Clive, it's not an unusual arrangement to interleave the windings by splitting the primary like that, it's actually very common. That technique maximises coupling and lowers leakage inductance, meaning the snubber network can be more modest as high voltage spikes on the transistor are reduced. It's important to use this technique on flyback style smps designs to avoid transistor/FET failure. Love the videos, keep them coming!
I love the built in mystery feature - an LED to show that it's on, hidden in an opaque case so only insects that somehow manage to get in the case can see it.
Common practice to place the secondary between a split primary it improves flux fits the windings and improves the efficiency in high frequency transformers. You should rewind it preferably with litz wire this will more than likely increase the dc resistance but inefect reduce the ac resistance or you could use a high voltage enameled wire if isolation is your primary concern. You won't have the space for much else.
@@keithking1985 I do understand that Acme's post has offended you, but that doesn't mean that you should insult him as well. A Litz is a braided wire consisting of several lacquer coated/insulated thin strands. This (according to Wikipedia) reduces the losses of high frequency currents.
+Nick Moore Technically between you and certain death is standing 7μm (7.5μm) of insulation on wire (Grade 1 magnet wire, I doubt they used higher grade), tape there is only as secondary insulation (and it is much thinner than 1 mm, more like 0.05mm), but that should be enough. Tape there is only so they can meet Class II (double insulated) standard (theoretically they could use one layer). 7 μm of insulation on magnet wire have breakdown voltage of over 1400V (IEC 60851.5.4.2).
The lack of insulation between windings is certainly worrying in view of the fact that the primary will have considerably more than the mains voltage on the collector end. Calling this Class 2 double insulation is a very doubtful claim. It's interesting to contrast this situation with a couple of UK electricians debating lengthily about whether an added 13A socket will comply with voltage drop, prospective fault current, earth resistance, cable temperature and safe zone routing requirements. And, ending-up using cable of treble the required rating just to be sure they are in compliance will all the regs. After which they decide that the back-box needs an earth wire adding in spite of already having earthed screws, and a separate 6mm earth wire run to an adjacent radiator pipe as well, just in case, complete with safety warning tag. Then realising the socket is 1.2mm nearer the floor than the disability access guidelines allow, so the whole thing has to be done again. After which a series of tests with specialist equipment is made and test certificates issued, then paperwork made out for filing with the local authority under the terms of the Part P Building Regulations. Then, someone comes along and plugs in an item like this.
As others have already stated, the split primary is actually very common on the “better” power supplies. The coupling improves efficiency and with the proper optimized ratio of turns between the two sections balances transient noise. One item that seemed to go without notice is that the circuit type is called Flyback. That means the stored energy in the core is mainly transferred to the secondary when the primary is switched off, (with a small portion dissipated in the snubber). That along with the turns ratio and the snubber network determines the raw secondary voltage and finally the optocoupler loop feedback fixes it. With this principle you can get universal primary voltage. Would not happen with Forward Converter without more complex PWM control. You can see the difference in the winding direction between the primary and secondary, which would also have deserved a note pair of Dots on the schematic.
That's quite a clever (and low cost) way of doing the feedback. For extra stable regulation, you could use a TL431 to control the LED for a precision reference. That zener will increase the output voltage as it heats up with time. Could be a bit dodgy.
I was hoping for a test, with or without flames. Interesting circuit, but seems good enough for what it does. Even some load tests would be in order. Good job of figuring it all out, it's not the most straight forward circuits.
The easiest way I've found to open devices like those without damaging the cases is this. Take a folded hand towel and place it on a hard floor. Take a hammer and wrap+tie another hand towel around it. Hold the device with one corner on the hand towel you placed on the floor and hit the opposite corner. Tap a few times and then rotate to apply force on the other two corners. You want to hold the device so that it is the tallest it can be so that the force is applied across opposite corners when hit with the hammer. You are basically turning the rectangular section that holds the two halves together into a parallelogram each time you hit the opposite corner with the hammer which will loosen the bonds in the plastic. It may take a few extra minutes but you can usually get them apart without any tool marks along the sides.
Oh yeah. I've watched a ton of your teardown videos, and this is the first one I've seen that has this type of power supply. I tore down one cheap chinese power supply (seemed otherwise good quality but had the same issue with poorly insulated transformer) and tried to figure out how it works a few months ago, but couldn't really. It was similar to this (same kind of control circuit), but for 5 volts, and lacked some features (such as the noise suppression choke at the input). Only now I realize how the circuit functions. (I know the basics of switched mode power supplies, but there are so many different designs of them...)
Hi bigclive, I love these "switching converter porn" videos of yours; I seem to have an insatiable appetite for seeing what these look like inside and understanding how they work :) The topology of this supply is the so-called "ringing choke" or sometimes "self-oscillating" converter. It's main distinction is that it's the lowest cost to build. The downside is that the switching frequency is not fixed, so the EMI filtering is harder to do (can't be tuned to maximize noise reduction, the noise harmonics keep moving). Also, it's only good for modest power levels. It seems to be pretty common in the "bargain" class of switching adapters. I did a full analysis a couple weeks back, you can see it here on EE.SE if you're interested: electronics.stackexchange.com/questions/214101/how-does-this-wall-wart-switcher-work. But I essentially ended up with exactly the analysis you did and the same open questions, e.g. "does the opto make it skip cycles or somehow modulate the length of the on-stroke?". I'll be watching all the rest of these you've done that I can find. Thanks a million for doing these :)
Wow, what service. I only just asked for the turns info on your last tear-down (ikea 3 output unit) 5 minutes ago. I wonder if the primary-secondary-primary sandwich has any filtering quality to it? Do you ever use phase dots to indicate polarity of coil/x'former windings?
Just realised I have a 5v / 2amp version of one of those, still wrapped up! :D I can't even remember what I bought it for :( I've always been a little cautious with the quality of these things though and so I do things like buying old internet routers JUST to get the power supply (which is usually of a good standard), for example the BT homehubs have 9v and 15v supplies and Netgear use some very stable 12v switchers. When I can't get a decent quality PSU I strip them down and wire in a thermal fuse wrapped in heat shrink, maybe it's just me, but I think all power supplies should have a thermal fuse in addition to any protection already built in.
Interleaving primary and secondary such as splitting the primary and sandwiching the secondary in the middle e.g. Primary-secondary-primary-aux is quite a common winding layer configuration in SMPS transformers. Better coupling from primary to secondary, less leakage inductance, less energy lost in the primary RCD snubber, higher efficiency etc. The downside is as you now have 2 insulation layers where primary and secondary meet the materials costs for insulation tape etc go up and the window fill factor decreases, but since they didn't bother with proper insulation here it probably doesn't matter to them. I'm slightly suprised they bothered in this cheap RCC flyback wallwart though. As you know they often don't and they didn't bother with proper primary-secondary insulation in the transformer. They must have a good reason to do so other than trying to slightly improve efficiency.
Clive when do you let us see that you after unwinding the transformer to rewind it as you would do that would be interesting to see and also very educational that would be my ultimate video 👌let me hear what you wonna do or ask the people what they think greetings from the Netherlands keep up the good work love your channel Dinant de Kruijff
Hehe, I picked up a handful of those screwdrivers from a big bowl by the till at Maplin. At 49p a pop, it was rude not to. I was super surprised when I found out they came with a bunch of bits in them too...bargain! At that price, not the kind of tool you want to use on your smartphone perhaps, but something one doesn't mind wrecking in the process of an aggressive dismantlement :-) And they're *so* light! :-D
Nice diagnostics of a power supply, thumbs up, it looks like this is also re-purposed power supply boards as there's LED that is not used. The LED inside the optocoupler is usually IR type with soft translucent silicon material between LED and the transistor. Have you tested how low DC voltage it starts to oscillate? The circuit looks similar to my PSU that started to oscillate with 4VDC input to mains and had correct USB output voltage with 12VDC supply. (It's quite interesting to see output LED glow with 9V battery stuck against the mains prongs, very different characteristics compared to linear power supplies.)
I have one of the supplied with a 5m strip i bought from a uk seller, they have a weird tendency to flash the led strip briefly about 1-2 seconds after turned off
+jusb1066 That seems to be quite a common effect. I think it's the result of the circuitry shutting down, but the residual charge in the electrolytic causing the boot circuitry on the switchmode driver to have another go.
I always wonder if the spec's on eBay wall wart power supply's are accurate especially if they've been repurposed, was the 2nd one you had exactly the same?
+Green Silver Both were identical and of dubious transformer isolation quality. And yet two others from different sources had random re-purposed PCBs in them that seemed to be from higher quality equipment.
+Slot1Gamer Sadly no. I wish it was Kapton. It;s a fairly generic tape that seems to be purely for use in these applications. It's barely a cut above dollar-store Sellotape.
+SuperBoobaloo It is indeed a lottery if you buy from China. That said, the ones for kitchen LED lighting in in-line plastic cases (I've reviewed three on here) did seem to have proper isolation. For plug-in packs you really need to go for a prominent brand in your home country. That does mean paying a more realistic price though. But it's crucial if you are supplying the power supplies to others.
iv seen the same windings on transformers teardowns on "Diode gone wild's" channel, where the secondary is between two of the primaries windings but not centre tapped to the board!!! but always on cheap chinese PSU's...
the lines on the one side reminds me of the first hard drive shove in style from the top bays I ever bought (alas only goes to a TB but hey who thought we'd need more :P) Never had a power supply go bad with those. Never took one apart don't seem to remember an LED on that but I have taken some apart and have found LED's that indeed glow . :( kinda like some usb sticks that should have an LED, and you take one apart and well gosh darn it , it has an LED. Course the reason for taking it apart was well it fell apart. Toshiba 3.0 16 gb stick literally fell apart. Nice Blue LED. didn't last long either and Toshiba won't honor a 3 year warranty sigh. (unless I call a # they never answer. smh) Usually that means Toshiba didn't actually make the stick but who knows. Has a Toshiba chip on it, but it doesn't look right (bent). Toshiba usb flash sticks overall are some of the worst I've ever come across so not a big surprise. Just never had one go from 16 gb to 15 mb before. :( Lost 16 gb of important info. Never doing that again.
I try not to use a small screwdriver to open stuff. Many years of jabbing myself with sharp pointy flat head bits. I find the larger bits dont ruin the case as much too.
+thingyee1118 you can get a "phone repair kit" or something similar really cheaply off ebay these days that have like 2 metal spudgers, 2 plastic spudgers a pair of sharp pointed tweezers, 6 mini screwdrivers including torx etc for about £4. That's what I use. The tweezers are actually really suprisingly good quality, comparable to my branded ESD safe tweezers.
Yes I own a case opening pry bar thing. Basically the phone repair kit. Works well for certain cases and saves painful fingernails. I generally use large flat head screwdriver though as I can twist it to pop it open. Oh the electrical / CAT rated screwdrivers are not as strong as mechanical one. They are quite soft.
Side thought. Is there much scope for swapping the zener with a different value to get a lower output voltage? Or is the turn ratio of the transformer the bigger player.
+James Sleeman I'm not sure how optimised the transformer is for the intended load, but I'd guess that changing the zener on the secondary side would lower the voltage that the opto-isolator cut in. You would probably still be limited to the 2A output if the supply is indeed capable of that.
+bigclivedotcom IME you can tweak them a bit but not say 5V to 12V, and the voltage regulation feedback is usually combined with current foldback which complicates things.
Yes, but more than a couple of volts either way and you'd need to add or subtract turns from the transformer secondary and maybe the feedback winding. Also you have to be careful not to exceed component tolerances like output filter caps and diodes. As the output voltage changes, so does the feedback voltage, which must be in a range that will power the switching circuit successfully. That is often overlooked!
+James Sleeman Based on tonight's fantastic teardown of a couple of single transistor switchmodes the scope for changing the voltage within reason by changing the zener is high. The wacky circuit design means that the capacitor on the primary pretty much mirrors the one on the secondary.
Did the calculations based on windings. Primary (first winding) = 32, Secondary (second winding) = 11, Primary (third winding) = 46. So Primary winding is 78 and secondary is 11. That means the voltage conversion is 16.94 Volts at 1.54 Amps. I noticed those supplies show ratings of 12 Volts at 2 Amps, which makes sense if they drop voltage it will increase Amperage. If we apply a factor of 75% from original figure it comes to 12.7 Volts at 1.925 Amps. I am curious to know if you attached a DMM and it showed 12.7 Volts as many do register right around that figure. Interesting transformer design....
Sierra Juliet -- Say what? (Root2 x 100 VAC - 2 volts for the bridge) x 11/78 = 19.66 volts max on the secondary winding, when the unit is lightly loaded. But the main input capacitor is so small that at 100 VAC input, I don't see how the unit can carry a 2-amp load. Running on 50 Hz *and* 100 VAC makes matters worse, as the main cap will only charge to about 139 volts and store 97 millijoules per 10-millisecond halfcycle, but it needs to deliver over 24 millijoules per millisecond to maintain the rated output voltage and current. While direct feedthrough to the transformer will work for part of each halfcycle, reducing the energy storage needed, the prospects for full rated output does not look good. Running on 60 Hz and 120 VAC is slightly better, as the main cap will charge to about 168 volts and store 140 mJ, enough energy for 5.8 ms of the 8.3 ms halfcycle. If run on 256 VAC, the main cap will charge to about 360 volts and store 648 mJ, which is enough energy to maintain the output for close to 27 ms. Maybe +bigclivedotcom would check my math? It would be interesting to see how the output regulation holds up while running the input through a variac.
I used to get the same kind of case allbeit with a euro plug with wifi routers and hard disk cases branded deltaco here in sweden. (Generic chinese stuff)
Do you know of any use (reuse) for a 110vac 12 to 22 watt fluorescent ballast, clive? (I'm rewiring the light to run on LED 3528 tape 2700k... also do you think 4000k is better for photography & human health -- I saw LED wholesalers call it "Neutral white" and I'd never heard of that. I like 2700-3000K myself but only because most I see are those nasty 5000K and horribly blueish -- The kind that look like they should only be used as parking lights.) I've watched a lot of your vids some time ago and you seem one of top 3 most knowledgeful on LED lights. You could surely school at least one of my electronic's college professors. When I have the time you're enjoyable to watch. Keep up the good work.
two primary layers with the secondary in the center is quite common. this is to reduce leakage inductance. the usage of multi strand reduces skin impedance. PS insulation is pretty bad in this unit you have there. i even think its illigal to sell it like this
You should buy one and take a look at it as it's really bad. [Heck i'd say it's unsafe.] Here is the one i bought [I got the 2Amp model]: www.ebay.co.uk/itm/1A-2A-3A-5A-6A-8A-10A-Power-Supply-Charger-AC-100-240V-DC12V-For-Led-Strip-Light/371656815343?ssPageName=STRK%3AMEBIDX%3AIT&var=640649573013&_trksid=p2057872.m2749.l2649 I have a photo of the insides of one and the new ones look nothing like the old ones.
Clive, it's not an unusual arrangement to interleave the windings by splitting the primary like that, it's actually very common. That technique maximises coupling and lowers leakage inductance, meaning the snubber network can be more modest as high voltage spikes on the transistor are reduced. It's important to use this technique on flyback style smps designs to avoid transistor/FET failure. Love the videos, keep them coming!
I love the built in mystery feature - an LED to show that it's on, hidden in an opaque case so only insects that somehow manage to get in the case can see it.
Excellent video, yet again. Interesting and informative. Keep up the good work, Sir!
Excelente schem analysis, thanks for sharing I feel more confident on smps day by day
That was truly a pleasure to watch, thank you.
Nice video, Clive. I enjoyed it a lot and learned something besides!
Very well explained! Appreciate the effort you are taking for giving such an awesome shower of information in your videos. :)
Thoroughly enjoyed your teachings.
Common practice to place the secondary between a split primary it improves flux fits the windings and improves the efficiency in high frequency transformers. You should rewind it preferably with litz wire this will more than likely increase the dc resistance but inefect reduce the ac resistance or you could use a high voltage enameled wire if isolation is your primary concern. You won't have the space for much else.
+Peter Jones Is this really worth bothering investing in to rewind with expensive Litz etc?
What is litz wire????
@@keithking1985
God gave you Google and a brain (possibly), so use them.
@@acmefixer1 FUCK YOU AND ALL BELONG TO YA!!! THAT'S WHAT THE COMMENT SECTION IS FOR.... DICK HEAD.....
@@keithking1985 I do understand that Acme's post has offended you, but that doesn't mean that you should insult him as well. A Litz is a braided wire consisting of several lacquer coated/insulated thin strands. This (according to Wikipedia) reduces the losses of high frequency currents.
Excellent, thank you Mr. Clive.
All that stands between us and certain death (of our phones and cameras) is 1mm of chinesium tape. What an age to be alive!
+Nick Moore Wow, I never thought of it like that, I'll get a rush every time I plug my phone in now!
+Nick Moore Technically between you and certain death is standing 7μm (7.5μm) of insulation on wire (Grade 1 magnet wire, I doubt they used higher grade), tape there is only as secondary insulation (and it is much thinner than 1 mm, more like 0.05mm), but that should be enough.
Tape there is only so they can meet Class II (double insulated) standard (theoretically they could use one layer).
7 μm of insulation on magnet wire have breakdown voltage of over 1400V (IEC 60851.5.4.2).
*+Nick Moore* You have to be utterly clueless to think the Chinese invented capitalism.
Awesome, always wondered what the theory of operation of this style of "unintelligent" switcher is.
So educative may GOD bless you
The lack of insulation between windings is certainly worrying in view of the fact that the primary will have considerably more than the mains voltage on the collector end. Calling this Class 2 double insulation is a very doubtful claim.
It's interesting to contrast this situation with a couple of UK electricians debating lengthily about whether an added 13A socket will comply with voltage drop, prospective fault current, earth resistance, cable temperature and safe zone routing requirements. And, ending-up using cable of treble the required rating just to be sure they are in compliance will all the regs. After which they decide that the back-box needs an earth wire adding in spite of already having earthed screws, and a separate 6mm earth wire run to an adjacent radiator pipe as well, just in case, complete with safety warning tag. Then realising the socket is 1.2mm nearer the floor than the disability access guidelines allow, so the whole thing has to be done again. After which a series of tests with specialist equipment is made and test certificates issued, then paperwork made out for filing with the local authority under the terms of the Part P Building Regulations.
Then, someone comes along and plugs in an item like this.
It is quite normal to split the primary into two windings with the secondary in between. This is to get better coupling between the windings.
As others have already stated, the split primary is actually very common on the “better” power supplies. The coupling improves efficiency and with the proper optimized ratio of turns between the two sections balances transient noise. One item that seemed to go without notice is that the circuit type is called Flyback. That means the stored energy in the core is mainly transferred to the secondary when the primary is switched off, (with a small portion dissipated in the snubber). That along with the turns ratio and the snubber network determines the raw secondary voltage and finally the optocoupler loop feedback fixes it. With this principle you can get universal primary voltage. Would not happen with Forward Converter without more complex PWM control. You can see the difference in the winding direction between the primary and secondary, which would also have deserved a note pair of Dots on the schematic.
That's quite a clever (and low cost) way of doing the feedback.
For extra stable regulation, you could use a TL431 to control the LED for a precision reference. That zener will increase the output voltage as it heats up with time. Could be a bit dodgy.
I was hoping for a test, with or without flames.
Interesting circuit, but seems good enough for what it does. Even some load tests would be in order.
Good job of figuring it all out, it's not the most straight forward circuits.
Thanks for the upload. ps You can soak the seam in petrol with a rag or brush. Then it should come apart.
The easiest way I've found to open devices like those without damaging the cases is this. Take a folded hand towel and place it on a hard floor. Take a hammer and wrap+tie another hand towel around it. Hold the device with one corner on the hand towel you placed on the floor and hit the opposite corner. Tap a few times and then rotate to apply force on the other two corners. You want to hold the device so that it is the tallest it can be so that the force is applied across opposite corners when hit with the hammer.
You are basically turning the rectangular section that holds the two halves together into a parallelogram each time you hit the opposite corner with the hammer which will loosen the bonds in the plastic. It may take a few extra minutes but you can usually get them apart without any tool marks along the sides.
10100rsn Great tip, I'll try this next time.
Oh yeah. I've watched a ton of your teardown videos, and this is the first one I've seen that has this type of power supply. I tore down one cheap chinese power supply (seemed otherwise good quality but had the same issue with poorly insulated transformer) and tried to figure out how it works a few months ago, but couldn't really. It was similar to this (same kind of control circuit), but for 5 volts, and lacked some features (such as the noise suppression choke at the input). Only now I realize how the circuit functions. (I know the basics of switched mode power supplies, but there are so many different designs of them...)
AFAIK, the double primary winding increases coupling between primary and secondary. ;)
Hi bigclive, I love these "switching converter porn" videos of yours; I seem to have an insatiable appetite for seeing what these look like inside and understanding how they work :) The topology of this supply is the so-called "ringing choke" or sometimes "self-oscillating" converter. It's main distinction is that it's the lowest cost to build. The downside is that the switching frequency is not fixed, so the EMI filtering is harder to do (can't be tuned to maximize noise reduction, the noise harmonics keep moving). Also, it's only good for modest power levels. It seems to be pretty common in the "bargain" class of switching adapters. I did a full analysis a couple weeks back, you can see it here on EE.SE if you're interested: electronics.stackexchange.com/questions/214101/how-does-this-wall-wart-switcher-work. But I essentially ended up with exactly the analysis you did and the same open questions, e.g. "does the opto make it skip cycles or somehow modulate the length of the on-stroke?". I'll be watching all the rest of these you've done that I can find. Thanks a million for doing these :)
I want to see you wind the transformer back
+MrBuck295 just play the video backwards. You wouldn't notice a difference :)
+El Neuro Quila Hey hey , never thought of that big thank you ;)
very informative video thumbs up
Hi! Wouldn't it be nicer if the interference capacitor was after(not before) the common mode choke forming a low-pass filter?
Wow, what service. I only just asked for the turns info on your last tear-down (ikea 3 output unit) 5 minutes ago. I wonder if the primary-secondary-primary sandwich has any filtering quality to it? Do you ever use phase dots to indicate polarity of coil/x'former windings?
Just realised I have a 5v / 2amp version of one of those, still wrapped up! :D I can't even remember what I bought it for :(
I've always been a little cautious with the quality of these things though and so I do things like buying old internet routers JUST to get the power supply (which is usually of a good standard), for example the BT homehubs have 9v and 15v supplies and Netgear use some very stable 12v switchers.
When I can't get a decent quality PSU I strip them down and wire in a thermal fuse wrapped in heat shrink, maybe it's just me, but I think all power supplies should have a thermal fuse in addition to any protection already built in.
What can break if the optocoupler is installed in the wrong polarity ?
I don't find the problem on a led driver which have this problem...
Interleaving primary and secondary such as splitting the primary and sandwiching the secondary in the middle e.g. Primary-secondary-primary-aux is quite a common winding layer configuration in SMPS transformers. Better coupling from primary to secondary, less leakage inductance, less energy lost in the primary RCD snubber, higher efficiency etc.
The downside is as you now have 2 insulation layers where primary and secondary meet the materials costs for insulation tape etc go up and the window fill factor decreases, but since they didn't bother with proper insulation here it probably doesn't matter to them.
I'm slightly suprised they bothered in this cheap RCC flyback wallwart though. As you know they often don't and they didn't bother with proper primary-secondary insulation in the transformer. They must have a good reason to do so other than trying to slightly improve efficiency.
I like watching your vids...
I just realized you are on the Isle of Man - the THREE LEGED TRANSISTOR - "Quocunque Jeceris Stabit"
They wind the primary in two sections to achieve better coupling between the primary and secondary while reducing EMI.
Less leakage inductance if you split the primary that way.
Haha, who remembers UK "lucky bags"? It's an age thang
+Markus Bates I used to hound my Dad for one every time he took me into the local newsagents in Brum. They were the days
Clive when do you let us see that you after unwinding the transformer to rewind it as you would do that would be interesting to see and also very educational that would be my ultimate video 👌let me hear what you wonna do or ask the people what they think greetings from the Netherlands keep up the good work love your channel Dinant de Kruijff
Hehe, I picked up a handful of those screwdrivers from a big bowl by the till at Maplin. At 49p a pop, it was rude not to. I was super surprised when I found out they came with a bunch of bits in them too...bargain! At that price, not the kind of tool you want to use on your smartphone perhaps, but something one doesn't mind wrecking in the process of an aggressive dismantlement :-) And they're *so* light! :-D
Makes a nice change to find one of those psu's that you can actually open without resorting to violence with saws and hammers.
I love the tool that you use to grab hold of tiny things like tape endings. What would you call it ?
***** thanks I've got them now
Nice diagnostics of a power supply, thumbs up, it looks like this is also re-purposed power supply boards as there's LED that is not used. The LED inside the optocoupler is usually IR type with soft translucent silicon material between LED and the transistor. Have you tested how low DC voltage it starts to oscillate? The circuit looks similar to my PSU that started to oscillate with 4VDC input to mains and had correct USB output voltage with 12VDC supply. (It's quite interesting to see output LED glow with 9V battery stuck against the mains prongs, very different characteristics compared to linear power supplies.)
Clive,how much is Diesel per Litre on the Isle of Mann?Sorry off topic but I'm curious.
I have one of the supplied with a 5m strip i bought from a uk seller, they have a weird tendency to flash the led strip briefly about 1-2 seconds after turned off
+jusb1066 That seems to be quite a common effect. I think it's the result of the circuitry shutting down, but the residual charge in the electrolytic causing the boot circuitry on the switchmode driver to have another go.
I always wonder if the spec's on eBay wall wart power supply's are accurate especially if they've been repurposed, was the 2nd one you had exactly the same?
+Green Silver Both were identical and of dubious transformer isolation quality. And yet two others from different sources had random re-purposed PCBs in them that seemed to be from higher quality equipment.
It is possible the windings were sectioned for the sake of better coupling
try using petrol / gasoline to (i think) dissolve glue/sealant used for the plastic enclosure
feedback secondary secondary primary.
ok. centre tab is often used for push pull supply's.
What is the yellow tape thats used on transformers? Is it some kind of kapton?
+Slot1Gamer Sadly no. I wish it was Kapton. It;s a fairly generic tape that seems to be purely for use in these applications. It's barely a cut above dollar-store Sellotape.
Do eleven turns of double-wire count as twenty-two turns (windings)?
No. Just a more flexible way of doing the eleven turns for higher current.
nice
Genius again. Hit the weak spots again and again. The Chinese will listen and learn.
Externally they look exactly like a cord I had for a flat bed scanner.
Do you have a link to the seller? Thanks
my same charger.. not working sir .. please tell me how to trace problem n rectify
धन्यवाद 🙏🙏
Clive - Any chance you can link us to a source of quality 5V 2A supplies? Buying power supplies seems to be such a lottery.
+SuperBoobaloo It is indeed a lottery if you buy from China. That said, the ones for kitchen LED lighting in in-line plastic cases (I've reviewed three on here) did seem to have proper isolation. For plug-in packs you really need to go for a prominent brand in your home country. That does mean paying a more realistic price though. But it's crucial if you are supplying the power supplies to others.
iv seen the same windings on transformers teardowns on "Diode gone wild's" channel, where the secondary is between two of the primaries windings but not centre tapped to the board!!! but always on cheap chinese PSU's...
the lines on the one side reminds me of the first hard drive shove in style from the top bays I ever bought (alas only goes to a TB but hey who thought we'd need more :P) Never had a power supply go bad with those. Never took one apart don't seem to remember an LED on that but I have taken some apart and have found LED's that indeed glow . :( kinda like some usb sticks that should have an LED, and you take one apart and well gosh darn it , it has an LED. Course the reason for taking it apart was well it fell apart. Toshiba 3.0 16 gb stick literally fell apart. Nice Blue LED. didn't last long either and Toshiba won't honor a 3 year warranty sigh. (unless I call a # they never answer. smh) Usually that means Toshiba didn't actually make the stick but who knows. Has a Toshiba chip on it, but it doesn't look right (bent). Toshiba usb flash sticks overall are some of the worst I've ever come across so not a big surprise. Just never had one go from 16 gb to 15 mb before. :( Lost 16 gb of important info. Never doing that again.
I try not to use a small screwdriver to open stuff. Many years of jabbing myself with sharp pointy flat head bits. I find the larger bits dont ruin the case as much too.
+thingyee1118 you can get a "phone repair kit" or something similar really cheaply off ebay these days that have like 2 metal spudgers, 2 plastic spudgers a pair of sharp pointed tweezers, 6 mini screwdrivers including torx etc for about £4. That's what I use. The tweezers are actually really suprisingly good quality, comparable to my branded ESD safe tweezers.
Yes I own a case opening pry bar thing. Basically the phone repair kit. Works well for certain cases and saves painful fingernails.
I generally use large flat head screwdriver though as I can twist it to pop it open.
Oh the electrical / CAT rated screwdrivers are not as strong as mechanical one. They are quite soft.
Side thought. Is there much scope for swapping the zener with a different value to get a lower output voltage? Or is the turn ratio of the transformer the bigger player.
+James Sleeman I'm not sure how optimised the transformer is for the intended load, but I'd guess that changing the zener on the secondary side would lower the voltage that the opto-isolator cut in. You would probably still be limited to the 2A output if the supply is indeed capable of that.
+bigclivedotcom IME you can tweak them a bit but not say 5V to 12V, and the voltage regulation feedback is usually combined with current foldback which complicates things.
Yes, but more than a couple of volts either way and you'd need to add or subtract turns from the transformer secondary and maybe the feedback winding. Also you have to be careful not to exceed component tolerances like output filter caps and diodes. As the output voltage changes, so does the feedback voltage, which must be in a range that will power the switching circuit successfully. That is often overlooked!
+ProdigalPorcupine Sorry- when discussing feedback voltage I meant the auxiliary winding that powers the switching circuitry.
+James Sleeman Based on tonight's fantastic teardown of a couple of single transistor switchmodes the scope for changing the voltage within reason by changing the zener is high. The wacky circuit design means that the capacitor on the primary pretty much mirrors the one on the secondary.
Did the calculations based on windings. Primary (first winding) = 32, Secondary (second winding) = 11, Primary (third winding) = 46. So Primary winding is 78 and secondary is 11. That means the voltage conversion is 16.94 Volts at 1.54 Amps. I noticed those supplies show ratings of 12 Volts at 2 Amps, which makes sense if they drop voltage it will increase Amperage. If we apply a factor of 75% from original figure it comes to 12.7 Volts at 1.925 Amps. I am curious to know if you attached a DMM and it showed 12.7 Volts as many do register right around that figure. Interesting transformer design....
Sierra Juliet -- Say what? (Root2 x 100 VAC - 2 volts for the bridge) x 11/78 = 19.66 volts max on the secondary winding, when the unit is lightly loaded. But the main input capacitor is so small that at 100 VAC input, I don't see how the unit can carry a 2-amp load. Running on 50 Hz *and* 100 VAC makes matters worse, as the main cap will only charge to about 139 volts and store 97 millijoules per 10-millisecond halfcycle, but it needs to deliver over 24 millijoules per millisecond to maintain the rated output voltage and current. While direct feedthrough to the transformer will work for part of each halfcycle, reducing the energy storage needed, the prospects for full rated output does not look good. Running on 60 Hz and 120 VAC is slightly better, as the main cap will charge to about 168 volts and store 140 mJ, enough energy for 5.8 ms of the 8.3 ms halfcycle. If run on 256 VAC, the main cap will charge to about 360 volts and store 648 mJ, which is enough energy to maintain the output for close to 27 ms.
Maybe +bigclivedotcom would check my math? It would be interesting to see how the output regulation holds up while running the input through a variac.
I used to get the same kind of case allbeit with a euro plug with wifi routers and hard disk cases branded deltaco here in sweden. (Generic chinese stuff)
Maybe the reason for the split primary is that it halves the potential across each of the windings? Just an idea.
+Allan Stirling It splits the mutual inductance over a larger area. It cuts down on inductive spikes between switching.
More simply method to open this package is , cover the P.S. by some cloth and smash few times by small hammer just in place where it welded together.
+ValsoRay Every thing opens better with "Mr. Smashy TM"
+ValsoRay Mike Harrison's method of wacking along the weld with a plastic screwdriver handle can work nicely and doesn't leave any scars.
+Graham Langley OK))
That red led would get nice and hot inside that case 🔥 and it's a dead give away that it's repurposed.
Do you know of any use (reuse) for a 110vac 12 to 22 watt fluorescent ballast, clive?
(I'm rewiring the light to run on LED 3528 tape 2700k... also do you think 4000k is better for photography & human health -- I saw LED wholesalers call it "Neutral white" and I'd never heard of that. I like 2700-3000K myself but only because most I see are those nasty 5000K and horribly blueish -- The kind that look like they should only be used as parking lights.)
I've watched a lot of your vids some time ago and you seem one of top 3 most knowledgeful on LED lights. You could surely school at least one of my electronic's college professors. When I have the time you're enjoyable to watch. Keep up the good work.
If you put those transformers in boiling water, the core pops right apart😜👍🏼
I find a dinner knife the best for popping cases big man
+joohop Most of my dinner knives are slightly bent at the end for that very reason.
+bigclivedotcom Ha ha haaaaaa Happy new year fella
graham , b, ham
two primary layers with the secondary in the center is quite common. this is to reduce leakage inductance. the usage of multi strand reduces skin impedance. PS insulation is pretty bad in this unit you have there. i even think its illigal to sell it like this
Your current 230vac or 115Vac
240V AC.
I guess not drilling a hole for the LED saves them 0.000047p per unit, though they could save anoter 0.000021p if they actually ditched the LED... :P
+twocvbloke Maybe not, if they are using the LED as part of the voltage regulation in conjunction with the zener diode.
+Firecul42 I don't think so, it's probably a board out of another PSU with a light-guide above the LED.
... Woo Hoo NO Adverts ...
I didn't see any insulation between the PCB and the joints on the back of the mains pins, which is worrying.
Sadly, all of the newer 12V 2A power supplys form ebay are all 100% guaranteed crap.
+Ben Chavez I halve the rating at best. They push them too far.
You should buy one and take a look at it as it's really bad. [Heck i'd say it's unsafe.]
Here is the one i bought [I got the 2Amp model]: www.ebay.co.uk/itm/1A-2A-3A-5A-6A-8A-10A-Power-Supply-Charger-AC-100-240V-DC12V-For-Led-Strip-Light/371656815343?ssPageName=STRK%3AMEBIDX%3AIT&var=640649573013&_trksid=p2057872.m2749.l2649
I have a photo of the insides of one and the new ones look nothing like the old ones.