@@nomusicrcthey're just CALLED Pentium, they're not actually the same as the pentiums from the 90s. Still, pitiful performance like all of Intel's low end/low power chips
Your content on your old html site I often visited, bought pcbs from you on many occasions, and eventually found your channel here. You are hands down one of my favorite creators and just wanted to say thank you for sharing your knowledge with the world.
Clive has provided a great deal of encouragement to get new people started on the path of electronics. His videos are always informative and also just as importantly, entertaining. I had completely forgotten about this chip even though I have loads of them buried in my `piling system`.
@@TheTemporalAnomalyhim and electroboom have helped me avoid death and also motivated me as a software engineer to be a better electrical engineer. Now I work at Ford for their ev division still dealing with software lol they can't trust us 😂
@@bigclivedotcom And Good To See You 🤗 Had to resubscribe, and never get notifications. But Always Fun ! ! ! ... how's that vibrator working , probably not the best forum to be asking 😂 🤣 😂 ,ill see myself out
Just finished designing a power supply to take 24VAC and through a diode bridge and into one of the MC34063s down to 12VDC. I have designed dozens of power supplies and avoided this for years because it was an OLD chip... once I used it for the first time a few years ago, I keep coming back to if for low current applications! As long as they keep making it, I'll keep using it.
Love it. Dave Jones covered this chip’s functionality and it’s data sheet several years back. Your discussion here in no way duplicates his video. Indeed it compliments it nicely.
Motorola Semiconductor Products started in Phoenix, AZ in 1949 and made some of the earliest transistors and ICs. In 1974, Motorola built its first chip plant in Austin, TX. Its microprocessors were made in Austin. Its transistors, analog ICs, and logic chips continued to be made in Phoenix. The Phoenix division was spun off as ON Semiconductor in 1999. The Austin division was spun off as Freescale in 2004. Freescale merged with NXP in 2015. So the MC34063 was effectively made by its original manufacturer as are many, many other "classic" devices.
@@acmefixer1 Some of the plants are gone. But my old Motorola data sheets show 3501 Ed Bluestein Blvd., which Google maps shows as an NXP location. NXP's web site says they have two wafer fabs in Austin and two in Chandler, AZ. Not sure if the Chandler fabs are former Motorola ones or new. More research to do. I'm somehow obsessed with semiconductor manufacturer history, which I've been following since the 1960s!
@@mikemike7001 Goo Street view shows two small businesses and a sign that says "Life Sciences & Technology Innovation District." Looks like the bldg behind it was torn down and was made into a parking lot. But yeah, across Semiconductor Dr. is the big NXP plant.
@@acmefixer1 Here in New England, many biotech companies also moved into former minicomputer company buildings. Will be interesting to see where the CHIPS and Science Act will help revive the semiconductor industry in Austin and elsewhere in the U.S.
The NXP plant in Chandler AZ at the intersection of Alma School and Knox has been there since the 1980s when I worked for Motorola. It seems to still be active - there are cars in the parking lot when I drive past.
I got a bunch of those at the dollar store. I thought they'd be good 12 volt to 5 volt supplies for ESP32 boards and such. Cool to know they can be more versatile. And the warning about pushing them too far.
The mystery missing diode was a 1N914 or similar. It basically serves as a fuse in case someone's power outlet is wired backward, or they install their battery incorrectly. It would create a dead short to blow any upstream fuse, or it would let the smoke out and shut down the circuit protecting the cell phone or other device plugged into it.
How would the adapter and device remain shut down after the backwards diode blows? I think it would only protect the output circuit if it blows the internal 1 ohm fuse before dying . Car batteries can dump out 70 A or more for the starter motor anyway, so will barely feel the short, and utility sockets (cigar lighter sockets) often have generous enough fuses to allow plugging in an external starter power source .
@@johndododoe1411 They would be relying on the 1 Ohm "fuse" burning out. It seems unlikely they would choose a small signal diode like a 1N914 which is only tated for 300mA, you usually see a rectifier diode like a 1N4001 or equivalent in this position.
@@johndododoe1411 You're right, the battery would't care, but the fuse in line with the outlet will. It could be that the 1R is faster to blow but I'm not sure who would win this race
When I was at college in 1996 doing my engineering course, I fell in love with the NE555 chip. This one I will have to study. Thank you, Clive, for bringing this to our attention. A true Scotsman knows their history.
The data sheet for that chip is a monster - mainly because it is collection of building blocks that can be used for boost, buck and buck-boost, not to mention voltage inversion, with current limits determined by external components. Its biggest limitation is the upper oscillator frequency, compared to modern chips the PWM is very slow, so inductor choice is limited and controlling ripple on the output is more difficult. In terms of current limit, the circuit you analysed is using the internal driver transistor to pass the current, so of course there can be thermal problems - if you use that to switch an external FET then heating of the chip becomes less of an issue and your switching device can have a chunky heatsink. Ultimately you can always use a current sense on the output and drive the shutdown pin to stop the oscillator as required, though really a more modern device with thermal protection would be a better choice at this stage.
Never has a cheap USB charger seemed so interesting! Loved the straightforward explaination of exactly how the circuit works. This mundane little device has made some fascinating content - thanks Clive.
The reason why the MC34063 is still around is because it's a commodity item, it's cheap! It's also available in the 8 pin DIP. I bought a tube of them recently. 👍
They aren't very good, though, as they are prone to subharmonic oscillation without very careful circuit design. They tend to come on, overshoot, switch off, let the voltage decay, then repeat. A lot of switch mode regulators can do the same thing but they usually use current mode control or other more advanced multipole compensation methods to try to avoid it because it's bad for EMI, regulation, squeal and other things.
I saved a bunch of them from some PLC modules that sank down when flood hit open pit coal mine. Modules became unusable from transformers going bad from water.
This explains so much. I had a car charger like this that destroyed a small lithium ion battery charger. And now I have the explanation of how it happened.
ngl from the title I was half expecting you to find a 7805 or similar linear regulator in there, but this was sooo much cooler. I'd surprisingly never come across this chip before, but will definitely be adding a tube to my next component order.
same! given the kind of stuff we're used to seeing Clive dissect, I figured an enclosed 7805 with no heatsink whatsoever... this was a pleasant surprise!
I had one of these circuits power a dashcam (700mA MAX) for about 5 years and it worked flawlessly, regardless of the ambient temperature (-23C lowers and +82C highest).
I used the extended thermal version of this regulator (MC33063) in so much projects I couldn't even count it. They are such a versatile workhorse if relative low power needed.
I tried to use one of those devices recently. It seemed nice and available when so many parts are hard to get. It's got dreadfully high on-resistance so very lossy. I tried adding an external transistor but had to replace it with a tps61088, which is far more effective.
They are a poor choice for any new design that isn't extremely cost sensitive. The losses in the bipolar switch is one issue, but also they only work up to around 100kHz so it demands larger inductors which incur either more loss or more cost. It also uses hysteretic control which is prone to creating RF interference at low loads (a very common complaint for these cheap car adapters).
They were designed back in the day, when the standard for getting 5V out 12V was to use 7805 with an external BJT and a heatsink. So you can imagine, that even getting efficiency from 40% to just above 80% was a huge deal. They seem to be "very lossy" now, because you have next two generations of switch-mode regulators available to choose from, but you wouldn't have those without the first gen. stuff like this.
Iirc, the MC34063 was originally introduced as a simplified version of the Fairchild µA78S40 which had been around for a few years before the '063 hit the market. Motorola made the µA78S40 for quite some time, too. Both parts are discussed in an ap note by Jade Alberkrack (I have a vague recollection of talking to Jade on the phone a looong time ago.) Almost all buck converters are subject to failing such that the input is connected directly to the output. edit Because these parts essentially regulate the ripple voltage on the output capacitor they are free from need for frequency compensation of an error amplifier. That made them a lot easier to use and prevents palpitations in those who get alarmed when presented with jω.
We have enough time in our lives to learn ANYTHING, our brains can conceive everything if you only try. I do not wish I was smart enough to get this channel and all the comments because I smart enough to know I can start by learning it. 🙂
Many car USB chargers (the ones with a single standard USB socket and a supposed 1-amp rating) are still a handful of discrete components. I haven’t taken apart a recent one from Dollar Tree, and maybe it’s time I spent $1.25 and put in a little easy work to see if they’re better now. At that price point… probably not!
I bought a few sticks of 34063 in PDIP (about 300 of them) and have used them for years. They are are actually very rugged, despite not officially having thermal limiting. The 0.22 ohm current limiting resistor always restricts the on-state current of the pass transistor inside the 34063, so they are quite safe in most normal operation. I have had them short curcuit outputs from time to time and never blew a chip. And it should never go over voltage if the feedback resistors are working because any time the output is greater than the feedback voltage the 34063 will not turn on its pass transistor. If the top feedback resistor goes open circuit you are in trouble, but that applies to every chip with feedback resistors for voltage reg.
Funnily enough, I have one of these car USB adaptors sitting in front of my keyboard that came with our second hand car that we bought back in 2014. The only difference is my one having a different sticker on it.
Thank You! Finally explained how these impulse power supplies work that I understood 100%. Sure there are numerous variants of these, but I got this one right. Tells much about the teacher
The missing diode will be for reverse polarity protection. If the supply is connected with the wrong polarity, the diode will be forward-biased and pass the best part of 12A through the 1 ohm resistor, which will do a convincing enough impression of a fuse.
The 1N5819 is there as a "Damper diode" to prevent the output voltage from going negative with respect to ground, which would destroy the chip. It's from the "High efficiency" series, called this because of it's very fast recovery time from saturation. Earlier damper diodes dissipated power during their slow recovery time reducing efficiency.
I use 34063 as negative voltage generator for OP Amps that need negative voltage rail. I used to hand-build such circuit but recently found few tiny PCB modules that do the same thing.
Oh i've played a lot with that chip. It's a great building block for building small boost converters for Nixie tubes and the like. It doesn't care if you build on a breadboard, where as more modern chips requires a nice PCB layout to work at all.
i had/have that same car charger and used it to charge my android (used it all day while working) for over a year. it actually held up better than lots of other ones i used
More high end chargers claim to have multiple layers of overvoltage/overcurrent protection, which means more failure modes, which works against holding up the longest.
@@johndododoe1411 i had opened it up years ago and seemed to remember it looking exactly like it. however since you asked i decided to dig it out and look again. turns out is not even close to the same circuit at all
This exact adapter is (or was) in Dollar Tree stores here in the US, I bought a bunch to use as handy 12V->5V regulator boards. Though the versions I had, used the PDIP package.
It's so common, that I even have a few in my parts bin! If the 1.5A built-in transistor is not enough, an external transistor can be driven by the output of this chip.
there is lots of chips that date way back in time that's still avail ect , there was sooo much new development of new chips that is still avail and made today from all those years ago , truly amassing
OH GOD NOT THAT FKING CHIP Working on my senior project in college ... my classmate was looking REALLY CLOSELY at the chip AS I WAS APPLYING POWER TO IT and it grenaded right in his face, with a fragment striking just below the eye. There was no obvious reason for the failure, either. IIRC, I replaced the chip and just tried again and it was fine. (I was stepping 24VAC, rectified to ~33VDC, down to 3.3VDC, so I was asking a bit from it) EDIT: I imagine that someone has said this already but you may as well consider ON Semiconductor as literally being Motorola because they... well they are.
I had a charger similar to this and, yes, it did blow up my expensive phone. I am not sure that an over-current can cause this chip to overheat as the current sensing through the 0R22 resistor reduces the chip output duty cycle, and hence average output current, if an excessive load is placed on the output. This is all well described in the Motorola AN920 application note, which also describes the uA78S40 switching regulator chip, the predecessor to the MC34063A.
Problem is the chip has no on die temperature sensing, so will operate till the output pass transistor goes short circuit from the die running over 200C for a few hours. The old process ones would run around 100 hours into a 2A before failing, the new die shrunk ones are a lot less robust though. Run the die hot and the layers continue to diffuse through the silicon, till the output collector diffuses the junction through to the bottom of the die, and then basically forms a diode from input to output, or the collector base junction diffuses to the point the metal layer shorts the collector to the base region.
In your diagram the oscilator input from the current sense resistor is labeled Ipk, supposedly peak current, meanining it detects when an overcurrent situation occurs and shuts down the oscilator.
I thought about mc34063 just by looking at the thumbnail and title. anyway I think it's time for it to go, the stabilization in those chargers is often pretty bad and output power is miniscule. It was probably good enough for phones 10 years ago, but not really now.
AFAIK, MC430 is that type of a chip which can be used as a boost, or buck voltage regulator, or it can even provide negative voltages as you need them in your design. Moreover, it also supports using an external transistor for switching the inductor in case of a bigger current.
This made me think of switching power supplies, but using a similar mechanism to buck/boost. I’ve been playing with electronic circuits since the late 70s and haven’t had the opportunity to play with this chip!
It'd be interesting to see if one can add a simple upgrade for thermal protection... e.g. something that would cut off the 12V input, based on a thermistor value.
@@bigclivedotcom Chinese XL4015 buckconverter chip has this kind of suggestion even in datasheet. Introducing over 3,3 V to feedback pin makes chip to go in standby.
Given the failure mode discribed.. Wouldn't it be a good idea to put a zener in that diod position to shunt overvoltage and potentially blow the 1R instead of the divice that is pluged in?
A suitably rated zener would add a couple of pennies to the cost, not acceptable in the bottom of the barrel market this is aimed at. This is also why there is no proper fuse on the input, cost reduced to the last fraction of a penny and reliability be damned.
@@ferrumignis I mean, yeah. True, nobody would spend a dime more than they have to on this kind of thing. But ppl like us could if we want to use it a little more savely
This chip has been on my jellybean list for years. A really sure way to bring on the short circuit failure mode is to use an inductor with an insufficient saturation current rating. It has a big brother, the 78S40 which has even more modular parts accessible via pins. An opamp with its own supply pin, a floating diode. The voltage reference has its own pin too.
thanks for mentioning the uA78S40, so i neednt do it, the diode in it is rather poor performer, any discrete 1A40V Schottky improves efficiency over it.
i love your videos big clive and im not at all educated on any of the atuff ya do . but iv always been fascinated with it and now im old i like to try andale my own stuff from other stiff.. Like power banks and 12 volt stuff since me and my dog live in out van its very satisfying to make something.. i haven't yet but im leaning slowly and you nake it easier to learn because you explaim ot.. thank you so much
Clive thx for sharing. There is a new rendition of this chip that is used in car chargers. It's still a 34063 IC under the hood but the V divider is internal and set for 5V. I guess that's optimization for you. Two less parts made somebody go woo-hoo. I don't care who you are these ICs are excellent for the ratings envelope that they possess. You can these at the dollar store. Then you get the diode, inductor, PCB, LED and IC for around a buck. Wind several turns over the inductor and you get either a fwd or flyback converter. Now that's a sweet deal. The clk in the IC runs all the time. The ratio of the I_dchg to I-chg of the timing cap is described in the datasheet as ~ 6:1. The Timing crkt is interesting. The I_sns increases the charge current to the timing cap many times so it has a much reduced charging time/current. Ok, so this ain't a multi-Mhz clocked switcher, but is stable and quite easy to work with. I'm not understanding why either fwd or flyback mode isn't utilized if you have a critical load that cannot tolerate a shorted buck output xsistor. The energy can only cross the Xfrmr while it switches. Take away the switching, shorted output Xsistor, and the output dies. Think of a buck reg as a PWM output that is low-passed that can pass the DC input to its output. My guess is that this requires more magnetic design than people want to deal with. If you choose fwd or flyback you get to take advantage of the VCEsat voltage instead of the darlington emitter follower. The voltage across the output xsistor is over 1.7V when combined with a small amount of maximum current soon exceeds the pkg thermal rating. 1.25W/1.3V= 960mA. Don't forget the TJmax is based on a TJ of a 150C. If you run a device at it's ultimate thermal limit you are actually performing a HALT test. That's Highly-Accelerated-Life-Test. So, don't do it. If you take the VCEsat @0.45V and I_max @ 1A that power level, 450mW, is below the pkg rating even for the SOIC pkg of 625mW. The PDIP-8 and DFN have power ratings of 1.25W. I seem to remember if you have a very expensive load that can be destroyed by an over-voltage occurrence, then for heavens sake use a crowbar. Thx for reading. That's all I got for now. Again, thx Clive.
This looks like the kind of chip that'd be fairly easy to make the building blocks of with discrete components. That'd be a fun project and learning tool :D
Jack st clair Kilby invented the computer chip ( IC ) August 1958, won the Nobel Prize in physics, 186,000 dollars, and the Nobel Prize is now called the KILBY AWARD , and Texas instruments who he worked for built a museum called the Kilby museum, in Texas. 🤔👍
one of these blew up my bluetooth speaker. they should not be allowed to make something that fails into an overvoltage state. i think it gave it full 12v
I just happen to have two of these 12/24V->5V device chargers on my desk, so I took one apart while watching 🙂. I must have an older version of the PCB (silkscreen version: LC-022?). D1 is populated (not on yours), but I don't have D2 on the PCB (efficiency diode). All other components seem similar, including the use of the MC34063.
With those devices can also create havoc with CAN Bus system by creating false error messages as I discovered with my Smart car. It gave indications that CPU had failed I turned the ignition off removed the charger and everything went back to normal and no more error messages. A great sigh of relief because I was some distance from home and the nearest Smart car service agent was about 60 to 70kms away.
Radio frequency interference is a common issue with this device due to the way it regulates, and the lack of any additional filtering in these cheap and nasty adaptors.
@@ferrumignis Agreed about the RF interference and really surprised at the havoc it caused. It makes you wonder what other problems a poorly surpressed device can cause. It's annoying that manufactures don't prosercuted for bad RFI interference and possible damage it can cause.
Rsc perhaps may moderate the operation of the oscillator to either shut it down on excess current draw or to vary the frequency of the oscillator as the load changes to better keep up with a changing load.
A part I used years ago for a nixie clock HT power supply: 12V in, 180V out. The data sheet says that's too much, but it's worked for years... These days I'm looking more at buck than boost and efficiency is a priority, but it's nice to know this is still around.
Uses a fixed frequency with a variable pulse width. You can modify for MUCH higher current with a mosfet follower and a bigger inductor. One thing I have used these for is for making constant voltage power converters for use with solar panels.❤
On is Motorola's chip division, just as NXP is Philips . So they are the original company with a new name and logo . Note that NXP was recently split into Nexperia and NXP as part of a failed attempt to merge with some foreign company .
I saw "iconic component" and guessed a 555. At least I was right that the 555 is possibly the most iconic component! (Though I think a 555 with a big inductor could probably work as a variable voltage divider?)
3:48 Would this still be a disaster with USB C PD capable devices? Since they can accept higher voltages though I have no idea what happens if a PD capable device was expecting 5V only to suddenly get 12 since it is more than 9 but less than 15, hopefully just blow a fuse on the charge port.
Yes .very good explanation of product.I live off grid Solar power 12 volt.Use 1 amp .2 amp use charger for Android device,s..Now available in Australia a 36 amp 12 volt charger .
The data sheet is more than a little vague about it but I believe that the 0.22 Ohm resistor sets the peak current limit at 0.3V/0.22. When this limit is hit, the switch is turned off early in the cycle. This almost protects the chip. If the inductor is a really good one, the catch diode is a good one and the output has a solid short, the inductor current will get kicked a little higher on each cycle until it burns the chip out. For cheap stuff the solution to this is to make the inductor not so good so that cycle to cycle the current runs back down to near zero. For better protection, there is a trick where you detect the extremely short ON time and tell a lie to the voltage comparator. Basically the circuit listens to the voltage on the 0.22 and if it sees a really sudden increase in the drop, a PNP turns on and charges up a capacitor that times the delay until it tries again.
I encountered an elastomeric (zebra) conductor today when opening up a cheap combination hygrometer/thermometer (got a dozen for less than $20). I was surprised they existed, let alone already a somewhat old technology!
I still use this chip for personal projects. I can fill in the pin numbers off the top of my head. A recent project uses 2. With 12V in, one makes -12V, the other with an external FET, makes +250V, for tube projects.
The voltage regulation is achieved by the action of the comparator. When the input of pin 5 gors above 1.25v the output of the comparator switches off and that in turn shuts off one input the the AND gate driving the set input of the RS ff. This basically shuts off the oscillation driving the output circuit until the voltage at pin 5 drops below 1.25v at which point the chip begins pulsing the output and the series inductor to build the output voltage again. Etc. Etc
Using an inductor/capacitor pair like that, instead of a capacitor is particular unusual. Most newer power supply chips, be them linear or switching use only a capacitor for the power draw line. If they use an inductor it is a small choke chosen just to smooth ripples.
When I saw the charge adapter and title, I had a feeling this would be about the MC34063. One of Dave Jones's earliest videos is a tutorial on all the maths for building a buck converter with this chip. ua-cam.com/video/qGp82xhybs4/v-deo.html
I recently dismantled one of these myself, actually yesterday, it has a CX8812 smps chip, these are higher current devices and the inductor was a toriodal type.
I love seeing older technology still being relevant in todays world :)
I've seen some cheap laptops at Costco with PENTIUM processors in them lol
@@nomusicrcthey're just CALLED Pentium, they're not actually the same as the pentiums from the 90s. Still, pitiful performance like all of Intel's low end/low power chips
@@qwertzy121212Yeah low end has changed a lot especially since AMD has been clapping Intel's Cheeks for a good number of years now
All circuits must contain a 555, an obsolete chip and at least 4 resistors and a capacitor that you cannot work out what they do or why they are there
@@qwertzy121212 So why would a company rename something that they've already released but it's different
Your content on your old html site I often visited, bought pcbs from you on many occasions, and eventually found your channel here. You are hands down one of my favorite creators and just wanted to say thank you for sharing your knowledge with the world.
Clive has provided a great deal of encouragement to get new people started on the path of electronics. His videos are always informative and also just as importantly, entertaining. I had completely forgotten about this chip even though I have loads of them buried in my `piling system`.
@@TheTemporalAnomalyhim and electroboom have helped me avoid death and also motivated me as a software engineer to be a better electrical engineer. Now I work at Ford for their ev division still dealing with software lol they can't trust us 😂
I taught him everything he knows.
There are some circuits that are timeless and will probably be with us for as long as we build electronics. Thanks Clive❤
No need to re-invent the wheel.....
Great to see older chips still in use. The 555 is the very first chip I ever messed with when learning electronics.
Followed by the 666 chip? 🤣
No, the 7555 succeeded it as the CMOS version @@two_tier_gary_rumain
@@two_tier_gary_rumain nah, 556. it's two 555 in a trenchcoat.
Yup 555 n 556 n l386 ♥️👻
That and the 741❤
I love that some how you haven't run out of random dollar store crap to reverse engineer. You rock Clive!
There's no end of tat in sight.
I'll Buy That For A Dollar 😂
... it's a dollar and a quarter now sir,.. Plus Tax
@@bigclivedotcom
And Good To See You 🤗
Had to resubscribe, and never get notifications. But Always Fun ! ! !
... how's that vibrator working , probably not the best forum to be asking 😂 🤣 😂 ,ill see myself out
@@bigclivedotcom😂
Which is a 25%increase if I might add which is ridiculously outrageous
Just finished designing a power supply to take 24VAC and through a diode bridge and into one of the MC34063s down to 12VDC. I have designed dozens of power supplies and avoided this for years because it was an OLD chip... once I used it for the first time a few years ago, I keep coming back to if for low current applications! As long as they keep making it, I'll keep using it.
Love it. Dave Jones covered this chip’s functionality and it’s data sheet several years back. Your discussion here in no way duplicates his video. Indeed it compliments it nicely.
That is so true! Great video here, bringing back fond memories of the past.
Motorola Semiconductor Products started in Phoenix, AZ in 1949 and made some of the earliest transistors and ICs. In 1974, Motorola built its first chip plant in Austin, TX. Its microprocessors were made in Austin. Its transistors, analog ICs, and logic chips continued to be made in Phoenix. The Phoenix division was spun off as ON Semiconductor in 1999. The Austin division was spun off as Freescale in 2004. Freescale merged with NXP in 2015. So the MC34063 was effectively made by its original manufacturer as are many, many other "classic" devices.
Looking at Goo maps, the Motorola plant on Ed Bluestein Bl in Austin is just a concrete pad overgrown with weeds. 😮
@@acmefixer1 Some of the plants are gone. But my old Motorola data sheets show 3501 Ed Bluestein Blvd., which Google maps shows as an NXP location. NXP's web site says they have two wafer fabs in Austin and two in Chandler, AZ. Not sure if the Chandler fabs are former Motorola ones or new. More research to do. I'm somehow obsessed with semiconductor manufacturer history, which I've been following since the 1960s!
@@mikemike7001
Goo Street view shows two small businesses and a sign that says "Life Sciences & Technology Innovation District." Looks like the bldg behind it was torn down and was made into a parking lot. But yeah, across Semiconductor Dr. is the big NXP plant.
@@acmefixer1 Here in New England, many biotech companies also moved into former minicomputer company buildings. Will be interesting to see where the CHIPS and Science Act will help revive the semiconductor industry in Austin and elsewhere in the U.S.
The NXP plant in Chandler AZ at the intersection of Alma School and Knox has been there since the 1980s when I worked for Motorola. It seems to still be active - there are cars in the parking lot when I drive past.
I got a bunch of those at the dollar store. I thought they'd be good 12 volt to 5 volt supplies for ESP32 boards and such. Cool to know they can be more versatile. And the warning about pushing them too far.
The mystery missing diode was a 1N914 or similar.
It basically serves as a fuse in case someone's power outlet is wired backward, or they install their battery incorrectly.
It would create a dead short to blow any upstream fuse, or it would let the smoke out and shut down the circuit protecting the cell phone or other device plugged into it.
How would the adapter and device remain shut down after the backwards diode blows? I think it would only protect the output circuit if it blows the internal 1 ohm fuse before dying . Car batteries can dump out 70 A or more for the starter motor anyway, so will barely feel the short, and utility sockets (cigar lighter sockets) often have generous enough fuses to allow plugging in an external starter power source .
@@johndododoe1411 They would be relying on the 1 Ohm "fuse" burning out. It seems unlikely they would choose a small signal diode like a 1N914 which is only tated for 300mA, you usually see a rectifier diode like a 1N4001 or equivalent in this position.
@@johndododoe1411 I think the idea is to burn the 1 ohm resistor. It's probably rated an 1/4 or 1/2 Watt.
That would make sense too
@@johndododoe1411
You're right, the battery would't care, but the fuse in line with the outlet will. It could be that the 1R is faster to blow but I'm not sure who would win this race
EEVBlog did a great walk-through and review on this years ago. Nice chips.
When I was at college in 1996 doing my engineering course, I fell in love with the NE555 chip. This one I will have to study. Thank you, Clive, for bringing this to our attention. A true Scotsman knows their history.
"A true Scotsman knows their history."/// I wonder what fallacy that is
The data sheet for that chip is a monster - mainly because it is collection of building blocks that can be used for boost, buck and buck-boost, not to mention voltage inversion, with current limits determined by external components. Its biggest limitation is the upper oscillator frequency, compared to modern chips the PWM is very slow, so inductor choice is limited and controlling ripple on the output is more difficult. In terms of current limit, the circuit you analysed is using the internal driver transistor to pass the current, so of course there can be thermal problems - if you use that to switch an external FET then heating of the chip becomes less of an issue and your switching device can have a chunky heatsink. Ultimately you can always use a current sense on the output and drive the shutdown pin to stop the oscillator as required, though really a more modern device with thermal protection would be a better choice at this stage.
Never has a cheap USB charger seemed so interesting! Loved the straightforward explaination of exactly how the circuit works. This mundane little device has made some fascinating content - thanks Clive.
The reason why the MC34063 is still around is because it's a commodity item, it's cheap! It's also available in the 8 pin DIP. I bought a tube of them recently. 👍
They aren't very good, though, as they are prone to subharmonic oscillation without very careful circuit design. They tend to come on, overshoot, switch off, let the voltage decay, then repeat.
A lot of switch mode regulators can do the same thing but they usually use current mode control or other more advanced multipole compensation methods to try to avoid it because it's bad for EMI, regulation, squeal and other things.
Not a thing wrong with it so long as you put some safety circuitry on the load side, like say, a crowbar, lol.
Yup yup, have used them, cheap they do EVERYTHING but yeah, does need some consideration.
maybe they are still being used in european washing machines and russian drones.
I saved a bunch of them from some PLC modules that sank down when flood hit open pit coal mine. Modules became unusable from transformers going bad from water.
dude you posted this right during my week of being hyperfixated on 555 timers. you've given me my kibble thank you
This explains so much. I had a car charger like this that destroyed a small lithium ion battery charger. And now I have the explanation of how it happened.
ngl from the title I was half expecting you to find a 7805 or similar linear regulator in there, but this was sooo much cooler. I'd surprisingly never come across this chip before, but will definitely be adding a tube to my next component order.
same! given the kind of stuff we're used to seeing Clive dissect, I figured an enclosed 7805 with no heatsink whatsoever... this was a pleasant surprise!
Cooler for real, 7805 regulator runs so hot!
I had one of these circuits power a dashcam (700mA MAX) for about 5 years and it worked flawlessly, regardless of the ambient temperature (-23C lowers and +82C highest).
82°C ambient?
@@shaunclarke94 yes, you get those high temps in a car which is sitting in the sun for too long.
I used the extended thermal version of this regulator (MC33063) in so much projects I couldn't even count it. They are such a versatile workhorse if relative low power needed.
I tried to use one of those devices recently. It seemed nice and available when so many parts are hard to get. It's got dreadfully high on-resistance so very lossy. I tried adding an external transistor but had to replace it with a tps61088, which is far more effective.
They are a poor choice for any new design that isn't extremely cost sensitive. The losses in the bipolar switch is one issue, but also they only work up to around 100kHz so it demands larger inductors which incur either more loss or more cost. It also uses hysteretic control which is prone to creating RF interference at low loads (a very common complaint for these cheap car adapters).
Not on resistance as it has a darlington BJT. VBE plus VCE drop.
They were designed back in the day, when the standard for getting 5V out 12V was to use 7805 with an external BJT and a heatsink. So you can imagine, that even getting efficiency from 40% to just above 80% was a huge deal. They seem to be "very lossy" now, because you have next two generations of switch-mode regulators available to choose from, but you wouldn't have those without the first gen. stuff like this.
I really enjoyed the clarity of your circuit analysis. Great video!
Iirc, the MC34063 was originally introduced as a simplified version of the Fairchild µA78S40 which had been around for a few years before the '063 hit the market. Motorola made the µA78S40 for quite some time, too. Both parts are discussed in an ap note by Jade Alberkrack (I have a vague recollection of talking to Jade on the phone a looong time ago.)
Almost all buck converters are subject to failing such that the input is connected directly to the output.
edit
Because these parts essentially regulate the ripple voltage on the output capacitor they are free from need for frequency compensation of an error amplifier. That made them a lot easier to use and prevents palpitations in those who get alarmed when presented with jω.
We have enough time in our lives to learn ANYTHING, our brains can conceive everything if you only try. I do not wish I was smart enough to get this channel and all the comments because I smart enough to know I can start by learning it. 🙂
That's a great title, I would never have watched a video about a car charger without it
Many car USB chargers (the ones with a single standard USB socket and a supposed 1-amp rating) are still a handful of discrete components. I haven’t taken apart a recent one from Dollar Tree, and maybe it’s time I spent $1.25 and put in a little easy work to see if they’re better now. At that price point… probably not!
I bought a few sticks of 34063 in PDIP (about 300 of them) and have used them for years.
They are are actually very rugged, despite not officially having thermal limiting.
The 0.22 ohm current limiting resistor always restricts the on-state current of the pass transistor inside the 34063, so they are quite safe in most normal operation.
I have had them short curcuit outputs from time to time and never blew a chip.
And it should never go over voltage if the feedback resistors are working because any time the output is greater than the feedback voltage the 34063 will not turn on its pass transistor. If the top feedback resistor goes open circuit you are in trouble, but that applies to every chip with feedback resistors for voltage reg.
Funnily enough, I have one of these car USB adaptors sitting in front of my keyboard that came with our second hand car that we bought back in 2014. The only difference is my one having a different sticker on it.
Thank You! Finally explained how these impulse power supplies work that I understood 100%. Sure there are numerous variants of these, but I got this one right. Tells much about the teacher
The missing diode will be for reverse polarity protection. If the supply is connected with the wrong polarity, the diode will be forward-biased and pass the best part of 12A through the 1 ohm resistor, which will do a convincing enough impression of a fuse.
... with added magic smoke for effect 😂
The 1N5819 is there as a "Damper diode" to prevent the output voltage from going negative with respect to ground, which would destroy the chip. It's from the "High efficiency" series, called this because of it's very fast recovery time from saturation. Earlier damper diodes dissipated power during their slow recovery time reducing efficiency.
We're still using MC33063/34063 in development and production (about 20k pieces a year) :D
I use 34063 as negative voltage generator for OP Amps that need negative voltage rail.
I used to hand-build such circuit but recently found few tiny PCB modules that do the same thing.
Oh i've played a lot with that chip. It's a great building block for building small boost converters for Nixie tubes and the like. It doesn't care if you build on a breadboard, where as more modern chips requires a nice PCB layout to work at all.
I absolutely love this! So informational, and so many little tidbits of gold that I was not aware of😊
I'd be curious to see the 5v output on a scope.
Yeah, I am wondering how smooth the output is? What's the ripple?
@@BobHannent Raspberry I hope.
@@shaunmorrissey7313 tasty
Always great to see a good design stick around. Makes me sad to see the Ne555 hate as it is perfectly good even if a little archaic.
i had/have that same car charger and used it to charge my android (used it all day while working) for over a year. it actually held up better than lots of other ones i used
The same circuit or just the same plastic casing?
An android phone will often test the charging device to see what current it can supply by gradually increasing the current and monitoring the voltage.
More high end chargers claim to have multiple layers of overvoltage/overcurrent protection, which means more failure modes, which works against holding up the longest.
@@johndododoe1411 i had opened it up years ago and seemed to remember it looking exactly like it. however since you asked i decided to dig it out and look again. turns out is not even close to the same circuit at all
This exact adapter is (or was) in Dollar Tree stores here in the US, I bought a bunch to use as handy 12V->5V regulator boards. Though the versions I had, used the PDIP package.
It's so common, that I even have a few in my parts bin!
If the 1.5A built-in transistor is not enough, an external transistor can be driven by the output of this chip.
I had one of those usb socket chargers. It heated up to a blazing temperature. I binned it!
there is lots of chips that date way back in time that's still avail ect , there was sooo much new development of new chips that is still avail and made today from all those years ago , truly amassing
It's so like the LM78S40
Used it for years for everything.
The oscillator runs freely and the Ipk modifies the duty cycle.
OH GOD NOT THAT FKING CHIP
Working on my senior project in college ... my classmate was looking REALLY CLOSELY at the chip AS I WAS APPLYING POWER TO IT and it grenaded right in his face, with a fragment striking just below the eye.
There was no obvious reason for the failure, either. IIRC, I replaced the chip and just tried again and it was fine. (I was stepping 24VAC, rectified to ~33VDC, down to 3.3VDC, so I was asking a bit from it)
EDIT: I imagine that someone has said this already but you may as well consider ON Semiconductor as literally being Motorola because they... well they are.
Brilliant use of older technology love it thanks Clive 😊
Thank you, Big Clive. I still don't have a clue what you're talking about, but I'm learning slowly :D
I had a charger similar to this and, yes, it did blow up my expensive phone.
I am not sure that an over-current can cause this chip to overheat as the current sensing through the 0R22 resistor reduces the chip output duty cycle, and hence average output current, if an excessive load is placed on the output. This is all well described in the Motorola AN920 application note, which also describes the uA78S40 switching regulator chip, the predecessor to the MC34063A.
Problem is the chip has no on die temperature sensing, so will operate till the output pass transistor goes short circuit from the die running over 200C for a few hours. The old process ones would run around 100 hours into a 2A before failing, the new die shrunk ones are a lot less robust though. Run the die hot and the layers continue to diffuse through the silicon, till the output collector diffuses the junction through to the bottom of the die, and then basically forms a diode from input to output, or the collector base junction diffuses to the point the metal layer shorts the collector to the base region.
I liked this video very much. Great to see this old chip. I have a request: could you indicate in schematic(s) the pin numbers on the chip rectangle?
i LOVE Your way to tell "It is Not a good thing" when 12 volts passes on to the 5v side :DDD
Legend ...
In your diagram the oscilator input from the current sense resistor is labeled Ipk, supposedly peak current, meanining it detects when an overcurrent situation occurs and shuts down the oscilator.
These kinds of chips are exactly what i like and need really. Saves me from filling in a mountain of paperwork when ON stops production.
I thought about mc34063 just by looking at the thumbnail and title.
anyway I think it's time for it to go, the stabilization in those chargers is often pretty bad and output power is miniscule. It was probably good enough for phones 10 years ago, but not really now.
AFAIK, MC430 is that type of a chip which can be used as a boost, or buck voltage regulator, or it can even provide negative voltages as you need them in your design.
Moreover, it also supports using an external transistor for switching the inductor in case of a bigger current.
This made me think of switching power supplies, but using a similar mechanism to buck/boost. I’ve been playing with electronic circuits since the late 70s and haven’t had the opportunity to play with this chip!
I was an electronics technician for 35 years.
It seems that I should have heard of this chip.
I haven't, so perhaps it's just as well I've retired!
It'd be interesting to see if one can add a simple upgrade for thermal protection... e.g. something that would cut off the 12V input, based on a thermistor value.
It could possibly be used to skew the voltage feedback circuit to lower the voltage.
@@bigclivedotcom Chinese XL4015 buckconverter chip has this kind of suggestion even in datasheet. Introducing over 3,3 V to feedback pin makes chip to go in standby.
I used one of these to generate a -25v 30mA power supply from 5v for a VFD tube. Most old car chargers. I'm talking mid 2000s will have this chip
I bought two of this exact one in 2014. The store had bins full in different colors.
Given the failure mode discribed..
Wouldn't it be a good idea to put a zener in that diod position to shunt overvoltage and potentially blow the 1R instead of the divice that is pluged in?
A suitably rated zener would add a couple of pennies to the cost, not acceptable in the bottom of the barrel market this is aimed at. This is also why there is no proper fuse on the input, cost reduced to the last fraction of a penny and reliability be damned.
@@ferrumignis
I mean, yeah.
True, nobody would spend a dime more than they have to on this kind of thing.
But ppl like us could if we want to use it a little more savely
This chip has been on my jellybean list for years. A really sure way to bring on the short circuit failure mode is to use an inductor with an insufficient saturation current rating.
It has a big brother, the 78S40 which has even more modular parts accessible via pins. An opamp with its own supply pin, a floating diode. The voltage reference has its own pin too.
thanks for mentioning the uA78S40, so i neednt do it, the diode in it is rather poor performer, any discrete 1A40V Schottky improves efficiency over it.
i love your videos big clive and im not at all educated on any of the atuff ya do . but iv always been fascinated with it and now im old i like to try andale my own stuff from other stiff.. Like power banks and 12 volt stuff since me and my dog live in out van its very satisfying to make something.. i haven't yet but im leaning slowly and you nake it easier to learn because you explaim ot.. thank you so much
Clive thx for sharing. There is a new rendition of this chip that is used in car chargers. It's still a 34063 IC under the hood but the V divider is internal and set for 5V. I guess that's optimization for you. Two less parts made somebody go woo-hoo.
I don't care who you are these ICs are excellent for the ratings envelope that they possess. You can these at the dollar store. Then you get the diode, inductor, PCB, LED and IC for around a buck. Wind several turns over the inductor and you get either a fwd or flyback converter. Now that's a sweet deal.
The clk in the IC runs all the time. The ratio of the I_dchg to I-chg of the timing cap is described in the datasheet as ~ 6:1.
The Timing crkt is interesting. The I_sns increases the charge current to the timing cap many times so it has a much reduced charging time/current.
Ok, so this ain't a multi-Mhz clocked switcher, but is stable and quite easy to work with.
I'm not understanding why either fwd or flyback mode isn't utilized if you have a critical load that cannot tolerate a shorted buck output xsistor. The energy can only cross the Xfrmr while it switches. Take away the switching, shorted output Xsistor, and the output dies. Think of a buck reg as a PWM output that is low-passed that can pass the DC input to its output. My guess is that this requires more magnetic design than people want to deal with. If you choose fwd or flyback you get to take advantage of the VCEsat voltage instead of the darlington emitter follower. The voltage across the output xsistor is over 1.7V when combined with a small amount of maximum current soon exceeds the pkg thermal rating. 1.25W/1.3V= 960mA. Don't forget the TJmax is based on a TJ of a 150C. If you run a device at it's ultimate thermal limit you are actually performing a HALT test. That's Highly-Accelerated-Life-Test. So, don't do it.
If you take the VCEsat @0.45V and I_max @ 1A that power level, 450mW, is below the pkg rating even for the SOIC pkg of 625mW. The PDIP-8 and DFN have power ratings of 1.25W.
I seem to remember if you have a very expensive load that can be destroyed by an over-voltage occurrence, then for heavens sake use a crowbar.
Thx for reading.
That's all I got for now.
Again, thx Clive.
This looks like the kind of chip that'd be fairly easy to make the building blocks of with discrete components. That'd be a fun project and learning tool :D
For extra bonus points, try doing it with valves!
Yes it is. Or you can also try the TL494.
Jack st clair Kilby invented the computer chip ( IC ) August 1958, won the Nobel Prize in physics, 186,000 dollars, and the Nobel Prize is now called the KILBY AWARD , and Texas instruments who he worked for built a museum called the Kilby museum, in Texas. 🤔👍
one of these blew up my bluetooth speaker. they should not be allowed to make something that fails into an overvoltage state. i think it gave it full 12v
I just happen to have two of these 12/24V->5V device chargers on my desk, so I took one apart while watching 🙂. I must have an older version of the PCB (silkscreen version: LC-022?). D1 is populated (not on yours), but I don't have D2 on the PCB (efficiency diode). All other components seem similar, including the use of the MC34063.
With those devices can also create havoc with CAN Bus system by creating false error messages as I discovered with my Smart car. It gave indications that CPU had failed I turned the ignition off removed the charger and everything went back to normal and no more error messages. A great sigh of relief because I was some distance from home and the nearest Smart car service agent was about 60 to 70kms away.
Radio frequency interference is a common issue with this device due to the way it regulates, and the lack of any additional filtering in these cheap and nasty adaptors.
@@ferrumignis Agreed about the RF interference and really surprised at the havoc it caused. It makes you wonder what other problems a poorly surpressed device can cause. It's annoying that manufactures don't prosercuted for bad RFI interference and possible damage it can cause.
Rsc perhaps may moderate the operation of the oscillator to either shut it down on excess current draw or to vary the frequency of the oscillator as the load changes to better keep up with a changing load.
Amazing R&D on a worthless charger. You deserve a Nobel prize nomination.
This circuit exists to really slowly charge your phone whilst interrupting the radio reception
Excellent stuff once again clive. Love the new bench 👍.
A part I used years ago for a nixie clock HT power supply: 12V in, 180V out. The data sheet says that's too much, but it's worked for years... These days I'm looking more at buck than boost and efficiency is a priority, but it's nice to know this is still around.
Neat! I have that exact device here somewhere too. Same case and all. Sticker on the front is slightly different. I don’t even remember buying it.
Like the 555, not only useful and versatile, but also cheap as chips!
I see what you did there!
Uses a fixed frequency with a variable pulse width. You can modify for MUCH higher current with a mosfet follower and a bigger inductor. One thing I have used these for is for making constant voltage power converters for use with solar panels.❤
Many thanks for another interesting and enlightening video, Sir!
Awesome to see older tech still being used nowadays
Things like this are why I have bins of old components and boards for scavenging parts for my home projects.
Everytime i see the pictures come on the screen I remember DrunkClive trying to measure voltage on it. ❤
On is Motorola's chip division, just as NXP is Philips . So they are the original company with a new name and logo . Note that NXP was recently split into Nexperia and NXP as part of a failed attempt to merge with some foreign company .
I saw "iconic component" and guessed a 555. At least I was right that the 555 is possibly the most iconic component! (Though I think a 555 with a big inductor could probably work as a variable voltage divider?)
😎,cool video mate👍, my favourite classic power supply IC is the TL494 ... 😁
3:48 Would this still be a disaster with USB C PD capable devices? Since they can accept higher voltages though I have no idea what happens if a PD capable device was expecting 5V only to suddenly get 12 since it is more than 9 but less than 15, hopefully just blow a fuse on the charge port.
I think a new device would handle the higher voltage with no issue.
Yes .very good explanation of product.I live off grid Solar power 12 volt.Use 1 amp .2 amp use charger for Android device,s..Now available in Australia a 36 amp 12 volt charger .
The data sheet is more than a little vague about it but I believe that the 0.22 Ohm resistor sets the peak current limit at 0.3V/0.22. When this limit is hit, the switch is turned off early in the cycle. This almost protects the chip. If the inductor is a really good one, the catch diode is a good one and the output has a solid short, the inductor current will get kicked a little higher on each cycle until it burns the chip out. For cheap stuff the solution to this is to make the inductor not so good so that cycle to cycle the current runs back down to near zero.
For better protection, there is a trick where you detect the extremely short ON time and tell a lie to the voltage comparator. Basically the circuit listens to the voltage on the 0.22 and if it sees a really sudden increase in the drop, a PNP turns on and charges up a capacitor that times the delay until it tries again.
Yes, you're right, this DID blow up my phone!
Im telling my kids this is scott manley
Lots of people think I am.
Clive will you look into the current sensing / timing capacitor part and let us know. It’s really bugging me not knowing how it works now. 😂 Plz Plz
Ooof. Switching voltage regulator at it's defined use... yet, quite an ingenious use of the chip.
Interesting chip design. I wonder if its possible to make a bit bigger discrete components version. 🤔
Wow, used since 1983? That's an impressive bit of design work
I encountered an elastomeric (zebra) conductor today when opening up a cheap combination hygrometer/thermometer (got a dozen for less than $20). I was surprised they existed, let alone already a somewhat old technology!
I still use this chip for personal projects. I can fill in the pin numbers off the top of my head.
A recent project uses 2. With 12V in, one makes -12V, the other with an external FET, makes +250V, for tube projects.
The voltage regulation is achieved by the action of the comparator. When the input of pin 5 gors above 1.25v the output of the comparator switches off and that in turn shuts off one input the the AND gate driving the set input of the RS ff. This basically shuts off the oscillation driving the output circuit until the voltage at pin 5 drops below 1.25v at which point the chip begins pulsing the output and the series inductor to build the output voltage again. Etc. Etc
Your electronic photographs are gorgeous, you should sell them as prints
Using an inductor/capacitor pair like that, instead of a capacitor is particular unusual. Most newer power supply chips, be them linear or switching use only a capacitor for the power draw line.
If they use an inductor it is a small choke chosen just to smooth ripples.
When I saw the charge adapter and title, I had a feeling this would be about the MC34063. One of Dave Jones's earliest videos is a tutorial on all the maths for building a buck converter with this chip.
ua-cam.com/video/qGp82xhybs4/v-deo.html
Got one from a dollar store, was missing the output capacitor, lucky I had one. still going strong.
I love you videos I wish I knew what you were talking about in extreme detail I want to learn!
I recently dismantled one of these myself, actually yesterday, it has a CX8812 smps chip, these are higher current devices and the inductor was a toriodal type.
ON Semiconductor was spun off from Motorola in the late 90s. Some of their part numbers on older components retained the old "MC" numbers.
I worked for both MOT and ON