Transistors 101 ep6: Biasing and Bootstrapping

[@brjplummer9415]

I pulled myself up by my shoe laces and Subscribed. From your video I discoverd that some of my brain cells and memory cells are still alive. Thank you. Brian

[@ognjencetkovic7719]

My mind got blown by the bootstraping technique. Such a good explanation

[@KludgesFromKevinsCave]

You'll see more about it moving forward. It tends to come up at odd times. I plan at some point to do a few videos about an old-school linear speaker amp (not the modern PWM stuff) and they almost always bootstrap the collector resistor on the driver transistor.

[@ognjencetkovic7719]

I am looking forward to see those videos ​@@KludgesFromKevinsCave

[@jdos2]

@@KludgesFromKevinsCave I've been searching KZbin for years for a simple process-based explanation on how to do overall circuit design - and the bootstrap technique you demonstrate (with additional experience shared) is exactly what I'm looking for. Thank you for the clarity and detailed explanation!

[@aayushhegde6234]

Grateful to you for sharing your knowledge openly.

[@KludgesFromKevinsCave]

Glad it was helpful!

[@mineown1861]

From bootstraps down to equivalents , thank you for a compact and bijou dive into the emitter follower .

[@KludgesFromKevinsCave]

My pleasure!

[@breedj1]

Nice video. I like the bootstrapping. Choosing the corner frequency of both the input and output filter the same, 2Hz here, will create a second order system. It is better to make one of these corner frequencies 10x lower to end up with two first order systems with poles well apart.

[@KludgesFromKevinsCave]

Right. In a superabundance of caution, I put *both* corner frequencies 10× below the 20 Hz-20 KHz range I wanted the amp to pass. Separating the poles would make the passband flatter, but I just made it excessively wide and used the flat part. I *do* plan to get into active filters as the channel progresses. Just trying to keep things brutally simple for now!

[@KludgesFromKevinsCave]

@danstiurca7963 The question made sense. The order refers to the number of pole pairs, and has nothing to do with feedback. What @breedj1 is suggesting makes sense if you're tuning filter response. It's just not necessary if you have the luxury of putting the corner frequency far away from the intended passband. You don't necessarily have to work out filter response in detail for every DC blocking capacitor that you use.

[@Robert.Chauval]

Thank you so much for this revisit to a long forgotten friend.... I havent had to get back to basics since 1980s and ive forgotten so much. Opamps made me lazy ... lol I was looking for exactly this follower technical discussion as i was literally planning to breadboard an emitter follower this week using BC549C as i plan to upgrade a fully discrete preamp from the early 1970s that has a very complex high impedance LCR output stage thats unbuffered .. explains a lot about the wildly varying listening reviews.. Your explanations and visual aid processes are right on the money... very well done.. You just got a new sub and i plan to use your series as reference material as ive been mentoring a young fella in the dark arts of audio amplifier design for the last 8 yrs and well discuss your work in detail. Thank you.. i look fwd to the next installment from you and applaud your generosity .. back to school for this ol fella.. 😊😊

[@KludgesFromKevinsCave]

Glad you're finding it useful! There's more in the works. I have several current mirrors (using DMMT5401/DMMT5551 matched pairs) on the breadboard right now - look for that in about 8 weeks (I always like to have that much in the queue.) Looking much farther ahead, I'm thinking of several creative ways to abuse analog IC's that require you to understand the schematic inside, which is why I'm starting from the basics. A lot of the creative things that people do with the humble 555 or the LM13700 are NOT in the data sheets!

[@Robert.Chauval]

@@KludgesFromKevinsCave awesome.. your efforts and especially your very watchable presentation skills will become a valuable resource for many old and new electronics enthusiasts.. I recently had this very discussion about understanding or at least acknowledging the internal schematics of an ic and not "black boxing" them. This came from questions of how a low value electrolytic filter cap on a dolby chip would change the operating points of its bjts due to non ideal dc current leakage - that shouldnt be there - increasing distortion and filter shaping issues etc... replacing these low value electros with film is the only sensible solution. Im looking forward to your next issue/s. Creative abuse of analog ics??. I cant wait LOL fantastic.. Thanks again for your efforts..

[@shawncalderon4950]

I'm new to electronics. I'm so glad I found this channel

[@KludgesFromKevinsCave]

Glad you're finding it helpful!

[@adammontgomery7980]

One of the harder concepts for me, especially when I started. I always end up chasing my tail

[@KludgesFromKevinsCave]

The real trick is to analyze the quiescent point first, and then analyze the signal path. It gets a ton easier once you realize that they really can be separated. You can often figure out the quiescent point just with "the emitter is a diode drop away from the base."

[@shawncalderon4950]

@@KludgesFromKevinsCave I'm saving these knowledge nuggets for they will become helpful, even necessary, to further my experience and understanding of electrical engineering. Thank you, sir!

[@ismailfaruqi9570]

Thank you so much sir, i'm having quite a trouble on understanding where did all those resistors and capacitors values come from when designing any circuit using transistor. You explained them really well. God bless you❤

[@KludgesFromKevinsCave]

Glad it helped!

[@Stelios.Posantzis]

Simply excellent tutorial!

[@M1les_Grey]

Always been trying to find a good explanation of what bootstrapping is and how it works, this was the only video that really explained it in detail!

[@KludgesFromKevinsCave]

@M1les_Grey Glad it helped! This is by far the most popular of my videos so far, and maybe that's why!

[@hardrocklobsterroll395]

Saw the intro and subbed already. You know I gotta hear from an EE graybeard. Thanks for the great info can’t wait to come back!!

[@KludgesFromKevinsCave]

Awesome! Thank you!

[@TheGorf]

Great video series. I did my EE back in the 90's and tend to pop back into videos like this to refresh my memory. I wasn't a big BJT fan. I specialized primarily in transmission theory and high energy switching and did a lot of work in the heart defibrillator world which was more about finding IGBT's big enough and fast enough. haha.

[@KludgesFromKevinsCave]

Glad you enjoyed it! I worked at GE Research for over thirty years, so I was around when the enormous IGBT's - the ones used in DC-DC converters for high voltage transmission lines - were being developed. (Wasn't my thing, I'm an instrumentation guy primarily.)

[@joelstyer5792]

Twenty plus years ago I designed a lot of video amplifiers with transistors, until high speed op amps and very low supply voltages came along. It was fun to look back on the transistor design. I must say this was an excellent, well thought out and executed tutorial. Discrete transistor circuit design is becoming a lost art, much like tube circuit design. Sadly, we no longer have the variety of choices in transistor selection either. My one curiosity, not really relevant to the video was which transistor you were using. My video world used to revolve around the 2N4124 (NPN) and 2N4126 (PNP). Thank you for the great video!

[@KludgesFromKevinsCave]

I happened to have picked a 2N4401 out of my parts box. It falls under my mental category of 'TUN', 'transistor, universal, NPN', - PN2222, 2N3904, and BC547 are the other ones that I tend to group under the heading of, "I need a transistor here, and don't care much about the characteristics'. All of those have >25 volt breakdown, >=100 mA collector current, beta>=100, total power >=0.1 W, GBP >= 100 MHz, which is "if I don't need a lot of speed or a lot of power, this will do". I like 2N4401/2N4403 for audio work because they're quiet. By the way, all of my mental category of 'universal transistors' have PNP complements: 2N4403, PN2907, 2N3906, BC557. The whole lot of them also have variants in SOT-23 packages (with the MMBT prefix instead of 2N, or BC847/857 instead of 547/557). Most of them also have matched-pair variants with the DMMT prefix. 2N5551/2N5401 sort of fall in that category, but with the advantage of handling much higher voltages. There might be a lot fewer varieties of transistor available than there used to be, but the ones left standing are good enough that I can generally find one that meets my needs. Like everyone else, I use opamps when appropriate. But being a cave man, I like to see how far I can get with jellybean parts.

[@joelstyer5792]

@@KludgesFromKevinsCave - I totally agree with your logic and like the TUN acronym. I had wanted to mention that perhaps my favorite small signal bipolar is the 2N5089 (or MMBT5089) NPN and its somewhat similar PNP, the 2N5087. The '89 has a beta of about 1200 (400 min) which can be very handy at times. Still in production too. For higher voltages, I lean toward the MPS series and other Motorola originated parts usually. Recently, I've done a few projects that needed higher voltages and the choices thin out greatly compared to the past. I'll have to look at the 2N5551/2N5401 again, been decades since I researched or used those. The 2N4124/2N4126 are also low noise designed for audio and video. I've rarely seen them used outside of professional video equipment, however. I have a lot of them as well as the MMBT equivalents, so I just used those when needed. Same with the 3904/3906/5089 devices. In the 1970s, there were a lot of transistor project books and the parts that were spec'd most often were the 2N5501 and the 2N4401/2N4403 pairs (along with the classic 2N3055).

[@KludgesFromKevinsCave]

Yeah. The next couple of episodes are about chasing gain, and the 2N5089 (5087) are mentioned specifically in episode 8 (due to drop October 18, I try to keep the queue full at least 6-8 weeks out). It's my go-to for "cheap superbeta". I don't list it under 'universal' because all that gain comes at the cost of Miller capacitance, noise, and instability. All of those can be tamed, of course, but the '5087/'5089 aren't 'the transistor I'll grab from the parts box for something quick and dirty.' I looked at a couple of my old project books, and they seemed to focus on the '2222/'2907 or '3904/'3906. But a lot of those project books were sponsored by the manufacturers, and maybe yours spec'ed different jellybeans from mine. I also turned up an ancient project book with the 2N107 and 2N170 - which were 'germanium transistors that failed inspection, but the hobbyists will put up with them.' The first transistor radio I ever built was a superregenerative with a 2N170 as the RF stage, a pair of Ge diodes as the detector, and a 2N107 as the headphone amp. It didn't work all that well, but the amazing thing was that it worked at all. I was a little kid, and proper lead dress was a mystery to me. When it started to oscillate, my grandmother could hear the howl in her hearing aid.

[@jmaguilarr]

es la mejor explicación de un bjt como seguidor que he visto , excelente

[@KludgesFromKevinsCave]

@@jmaguilarr ¡Gracias! Me alegro de que te haya resultado útil.

[@louco2]

This was amazing! Tank you so much for taking the time, also love your cave very much 🙂

[@KludgesFromKevinsCave]

Glad you enjoyed it! The cave is a bit of a mess, but my wife and I tend to be pack rats.

[@copernicofelinis]

Finally, someone who places the input coupling capacitor with the correct polarity! Have you noticed they are reversed in most introductory books?

[@KludgesFromKevinsCave]

I hadn't, but I've seen that sort of rookie mistake in a lot of KZbin videos, making me say to myself, "has this person actually _built_ the thing? Of course, the correct polarity depends on the biasing of the previous stage. You haven't lived until you've set off the smoke alarm with an exploding capacitor!

[@copernicofelinis]

@@KludgesFromKevinsCave Malvino's is one such book. But even Millman & Halkias have reversed biased input caps (I had to verify with a simulation, I mean... Millman!) I wonder if these reversals were the actual author's fault, or were introduced by the artist who created the circuits for publication.

[@GodwinOAkoma]

A very big thanks for this wonderful explanation about transistors biasing 👍

[@KludgesFromKevinsCave]

Thanks for watching! I enjoy teaching,

[@ezion]

Your videos are great. Thank you for your knowledge and efforts.

[@KludgesFromKevinsCave]

@@ezion glad you like them! I enjoy teaching, so this is fun for me.

[@ivocapella4541]

Excelente explicação. Vou assistir de novo, analisar com cuidado, testar no simulador e montar para verificar os resultados.

[@KludgesFromKevinsCave]

Obrigado!

[@uricohen5463]

I understand that your videos only for engineers , so sad , because the videos are amazing . Thank you sir

[@KludgesFromKevinsCave]

@@uricohen5463 If you are understanding them, the videos are for you! They're aimed at tinkerers and students, too!

[@sergius4691]

I stumbled upon a gem of a KZbin channel.

[@KludgesFromKevinsCave]

@sergius4691 Thanks! Glad you like it!

[@xaviergonzalez5828]

New subscriber! This video is gold for me! Thanks!

[@KludgesFromKevinsCave]

Welcome! Glad you found it helpful!

[@Ek1Tyr]

Excellent explanation and simulation! Subscribed Thank you also for sharing the knowledge

[@KludgesFromKevinsCave]

And the simulation is verified on the breadboard! The simulated amplifier is actually built and measured. Glad it helped you!

[@eduardocastroisaac1630]

Thanks Kevin 🎉

[@KludgesFromKevinsCave]

My pleasure!

[@AliMonzet]

Thank you When do we use electrolytic cap and when not ?

[@KludgesFromKevinsCave]

The big advantage of electrolytics is that they have a lot of capacitance. If you need tens to 10000's of microfarads, they're by far the cheapest alternative. But the capacitance isn't all that predictable. The 'good' ones often vary by 20%, the 'cheap' ones are often +80/-20%. When you're using one for an audio coupling capacitor, or a power supply filter capacitor, you don't care as long as they're big enough. They're completely inappropriate for anything where you're using an RC or LC time constant to establish a time or frequency, because they don´t have a precise value. For the timing applications you need either polymer dielectric (polypropylene or perhaps polyester) or C0G ceramic. Electrolytics also are resistive and inductive - which is to say, they're slow. If you're using them to filter a power rail, you almost certainly need smaller ceramic capacitors distributed around the board, and it isn't uncommon to see a rail with a 1000 µF aluminum electrolytic, a 10 µf tantalum, and multiple 0.1 µF ceramic and 470 pF mica capacitors, filtering out transients at different frequencies.

[@marijntopgear]

Thank you for the explanation, very well done! You’ve got a new subscriber.

[@KludgesFromKevinsCave]

Glad it helped!

[@AK-vx4dy]

First class (educational) material

[@KludgesFromKevinsCave]

@@AK-vx4dy glad you're finding it helpful!

[@AK-vx4dy]

@@KludgesFromKevinsCave Yes, although I graduated from technical school in electronics (but mor digital profile), your educational skills and clear practical presentations are outstanding.

[@KludgesFromKevinsCave]

I left the academic life for industry thirty-odd years ago. Now that I'm semi-retired, I've begun to miss teaching. KZbin is a _big_ classroom!

[@AK-vx4dy]

@@KludgesFromKevinsCave Good for us, I love teachers with industry experience, my favourite in school was like that.

[@sahhaf1234]

Thank you very much for this explanation of bootstrapping. Could you also explain the audio amplifier version of bootstrapping? I keep on seeing it on push-pull output stages but cannot understand how it works..

[@KludgesFromKevinsCave]

I have an ever-growing list of 'ideas for future videos.' Your question just got added to it. ;-) The short version: It works just the same as what I showed here. The push-pull stage is a pair of emitter followers (you can kind of ignore the ballast resistors between the two emitters, and the feedback resistor). The resistor that the bootstrap wants to make disappear is the collector resistor of the driver stage. It gets split with the output fed back into the middle of it by a bootstrap capacitor. (Hard to explain without a chalkboard...)

[@sahhaf1234]

@@KludgesFromKevinsCave Hmmm. Let me think about this.. Thanks a lot for the explanation...

[@wizardofboz76]

This was great, thank you!

[@KludgesFromKevinsCave]

Glad it was helpful!

[@Starlite4321]

Thanks for the great video. But I'm totally lost at one point. Could you please expand on this a little ? At about 14:30 discussing the bootstrap capacitor, you say, "But look at the situation at AC. The base and the emitter of the transistor are one diode drop apart. There's no AC voltage between them. And we'll choose the new capacitor so it will look like a short circuit at signal frequencies. ... The resistor R3 therefore has both ends at the same AC voltage." I can't see why R3 has no voltage and isn't conducting current (AC and DC) via R2, and I definitely can't see that "the base and emitter are a diode drop apart" (presumably via the transistor itself, what else could it be ?). That's the same as saying the transistor isn't doing anything, right ? Then where's the amplification from source to load coming from ? Thanks, Russ

[@KludgesFromKevinsCave]

An emitter follower has no voltage gain - the voltage at the emitter is a constant diode drop below the base. But there's more to life than voltage. Consider a source that looks like an AC voltage running through a 10k resistor. Now I want to drive a 100 ohm load. The voltage at the load is going to be a little less than 1% of the voltage at the source. But if I put an emitter follower in between, the output impedance of the follower will be reduced by a factor of the transistor's beta. If beta=100, the output impedance will be 100 ohms, and the output voltage will now be half the source voltage. If I use a Darlington, or a second follower, I'll get another factor of beta, and the output impedance will be just 1 ohm (I'm oversimplifying here, because the intrinsic resistance of the emitter comes into play here), and we'll lose only 1% of the driving voltage. So R2, in our simple model, does indeed have no voltage across it, but the transistor is delivering power to the load that the source cannot deliver.

[@Starlite4321]

@@KludgesFromKevinsCave Thank you for the reply but now I'm TOTALLY lost. You say, "But if I put an emitter follower in between, the output impedance of the follower will be reduced by a factor of the transistor's beta " - huh ? I can't quite make sense of that sentence. Do you mean the load resistance the source sees is increased by a factor of the follower transistor's beta (plus one) ? Sorry, totally lost.

[@KludgesFromKevinsCave]

@Starlite4321 the impedance of the source, divided by the transistor's beta, will be the output impedance of the follower. The follower is an impedance transformer. It presents a much higher impedance to the source than the naked load would, and a much lower impedance to the load than the naked source would. There's a longer explanation in episode 4.

[@profdc9501]

Does bootstrapping sometimes reduce the stability of the amplifier by providing positive feedback? Bootstrapping is used sometimes when the fT of a transistor is barely adequate, but it has to be weighed against the potential reduced stability.

[@KludgesFromKevinsCave]

Bootstrapping in this configuration can indeed reduce stability, but it's done pretty routinely in class-B audio amps. In this configuration with cheap transistors, you have plenty of phase margin. I don't think I've ever had a bootstrapped power amp start oscillating on me.

[@cobar5342]

Thank you for a very good video

[@KludgesFromKevinsCave]

Glad you liked it!

[@axeljespersen8307]

Very good video!

[@KludgesFromKevinsCave]

Glad you liked it!

[@freedomisfood6966]

Hi Kevin. When creat pwm and attach even to the an mosfet without drain to source electricity in pwm creat spikes ????I know it is for capacitor capacity in mosfet .but how remove it.

[@KludgesFromKevinsCave]

Hmm, you're probably asking the wrong guy, since I do mostly instrumentation rather than power electronics. (Or at least I'm guessing that's what you're asking about.) But I can remember a few points to consider. The spike comes from the switching transient. When you change the state of a big MOSFET, you're driving a lot of gate capacitance - perhaps 1 nF or so. Actually, for design it's better to think of it in terms of charge - perhaps the gate, charged to 10 V is carrying 10 nC. You're trying to dump all that charge, or charge it back up again, in a few nanoseconds, so you'll need several amperes of current for that brief pulse. Any stray inductance in the circuit is going to cause a huge voltage spike when that current turns off. Some suggestions: Use a commercial, packaged gate driver like a TC4420/TC4429 rather than trying to make your own pulse amplifier. Make sure you bypass its power pins with a good, low-ESL capacitor. (An X7R capacitor of 100 nF or so). Use wide PC traces, or copper pours, and keep the current paths short. Make sure the ground plane is continuous. If you're using a bridge configuration, you need to protect the output of the gate driver IC with a reverse-biased Schottky diode (like a 1N5819) to ground. Otherwise, the driver can latch up because source inductance can pull the output below ground. The diode allows the reverse current to freewheel. If you don't need the fastest possible switching time, experiment with a resistor between driver and gate. Somewhere between 5 and 100 ohms is usually the sweet spot. Bigger R is slower turn-on and turn-off. Often it's just the turn-on that you want to slow down, so a 1N5819 across the resistor can effectively take it out of the circuit on the turn-off side. (Or you can use something like a TC4421/TC4422, which has separate pull-up and pull-down outputs that can use resistors of different values.) There's more good information, including a table of part numbers and specs for gate driver IC's, in Chapter 3x of _The Art of Electronics: the X chapters_ by Horowitz and Hill. (All three volumes of _The Art of Electronics_ are packed with good information!) Or just buy your power supply off the shelf. For a lot of stuff, I can get a supply from an outfit like MeanWell for cheap enough that it's not worth the effort to design one.

[@leandrocarg]

Very nice!!

[@KludgesFromKevinsCave]

Thanks!

[@biswajit681]

Wow!! Subscribed

[@AnalogDude_]

Very good explication.

[@KludgesFromKevinsCave]

Glad it was helpful!

[@gaborungvari784]

This is world class presentation! Thank you Sir! You have +1 subcriber :)

[@KludgesFromKevinsCave]

Thanks and welcome! Glad you found it helpful!

[@theoryandapplication7197]

thank you sir

[@KludgesFromKevinsCave]

Glad you find it helpful! I enjoy teaching.

[@keylanoslokj1806]

15:00 can you explain to my dummy brain why the capacitor in the middle will have equal current in both ends?

[@KludgesFromKevinsCave]

Any two-legged component has equal current at both ends. That's Kirchhoff's current law at work. But I think you meant to ask, 'why does it have equal _voltage_ at both ends?' Well, it doesn't. It has a DC voltage across it. But that DC voltage is set by the biasing and doesn't change when you apply a signal. The trick here is that we've split the analysis. We've already worked out all the biasing, which is the DC voltages and currents that will be present when no signal is applied. Now we're considering only the small-signal voltages and currents. The terms, 'signal voltage' (or 'AC voltage') and 'signal current' refer to the difference between the bias point and what you see with the signal applied. We're just subtracting out the constant bias voltages from every node of the circuit and looking at these differences. Applying a small signal won't affect the 1 diode drop between base and emitter of the transistor - it's going to be the same value (typically 600--700 mV) whatever small signal it sees, so there's no signal voltage difference between those terminals, only DC voltage. At signal frequencies, we can look at the base-emitter junction as a short circuit. The _signal_ voltage at the base will equal the _signal_ voltage at the emitter. And we've chosen the capacitor to be so big that it looks like a short circuit at signal frequencies. There's no signal voltage across it, only the DC difference between the voltage on the bias divider and the voltage at the emitter. Whatever signal voltage there is at the emitter will appear at both ends of the capacitor. So at signal frequencies, the transistor base, the transistor emitter, and the R1-R2-R3 junction all are at the same potential - there's only a DC difference among any of them. R3 carries only DC current, no signal current, because there's no signal voltage across it. Since it doesn't any carry any signal current, we can treat it as an open circuit for the purposes of small signal analysis. Which means that we've successfully isolated the bias divider from the base circuit at signal frequencies.

[@keylanoslokj1806]

@@KludgesFromKevinsCave your channel, presentation, teaching and character are a blessing. Subscribed!

[@EEverse-po4vu]

@keylanoslokj1806 Current-in equals current-out that is Kirchhoff’s current law.

[@keylanoslokj1806]

@@EEverse-po4vu no,he understood the question

[@HipocratesAG]

🙏

[@ahmedalami-f3c]

❤

[@JimmyDanieley]

Ty

[@KludgesFromKevinsCave]

My pleasure.

[@ahmedalami-f3c]

good

[@javiermac381]

If all the resistors are adjusted, Can Vcc be at 3volts?

[@KludgesFromKevinsCave]

There's no reason it can't be. I think I've given you enough examples that you can run the numbers and try it yourself. (Look at it in a simulator, and then try it at the bench. Generic BJT's are cheap.) Of course, you'll be really limited with what signal levels you can accomodate - life is full of tradeoffs, and you'll be spending at least about a third of your supply voltage on the base-emitter drop and the pull-down resistor. I don't know your application, but A CMOS op-amp might be a better choice at that low a voltage - there are ones out there that can run off quite tiny supplies indeed. I suspect that if you're asking about a 3V supply, you''ll also be concerned about battery life. A micropower op-amp like the LMC6442A might fit your needs. Its supply span is as low as 2.2V, its inputs and outputs can swing very close to the supply rails, and it draws only 2 µA of quiescent current. (TLV2401 is even less power-hungry - 900 nA of quiescent current - and has "beyond the rails" operation on the positive side.)

[@javiermac381]

Thank you! That's a very nice reply! Will be taking a good look at the two Op-Amp, you suggested! ❤️💐💐💐💐😬

[@1234Tubie]

Biasing...'sipping'...😂

[@KludgesFromKevinsCave]

@@1234Tubie mmmmm... coffee!

[@SatExam-gl9eq]

Grateful to you for sharing your knowledge openly.

[@KludgesFromKevinsCave]

Glad it was helpful!