Darlington output stages

A Darlington configuration is a little used technique in amplifier design. What is it and how does it work?

[@SLCVideoProductions]

Took me back about 40 years to my EE101 class. Enjoyed your explanation and how you were modifying your drawings to make them more correct as you went along- like the true engineer that you are!

[@marianneoelund2940]

Paul: Darlington output stages ARE emitter-follower, with voltage gain near unity. They DO NOT convert "little wiggles" into "big wiggles" unless you are talking about current rather than voltage.

[@LordVictorHalgaard]

Would absolutely *love* if you did something like a "crash course on amplifier circuits" - and don't be afraid do go on for longer on our behalf! I would happily spend the time watching, to get a basic understanding of how to look at the insides of an amplifier, and actually understand what I'm looking at - even if just at a rudimentary level!

[@bernhardmichaelfux308]

OMG! This is the Kind of Video I missed so much from you! Technical information instead of: " A Darlington Transistor is a Transistor that uses a Darlington Architecture to work.... " This reminds me of the Servo-function of the Woofers RS5... ;-)

[@marcusm5127]

Nice, love seeing thease more technical videos.

[@jcolonna12]

I love it when you draw the ckts, it helps me understand what’s going on

[@edholmwood]

If I remember correctly the freq. response on that HK amp was about DC to light. 1Hz to 150KHz. I think those old HKs sound pretty good. I have a couple.

[@marcusm5127]

Haha that is the best thing I have heard in a long time. I think I will steal it.

[@PanAmStyle]

I have a 660. It’s great except fo some of the issues. Granted it’s pretty old, so I’m not holding any grudge.

[@wishusknight3009]

Darlington designs are generally very fussy and fickle, but when done right are fantastic. And can be *very* stable.

[@PanAmStyle]

I have the Harman Kardon PM660, from the same era. I wonder if it is also a Darlington design. It’s a good sounding amp but I have continuing problems with the front end, as one channel drops out unless pure direct is engaged. It’s in my workshop setup, and I’ve substituted a different vintage amp for the time being.

[@morganrowe]

Love the electronics side of this game! Thanks for sharing your knowledge ✌️💜

[@Grassy_Gnoll]

"Hey Google!" Videos stops. Good job, Paul!

[@QoraxAudio]

😂 Never turn your google device on when you don't need to use it!

[@chadbarker4023]

Cool. Learn something new everyday! Thanks Paul

[@technoxtreme178]

LOL, when I first saw your title I thought you were now experimenting with fission power for amps ... "Darlington" is also the name of our nuclear power plant here in Ontario!

[@NeilDSouza7]

What a 'Darling' Paul is !!! Tip : Go to settings gear in KZbin switch playback speed to 2 .... Enjoy Paul at 2x !!!

[@garydurler5603]

I had 2 Soundstream car amps in the early 1980's and they had Darlington outputs. They were fabulous. Supposed to have been designed by Nelson Pass,

[@wishusknight3009]

I still have my Rubi4 from then as well. Still use it after it sat for a number of years, had it restored with new caps recently. Best damn sounding car amp I have ever heard.

[@Hondaguru1122]

The best audio professor ;). Love when Paul hits the drawing board

[@marianneoelund2940]

The worst audio professor, actually. Too many detail errors and mis-statements.

[@Hondaguru1122]

@@marianneoelund2940 so why do you watch? Start your own business and make a youtube aeries so you can teach us all then ;)

[@marianneoelund2940]

@@Hondaguru1122 Someone needs to keep an eye on Paul's shenanigans. There are already plenty of good electronics series on KZbin. Check out *Mr Carlson's Lab* or for discussions and demonstrations about simple solid-state amps, look at *JohnAudioTech* or on more advanced topics, *EEVblog*

[@Hondaguru1122]

@@marianneoelund2940 like I said, please make your own KZbin channel let’s see what you can bring to the table ;) ! Good night 😴

[@sebastianbachert9528]

Thank you for answering my question! Stay safe.

[@havbeen18]

Hi Paul great to see you at the white board so much easier to explain. You missed the point that the Beta, as you call it is all current gain those little wiggles don't get any bigger. The other plus is increased input impedance meaning the drive circuit can be made simpler. Many amps around the age of the HK used darlington output transistors which being no longer made are difficult to find. Treat it kindly.

[@joaomantunes]

I missed these white board technical explanations! :)

[@woopygoman]

"No free lunch" is a pretty good life rule not just in engineering.

[@brandonburr4900]

Paul. A white board video! Thank you! Please do more like these videos! Perhaps darin could do one explaining some stuff to like local and global feedback in circuits. Thanks!

[@akirchner3]

Great video. One nit: in referring to the total gain of the compound transistor, you said something like "a gain of 100 plus a gain of 100 give a lot of gain." This might be interpreted as saying that the total gain is the sum of the two gains when in fact it's the product (even higher). I know you know this but your audience might get confused.

[@tomterrific9459]

To be precise, overall gain is not actually the product of the two gains. The overall gain of a two transistor Darlington transistor connection is the product of the two gains *_PLUS_* the sum of the two gains. Two transistors with Beta of 100 and 40 connected Darlington would have an overall current gain of 4140, not 4000.

[@Joshualbm]

That HK 680 is a brilliant design with incredible specs for its price range. It'll drive almost any loudspeaker with a superior amount of juice while maintaining very respectable levels of distortion in the range of 0.08%. at full 8O Ohm output. It's a discrete design with quality parts, dual monaural layout and all sorts of nifty circuits. The transient peak potential is pretty astonishing too. As many speaker crossover loads can vary widely from specs while being drive hard, the HK680 can really keep up. In fact test bursts of up to 680 watts into 1 Ohm were recorded in independent testing. Maybe that's how they came up with the model number. So rather than looking into the reason HK might have used this Darlington design, Paul dismisses it universally as not good sounding, and according to his preferences, undesirable. So the guy who wrote in now has that bug in his ear. All this for equipment which produce nearly inaudible levels of distortion for most listening levels, at a reasonable price. It's time to bring double-blind testing into the arena.

[@sebastianbachert9528]

Thanks Joshua, I'm very pleased with my HK680. I've bought it 7 years ago. I'm running Sansui LM220-Speakers. The amp has been checked by a professional, some caps had been changed. I see no reason to switch to another amp. In combination with my Harman/Kardon HD760 CD-transport I obtain great soundstage and a very pleasing bass. Got the amp for 500€ and the transport for 100€. A real bargain for the given price/sound.

[@QoraxAudio]

I like these whiteboard sessions 👍

[@suprgx472]

Hey Paul, Off topic of this video but was looking over your recommend music . All great but was wondering if you would listen to Seal - Heavenly and talk about the bass on that track(as I know you like 2 subs and your system is Crazy). OMG!!

[@indopleaser]

always wanted and explanation of triple darlington output. was familiar hearing the term in 90’s car audio. that leads to to a question. can you do a video on anything Car Audio related? really interested in box design and maybe ways the oem companies use analysis software. do they use cfd like on a race car aerodynamics simulation. wouldnt they be similar? i like f1. max baltimore

[@davidtomsett]

The PMC amp uses a Darlington topography.

[@ARGBlackCloud]

Macintosh MC 250 uses Darlington output's and they sound pretty damn good !!

[@SJMessinwithBoats]

There’s been a ton of Darlington power packs put in stereos. My Marantz has them for the rear channels.

[@zulumax1]

Pioneer SX-780 receiver uses Darlington power pack STK-0050. Just transistors in an epoxy case, cheaper to mass produce on a production line. Just marketing trying to make topography a sales gimmick, and cut costs too.

[@BIGRIP87]

i love your paying lunch part

[@tertia0011]

Darlington configuration BJ transistors (often in single package) used in phase splitting quasi-complementary push-pull transformerless class B power amplifier is more efficient than class A amplifier. Darlington transistor configuration is like single transistor with very large current gain (current gain is product of the current gains of each transistor).

[@tertia0011]

@G Guest Not according to 'Electronic Devices & Circuit Theory' by Boylestad & Nashelsky. AC current gain of transistor is change in collector current divided by change in base current with Vce held constant. The collector current of the first transistor becomes base current of second transistor. Total current gain is product of each transistor's current gain. Total gain B = B1 * B2. If gain of each is 100, B = B^2 = 10,000.

[@tertia0011]

However - B&N makes an equivalence between Ic and Ie which is not strictly correct as ie = ic + ib. In which case, yes, B=((B2+1)(b1+1)ib1)/ib1=(B2+1)(B1+1)=B1*B2+B1+B2.Yet as B1*B2 is >> (B1+B2), B is simplifed to B1*B2. 10,000 >> 200.

[@tomterrific9459]

@@tertia0011 I am not sure why people find this so confusing. Yes, a lot of people say this, and a lot of people are wrong. See how far you would get in electronic engineering by using that. The current gain of a two transistor Darlington stage is (B1 x B2) + (B1 + B2). *_That is the formula_* . en.wikipedia.org/wiki/Darlington_transistor

[@tertia0011]

@@tomterrific9459 My reference is 'Electronic Devices and Circuit Theory' by Boylestad & Nashelsky, an upper level electronic engineering text book. Wikipedia? - 'See how far you would get in electronic engineering by using that.' Given (B1 x B2) >> (B1 + B2) I'm happy with B&N's formula as given. No confusion.

[@terrysteven3528]

@@tomterrific9459 the correct formula is actually (b1+1)(b2+1)=b1*b2+b1+b2+1 . Eg (100+1)(50+1)=5,151=50*100+100+50+1. Wikipedia forgot the 1 on the end so technically Wikipedia is wrong. To their credit they do mention b1*b2 is a good approximation for large beta values.

[@richardbrobeck2384]

I use A Harmon kardon 5 channel power amp for my home theater mine looks just like that HK 680 but it is a poweramp only .

[@turbo5483]

I have Phoenix Gold car amplifiers that have triple darlington

[@LaGataNegra3073]

Dont worry Paul. We all use our fingers to erase on a dry erase board. and then secondly, we all use the eraser after.

[@jamesderby4522]

More whiteboard remedial electronics please

[@thisisnev]

I wish Paul could have gone into the actual advantages and disadvantages of Darlington circuits. "I just don't like the way they sound" isn't very informative.

[@SJMessinwithBoats]

These chips have been used by so many manufacturers. Been around for 40 years. Class AB, Low THD, wide frequency response, good slew rate and most the time you can just pop in another chip and go. Also, low price.

[@johnsweda2999]

I think he means because Darlington pair has to be matched up other you get ringing so it's a lot of sorting out of transistors time-consuming measuring them

[@SJMessinwithBoats]

@@johnsweda2999 ok, but for sure they are just not “audiophile” level.

[@scottyo64]

Seemed to me he did

[@marianneoelund2940]

There's a lot of confusion here. Some people think the term "Darlington" only refers to two transistors in a single package, connected in the Darlington configuration. That is not true. Most Darlington pairs or triples are implemented as separate single transistors - and you will find them in virtually EVERY power amplifier that has a BJT output stage. If Paul thinks he doesn't like the sound of Darlingtons, he needs to go back and look at the schematics of their BJT-output power amp designs.

[@janinapalmer8368]

Beta as such is not used in data sheets .. the small signal current gain is referred to as hfe

[@chefchutardo5215]

Limited frequency response ? That amp plays from 1hz to 200khz. Ive always found harman Kardon to be the best sounding amps. And they dont overate the wattage like many others.

[@303nitzubishi4]

Seems like a clever way to get better dynamic range out of some inexpensive components, with basic preamp and power supply architecture. Am I wrong?

[@Sams911]

Is it ever better to keep turning on and off your audio equipment vs just leaving it on all the time? (not factoring in electrical costs)

[@leekumiega9268]

If you don't care about electrical costs then it is better to leave them on , but most people do care about electrical costs and then it depends on the class of the amplifier . Class A ( like a space heater) and AB draw too much current while idle while class H draws less since it idles on a low voltage rail , now with class D and G some manufactures recommend leaving them on . My Sunfire receiver which is class G has a stand by button that turns off the display and output relays but keeps the electronics on since it like class D draw little power when left on. Now with tube amps the filaments will burn out if left on all the time unless it has a standby switch that keeps the filaments at half power.

[@bobross6802]

A 25 yr old amp probably has some aging electrolytic capacitors that need a little time and heat to reform ...

[@marianneoelund2940]

Reforming does not occur every time the unit is turned on. Forming is good for many years. Problems with aging electrolytics aren't related to forming (unless they've been sitting unused for many years). The main problem is increasing ESR, due to loss of moisture from the electrolyte.

[@zulumax1]

@@marianneoelund2940 The ESR of a lossy capacitor will go down with heat. When it cools down the ESR goes back up. Electrolytic capacitors that is. Forming lasts for at least a few years, but if it has not been plugged in for say 10 years of more, it should be looked at by a competent technician. It will need to be brought up slow with special equipment or the capacitors could develop excessive heat reforming and may damage them, vent gas, or in worst case explode.

[@marianneoelund2940]

@@zulumax1 Typically, ESR only decreases 10-20% over the temperature range of 20C to 40C. Good thermal design shouldn't be allowing electrolytics to run hotter than that. If a capacitor is old enough that you can hear the effect of a 20% change in ESR, then I'd say it's well past its useful life.

[@pavanu2427]

How transistor amplify works?

[@finscreenname]

So it's basically a solenoid? Like on the side of a car starter.

[@shahidyt]

Darlington is a not so glamorous place in the north of England

[@niclaskarlin]

You have a decent race track, right?

[@shahidyt]

@@niclaskarlin I believe they do. Perhaps I’m being a bit harsh, I’ve never been Darlo, but I managed to avoid going a number of times. My employer has an office there.

[@Enemji]

Mine amp spec sheet states it is a 3 stage inverted darlington

[@ThinkingBetter]

Hey Paul, what is a Darlington transistor...and the PS Audio speaker of the future responded with Paul's voice.

[@ThinkingBetter]

​@Fat Rat "Paul, what is a phono stage?"...and an actual answer comes that takes long to complete involving various stories. "Alexa, what is a phono stage?"...and Alexa says "Here is something I found on the web. According to diyaudio.com bla bla bla". Actually, the latter is how Alexa answers today.

[@ThinkingBetter]

@Fat Rat I’m familiar with the inner workings of the Alexa SDK and how it works concerning the WWE (wake word engine detecting the Alexa word), the ASR (automated speech recognition turning voice to text) and NLU (natural language understanding doing the text to intelligent actions) services of Amazon. I know some of the people working on this at Amazon. Yes, the system captures your voice after you say the wake word, and what you said within a short time window is stored in the cloud. Thus, if you say Alexa, my credit card information bla bla bla...it will exist in the Amazon AVS cloud service as a recording under strict account security access. I have 9 of these in my house. Just never say a secret right after you say the wake word. The audio stream (16kHz, 16 bits PCM in mono of your voice) is unable to get recorded until you wake up these devices and it stops quickly when you don’t speak. Jeff Bezos is quite clear about keeping the system safe for users as privacy violations can damage their huge business. But, yes, things can potentially get hacked by some Russians, Chinese or whoever we like to blame. I don’t worry about Amazon or Google with their many engineers (unable to keep secrets) doing illegitimate things but worry about hackers finding a way into such products.

[@ThinkingBetter]

@Fat Rat Well, I have gone through my own recordings and have found some amusing cases but nothing bad. I have security cameras that also get recorded in the cloud that are much more concerning because they activate on simple motion and also the audio is recorded. I had cameras that got hacked some years ago. My biggest worry is hacking and especially those Shenzhen China developed cameras are where people need to worry. I stopped using those.

[@PanAmStyle]

Darlington is a nuclear generating station on the north shore of Lake Ontario 😂

[@graxjpg]

When are you gonna talk about emitter follower setups? When you get a question about it?

[@chrisharper2658]

Darlington devices are three terminal configurations so showing multiple emitter resistors makes no sense. The sum of all the base currents will flow through the final output's emitter.

[@chrisharper2658]

@Douglas Blake Please provide the part number of such an example as I've never seen one that would still be considered to be a darlington.

[@marianneoelund2940]

Paul was being sloppy with his diagram. In almost all cases, a Darlington pair will have a base-emitter resistor for each transistor, to ensure and help speed up turn-off by providing a discharge path for the gate charge. Almost every commercial Darlington pair built into a single package includes the base-emitter resistors. I challenge you to find one that doesn't.

[@chrisharper2658]

@Douglas Blake Thank you, I stand corrected. The resistors aid in turn off.

[@rutvinchaudhari3185]

Thank you so much

[@johnnytoobad7785]

True with class-D. They need to "warm up" before they sound decent.

[@ThinkingBetter]

Depends on the circuit. Class-D amps aren't generating much heat and the circuit heavily relies on feedback. Amps that operate without feedback where component temperature coefficients play a bigger role are worse.

[@wishusknight3009]

Chances are it is the input stage that needs to warm up, not the output stage.

[@donpayne1040]

Could ANYONE pint out an alternative to a darlington configuration? In a PSU PCB format, say

[@donpayne1040]

@Douglas Blake Ok, thanks.

[@hom2fu]

Professor Paul going to the board.

[@infinite1der]

More white board sessions, please!

[@lynnpoole7830]

So why does it take a while to ''warm up''?

[@jamesplotkin4674]

@Douglas Blake I agree... re-cap the power supply and it will perform nicely.

[@marianneoelund2940]

@@jamesplotkin4674 Not necessarily. Output-stage biasing circuitry is far more important. If there is that much audible change on warm-up, the owner should have a technician check it. I once serviced a small receiver for a friend, which was producing a small amount of audible distortion until it warmed up. Turned out to be poor soldering on a compensation capacitor, which was allowing the output stage to produce short bursts of RF oscillation at the crossover point.

[@jamesplotkin4674]

@@marianneoelund2940 You make a good point ;-)

[@marianneoelund2940]

@Douglas Blake The RF was rather low amplitude, less than 1V p-p if I remember correctly. I couldn't hear the distortion myself, but my friend, who is legally blind and has amazing hearing, had been complaining about it for quite some time before I finally agreed to have a look at the receiver with a scope.

[@jamesplotkin4674]

@Douglas Blake I never said that.

[@vitadelicatus]

Its a noise amplifier.

[@genez429]

You wanna good laugh? Ask Alexa when her birthday is.

[@jeffbean2409]

I am supposing you know that current flows from negative to positive in actuality ~

[@zulumax1]

Well literally yes, but many of my instructors taught circuit flow which is different from current flow. Most confusing isn't it.

[@zulumax1]

Mr.Carlson's Lab discussed and explained this very well, better than I can. Just remember you can have DC current and an AC signal on the same wire doing two different jobs. Block diagrams are sometimes helpful to understand more complex circuits, that way you are just looking at a logical flow.

[@scottdc6971]

Wow you really mauled that one. How about it’s a single device that has two transistors inside giving a higher gain than a single device. I do agree they don’t sound nice. Good in a fan control circuit though (TIP122).

[@marianneoelund2940]

@Douglas Blake Yes, it's a little annoying that folks around here (including Paul?) seem to think that "Darlington" only refers to multiple transistors in a single package.

[@TheMirolab]

@@marianneoelund2940 No, I'm quite sure Paul knows that it's describing the circuit configuration. It just happens that they've "integrated" darlington pairs into single a single device, just like an Op-Amp does not have to be a small IC chip. It can be a small circuit board populated with all discrete components, like in lots of pro audio equipment.

[@marianneoelund2940]

@@TheMirolab I certainly agree with you, but judging by Paul's statements under a comment I made, he does not understand that a pair of discrete transistors can also be connected as a Darlington pair.

[@marianneoelund2940]

I cannot for the life of me, understand why you call Darlingtons a "little used technique." Every BJT output stage of moderate to high power uses at least a Darlington pair, if not a triple. Less common is the complementary or Sziklai Darlington pair. A Darlington pair may be built into a single package, or may be implemented with two discrete transistors. Only the smallest linear BJT amplifiers can get away with single output transistors.

[@geoff37s38]

Unfortunately, despite claiming to be an engineer, Paul is a salesman and his technical explanations should be taken with a pinch of salt.

[@QoraxAudio]

@@geoff37s38 Exactly, he does know a lot of EEs and even set up PS Audio with a friend who's engineer, that's why he's able to explain these basics.

[@Paulmcgowanpsaudio]

Mmmmm, not sure that's true. You may be confusing the emitter followers commonly used with Darlingtons. I've designed and seen schematics for hundreds of amplifiers and in nearly every case, the standard emitter follower pre-driver, driver, arrangement is used. Rarely Darlingtons.

[@marianneoelund2940]

@@Paulmcgowanpsaudio What's the difference between a Darlington pair, and the driver/output transistor connection where the collectors are tied together and the driver emitter goes to the output-transistor base? I don't think you understand how generally the term "Darlington pair" applies. It's not limited to a transistor pair fabricated in a single package.

[@Paulmcgowanpsaudio]

@@marianneoelund2940 :). Well, one of us doesn't, that's for sure. A Darlington typically is considered one transistor (a compound transistor) where an emitter follower driver is considered multiple. The biggest differences are that generally, in a Darlington, there's no emitter resistor on the first transistor. The current is set from the base of the transistor it is feeding and the arrangement is used to increase the beta of the overall compound transistor. An emitter follower on the other hand, stands on its own as a driver/buffer and has its own emitter resistor that sets the current through it. In most amplifier arrangements we use a complimentary pair for pre-drivers and output drivers. You could, at any point, tap off the output of the emitter follower and draw signal current - something you cannot do with the Dralington arrangement. From Wikipedia for you. "A Darlington pair behaves like a single transistor, meaning it has one base, collector, and emitter. It typically creates a high current gain (approximately the product of the gains of the two transistors, due to the fact that their β values multiply together).

[@Synthematix]

Lets put it this way, darlington pairs are the only way of getting really high quality 2channel stereo sound out of an AV amplifier, theres no contest between my sony and my yamaha, the yamaha in 2ch stereo mode sounds completely lifeless, hardly any bass and very poor separation.

[@soniclab-cnc]

my wife weighs a ton... darling-ton

[@chefchutardo5215]

I hope your wife doesnt read the comments on Pauls Videos 🤔

[@marianneoelund2940]

You're risking being darling-gone for a while.

[@QoraxAudio]

Attaboy! 👍😂

[@QoraxAudio]

@Fat Rat Working from home like a good citizen! 👍

[@QoraxAudio]

@Fat Rat You should see this: kzbin.info/www/bejne/mpvIaKierrxladE Things like that are great to get through corona times 😂

[@markknight3983]

I wish he could’ve been my Physics teacher 😀

[@marianneoelund2940]

That would my version of a nightmare. Paul rarely gets the details straight.

[@bayard1332]

Oh sure... and now tell me what a 3 step inverted darlington circuit is... (my Integra unit....)

[@Enemji]

My Onkyo A-9150 also has the same + a slew rate of more than 500 V/µs avaustralia.com.au/onkyo-a-9150-integrated-stereo-amplifier/

[@zulumax1]

Marketing guys loved to throw the techno babble about liberally in the sales brochures. A quote from the sales brochure for my vintage Sansui AU-7700: "A differential amplifier and a 3 stage Darlington system designed in a parallel push-pull OCL pure complementary output circuit arrangement."

[@iwliwl]

Hallo Sebastian auch von Leipzig.

[@sebastianbachert9528]

Was für ein Zufall!

[@iwliwl]

@@sebastianbachert9528 Das in Leipzig ein paar im Thema sind wundert mich nicht.

[@sebastianbachert9528]

Ich habe bisher nicht viele "Audiophile" in Leipzig persönlich kennengelernt. Aber bei 600000 Einwohnern wird schon der ein oder andere Paul's Kanal verfolgen ;)

[@josefbuckland]

MRS GOOGLE? Who is she married to?

[@nicoras8803]

The question was why, not what.

[@FSXgta]

He did say what. I think you need a hearing aid

[@thisisnev]

@@FSXgta And 'what' wasn't the appropriate answer to a 'why' question. Maybe you need a reading aid? ;¬)