid too! My grandad had a book about science that I loved but it was from the 1930s, and it showed a big mechanical rectifier and I thought "there must be a better way". I wonder how many times the bridge rectifier has been independently discovered!
@@ProtonOne11 You missed the point. To an 11 or 12 year old kid the "discovery" of a full wave bridge rectifier made from 4 discrete diodes was like gold. Yes, you loose a ton of voltage across the diodes, but again, you missed the point of the "discovery". As for 2 brain cells, maybe you should post again when you find them. Be well.
Totally outside the scope of this video, I will just say how cool it is to see traditional draftwork done so well. Being able to share one's mind through clear words and clear illustration is becoming a lost art. It shows a respect for one's audience and for oneself. I thank you.
When you go above 10A these things start looking real good. Above 25A, especially for naturally cooled devices, Ideal diodes/ active rectification is almost a must have.
I was thinking that it was probably better for higher current loads as well. But what would be the use case for higher current ac to dc loads that this would make use of this?
@@siimanderson4549 Good examples, I knew there would be some good uses I just couldn't think of one right off the top of my head. Thinking of high current inductive loads my mind went straight to motors and with ac an vfd setup would be better so I was drawing a blank
This is really interesting! I've read that some modern cars have active rectification in their alternators to reduce power loss. This really makes sense with the low voltage and high currents in a car. 12-14V and 100Amps or more is not uncommon. The three phase windings in a car alternator have six diodes, and if they drop 0,6V each at 100A, that's a loss of 360 Watts to blow away with the cooling fan!
Well, the good news is that in full wave 3 phase rectification using 6 diodes only 2 diodes (or no diodes) are conducting at any given time so, using your numbers the power being dissipated would be (0.6V X 2) X 100 Amps or 120 watts. It doesn't really work that way though... The bad news is that at high currents silicon rectifiers drop much more than 0.6 volts. It is not uncommon to see 1.0 volt to 1.5 volts (look at the Vf curves on a 1N400X data sheet...) forward voltage drop across a silicon diode when conducting high currents - which is the case in most rectification systems because the conduction angle of the diodes is much less than 180 degrees (single phase) or 60 degrees in a 3 phase system. Since the diodes are not conducting continuously, and the output current is a continuous 100 amps (in your example) it follows that when the diodes ARE conducting the peak current must be MUCH higher than 100 amps in order to provide an average 100 amp output current. These high current pulses (near the maximum voltage of the AC input) result in a much higher RMS current in the rectifiers than you would expect which causes a lot more heating than you would expect - requiring a much larger diode than you would expect if you are to have a reliable circuit.
@@Vincent_Sullivan Yes, you're right about how the diodes are conducting during the 3-phase cycle! My maths may fall apart but I still think there's a lot of losses in a 12V full wave rectifier with ordinary diodes, just the way you describe. :)
@@Vincent_Sullivan 1n4007 isn't exactly a high current diode. With appropriately overrated diodes there shouldn't be more than 0.7v drop, and much less with Schottky diodes. And your note of the diodes only conducting for a lower conduction angle isn't really relevant because that is the angle of each diode. The total angle is much larger. Newer alternators even use four phase setups which increase the total conduction angle even more. For a 100a output I would expect 200a diodes would work just fine. What would you think?
Nice magic eye for your avatar. I love it. I forget the channel, but it reminds me of a video I watched the other day where somebody tried making a modern version of a magic eye using entirely too many LEDs, and a diffuser to soften the light. The end result looked nice, but it was far from practical. And I'm sure some Chinese place has already made some janky LED based magic eyes ages ago.
A million years ago, battery chargers used something called a "synchronous rectifier." A solenoid, driven by the magnetic field within the step-down transformer, moved a set of switch contacts back and forth in phase with the AC (hence "synchronous") and these contacts literally switched polarity of the AC connections between the transformer and the battery being charged. This provided the rectification. The whole thing, while noisy, was very efficient. There were no silicon diodes then.
@@elgranlunatiko Not long. They proved unreliable and the switch contacts had to be filed often. But, we're talking about the 1930's, so the only other option was huge, expensive, and grossly inefficient oxide rectifiers.
@@npenim All cars used to use such regulators. They were simply a couple of relays with carefully calibrated pull-in voltages, and these relays would switch on and off in response to the output of the alternator, which then switched resistors in and out of the field coil circuit of the alternator, thus more or less regulating the voltage.
Trap that a lot of people get into: rectifier with a capacitor and some load can draw current peaks easily reaching 7-8x the load current, a 10A rectifier will burn out on a 10A load, the better the smoothing the higher the current peaks
Finally! Someone who actually understands how rectification systems work! Trombo is 100% correct. The larger the capacitor the shorter the conduction angle of the rectifiers and thus the higher the peak currents which means more heating than expected in the rectifiers. My rule of thumb is for a full wave bridge with 10 amps of output current you need about a 30-35 amp rectifier. If the circuit is fed by a transformer you also need a larger transformer than you would expect for the same reason.
Sure about that? A 10x pulse should be high but ok if short enough. But trying to pull 10A through a 10A diode, I would be Very careful about designing the cooling of that, since it will likely be some 20-30watts of heat being generated
In this world of cost cutting, most devices are likely to stick with an FBR just because of the lower parts count and cheaper components. There are designs out there living off of single diode rectifiers still.
Funnily enough single diode is still more efficient. If you ignore the terrible power factor due to the current spike from the capacitance at least. So, if you can get away with it, it's worth considering. Interestingly, many switch mode power supplies use a single diode on their secondary for exactly this reason. It's AC from the coil, but they only use half of it.
This is honestly the best channel on youtube for me. Educational, interesting and with a good sense of humour.... everything yt should be. Keep it up mate!
I have been designing these active rectifier for a long time and they are amazing.....until the controller is destabilized for some reason. Over time, I learned that every attempt to throw it pulses, noise, etc would reveal any stability issues. The end result is a very high-current design with no need for a fan and many times, no need for a heat sink of any kind. Totally worth it.
Interestingly, I was watching another video about these things and power factor correction earlier today. Your comment reminded me of FesZ Electronics showing that an AC side inductor can cause Killo-Volt spikes if unplugged at the wrong time. It makes sense, but man I wish my EE courses were this informative. th-cam.com/video/arvNukW-jeo/w-d-xo.html
@cat Absolutely, you just need to use MOSFETs with lower drive voltages, or use gate drivers that utilize a bootstrap circuit, which is a charge pump that will increase the gate-source voltage of a FET.
I designed an active full bridge rectifier 35 years ago for use in an x-ray machine. It was used in a low voltage sensing circuit where diode drops would have been unacceptable. These have been around a long time. By the way the original rectifiers were called commutators. All generators are really alternators, add a commutator instead of slip rings and you get DC instead of AC. There were giant synchronous commutators used in the old days to get high DC current for subway trains.
Except that a commutator is used to periodically reverse the current, it is a type of switch, and is nothing like a rectifier. A commutator is the part in a DC brushed motor, where the brushes connect to the circuits or coils, as the motor turns, the next coil gets an opposing current to the previous one, and so it makes the motor work. It is not what you described here and does not do what you described here. What you described, is a bridge rectifier with extra components, and not a commutator.
@@dk6578 He is talking about Mechanical rectifiers... which are also known as commutators. Yes it is a part in a DC brushed motor but it IS a rectifier. "serving as a mechanical rectifier to convert the alternating current from the windings to unidirectional direct current in the external load circuit."
@@DarkAttack14 But it isn't a rectifier. A commutator periodically reverses the current, a rectifier doesn't do that. And a brushed motor has no such thing in it. I know, I've opened enough of them.
@@dk6578 look up mechanical rectifier, commutators are a form of rectifier which isn't debatable lol. "The commutator acts as a mechanical rectifier or inverter which changes the polarity of supply/generated emf. In DC generator, the commutator converts the alternating current into the direct current"
@@DarkAttack14 You do understand that a rectifier only lets the power flow in one direction, correct? Because if you understand that, then you would also know that a commutator can not be a rectifier. A rectifier is not the same as an inverter. A rectifier takes AC and gives you DC with just one component. An inverter takes AC and gives you DC using a lot more components. You were talking about brushless motors, not a generator, not exactly the same thing since a generator can have circuitry to boost or buck the electricity. There is nothing in a brushed motor which switches DC to AC or AC to DC. There's the stator which is wires and an iron core, then there's the magnets, and the shaft, and the brushes and the contacts, nothing else. You talk smart, but you seem to actually know nothing.
I LIKE the presentation process of this video...your video segments are assembled in such a visual manner with the voiceover so well matched with the visual narrative that I could better understand. I am a very nontechnical individual with high curiosity but low technical knowledge. Your use of actual data sheets and highlighting the line with the specific element discussed is an example. THANKS so much.
For very low AC voltages such as e.g. the 6V of a bicycle dynamo, an active rectifier is absolutely necessary to achieve sufficient efficiency in the rectification.
I charge my Phonefrom my bicycle dinamo . And dont have a fancy rectifier. I use 4 diodes and a cap. And a step up / down converter. It's simple and effective. And the output is a usb connector. It even fits in a 3cm long electrical pvc pipe.
@@mickgibson370 Aren't most phone batteries 3.7 vdc? Seems like there's plenty of overhead here. There are places where the increased efficiency is warranted, and places where it is not - it's a matter of price/performance/reliability, etc.
As industrial's sensor design, I use that from the last 11 years for being mandatory. That little loss from rectifiers play a big role in some remote applications, specially when temperature is a problem.
@Christopher Grant sorry for the delay. Sourcing AC to sensors directly isn't common, except from sensors with NO/NC to be used in cheap chinese machinery (and we produce those, because of our clients' needs). Anyway those aren't critical. But we sell sensors used in laboratories where temperature and humidity fluctuations are a big problem (turtle eggs for example, are very sensible to those controls and a 2°C fast fluctuate can represent the failure to some sensible species, so they told us). I don't know two damns about those applications, but the specs we must follow for those stuff force us to design very thermally optimized products, and there is this kinda of losses play a big role. And batteries aren't an option, because people simply forget about replacing it (or so they tell us when specifying).
Your videos are the best presentations of electronic concepts. I love the way you draw schematics & add written explanations to accompany your commentary. Many thanks
I've seen mosfets used in the output of some USB chargers as a diode replacement since those 1-3A loads are a significant loss across that component at those voltages. Pretty smart use of them.
So did I. Using fast Shottkys is usual in PSU, getting the efficiency >85%. Active rectifiers lift it up to >90% or better. So, if you get an PSU with >90% efficiency, you can be sure to find active rectifiers for the output.
@@diegoyonamine8943 nahhhh you thought it so gonna let you keep that response!! Also can't wait for electrosboom response he he ^^. I don't usually eat this but umm *gets up and starts making pop corn* this gonna be gooooood!!! ^^
@@PhantomWorksStudios Yes, agreed. Both positive ContentCreators ... making fun into science educational. LOL . I may explain myself tho.. i suggested the edit/ad instead of spamming same other comment... but it's all gud 😂
It's a great idea. i had this electronic rectifier in the early 90's to develop a high power SMPS. The biggest plus is that you can ramp the main Voltage slowly to avoid peaks at ramp-up and control the PFC. For high current very good.. for mainstream electronics too expensive
FBR may also be a better option in adverse thermal environments. You end up with less variation in output characteristics, which makes handling that variation safely much easier. Even with well-designed power supplies, you might be amazed how differently a device might act at -40C vs. at 85C
@@erenoz2910 That was my main concern. MOSFET operation is highly dependent on parasitic characteristics (mainly capacitances between different nodes). These characteristics are typically relatively susceptible to thermal variation, which can make building an efficient controller which works across temperatures challenging. An FBR (Full bridge rectifier) is going to exhibit some variation in performance over device temperature, but it is comparatively easy to deal with since it isn't going to mess with a drive circuit... More notably, the FBR a single part as compared to the IC, four MOSFETs, and a handful of capacitors. I tend to like simple. It is less likely to break, and easier to troubleshoot if it does break. In some applications it may be possible to use a Schottky FBR, and get almost as good of performance as you would get from active rectification, but the reason that we tend to use traditional diodes for that application is that you don't really get much for the development time needed to implement a different solution... and that isn't taking into account the reality that higher part counts take more space on the PCB and increase the number of potential points of failure... In production that is going to mean a higher fallout rate, which drives up costs and reduces customer satisfaction. I would generally save the MOSFETs for applications where you need to have the switching (like an H-Bridge). I can imagine a few places where extreme efficiency may be desirable, but just not that many where it really overcomes the tradeoffs.
You are adding a lot of complexity to solve a problem that may not be an issue. But it’s interesting and educational to watch you do something in an unconventional manner.
I think with the capacitors fully charged, and when the MOSFETs in a bridge start to open, you have higher voltage on DC side an lower voltage on AC side. That makes a path for capacitors to discharge. So all the process is actually pumping/draining the capacitors. You still need a diode, to prevent that backflow.
i have been wanting such a rectifier for a long time, actually - it can pump electricity back into the grid, which is sometimes useful (for example if you're speed-controlling a dc motor with an autotransformer followed by such a rectifier, turning the motor speed down will actually cause regenerative braking). A huge inductor before the capacitor should be able to take care of these recirculating currents.
@@victortitov1740 If you connected a synchronous rectifier directly to a DC motor I would expect the losses to be higher due to the massive AC component of the voltage waveform applied to the motor. For your speed control application I would just use regular diodes. Instead of regen braking when you turn down the knob it will just coast.
@@user-td3yi1mq7p because the bumpy DC one gets out of a capacitor-less rectifier will be forced upon the capacitor. That is of course only for the rectifier that opens and closes the mosfets synchronized to (and only to) mains voltage polarity. The rectifiers GreatScott shows are reacting to voltage drop across each "diode" instead, so they behave essentially exactly as normal diodes.
When I attended engineering school in the early 1970s we always called it a full wave bridge rectifier. I never heard it called a full bridge rectifier until fairly recently.
@@peterduxbury927 "this is because you could make a half-wave Bridge Rectifier" A bridge is always a diamond shape (although it can be portrayed rectangular). Think Wheatstone Bridge for example. A center-tapped transformer can give full wave rectification with two diodes, but each winding is actually only half wave with one winding having current in the positive half cycle and the other winding current in the negative half cycle. That's fairly common but I wonder about the presence of pulsating DC in each half of the secondary.
I use bridge rectifiers a lot for compact polarity protection in low current applications, and also to exploit the 1v drop to avoid the need for regulation in some cases.
Most full-bridge rectifier packages are built with simple silicon diodes with a high voltage drop - that's why when I built my 120VAC input lab power supply I assembled my own rectifier using Schottky diodes (MBR40200PT ASEMI TO-247/3P), which cut the voltage drop in ~half and increased efficiency significantly. Using a larger than needed size diode also helps with efficiency. I still like the simple diode solution because it's 'bulletproof'.
With one observation ... that IC only starts to work at approximately 9v AC and can only handle up to around 72v AC , so it's not really usable with 110/230v AC and the 9v threshold is a bit of a pain if you want to use it on the secondary side of a transformer... it cuts quite a bit of the sine wave. Last but not least, it's Linear / Analog now, so expensive...
@@mariushmedias LT4320 is a fun but useless toy sadly. The high cost kills what few applications fit within its voltage window. It's also useless as a secondary-side SMPS rectifier due to the 400 Hz max frequency, if memory serves.
Your conclusion is right on! I designed and built my own active rectifier system using comparators and mosfet drivers etc. It worked extremely well, BUT, was only more efficient in a rather narrow range of power than a simple bridge rectifier.
I think, through some sort of protection kept forward, you can try going for phase controlled conduction mode as well, which I believe is industrially done through IGBTs...
the older i get the more and more I like you videos. especially this one feels like a electrical circuit lab in uni. I kinda miss the labs, even though I am not good and I don't like working with circuits. They ended up being a hobby for me.
I think you forgot about reliability. Diodes themselves are typically very reliable and can work for years without a problem. This active rectifier adds additional ICs in the main power path. I doubt it will be more reliable, probably much less. Additionally as you mentioned - this will only make sense if a lot of current is flowing through the rectifier, so it will be a very niche device.
@@internetisinteresting7720 Because you might need to squeeze more efficiency out of the device, and don't want to waste power dumping lots of current through diodes?
Many Hybrid and Electric vehicles make use of active rectification. The 3-phase bridge has a fair bit of work to do when you consider field oriented / space vector control and then switching to regen on over-run. 🤔
Good video. Active rectifiers have been known since the 1980s, and we used these in some satellite applications where low voltage batteries were being used. The main problem with FETs is that the higher the voltage standoff, the higher the source/drain resistance. FETs now are up to about 1000+volts and IGBTs are over 4000+volts, but in both cases the loss voltage across the device is almost proportional, so still requires roughly the same cooling power, for the given application. We did railroad applications at 2000+volts, 500+amps, where we had to use antifreeze liquid for cooling, in specially milled tubes inside of the heatsinks. Freon and chilled water cooling was also tried, but these has significant physical disadvantages, and had poor heat transfer under unusual or incorrect operating conditions, often resulting in catastrophic failures. Our ten-ton experimental 6-wheel truck would often burst into flames, with such failures. In thermal world, "there is no free lunch". You must cool adequately, for any given power level, period, or burst into flames. Your example is good for a "few" watts of loss, and then it will cook itself, since the heat outflow is very low.
I'm definitely starting to see more active rectifiers in the output stage of USB phone chargers, where they just used a single diode in the past. I guess this is where you would see the greatest efficiency benefit. These active rectifiers kind of remind me of the old fashioned mechanical rectifiers.
PC power supply is a good application too. They already use a center taped winding to reduce the diode losses but the remaining 4 ones are on a rather big radiator.
These broke into big inverter drives years back. For three phase high power you can twiddle the active switching in real time to minimize distortion on both sides. Also they let you run the drive backwards during braking. Turning high inertia loads into massive generators. I think ABB called it an ‘active front end’, A-B and Siemens have their own names for the same thing. Hella cool.
This is how HVDC converters work on the power grid! You take 3-phase AC, run it through a wye and a delta transformer to get 12 poles (6 pairs of phase). You then feed them to an active element, triggering at the right point. In the old days, it was mercury valves, then it moved to thyristors, and these days they're often IGBT (Insulated Gate Bipolar Transistors). For HVDC applications, these are huge, dwarfing a human. Voltage differences of up to a million volts kinda needs some room. Modern designs use stacks of them for each pole, switched at different times to match sine waves, minimizing ripple in magnitude and raising it in frequency, so it can be filtered with smaller capacitors + inductors. The 3.4 GW Pacific Intertie uses thyristors now, originally mercury valves when it was placed into service in 1970 (originally at 1.4 GW). The newer Transbay link under San Francisci bay uses IGBTs. HVDC links between England and Europe, Tasmania and South Australia, US/Canada and US/Mexico. In some places, there are back-to-back pairs to isolate different AC grids from each other; Japan has 4 like this, with 60 Hz on one side and 50 Hz on the other. It's the same circuitry on each side, whether used as an inverter or an AC-DC converter. Which direction the power flows comes down to the phase angle of the triggering signals. So solar and wind use them as inverters, while grid batteries use them in both modes.
@@BobKerns4111 oh yeah. HVDC is hell of cool! Crazy phase shift TX’s and massive multi pulse stuff, and the voltages, Jesus wept the voltages. I mean I get cagey working up to 5kvac, but it’s still just AC. And DC had its own quirks and foibles, but, getting up to 100’s of KV? Auuuugh that is a job for someone with more guts (and brains) than me. Cool shit though. Thx for the info!
I can understand why you were so nervous about causing a short circuit, main can be very scary! Once, when I was 13, I was playing around with a computer power supply, and I accidentally, switched it from 240V to 120V and turned it on. As I am in Australia, our mains is 240V… I think you know what happened next. A huge bang and a flash of light later, I had learned what that switch on a power supply did and never made the same mistake again XD
Yeah, there is a slight problem with capacitive loads, but if you suddently start drawing several amps, you could face the exact same issue (aka accidentally building a bomb). Also it's hard to have a very wide supported voltage range when making them. Just like synchronous rectifiers that you can find at the output of very high efficiency switching power supplies, they are stuff you better don't mess with unless you really really know what you are doing. I know what I am doing enough to create my own reliable power supply, but I'd still not feel comfortable building this kind of things...
the purpose of an active rectifier is to overcome the nonideal rectification behavior of an individual diode. the MOSFET drivers are called ideal diodes when used for rectification.
You are a wonderful teacher. Much of what you did was above my level. However your presentation is delightful. The neatness and colours ... Thank you so much for your teaching. David Lixenberg
Interesting. I wasn't aware that NXP picked up my idea from 2008 (See Christian Rausch, "Greener Rectifier Loses The Diodes, Adds Power MOSFETs, Efficiency", EDN, 2008, October Issue) and made an integrated circuit from it. 😁
Such a low drop bridge rectifier is basically essential, as you mentioned, in very low voltage applications where you don't want to lose voltage and/or waste power. BUT in this case you don't need such a complicated circuit. For a 3V or less application I designed a rectifier which basically consists of a flip flop: you apply the ac voltage on the "outputs" of the not gates and get the rectified voltage on the "vcc" and "gnd" terminals of the gates. Each not gate is made of a PMOS and a nmos transistor. At first the circuit starts because of the parasitic bulk source diodes which constitute a traditional bridge, then the gate source of each MOSFET is driven properly by the output voltage. By the way, what I would really love is an active zero-drop diode (MOSFET with it's own driver) to be used in the secondary of low-voltage-output switching power supply, where the voltage drop of the diode is causing a big loss in efficiency
7:31 The simplest and most likely answer may be to observe that a capacitor initially looks like a dead short. I can't see the size of the caps you're using, but they look big enough that it will look like a short for long enough to blow that fuse. This will be at least partly why the datasheet stipulates the use of a boost converter on the rectifier's output. A naïve boost converter will present a mostly inductive load at its input, so there is no high initial inrush current in quite the same way. Real boost converters will still have some C on their input in order to control EMC, but not as much as you're trying to load the rectifier with. Beyond that, another possible answer, and this is just a guess, is that these IRF740 FETs are just too slow and chunky for this application. As you noted, if a third FET is still partially conducting, you'll get a low impedance path that might look like enough of a short to blow the fuse. There are two parameters worth looking at that might explain how that situation could arise, even with the TEA2208 controller: gate charge and conduction recovery time. As you probably know, 1) a FET's gate is a capacitor and 2) the gate voltage that matters is that relative to the FET's source pin (ie Vgs), which is not always ground. Especially for the high side FETs, the controller has to be able to drive a sufficient Vgs relative to its own ground to turn it on†, but that voltage will be different from that relative to the controller's own ground. So, usually, the controller needs a charge pump to get the required Vgs. The current capacity of this charge pump affects how quickly it can turn the FET on and off, so a FET that stores more charge on its gate will take longer to switch. Contrast the specs for the IRF740 with a better power FET like On Semi NTMFS5H600NL. Recovery time is probably not relevant for a sinusoidal input at these frequencies, but it's worth remembering that a FET doesn't stop conducting immediately after Vgs goes away. † "on" can mean "in linear/triode region" or "in saturation". These terms are also used with bipolar transistors but, unhelpfully, their meanings are reversed. For FETs used as switches, you want the smallest possible drop across the FET itself, which means operating the FET in its triode region (Vgs > Vt, Vds < Vgs-Vt, noting again that Vds the voltage across the channel, not relative to ground).
Integrating a PFC with the active rectifier might make the increased complexity worth it. Yep, with 1A at mains input voltage you get 1W or 0.5-1% loss. It's usually good enough.
@@johndododoe1411 I think that's why the HV bulk capacitor after the PFC stage is used. Storing energy on the low voltage secondary side is difficult, and you have to supply power for 2ms of the cycle where there's very little input voltage.
A good place to start in deciding if they are worth the cost is comparing the diode drop to the supply voltage. 1.4V to mains voltage is a very small fraction, so the potential benefit is also quite small. And the active circuitry requires some power itself. Below some current threshold, that would be greater than the loss in the diodes themselves.
If you're planing on further improving power efficiency don't forget consider lowering the resistance of your wiring & the ESR of your power MOSFETs (some power MOSFETs only have a few dozens of µOhms, some of them are pretty new & expensive & involve fancy semiconductor materials like SiC or GaN). Also I've noticed that the quality & ESR of the smoothing caps matters - especially the big electrolytic ones.
I've always made it a point to have both bulk and filter caps (electrolytics and ceramics) in parallel - the bulk caps deals with the higher discharge currents while the ceramics deal with the noise from the rectifier. That being said, if you can simplify your decoupling and reduce part count with a higher-performance component - it's also worth it.
I also am one of your faithful students that comprehend through electric training, the Power Factor, now very well discussed in electronics. The current leads the voltage which is slower to flow, as a brown out, until the Power Factor is corrected with capacitance to have the current & voltage be supplied to the load. An Inductive load has a Power Factor correction to be made. ♥
From a purely theoretical standpoint, the fact that there are more components involved with the active circuit means that the MTBF (Mean Time Before Failure) will be shorter with the switching circuit than with the diodes. However, in real life it comes down to the specific design of each circuit and how hot each of the components get - typically a 10 degree C rise means roughly half the service life. Assuming there is enough space and / or airflow to keep everything cool enough, it shouldn't be hard to get the circuit to last for decades of normal use. For anything attached to AC mains however, a big killer is lightning surges that make it into the house. So the limiting factor is how much surge protection is available before those circuits, and the voltage ratings of the devices attached to the mains.
The reason for the part that only controls the low side is likely for use with some kind of semi-bridgeless boost pfc topology. For those circuits you do not need to rectify the upper half since the line and neutral are tied to inductors directly.
This reminds me exactly of when I was learning about H-Bridges in one of my uni classes where instead of acting as a rectifier, the MOSFET gates were controlled by a micro controller at a set frequency and we were effectively inverting a DC source to alternating AC. We also had to worry about ‘dead time’ or essentially, the necessary length of time from the microcontroller where all MOSFETS were closed between cycles so as to not cause a short, great nostalgia.
I thought of the same. A while back I was thinking about building something to convert single-phase to three-phase AC. It used six MOSFETs in the inverter to make three square-waves and an Arduino to control them.
What surprises me is that after decades of electronics and semi conductors, there's still novel ways of rectification being found. I would have thought that by now there's one tried and tested and best solution for every usage scenario established.
A warning for everyone: the configuration at 5:42 is dangerous as it's referenced to the mains. An autotransformer DOES NOT ISOLATE FROM THE MAINS! It's also a good way to blow up your oscilloscope.
@@thiagoamericano1412 That does not remove the risk of mains power. An autotransformer, and usually every variac in general just does NOT isolate from the mains.
I've used the mosfet approach to rectify the output from a bicycle hub dynamo to power a GPS tracker. Diodes would be too wasteful - it's only 6V at 3 Watt power max.
I have never used this type of circuit in my designs for power supplies. Purely reactive components on the outputs will be difficult for this system to control. Capacitance across the output without any real restive impedance ( I know redundant) will cause instability in the active circuit. PFC will improve the ability to of the active rectifier to control the system, and this alone would limit me using this in my designs for most of my applications. As some have said in the comments, at higher currents this system would be very helpful in controlling the forward voltage drop power losses in a standard diode. However the costs in switching noise, board space and limited output options will relegate this to a narrow family of applications. The explanations were very well done. I would suggest however the true name of this system is a Full Wave Bridge rectifier, the other option is a Half Wave rectifier.
I learned that I will continue to use full bridge rectifiers and lead based solder as somethings just work with out added complexity and points of failure. Keeping things simple makes for greater reliability as a rule but not always as efficient as more complex circuit configurations.
I realize I am focusing on the wrong part of the story but, man that is some sweet handwriting. And most people go their whole lives without using the phrase "Luckily, nothing blew up."
Fascinating. Definitely seems like a better idea for things with huge draw: maybe PC power supplies? (Which often have PFC anyway). I was going to say that DC charging of EVs might want this too, but the high speed DC chargers have big cabinets of electronics so they probably have something even more sophisticated. But, maybe the onboard charger in an EV for connection to Level 1/Level 2 AC service equipment uses these. My relatively new furnace has an "electrically commutated motor" in the blower, which is basically a DC brushless motor. Maybe it could also be a use for these, since the blower is a fairly large draw. (Or inverter type heat pumps...)
Outside of special applications or high power devices, these seem unnecessarily complex. More circuitry means more chances for something to fail. However, for high power applications these do seem like a pretty nice way to reduce the need for active heat dissipation on the input side, but depending on the amount of power conditioning after the rectification process or the amount of heat dissipated by the power supply total, the amount of benefit gained might not be significant enough to justify the additional manufacturing cost. Still a pretty cool circuit nonetheless.
Not just high power. Less power loss means less heat generated, means you might avoid a cooling fan. Active rectification is very common with modern switched mode power supplies
@@matthiaswandel Fancy seeing you here! I agree with you. I just meant the cost benefit ratio would be larger in high power devices, lower power devices would also benefit albeit to a lesser degree. If I could remove a cooling fan, and the money saved on that was higher than the increased cost of the circuitry, this component would definitely be a viable option. I was just taught to always to minimize complexity; if you have two solutions that work use the simpler one, because it minimizes failure points and reduces cost to the end user. When I wrote that response, a rectifier costs around two dollars, and the active rectifier cost nearly ten. If the benefit gained *is* significant enough to justify the additional cost, then its worth consideration for sure. The customer is the one paying for the additional cost, so if they are willing to pay more and the outcome is greater reliability, its a win win.
Great video, you didn't mention schottky diodes within a full bridge rectifier? I've not seen them packaged this way as they are with regular bridge rectifiers but it would be interesting to know their efficiency also. keep up the great work!
Is it because schottky diodes have greater reverse voltage leakage making the whole thing short together? I don't know that's why we're here but I looked it up and it's only like 40V almost like a zener so for low voltage maybe. For high voltage, I'm thinking a hard NO. Feel free to correct me if I'm wrong.
@@foxfyre3600 Correct me if I'm wrong, but wrong you are (I think). Basically all SiC power diodes are schottky junctions and they sport some of the highest breakdown voltages available. They are not very popular due to their cost vs. an active rectification solution and negligible efficiency gains attributable to lower forward knee which matters least at high voltages.
Would have to distinguish a schottky's emission coef. given a (probably) unknown saturation current by measuring based on SiC thermal voltage among other things. To know for sure I'd just hunt down a proper model with known process parameters and calculate.
That looks like a good match for things like wind/water power solutions, where you have low AC voltages and even a small increase in efficiency can make a big difference. Does that make sense?
I would be remiss to not point out that even in the test where the active rectifier was noticeably better the FBR only fell to 96-97% efficiency - not anything worth stressing about.
TEA2208 is £6.90 at Mouser and the LT4320 is a similar price. My use case is a 0.25V power supply for electrolytic purification of Copper. It's not so simple to design a switch mode power supply for that either. Going back in time is a tunnel diode transmitter which requires a similar voltage where dropping it from 1.5V is very inefficient. For these sort of applications one needs to roll ones own circuit.
You should do another one on DC-DC power supplies. My company loves to use them. We'll send out 24Vdc via a cable and then use them on a board at the end of the cable to then convert it to 5v and/or 12v. We even had a few systems run on 56Vdc as a mains voltage then converted down.. For me those DC-DC PS are a common issue, we use an IC version soldered on boards.
I think the most interesting thing is the MOSFET driver ICs (TEA2206 & TEA2208) while being active, meaning they require a power supply, they use the AC mains by charging a capacitor on the initial cycle. The data sheet says the body diode in the MOSFET is used initially in the 2208 as a traditional rectifier to charge the capacitor and power the IC. It would be neat to look closer at exactly how this action happens. Nice video!
Very nicely explained and drawn. You explain things so smoothly and calmly. Half the time I was expecting you to get shocked, burnt, bang your head, etc. Then I realized that was because every time Mehdi connected something, bad things happen. Keep up the good work.😊
The most fun I had with a bridge rectifier was modifying my Mom's "Light and Motion" Christmas tree ornaments. These had any number of motorized operations to bring the ornament to life, but if you had a number of them on the tree, they got noisy. So, I put a pair of diodes on each ornament so that one polarity went through the ornament, and the opposite bypassed the ornament. Then I ran all the power going to the tree through a full wave rectifier, followed by a double pole switch that allowed me to reverse the output polarity. Viola, in the middle switch position, the tree was unlit. In another, it lit up and in the third, it lit up and the motorized ornaments came to life. Mom thought it was magic.
Hello! The main disadvantage of a diode rectifirer - generation of harmonics in an AC main network.For example - when the average current of a load is 1A , the peack current of a recifier may be greater that 5A. When we want to solve this problem - we need a PFC. Unfortunately, there isn`t any PFC on Chinese Ali Express.
@GreatScott - if I'm not mistaken, that MOSFET network with H-design & switching controller is nearly identical to the design of a pure-sine DC/AC inverter, many of which use the EGS002 controller board that's everywhere on the internet these days, and built into many inverter products on the market.
OK, now I have to do another project. You can just make a little bias circuit for each one to turn it on when it goes forward bias, right? With zeners to protect the gate from over voltage either way. I am super busy right now with electro-chemistry, but I will try one very soon.
I think during the capacitive load test the capacitor bank was unable to discharge properly, it kept on charging, adding a bleeder resistance to it should resolve that, alongside, also utilizing the body diode during some period of the supply voltage. This means that the mosfets should be triggered at a certain value of alpha, which can be adjusted by the time constant of the load.
In a telecom system, the input voltage is -48V DC (nominal). There are two power busses for multiple power feeds for reliability. That means power can be drawn from either. MOSFETs were used to make that selection of power busses. All was good until someone connected one of the power busses backwards. The boards which were already powered up were fine, but any new board installed would blow the fuses in the board, because multiple MOSFET were enabled at one time before the power controller could pick the right one. Moral to the story is that the MOSFET controller may need to be powered by a full wave rectifier before they can be turned on, but there may still be a race condition like we had.
When I was a kid tinkering with electronics I "discovered" the concept of 4 diodes arranged such that no matter how I hooked up a battery to a project I would always get the polarity right and not destroy anything. I though I had the million dollar idea and wondered why no one hadn't thought of it already...
been there done that.
the disappointment that it does exist was painful T-T
Keep doing that! Once you "discover" something, the first you want to do is research. You never know, but you might hit a jackpot with something someday, and be sure to document it along the way.
If you basically loose half of your battery voltage with the forward voltage drop of the diode, it just does not make economic sense. A simple [+] sign indicating the polarity and 2 brain cells to figure out how to install the battery correctly solve that problem for most applications just fine.
I "discovered" the bridge rectifier as a kid too! My grandad had a book about science that I loved but it was from the 1930s, and it showed a big mechanical rectifier and I thought "there must be a better way". I wonder how many times the bridge rectifier has been independently discovered!
@@ProtonOne11 You missed the point. To an 11 or 12 year old kid the "discovery" of a full wave bridge rectifier made from 4 discrete diodes was like gold. Yes, you loose a ton of voltage across the diodes, but again, you missed the point of the "discovery". As for 2 brain cells, maybe you should post again when you find them. Be well.
Totally outside the scope of this video, I will just say how cool it is to see traditional draftwork done so well.
Being able to share one's mind through clear words and clear illustration is becoming a lost art. It shows a respect for one's audience and for oneself. I thank you.
I, too, am always impressed by how well-drawn his diagrams are... almost DIN - standards!
When you go above 10A these things start looking real good. Above 25A, especially for naturally cooled devices, Ideal diodes/ active rectification is almost a must have.
Thanks for the feedback :-)
I was thinking that it was probably better for higher current loads as well. But what would be the use case for higher current ac to dc loads that this would make use of this?
@@D3M3NT3Dstrang3r Any kind of power supply. Say you have a 120V server grade power supply with a 5kW output. It is not uncommon.
@@D3M3NT3Dstrang3r Welding machines, maybe? DC fast charging for cars definately.
@@siimanderson4549 Good examples, I knew there would be some good uses I just couldn't think of one right off the top of my head. Thinking of high current inductive loads my mind went straight to motors and with ac an vfd setup would be better so I was drawing a blank
This is really interesting! I've read that some modern cars have active rectification in their alternators to reduce power loss. This really makes sense with the low voltage and high currents in a car. 12-14V and 100Amps or more is not uncommon. The three phase windings in a car alternator have six diodes, and if they drop 0,6V each at 100A, that's a loss of 360 Watts to blow away with the cooling fan!
Well, the good news is that in full wave 3 phase rectification using 6 diodes only 2 diodes (or no diodes) are conducting at any given time so, using your numbers the power being dissipated would be (0.6V X 2) X 100 Amps or 120 watts. It doesn't really work that way though... The bad news is that at high currents silicon rectifiers drop much more than 0.6 volts. It is not uncommon to see 1.0 volt to 1.5 volts (look at the Vf curves on a 1N400X data sheet...) forward voltage drop across a silicon diode when conducting high currents - which is the case in most rectification systems because the conduction angle of the diodes is much less than 180 degrees (single phase) or 60 degrees in a 3 phase system. Since the diodes are not conducting continuously, and the output current is a continuous 100 amps (in your example) it follows that when the diodes ARE conducting the peak current must be MUCH higher than 100 amps in order to provide an average 100 amp output current. These high current pulses (near the maximum voltage of the AC input) result in a much higher RMS current in the rectifiers than you would expect which causes a lot more heating than you would expect - requiring a much larger diode than you would expect if you are to have a reliable circuit.
@@Vincent_Sullivan Yes, you're right about how the diodes are conducting during the 3-phase cycle! My maths may fall apart but I still think there's a lot of losses in a 12V full wave rectifier with ordinary diodes, just the way you describe. :)
@@Vincent_Sullivan 1n4007 isn't exactly a high current diode. With appropriately overrated diodes there shouldn't be more than 0.7v drop, and much less with Schottky diodes. And your note of the diodes only conducting for a lower conduction angle isn't really relevant because that is the angle of each diode. The total angle is much larger. Newer alternators even use four phase setups which increase the total conduction angle even more. For a 100a output I would expect 200a diodes would work just fine. What would you think?
using a wheelchair is a massive power loss. 4 wheels are inherently and physically unstable and wasteful
Nice magic eye for your avatar. I love it. I forget the channel, but it reminds me of a video I watched the other day where somebody tried making a modern version of a magic eye using entirely too many LEDs, and a diffuser to soften the light. The end result looked nice, but it was far from practical. And I'm sure some Chinese place has already made some janky LED based magic eyes ages ago.
A million years ago, battery chargers used something called a "synchronous rectifier." A solenoid, driven by the magnetic field within the step-down transformer, moved a set of switch contacts back and forth in phase with the AC (hence "synchronous") and these contacts literally switched polarity of the AC connections between the transformer and the battery being charged. This provided the rectification. The whole thing, while noisy, was very efficient. There were no silicon diodes then.
Interesting... But I wonder what would be the load life of such switches?
An aircraft-type motor commutator is almost silent, with far longer life and probably just as efficient. But too expensive for a home charger.
@@elgranlunatiko Not long. They proved unreliable and the switch contacts had to be filed often. But, we're talking about the 1930's, so the only other option was huge, expensive, and grossly inefficient oxide rectifiers.
@@jmi5969 Yes, and there was no rotating component on the battery charger for this to work.
@@npenim All cars used to use such regulators. They were simply a couple of relays with carefully calibrated pull-in voltages, and these relays would switch on and off in response to the output of the alternator, which then switched resistors in and out of the field coil circuit of the alternator, thus more or less regulating the voltage.
Trap that a lot of people get into: rectifier with a capacitor and some load can draw current peaks easily reaching 7-8x the load current, a 10A rectifier will burn out on a 10A load, the better the smoothing the higher the current peaks
Finally! Someone who actually understands how rectification systems work! Trombo is 100% correct. The larger the capacitor the shorter the conduction angle of the rectifiers and thus the higher the peak currents which means more heating than expected in the rectifiers. My rule of thumb is for a full wave bridge with 10 amps of output current you need about a 30-35 amp rectifier. If the circuit is fed by a transformer you also need a larger transformer than you would expect for the same reason.
@@Vincent_Sullivan obviously, if I was taught anything by experiments, it's that to overspec absolutely everything possible. And of course, fuses.
Thank you, David Lixenberg.
Sure about that? A 10x pulse should be high but ok if short enough. But trying to pull 10A through a 10A diode, I would be Very careful about designing the cooling of that, since it will likely be some 20-30watts of heat being generated
@@Vincent_Sullivan understanding is dangerous to the economy.
how is everyone meant to keep buying new products if theyre not designed to fail?
Now eagerly waiting for the response video by electroboom . Its going to be too much fun :D
Looking forward to it :-)
I can hear the burning sound effects😂
@@sudman08 😂
I think simplicity and efficiency are the big differences, but they're conceptually similar.
That needs to be a video with high voltage and at least 1 spark gap somewhere for a ZAP + BOOM effect
As an electroboom subscriber, i was very confused when you had 4 capacitors connected on the screen and none of them had blown up yet
In this world of cost cutting, most devices are likely to stick with an FBR just because of the lower parts count and cheaper components. There are designs out there living off of single diode rectifiers still.
Funnily enough single diode is still more efficient. If you ignore the terrible power factor due to the current spike from the capacitance at least. So, if you can get away with it, it's worth considering.
Interestingly, many switch mode power supplies use a single diode on their secondary for exactly this reason. It's AC from the coil, but they only use half of it.
Ah, the single diode garbridge rectifier!
Like the secondary side of many switching supplies. Single diode and capacitors.
Most cheap things use just a single diode for ac-dc conversion.
Yeah, but tomorrow this will be a no brainer solution.
This is honestly the best channel on youtube for me. Educational, interesting and with a good sense of humour.... everything yt should be. Keep it up mate!
I have been designing these active rectifier for a long time and they are amazing.....until the controller is destabilized for some reason. Over time, I learned that every attempt to throw it pulses, noise, etc would reveal any stability issues.
The end result is a very high-current design with no need for a fan and many times, no need for a heat sink of any kind. Totally worth it.
Interestingly, I was watching another video about these things and power factor correction earlier today. Your comment reminded me of FesZ Electronics showing that an AC side inductor can cause Killo-Volt spikes if unplugged at the wrong time.
It makes sense, but man I wish my EE courses were this informative.
th-cam.com/video/arvNukW-jeo/w-d-xo.html
@cat Absolutely, you just need to use MOSFETs with lower drive voltages, or use gate drivers that utilize a bootstrap circuit, which is a charge pump that will increase the gate-source voltage of a FET.
I designed an active full bridge rectifier 35 years ago for use in an x-ray machine. It was used in a low voltage sensing circuit where diode drops would have been unacceptable. These have been around a long time. By the way the original rectifiers were called commutators. All generators are really alternators, add a commutator instead of slip rings and you get DC instead of AC. There were giant synchronous commutators used in the old days to get high DC current for subway trains.
Except that a commutator is used to periodically reverse the current, it is a type of switch, and is nothing like a rectifier. A commutator is the part in a DC brushed motor, where the brushes connect to the circuits or coils, as the motor turns, the next coil gets an opposing current to the previous one, and so it makes the motor work. It is not what you described here and does not do what you described here. What you described, is a bridge rectifier with extra components, and not a commutator.
@@dk6578 He is talking about Mechanical rectifiers... which are also known as commutators. Yes it is a part in a DC brushed motor but it IS a rectifier. "serving as a mechanical rectifier to convert the alternating current from the windings to unidirectional direct current in the external load circuit."
@@DarkAttack14 But it isn't a rectifier. A commutator periodically reverses the current, a rectifier doesn't do that. And a brushed motor has no such thing in it. I know, I've opened enough of them.
@@dk6578 look up mechanical rectifier, commutators are a form of rectifier which isn't debatable lol. "The commutator acts as a mechanical rectifier or inverter which changes the polarity of supply/generated emf. In DC generator, the commutator converts the alternating current into the direct current"
@@DarkAttack14 You do understand that a rectifier only lets the power flow in one direction, correct? Because if you understand that, then you would also know that a commutator can not be a rectifier. A rectifier is not the same as an inverter. A rectifier takes AC and gives you DC with just one component. An inverter takes AC and gives you DC using a lot more components. You were talking about brushless motors, not a generator, not exactly the same thing since a generator can have circuitry to boost or buck the electricity. There is nothing in a brushed motor which switches DC to AC or AC to DC. There's the stator which is wires and an iron core, then there's the magnets, and the shaft, and the brushes and the contacts, nothing else. You talk smart, but you seem to actually know nothing.
The *FUUUUUUUUUUUUUUUL BRIDGE RECTIFIER* will never be replaced!
... and great video as always👍👍
elctrobom
What's Mehdi's fascination with the full bridge rectifier? I never heard him explain it.
Fool bridge rectified
Is that the "joke"? Fool instead of full? Wow, that's lame :)
Yeah boom baby boom!
I LIKE the presentation process of this video...your video segments are assembled in such a visual manner with the voiceover so well matched with the visual narrative that I could better understand. I am a very nontechnical individual with high curiosity but low technical knowledge. Your use of actual data sheets and highlighting the line with the specific element discussed is an example. THANKS so much.
For very low AC voltages such as e.g. the 6V of a bicycle dynamo, an active rectifier is absolutely necessary to achieve sufficient efficiency in the rectification.
very true! I've been using mosfet rectification for my bike dynamo circuits for many years!
I charge my Phonefrom my bicycle dinamo .
And dont have a fancy rectifier.
I use 4 diodes and a cap.
And a step up / down converter.
It's simple and effective.
And the output is a usb connector.
It even fits in a 3cm long electrical pvc pipe.
@@micheltbooltink But the full wave rectifier goggles .7 volts that you can use!
@@mickgibson370 Aren't most phone batteries 3.7 vdc? Seems like there's plenty of overhead here. There are places where the increased efficiency is warranted, and places where it is not - it's a matter of price/performance/reliability, etc.
@@mickgibson370 1.4V
As industrial's sensor design, I use that from the last 11 years for being mandatory.
That little loss from rectifiers play a big role in some remote applications, specially when temperature is a problem.
@Christopher Grant sorry for the delay. Sourcing AC to sensors directly isn't common, except from sensors with NO/NC to be used in cheap chinese machinery (and we produce those, because of our clients' needs).
Anyway those aren't critical.
But we sell sensors used in laboratories where temperature and humidity fluctuations are a big problem (turtle eggs for example, are very sensible to those controls and a 2°C fast fluctuate can represent the failure to some sensible species, so they told us).
I don't know two damns about those applications, but the specs we must follow for those stuff force us to design very thermally optimized products, and there is this kinda of losses play a big role.
And batteries aren't an option, because people simply forget about replacing it (or so they tell us when specifying).
Greatscott: This video is sponsored by altium designer
Also Greatscott: *proceeds to use EasyEDA to make schematic*
irony
Spotted !
Ouch :(
Even Altium is amazing, for the most hobbists like us paying for an expensiiiiiive altium license it's not worth it tho
That's how advertising works...
Your videos are the best presentations of electronic concepts. I love the way you draw schematics & add written explanations to accompany your commentary. Many thanks
Quite interesting, I like your new method of less videos but high quality. You seem very happy and relaxed- keep it up man! As always, great video 🤙🏻
Thanks :-)
R u a patreon
How is your comment 12 days old?
Were you able to attend to Stephen Hawking's party?
@@scienceteam9254 magic
@@RMquickbit Did you have a party with Dr. Hawking though?
I've seen mosfets used in the output of some USB chargers as a diode replacement since those 1-3A loads are a significant loss across that component at those voltages. Pretty smart use of them.
So did I. Using fast Shottkys is usual in PSU, getting the efficiency >85%. Active rectifiers lift it up to >90% or better. So, if you get an PSU with >90% efficiency, you can be sure to find active rectifiers for the output.
Saw "sorry electroboom" and "greatscot" so had to click!!! Also would love to hear electroboom response on this 😂😆
This is where the rivalry starts 😂
Indeed, and I propose more! What devilish response would come from ElectroBoom! Phantom Add that into your comment mah dud ! 😂
@@diegoyonamine8943 nahhhh you thought it so gonna let you keep that response!! Also can't wait for electrosboom response he he ^^. I don't usually eat this but umm *gets up and starts making pop corn* this gonna be gooooood!!! ^^
@@PhantomWorksStudios Yes, agreed. Both positive ContentCreators ... making fun into science educational. LOL . I may explain myself tho.. i suggested the edit/ad instead of spamming same other comment... but it's all gud 😂
@@atari_jhn
start the intro....
😂😂😂😂
It's a great idea. i had this electronic rectifier in the early 90's to develop a high power SMPS. The biggest plus is that you can ramp the main Voltage slowly to avoid peaks at ramp-up and control the PFC. For high current very good.. for mainstream electronics too expensive
FBR may also be a better option in adverse thermal environments. You end up with less variation in output characteristics, which makes handling that variation safely much easier. Even with well-designed power supplies, you might be amazed how differently a device might act at -40C vs. at 85C
Wouldn't a MOSFET behave just as badly in extreme temperatures?
@@erenoz2910 That was my main concern. MOSFET operation is highly dependent on parasitic characteristics (mainly capacitances between different nodes). These characteristics are typically relatively susceptible to thermal variation, which can make building an efficient controller which works across temperatures challenging. An FBR (Full bridge rectifier) is going to exhibit some variation in performance over device temperature, but it is comparatively easy to deal with since it isn't going to mess with a drive circuit... More notably, the FBR a single part as compared to the IC, four MOSFETs, and a handful of capacitors. I tend to like simple. It is less likely to break, and easier to troubleshoot if it does break. In some applications it may be possible to use a Schottky FBR, and get almost as good of performance as you would get from active rectification, but the reason that we tend to use traditional diodes for that application is that you don't really get much for the development time needed to implement a different solution... and that isn't taking into account the reality that higher part counts take more space on the PCB and increase the number of potential points of failure... In production that is going to mean a higher fallout rate, which drives up costs and reduces customer satisfaction.
I would generally save the MOSFETs for applications where you need to have the switching (like an H-Bridge). I can imagine a few places where extreme efficiency may be desirable, but just not that many where it really overcomes the tradeoffs.
You are adding a lot of complexity to solve a problem that may not be an issue. But it’s interesting and educational to watch you do something in an unconventional manner.
I think with the capacitors fully charged, and when the MOSFETs in a bridge start to open, you have higher voltage on DC side an lower voltage on AC side. That makes a path for capacitors to discharge. So all the process is actually pumping/draining the capacitors. You still need a diode, to prevent that backflow.
i have been wanting such a rectifier for a long time, actually - it can pump electricity back into the grid, which is sometimes useful (for example if you're speed-controlling a dc motor with an autotransformer followed by such a rectifier, turning the motor speed down will actually cause regenerative braking). A huge inductor before the capacitor should be able to take care of these recirculating currents.
@@victortitov1740 If you connected a synchronous rectifier directly to a DC motor I would expect the losses to be higher due to the massive AC component of the voltage waveform applied to the motor. For your speed control application I would just use regular diodes. Instead of regen braking when you turn down the knob it will just coast.
@@eDoc2020 Why would an active rectifier have a bigger AC component than a diode rectifier in this case?
@@user-td3yi1mq7p because the bumpy DC one gets out of a capacitor-less rectifier will be forced upon the capacitor. That is of course only for the rectifier that opens and closes the mosfets synchronized to (and only to) mains voltage polarity. The rectifiers GreatScott shows are reacting to voltage drop across each "diode" instead, so they behave essentially exactly as normal diodes.
@@eDoc2020 well, that's how it operates right now, and i sometimes wish i had a brake in the thing. Might as well be a regen brake.
When I attended engineering school in the early 1970s we always called it a full wave bridge rectifier. I never heard it called a full bridge rectifier until fairly recently.
It's called 'full bridge' just through ignorance.
I think that you are quite correct, this is because you could make a half-wave Bridge Rectifier - if this is sufficient for your project.
@@peterduxbury927 "this is because you could make a half-wave Bridge Rectifier"
A bridge is always a diamond shape (although it can be portrayed rectangular). Think Wheatstone Bridge for example. A center-tapped transformer can give full wave rectification with two diodes, but each winding is actually only half wave with one winding having current in the positive half cycle and the other winding current in the negative half cycle. That's fairly common but I wonder about the presence of pulsating DC in each half of the secondary.
I use bridge rectifiers a lot for compact polarity protection in low current applications, and also to exploit the 1v drop to avoid the need for regulation in some cases.
Most full-bridge rectifier packages are built with simple silicon diodes with a high voltage drop - that's why when I built my 120VAC input lab power supply I assembled my own rectifier using Schottky diodes (MBR40200PT ASEMI TO-247/3P), which cut the voltage drop in ~half and increased efficiency significantly. Using a larger than needed size diode also helps with efficiency. I still like the simple diode solution because it's 'bulletproof'.
You can use LT4320 as well, it's way more simple and use (pretty much) the same components.
(it's for lower voltages thou)
With one observation ... that IC only starts to work at approximately 9v AC and can only handle up to around 72v AC , so it's not really usable with 110/230v AC and the 9v threshold is a bit of a pain if you want to use it on the secondary side of a transformer... it cuts quite a bit of the sine wave. Last but not least, it's Linear / Analog now, so expensive...
@@mariushmedias LT4320 is a fun but useless toy sadly. The high cost kills what few applications fit within its voltage window. It's also useless as a secondary-side SMPS rectifier due to the 400 Hz max frequency, if memory serves.
Your conclusion is right on! I designed and built my own active rectifier system using comparators and mosfet drivers etc. It worked extremely well, BUT, was only more efficient in a rather narrow range of power than a simple bridge rectifier.
I think, through some sort of protection kept forward, you can try going for phase controlled conduction mode as well, which I believe is industrially done through IGBTs...
the older i get the more and more I like you videos. especially this one feels like a electrical circuit lab in uni. I kinda miss the labs, even though I am not good and I don't like working with circuits. They ended up being a hobby for me.
I think you forgot about reliability. Diodes themselves are typically very reliable and can work for years without a problem. This active rectifier adds additional ICs in the main power path. I doubt it will be more reliable, probably much less. Additionally as you mentioned - this will only make sense if a lot of current is flowing through the rectifier, so it will be a very niche device.
Exactly!!!!! Why to overcomplicate something that works passively and has less weaks points
Size is also an important factor for designers. The active rectifier is much larger than the FBR
@@internetisinteresting7720 Because you might need to squeeze more efficiency out of the device, and don't want to waste power dumping lots of current through diodes?
Your handwriting and drawing is *beautiful* and a real pleasure to watch. :-)
Many Hybrid and Electric vehicles make use of active rectification. The 3-phase bridge has a fair bit of work to do when you consider field oriented / space vector control and then switching to regen on over-run. 🤔
Good video. Active rectifiers have been known since the 1980s, and we used these in some satellite applications where low voltage batteries were being used. The main problem with FETs is that the higher the voltage standoff, the higher the source/drain resistance. FETs now are up to about 1000+volts and IGBTs are over 4000+volts, but in both cases the loss voltage across the device is almost proportional, so still requires roughly the same cooling power, for the given application. We did railroad applications at 2000+volts, 500+amps, where we had to use antifreeze liquid for cooling, in specially milled tubes inside of the heatsinks. Freon and chilled water cooling was also tried, but these has significant physical disadvantages, and had poor heat transfer under unusual or incorrect operating conditions, often resulting in catastrophic failures. Our ten-ton experimental 6-wheel truck would often burst into flames, with such failures. In thermal world, "there is no free lunch". You must cool adequately, for any given power level, period, or burst into flames. Your example is good for a "few" watts of loss, and then it will cook itself, since the heat outflow is very low.
I'm definitely starting to see more active rectifiers in the output stage of USB phone chargers, where they just used a single diode in the past. I guess this is where you would see the greatest efficiency benefit. These active rectifiers kind of remind me of the old fashioned mechanical rectifiers.
PC power supply is a good application too.
They already use a center taped winding to reduce the diode losses but the remaining 4 ones are on a rather big radiator.
Pretty cool! As a former electronics engineer, I enjoyed watching this.
These broke into big inverter drives years back. For three phase high power you can twiddle the active switching in real time to minimize distortion on both sides. Also they let you run the drive backwards during braking. Turning high inertia loads into massive generators. I think ABB called it an ‘active front end’, A-B and Siemens have their own names for the same thing. Hella cool.
This is how HVDC converters work on the power grid! You take 3-phase AC, run it through a wye and a delta transformer to get 12 poles (6 pairs of phase). You then feed them to an active element, triggering at the right point. In the old days, it was mercury valves, then it moved to thyristors, and these days they're often IGBT (Insulated Gate Bipolar Transistors). For HVDC applications, these are huge, dwarfing a human. Voltage differences of up to a million volts kinda needs some room. Modern designs use stacks of them for each pole, switched at different times to match sine waves, minimizing ripple in magnitude and raising it in frequency, so it can be filtered with smaller capacitors + inductors. The 3.4 GW Pacific Intertie uses thyristors now, originally mercury valves when it was placed into service in 1970 (originally at 1.4 GW). The newer Transbay link under San Francisci bay uses IGBTs. HVDC links between England and Europe, Tasmania and South Australia, US/Canada and US/Mexico. In some places, there are back-to-back pairs to isolate different AC grids from each other; Japan has 4 like this, with 60 Hz on one side and 50 Hz on the other. It's the same circuitry on each side, whether used as an inverter or an AC-DC converter. Which direction the power flows comes down to the phase angle of the triggering signals. So solar and wind use them as inverters, while grid batteries use them in both modes.
@@BobKerns4111 oh yeah. HVDC is hell of cool! Crazy phase shift TX’s and massive multi pulse stuff, and the voltages, Jesus wept the voltages. I mean I get cagey working up to 5kvac, but it’s still just AC. And DC had its own quirks and foibles, but, getting up to 100’s of KV? Auuuugh that is a job for someone with more guts (and brains) than me.
Cool shit though. Thx for the info!
I can understand why you were so nervous about causing a short circuit, main can be very scary! Once, when I was 13, I was playing around with a computer power supply, and I accidentally, switched it from 240V to 120V and turned it on. As I am in Australia, our mains is 240V… I think you know what happened next. A huge bang and a flash of light later, I had learned what that switch on a power supply did and never made the same mistake again XD
Yeah, there is a slight problem with capacitive loads, but if you suddently start drawing several amps, you could face the exact same issue (aka accidentally building a bomb).
Also it's hard to have a very wide supported voltage range when making them.
Just like synchronous rectifiers that you can find at the output of very high efficiency switching power supplies, they are stuff you better don't mess with unless you really really know what you are doing.
I know what I am doing enough to create my own reliable power supply, but I'd still not feel comfortable building this kind of things...
A very informative comparison!
Glad you liked it!
the purpose of an active rectifier is to overcome the nonideal rectification behavior of an individual diode. the MOSFET drivers are called ideal diodes when used for rectification.
I know it’s a simple thing, but I’ve always loved your pen markups in the videos.
Can we all take a moment to appreciate the FULL BRIDGE RECTIFIER it's an amazing invention that helps us to make our life we easy.
True
It's so simple yet effective
You are a wonderful teacher. Much of what you did was above my level. However your presentation is delightful. The neatness and colours ...
Thank you so much for your teaching.
David Lixenberg
Interesting. I wasn't aware that NXP picked up my idea from 2008 (See Christian Rausch, "Greener Rectifier Loses The Diodes, Adds Power MOSFETs, Efficiency", EDN, 2008, October Issue) and made an integrated circuit from it.
😁
Such a low drop bridge rectifier is basically essential, as you mentioned, in very low voltage applications where you don't want to lose voltage and/or waste power. BUT in this case you don't need such a complicated circuit. For a 3V or less application I designed a rectifier which basically consists of a flip flop: you apply the ac voltage on the "outputs" of the not gates and get the rectified voltage on the "vcc" and "gnd" terminals of the gates. Each not gate is made of a PMOS and a nmos transistor. At first the circuit starts because of the parasitic bulk source diodes which constitute a traditional bridge, then the gate source of each MOSFET is driven properly by the output voltage.
By the way, what I would really love is an active zero-drop diode (MOSFET with it's own driver) to be used in the secondary of low-voltage-output switching power supply, where the voltage drop of the diode is causing a big loss in efficiency
7:31 The simplest and most likely answer may be to observe that a capacitor initially looks like a dead short. I can't see the size of the caps you're using, but they look big enough that it will look like a short for long enough to blow that fuse. This will be at least partly why the datasheet stipulates the use of a boost converter on the rectifier's output. A naïve boost converter will present a mostly inductive load at its input, so there is no high initial inrush current in quite the same way. Real boost converters will still have some C on their input in order to control EMC, but not as much as you're trying to load the rectifier with.
Beyond that, another possible answer, and this is just a guess, is that these IRF740 FETs are just too slow and chunky for this application. As you noted, if a third FET is still partially conducting, you'll get a low impedance path that might look like enough of a short to blow the fuse. There are two parameters worth looking at that might explain how that situation could arise, even with the TEA2208 controller: gate charge and conduction recovery time.
As you probably know, 1) a FET's gate is a capacitor and 2) the gate voltage that matters is that relative to the FET's source pin (ie Vgs), which is not always ground. Especially for the high side FETs, the controller has to be able to drive a sufficient Vgs relative to its own ground to turn it on†, but that voltage will be different from that relative to the controller's own ground. So, usually, the controller needs a charge pump to get the required Vgs. The current capacity of this charge pump affects how quickly it can turn the FET on and off, so a FET that stores more charge on its gate will take longer to switch. Contrast the specs for the IRF740 with a better power FET like On Semi NTMFS5H600NL.
Recovery time is probably not relevant for a sinusoidal input at these frequencies, but it's worth remembering that a FET doesn't stop conducting immediately after Vgs goes away.
† "on" can mean "in linear/triode region" or "in saturation". These terms are also used with bipolar transistors but, unhelpfully, their meanings are reversed. For FETs used as switches, you want the smallest possible drop across the FET itself, which means operating the FET in its triode region (Vgs > Vt, Vds < Vgs-Vt, noting again that Vds the voltage across the channel, not relative to ground).
The most impressive thing about this video is your penmanship.
Integrating a PFC with the active rectifier might make the increased complexity worth it.
Yep, with 1A at mains input voltage you get 1W or 0.5-1% loss. It's usually good enough.
@@johndododoe1411 I think that's why the HV bulk capacitor after the PFC stage is used. Storing energy on the low voltage secondary side is difficult, and you have to supply power for 2ms of the cycle where there's very little input voltage.
yes the aim is to create a pfc without diodes. Power supply with no active pcf is not acceptable now.
A good place to start in deciding if they are worth the cost is comparing the diode drop to the supply voltage. 1.4V to mains voltage is a very small fraction, so the potential benefit is also quite small. And the active circuitry requires some power itself. Below some current threshold, that would be greater than the loss in the diodes themselves.
If you're planing on further improving power efficiency don't forget consider lowering the resistance of your wiring & the ESR of your power MOSFETs (some power MOSFETs only have a few dozens of µOhms, some of them are pretty new & expensive & involve fancy semiconductor materials like SiC or GaN). Also I've noticed that the quality & ESR of the smoothing caps matters - especially the big electrolytic ones.
I've always made it a point to have both bulk and filter caps (electrolytics and ceramics) in parallel - the bulk caps deals with the higher discharge currents while the ceramics deal with the noise from the rectifier.
That being said, if you can simplify your decoupling and reduce part count with a higher-performance component - it's also worth it.
I also am one of your faithful students that comprehend through electric training, the Power Factor, now very well discussed in electronics. The current leads the voltage which is slower to flow, as a brown out, until the Power Factor is corrected with capacitance to have the current & voltage be supplied to the load. An Inductive load has a Power Factor correction to be made. ♥
Any idea of their longevity? Meaning; will they last as long as or longer than the traditional diode design?
From a purely theoretical standpoint, the fact that there are more components involved with the active circuit means that the MTBF (Mean Time Before Failure) will be shorter with the switching circuit than with the diodes. However, in real life it comes down to the specific design of each circuit and how hot each of the components get - typically a 10 degree C rise means roughly half the service life. Assuming there is enough space and / or airflow to keep everything cool enough, it shouldn't be hard to get the circuit to last for decades of normal use.
For anything attached to AC mains however, a big killer is lightning surges that make it into the house. So the limiting factor is how much surge protection is available before those circuits, and the voltage ratings of the devices attached to the mains.
I highly doubt it. I think they would likely fail much more quickly.
The reason for the part that only controls the low side is likely for use with some kind of semi-bridgeless boost pfc topology. For those circuits you do not need to rectify the upper half since the line and neutral are tied to inductors directly.
This reminds me exactly of when I was learning about H-Bridges in one of my uni classes where instead of acting as a rectifier, the MOSFET gates were controlled by a micro controller at a set frequency and we were effectively inverting a DC source to alternating AC. We also had to worry about ‘dead time’ or essentially, the necessary length of time from the microcontroller where all MOSFETS were closed between cycles so as to not cause a short, great nostalgia.
I thought of the same. A while back I was thinking about building something to convert single-phase to three-phase AC. It used six MOSFETs in the inverter to make three square-waves and an Arduino to control them.
What surprises me is that after decades of electronics and semi conductors, there's still novel ways of rectification being found. I would have thought that by now there's one tried and tested and best solution for every usage scenario established.
A warning for everyone: the configuration at 5:42 is dangerous as it's referenced to the mains. An autotransformer DOES NOT ISOLATE FROM THE MAINS!
It's also a good way to blow up your oscilloscope.
I think his oscilloscope has isolated channels
@@thiagoamericano1412 That does not remove the risk of mains power. An autotransformer, and usually every variac in general just does NOT isolate from the mains.
Totally correct.
I'm glad you rectified this situation.
I've used the mosfet approach to rectify the output from a bicycle hub dynamo to power a GPS tracker. Diodes would be too wasteful - it's only 6V at 3 Watt power max.
I have never used this type of circuit in my designs for power supplies. Purely reactive components on the outputs will be difficult for this system to control. Capacitance across the output without any real restive impedance ( I know redundant) will cause instability in the active circuit. PFC will improve the ability to of the active rectifier to control the system, and this alone would limit me using this in my designs for most of my applications. As some have said in the comments, at higher currents this system would be very helpful in controlling the forward voltage drop power losses in a standard diode. However the costs in switching noise, board space and limited output options will relegate this to a narrow family of applications.
The explanations were very well done. I would suggest however the true name of this system is a Full Wave Bridge rectifier, the other option is a Half Wave rectifier.
I learned that I will continue to use full bridge rectifiers and lead based solder as somethings just work with out added complexity and points of failure. Keeping things simple makes for greater reliability as a rule but not always as efficient as more complex circuit configurations.
I realize I am focusing on the wrong part of the story but, man that is some sweet handwriting.
And most people go their whole lives without using the phrase "Luckily, nothing blew up."
The bridge rectifier chip makes for a more compact board design. I however prefer four diodes which is cheaper and easier to upgrade if needed.
I always marveled at the simplicity of the full-bridge rectifier.
Fascinating. Definitely seems like a better idea for things with huge draw: maybe PC power supplies? (Which often have PFC anyway). I was going to say that DC charging of EVs might want this too, but the high speed DC chargers have big cabinets of electronics so they probably have something even more sophisticated. But, maybe the onboard charger in an EV for connection to Level 1/Level 2 AC service equipment uses these. My relatively new furnace has an "electrically commutated motor" in the blower, which is basically a DC brushless motor. Maybe it could also be a use for these, since the blower is a fairly large draw. (Or inverter type heat pumps...)
0.9% might not sound good but given how many devices could use this across the country and world, that's an unbelievable amount of energy
Outside of special applications or high power devices, these seem unnecessarily complex. More circuitry means more chances for something to fail. However, for high power applications these do seem like a pretty nice way to reduce the need for active heat dissipation on the input side, but depending on the amount of power conditioning after the rectification process or the amount of heat dissipated by the power supply total, the amount of benefit gained might not be significant enough to justify the additional manufacturing cost. Still a pretty cool circuit nonetheless.
Not just high power. Less power loss means less heat generated, means you might avoid a cooling fan. Active rectification is very common with modern switched mode power supplies
@@matthiaswandel Fancy seeing you here! I agree with you. I just meant the cost benefit ratio would be larger in high power devices, lower power devices would also benefit albeit to a lesser degree. If I could remove a cooling fan, and the money saved on that was higher than the increased cost of the circuitry, this component would definitely be a viable option. I was just taught to always to minimize complexity; if you have two solutions that work use the simpler one, because it minimizes failure points and reduces cost to the end user. When I wrote that response, a rectifier costs around two dollars, and the active rectifier cost nearly ten. If the benefit gained *is* significant enough to justify the additional cost, then its worth consideration for sure. The customer is the one paying for the additional cost, so if they are willing to pay more and the outcome is greater reliability, its a win win.
"or high power devices"
High *current.
Nicely done fella. Look forward to see more great videos from you.
Interesting to see how it handles noisy mains supply that isn't a perfectly sine wave.
And variable frequency input like in wind turbines cars or bicycle. .
Multi phase alternators could be a more reliable and simpler solution.
Must say the drawing and writing are a joy to look at. Thanks.
Great video, you didn't mention schottky diodes within a full bridge rectifier? I've not seen them packaged this way as they are with regular bridge rectifiers but it would be interesting to know their efficiency also. keep up the great work!
Is it because schottky diodes have greater reverse voltage leakage making the whole thing short together? I don't know that's why we're here but I looked it up and it's only like 40V almost like a zener so for low voltage maybe. For high voltage, I'm thinking a hard NO. Feel free to correct me if I'm wrong.
@@foxfyre3600 Correct me if I'm wrong, but wrong you are (I think). Basically all SiC power diodes are schottky junctions and they sport some of the highest breakdown voltages available. They are not very popular due to their cost vs. an active rectification solution and negligible efficiency gains attributable to lower forward knee which matters least at high voltages.
@@Ormaaj I said I wasn't sure but the way to know is to test it for yourself
Would have to distinguish a schottky's emission coef. given a (probably) unknown saturation current by measuring based on SiC thermal voltage among other things.
To know for sure I'd just hunt down a proper model with known process parameters and calculate.
That reference to ElectroBoom is priceless XD
00:02 man, that might get you a copyright strike ;)
Full Bridge Rectifier 😂
Don’t worry, the *FULL BRIDGE RECTIFIER* will rectify the situation.
That looks like a good match for things like wind/water power solutions, where you have low AC voltages and even a small increase in efficiency can make a big difference.
Does that make sense?
I would be remiss to not point out that even in the test where the active rectifier was noticeably better the FBR only fell to 96-97% efficiency - not anything worth stressing about.
TEA2208 is £6.90 at Mouser and the LT4320 is a similar price. My use case is a 0.25V power supply for electrolytic purification of Copper. It's not so simple to design a switch mode power supply for that either. Going back in time is a tunnel diode transmitter which requires a similar voltage where dropping it from 1.5V is very inefficient. For these sort of applications one needs to roll ones own circuit.
Electro boom enter the chat
I'm still amazed at how neat your perf board wiring is.
Mehdi's HVDC video's got THE MOTHER OF FUUUUUULEST BRIDGE RECTIFIER.That's got around 12 diodes.Might not be an easy challenge replacing diodes.
And I forgot, you're left handed. Super skills mr.
Very interesting video thank you. Can we conclude from this that the good old bridge rectifier was way ahead of its time.😉
You should do another one on DC-DC power supplies. My company loves to use them. We'll send out 24Vdc via a cable and then use them on a board at the end of the cable to then convert it to 5v and/or 12v. We even had a few systems run on 56Vdc as a mains voltage then converted down.. For me those DC-DC PS are a common issue, we use an IC version soldered on boards.
Thanks!
Thanks for the support :-)
I think the most interesting thing is the MOSFET driver ICs (TEA2206 & TEA2208) while being active, meaning they require a power supply, they use the AC mains by charging a capacitor on the initial cycle. The data sheet says the body diode in the MOSFET is used initially in the 2208 as a traditional rectifier to charge the capacitor and power the IC. It would be neat to look closer at exactly how this action happens. Nice video!
learned this new thing, thanks to GREAT 👍👍SCOTT !!
Very nicely explained and drawn. You explain things so smoothly and calmly. Half the time I was expecting you to get shocked, burnt, bang your head, etc. Then I realized that was because every time Mehdi connected something, bad things happen. Keep up the good work.😊
Having seen the hand writing, I am not at all surprised to see the ultra neat work area!
The most fun I had with a bridge rectifier was modifying my Mom's "Light and Motion" Christmas tree ornaments. These had any number of motorized operations to bring the ornament to life, but if you had a number of them on the tree, they got noisy. So, I put a pair of diodes on each ornament so that one polarity went through the ornament, and the opposite bypassed the ornament. Then I ran all the power going to the tree through a full wave rectifier, followed by a double pole switch that allowed me to reverse the output polarity. Viola, in the middle switch position, the tree was unlit. In another, it lit up and in the third, it lit up and the motorized ornaments came to life. Mom thought it was magic.
Hello! The main disadvantage of a diode rectifirer - generation of harmonics in an AC main network.For example - when the average current of a load is 1A , the peack current of a recifier may be greater that 5A. When we want to solve this problem - we need a PFC. Unfortunately, there isn`t any PFC on Chinese Ali Express.
@GreatScott - if I'm not mistaken, that MOSFET network with H-design & switching controller is nearly identical to the design of a pure-sine DC/AC inverter, many of which use the EGS002 controller board that's everywhere on the internet these days, and built into many inverter products on the market.
Full Bridge Rectifier is like the humble Wheel...
always improved on but never obsolete
OK, now I have to do another project. You can just make a little bias circuit for each one to turn it on when it goes forward bias, right? With zeners to protect the gate from over voltage either way. I am super busy right now with electro-chemistry, but I will try one very soon.
Your videos are like a masterclass with a lot of knowledge!!
I think during the capacitive load test the capacitor bank was unable to discharge properly, it kept on charging, adding a bleeder resistance to it should resolve that, alongside, also utilizing the body diode during some period of the supply voltage. This means that the mosfets should be triggered at a certain value of alpha, which can be adjusted by the time constant of the load.
Definitely do a video on boost pfc. I never learned it in school and yet it seems to be everywhere in industry.
In a telecom system, the input voltage is -48V DC (nominal). There are two power busses for multiple power feeds for reliability. That means power can be drawn from either. MOSFETs were used to make that selection of power busses. All was good until someone connected one of the power busses backwards. The boards which were already powered up were fine, but any new board installed would blow the fuses in the board, because multiple MOSFET were enabled at one time before the power controller could pick the right one. Moral to the story is that the MOSFET controller may need to be powered by a full wave rectifier before they can be turned on, but there may still be a race condition like we had.
7:33 I wonder, if you had of added a regular full bridge rectifier in between the capacitors, would of that prevented it from a short circuit..?
I was afraid someone was tryna start some beef with ElectroBOOM but then I realized it was Greatscott!!