EEVblog

[@leeYT321987]

Hey Dave. I just wanted to say that you are freaking AWESOME! I'm somewhat of a beginner hobbyist. I just cracked open a car charger someone gave me, searched the chip (mc34063), and BAM -- Dave's got a ~40-minute, step-by-step video all about how to design for this chip!!!! So cool you are. You've taught me so much. If you were here, I'd shake your hand. Thank you 1000 times. I love your attitude. You are a pretty cool guy. My love to you and yours. Have a wonderful life!

[@tedvanmatje]

A golden oldie, which I'm watching again. Eight years ago, which seems like a long time, but feels like it was only the other week. Eight years ago, my 'journey' began - where a hobby became something else and has led me to retrade into a the holiest of holy trades and I couldn't be happier. Dave - thank you mate for guiding the way and making the seemingly impossible become reality! Thanks for 'passing it forward'...I owe you one! :)

[@tncbbthositg]

I'm really enjoying your videos. I feel like I'm watching the Steve Erwin of electrical engineering. Super entertaining!

[@fizzicist7678]

11:40 FORMULAS ON THE DATASHEET? Man I do love me when the components are THIS detailed.

[@thoughtyness]

Don't you wish everything came with a datasheet?

[@ciarfah]

Thoughtyness Wish my girlfriend had one

[@MrOpenGL]

MC34063 is like the NE555 for DC-DC converter controllers :-)

[@joshuanorris5860]

MrOpenGL is the 555 a mosfet?

[@Vikramslm]

The 34063 is more popular than the 555. There are timer peripherals in every microcontroller and the 555 is not as popular as before.

[@2003vito]

@@joshuanorris5860 555 is a timer or something like that

[@samihalter]

Just a quick note, when you are doing the calculations, you are calculating Lmin and Cmin, that means you have to use at LEAST these values using IDEAL components. You almost never go, or round, to a lesser values. Real components have ESR and a lot of stuff, which means, even in theory, you are going to need components with more values than calculated, not mentioning the tolerance, so for the example in the video, use a 47uH inductor, and 100uF capacitor, and you'll have results that comply with your specs. Great video, though.

[@TomTRobot]

One thing beginners need to watch for when using the MC34063 datasheet for a simple BOOST converter is the 40V maximum voltage allowed using the internal switch. Even though the formulas will give the user the component values for a 12V to 60V converter, for example, the result would exceed the maximum rating (Switch Collector Voltage VC(switch) 40 Vdc). Additional margin is required to allow for the drop across the Schottky diode as well as overshoot. (Consequently, a practical limit for the output of the simple boost using the 34063 is typically 35V.) If you want higher voltages you need to go to designs that augment the MC34063 with external transistors. Bottom line is that the formulas themselves don't include the checks that the input values and results are compatible with the device ratings. So after running through the steps in formulas the designer must always check that the input values and results are compatible with the maximum ratings.

[@11rmax95]

I've been looking EVERYWHERE for this video! I'm glad you made one! It's awesome

[@scin3759]

Great video Dave. Opens up possibilities.

[@ricardonunes6724]

One of the reasons why the ripple was higher than what you calculated might be that the formula used doesn't take into account the ESR. At 100 kHz the impedance of a 100 uF ceramic capacitor is dominated by the ESR.

[@moriambar6533]

Wow I did not remember how wonderful these videos were

[@pedrohenriquegoncalvesinaz2261]

Once again, thanks for sharing knowledge and giving us hope! You're definitively awesome!

[@playvectrex]

Great video dave! I've actually used the MC34063A forever in just about all of my designs. It's a great cheap, and it's cheap. If you want higher frequency though obviously it has a limitation of only 100kHz as you mention. I made a nice spreadsheet for calculating all of the values if anyone is interested.

[@jasonwarbird]

We need a "We won't go in to that" t-shirt!!!!!!!

[@SE45CX]

Also pretty useful if you don't know shit but still wanna sound like a smartass.

[@edinfific2576]

@@SE45CX I've met quite a few people like that. They're a minority, but a very annoying one.

[@criscros7]

Good video, as always. To make your lives easier: The best MC34063 calculator I could find (after 10 secs of googling and 3 mins of trying some of them out) is the one by rau-deaver. It's only for the step-up config, though, but there are more (typically less detailed) online calculators. The one on nomad.ee has the advantage of optimizing both R1 and R2 automatically. Another one is dedicated to Dave Jones. :)

[@jefflambricks]

Very helpful Dave. This and your other video about how a Switch Mode Power Supply Is basically a Linear regulator (they both use a noninverting amplifier just as on your shirt) is just awesome!

[@vibiance7197]

Hi. Noticed a subtle error in your calculation for R2. Note it should be: R2 = R1 * ( Vout/1.25 - 1) Note the parentheses

[@samuraidriver4x4]

Very good explanation, even for someone like me that's very new to electronics tinkering.

[@jkobain]

Damn, seems like a good chip for building and learning circuits. Thanks a lot!

[@CSSIandAssociate]

Thanks Dave, I have to admit for me this is one of the best vids yet. Scrounging for the chip inside a car adapter. I took apart that from Sirius Sat Radio no go there. Then, I opened up one from a Nintendo DS Score! What I got out of this the most was that repurposing old or unused electronic items you get quite the spare parts bin. Also, using the data sheet to start your diagraming and mathing out the circuit thanks. Also, I made note of the 4 DMMs I added that tidbit to my own Electronics note book. I have three and a O Scope 50MHZ. There are so many uses for this. Thanks again. This could

[@shannongibson325]

Great to see that Avasva has new instructions to save my money and energy to build it.

[@bertblankenstein3738]

Thanks Dave! I will keep my eyes open for this chip and experiment with it.

[@mosfet500]

I have boxes of these! About 20 years ago I got samples from ON Semi, they sent me 25 DIP8 and 25 SO8. This video is 10 years old and there are much better devices today (2021) using much higher freq. and much smaller inductors. 100khz is just too low today and now you get cheap chips with current limiting. If you go to LT's site you can even download their simulator. I just checked Proteus and you can simulate the 34063A in their simulator. Thanks Dave, it's still a fun chip.

[@darinwhite5475]

Very helpful, Dave. Thanks a lot.

[@Madeinoz1967]

very timely for me looking to replace a linear P/S for one of my project with a Switched mode, awesome thanks Dave!!

[@agstechnicalsupport]

This video is a very good resource to supplement the content written in the book "The Art of Electronics" by Paul Horowitz and Winfield Hill on MC34063. This video makes it much easier to understand what is explained in the book.

[@pratherat]

Wouldn't it be nice if datasheets had an embedded Excel spreadsheet?

[@xmenxwk]

+pratherat Thats an excellent idea.

[@fizzicist7678]

HE! EXCELlent...I'll go to the corner.

[@catalin3407]

pratherat Some ICs do have one. Whenever you got an IC, following that just search excel with that IC partnumber and you might be lucky.

[@giomjava]

Wow, Dave! Thank you so much! :) Your videos have greatly helped me undertand the SMPS! Also, The Amp Hour rocks! Cheers from republic of Georgia!

[@andymouse]

This stuff is pure gold.

[@21thTek]

Thanks fo all the Knowledge !! ⭐⭐⭐⭐⭐

[@bryansalamat4505]

Good informative video. Thanks for sharing, I have some clarifications though: The real power (in Watts) will be not be equal to the product of the measured voltage and current since the values of current and voltages are not a a constant DC. The real power (in Watts) equal to the product of V and I is applicable only to a constant DC voltage and current. The calculation of real power will be very complicated as you need to break down the signal into an equivalent Fourier series and calculate each component and adding the result; taking into account the power factor of each component sine waves. This will be unpractical though. A Wattmeter with true rms capability would be better.

[@mrkv4k]

Ahhh makíng me proud as ON took over old TESLA in here in Czech rep. and MC34063 was made about 20km from my home...

[@uN1Qu3DZ]

or if you can't get any fancy controllers and don't want to use the tl494/uc3842, you can make your own. i built my own smps controller using a 555 and a comparator. it drives a 2-transistor forward converter putting out 400w. works in pulse skipping mode like some of the power integration chips. great efficiency, but you do have to varnish the transformer so it doesn't whine. but it's got lightning fast response, and no compensation required because there's no error amp in the first place!

[@KostadinIVKostadinov]

24:00 You forgot a parenthesis. Great job anyway :D

[@Ryan_Smyth]

+Kostadin IV Kostadinov Glad I'm not the only one that saw that. You've verified that I'm not insane. :)

[@philbx1]

Hey Dave, thanks much for the hard work you've put into this. Magic! I had some fun recently with a 34063 Oatley Electronics K207A LED kit using 6 LEDs. Works fine now, but took a bit of research which included speaking with Branko :-) BTW, I tried an 'online' 34063 calculator and came up with the following: Strange value for Ct for a start... Ct=3028 pF Ipk=823 mA Rsc=0.365 Ohm Lmin=32 uH Co=68 uF R=180 Ohm R1=1k R2=11k (15V)

[@rlewis1946]

Thanks for your many fascinating videos! We have all learned a lot of good stuff from your lessons! A Big Fan - RL Atlanta, Georgia, USA

[@EEVblog]

@killerman4ever Correctamundo. Long videos like this take some time many hours to process. And KZbin is retarded enough to make it public before it's really finished processing.

[@dhadumia]

Great Video and a Nice caption on the T-shirt. 👍

[@Datzfunk]

Great video Dave. I just wanted to clarify something that caused me a bit of confusion. To calculate the timing capacitor, you have to convert the 10 micro seconds to pico seconds to arrive at the proper decimal place for pico farad.

[@buffplums]

Dave thanks for this. Really appreciate it. Not being unkind to Indians but it seems there are so many Indians just copying other peoples ideas and not knowing or understanding the theory themselves.

[@MikeJohnMentzer]

Hell yeah I agree Most KZbin educators of my country suck They are more of " exam oriented" and just want viewed to score " marks" They just come up with a paper and pen and draw stuff and they probably don't know what that thing actually is Neither do they actually show the thing with oscilloscopes and all as if it's something that doesn't exist in real No Indian would do a video like what Dave did ! No wonder why my country is among least innovative ones 😖

[@RRROBERT1990]

Excelent job on explaining the formulas!

[@derofromdown-under2832]

G'day Ken, an absolutely bonza vid. This has helped me immensely, as I've been clinging to old school tech by using a mains tranny setup with a V-reg configuration. Since doing my trade over 30 years ago, I've been very reluctant to switch over to SMPS design. You've changed that for me, thank you... One thing I should mention though: It would have been great to show us the total max input current that was needed for your design. Though I did cheat by pausing the vid at 32:46 to see that you plotted it at ~ 532mA. I felt this was important, due to knowing what external supply current was needed, say from a Plug-pack as an example... Once again mate, thank you so much for these videos... Take care :)

[@pjlegault6153]

Chips with MOSFET switches instead of the bipolar junction transistors found in the example MC34063 IC's are going to be preferred whenever better efficiency and higher currents are needed.

[@Laogeodritt]

Given the 1.25V ref, you can set the output voltage by feeding a certain fraction of the desired output to the comparator, hence the divider R1, R2. Too small resistors draws too much current from your output (hence loss of efficiency, loss of output current rating to the real load). Big values cause slower response of the comparator (possibly instability) due to large time constant with parasitic capacitances, reduced noise immunity. General rule, values in the 10k to 1M range are good.

[@RandyLott]

I actually like data sheets with lots of formulas. It makes your application more specific.

[@EEVblog]

@HillOrStream Annotation and link added to the video. It's #90

[@rftek]

Thanks Dave! Great video.

[@Sarahann2712]

I am a very long l

[@EEVblog]

@Schmiki24 It's for sale on my Merchandise page!

[@michaelwatts1597]

Very nice job, Dave!

[@respawnnnn]

Excellent VIdeo. Thank you very much.

[@crocop246]

Great video, very usefull, this IC rocks because its 1.5 A output is enough for driving some more powerfull mosfets, and is very cheap and easy to use.

[@gnagyusa]

Great video! 60% efficiency is crap though. I don't use anything below 90% in my designs. The products are small and there's not much room for cooling, so I can't afford a lot of heat dissipation, or wasting battery power...

[@freda5344]

if you use external transistor you can improve on that greatly. If you want greater current you usually do that anyway, because 34063 has no friggin thermal overload protection! The Vsat of ~0.5V is not so good, external mosfet you can get that down to 0.1V

[@EEVblog]

@leeYT321987 Glad you found it helpful, thanks.

[@GBiv78]

Dave I would have been interested to see you put up the efficiency curve for the Nokia charger that you showed to see how a commercial product compares.

[@Mike-ry4ti]

TL494 is my favourite by a long way

[@peterpanpixyland]

Thanks for the great tutorial! I've used some cheap boost regs based on a China made B6287d chip, and I've used them to get more life from 9V batteries when powering guitar effects pedals. The idea works well enough, but even 20 mV PP noise is too much for those high gain audio circuits. The good news is I've had much better luck with RC filters than LC for light loads. A 47 ohm resistor and a 50uF cap can make that ugly spiky ripple become immeasurable, and more important, inaudible. But I'm tempted now to try moving the feedback to a point beyond the filter. This would make the response time to load changes much slower, but for fairly constant loads, are there any other downsides? I'm thinking yes, because I never seem to see any boost circuits with the feedback taken from after an external filter, so there must be an awful reason not to do it?

[@EEVblog]

@TerminalJack505 One Hung Low is my favorite brand!

[@pir869]

WHAT ABOUT HUFLUNGDUNG

[@fknrdcls]

@Dibblah1900 Great suggestion, board layout with SMPS can become very important, especially when you start talking higher current designs.

[@janslos]

Thanks David, as always a very good tutorial.

[@hellterminator]

So, 7 years later, is this still an industry standard chip, or is there now something better/more ubiquitous?

[@UpcycleElectronics]

hellterminator They are still around and dirt cheap. I'm still sacrificing them to the magic smoke gods. They kinda suck. You can only use them for low power stuff, usually 500-700mA output current max. I'm currently experimenting with current pass circuits with a 34063 but the Darlington output configuration is not suitable for driving a power MOSFET. There are a bunch of controllers out there than can do a better job but I don't think there are better 8 pin "jellybeans." There is the more complicated TL494/TL594/KA7500 PWM controller. It is probably the second most common SMPS controller IC I've seen.

[@ryanroth9590]

The UCx84x series is quite popular as well.

[@SmithKerona]

My favorite switching converter IC is the SG3525A from ST or ON SEMI and is a lot eaiser to use than TL494. It is used a lot in high power DC to DC converters.

[@EEVblog]

@Chimpalimp1001 Caps and inductors have a large tolerance anyway, it's not that critical.

[@MrGayguru]

I use LM2576hv - because it is 60v input and 3a output chip. It is more expencive, but allow do the same invertors boards for 12v-48V

[@bac302]

+Miru Mir or you could just add a mosfet

[@AliensInc.]

bac302: Was just thinking the same thing. Need about 2 amps so I was about to ask just that. Thx

[@nxxxxzn]

ordered some MC34063A and SR540 :]

[@LaraSchilling]

The shirt, DO WANT. Need to stop spending on expensive meals and get your shirts! We're up to switch-mode ps's in two weeks. Exam on wednesday for basic power supplies, then we move onto the specifics Very easy subject, We're doing well.

[@Snijele]

Awesome shirt Dave!

[@aaronsnoswell]

On Dave's website!

[@FiveFishAudio]

The chip could be overheating. Try putting a TO-220 heatsink on top of the PDIP chip and see if that works better. (you can epoxy the heatsink to the chip).

[@EEVblog]

@mrmatz408 I explain it in a previous blog on dummy loads.

[@paulbendel]

TI examples show 87% and 83% efficiency. If you want greater efficiceny do you first tweak component values? If you still want greater efficiency do you let it have a greater ripple and then add a second order filter?

[@radu3g]

It would be useful to add a video with calculations with MC34063 for more than 1.5 amp, mosfet or bipolar driver.

[@manomurthy100]

u made it simplest ever..........thanks a lot

[@amoghjain]

in regards to 17:48 when you said switching frequency of the device is 100khz, but as shown at 17:12 ocillator frequency is about 33Hz. are they different?

[@RSP13]

Thanks, Dave!

[@Jpborus]

Great tutorial thanks a lot! Could I follow this tutorial for building a converter that delivers 100V output? What would I need to watch out for? Thanks Dave and anyone who has experience with DC DC converters.

[@mharding1258]

T-shirt is great. Love it.

[@Dibblah1900]

Switch mode power supplies are great. However, board layout can become more complex due to the higher frequencies. I would love to see more information on board layouts - Lord_of_the_board.pdf seems a good intro to me, but I'm not very skilled in the art. Cheers, Allan.

[@GeorgeTsiros]

100 mV ripple at 15 V that's really low. Bought an SMPS wall-wart, it cost like $30 it has 300 mV ripple at 5 V when pushing 180 mA. Ridiculous. And it's supposed to do 3 A at that output voltage? Not even testing that.

[@TaiViinikka]

+George Tsiros I read your experience with a cheap walmart walwart :) and almost gave it a thumbs-down because that's just terrible. Then I realized the downvote wouldn't reflect on Walmart.

[@grahamsayers5136]

Measurement Question. For pure constant DC levels Iconst*Vconst=Pavg. For a constant DC Voltage where the I signal varies with time, Iavg * Vconst = Pavg still works. However, for signals where both V and I vary with time, Iavg * Vavg is not equal to Pavg. The calculation for avg power requires you multiply the V and I waveforms together point by point, integrate, then divide by time interval. For a switching power supply like this, there is some amount of ripple on Vin,Iin,Vout, and Iout. What criteria can you use to justify approximating Pavg as Iavg*Vavg? I can tell you if you are dealing with square waves (pulsing a voltage to a load for example), Iavg*Vavg fails miserably.

[@11rmax95]

Could someone explain why the Darlington configuration voltage drop should only be accounted for in the stop down converter

[@freda5344]

just look at how the Q1 Q2 internal transistors must be configured

[@EEVblog]

@electronicfanatic1 It's just a regular resistor (usually low value) that happens to be used to measure current. Nothing special.

[@EEVblog]

@DagGirl Power Supplies is a pretty cruisy subject!

[@petersage5157]

The optional LC filter is where you'd place a current pass transistor if you needed more even lower noise, right?

[@tanishqbhaiji103]

Doesn’t drive MOSFETs very nicely, only really works with BJTs and has very low efficiency. Nowhere near sensible for a battery powered converter. MPS 2307 is a great choice for a buck regulator nowadays 3-4A 24v 1.8MHz+ fully synchronous so very very low dropout and very high efficiency with very low ripple like 20-30mVPP with 47uF MLCCs. 2 dollars shipped.

[@phrankus2009]

Would be cool to rig this design such that ALL of the RCL components are variable, within a "compatible range" .. .. Would love to be able to dial in various values and to see, in real time, what the range of effects are, for a "tweak" of one, or several components. ... My hope would be that I can find those values which reduce ripple (and noise) the most.

[@ivanv754]

@MrTranxistor It is a type of diode that can be used for very fast switching and has a very low forward voltage. Normal diodes have a limit on the speed they can switch, Schottky diodes are much faster.

[@EEVblog]

@rlewis1946 Thanks, I still enjoy doing them!

[@RandyLott]

@MrTranxistor A Schottky diode is a "fast" diode.

[@robertpeters9438]

So what we need is a multimeter that can multiplex 4 input sets and read out on a tablet via bluetooth?

[@joshuanorris5860]

Why wouldent i want a larger inductor? It wouldent be more efficient? I thought there was no replacement for displacement :P

[@vankiu]

that shirt rocks :) I would love to get my hands on some..

[@gdgdgddgdfgdgfdggddf6355]

Hello. Nice video. Thanks for this explanations.

[@JohnUsp]

31:46 - A good reason to buy more DMMs ;)

[@AliensInc.]

People say it"s extreamly important to shorten the cables and also the layout when trying this on a breadboard, but this is 100kHz and I've seen FM transmitters for 1MHz and it cables hanging everywhere and that is working, also a UTer said that. I wonder since I just can't get this to work on breadboard as I really wanna test before I make a PCB for it. Also later adding a MOSFET for more power but as I can't even get the 34063 to work...... What to do? I'm going for a 12V to 24V boost for 500mA as that will be enuf for powering the later added MOSFET.

[@jr8699]

Hi, do you have video about SEPIC anywhere? Thanks!

[@MrPeetersmark]

How much current should the Inductor be able to handle? would it have to handle the entire load of the output? or is it very small because it's just acting as the oscillator? Thanks for all the great videos.

[@uN1Qu3DZ]

@toffer1220 maybe you don't need videos. all you need is "Switching Power Supply Design, Third Edition". I learned pretty much everything i know about SMPS supplies from that book.

[@DavidLindes]

33:44 - I don't know if you've already done this somewhere, but I'll just say I'd be interested in knowing more (even than you get to) on how to tune efficiency... e.g. for, say, a small energy harvesting circuit or something, where efficiency makes a difference in possible/available duty cycle for any (relatively) high-power operations of a microcontroller on the other end, say.

[@Sentinel-1]

If anyone is looking for simpler and modern solution for DC-DC switching converter, check LM2576 and LM2577.

[@freda5344]

hardly a jelly bean price, half sw speed of 34063, not as versatile. If only there was a 34063 update with thermal overload, that would be perfect!

[@bryan13579]

big fan, nice vids, great tutorials, all just great

[@burski0954]

Love it!