The moment you said "Check out the links in the description" and I saw that manila background and the dual 2N3904 with the two 1N4148 forming the voltage divider I was like.. Yep, read that page already! That's a great reference for creating virtual grounds using essentially linear components and no switching or charge pumps.
@@greatscottlab......Dave Plumber from Dave's garage could use your help......he is the retired Microsoft software engineer and tons more cooler stuff.....he's got a problem with negative voltages and addressable leds
I am an electrician. Ive made my own way in my company by proving I could reverse engineer and repair the power supplies in our specialized computers. The main thing that breaks is -12v. Its been interesting seeing the different ways people get there as there are several generations of power supplies and each generation is completely different from the last
Another way to create negative voltage is also to: - use one diode as a voltage dropper, okay for low power IC that require small biasing to allow output voltage go to GND - use isolated power supply and connect them in series, for example use isolated 5V-5V module and connect it +out pin to -in pin to create +5V and -5V
@@vincentschumann937 yes...only good for line level stuff...op amp logic.... but enough to run headphones i guess...nothing more....or just a ref voltage for whatever your use case
Minor correction here... It's not when you "tie +12 and -12 rails together" that you get 24v, because if you did that then you would short out the power supply, possibly causing damage to it... Instead, it is rather when you measure across or put a load across the -12 and +12v output terminals of the power supply, that THAT'S when you get 24 volts.
important to note that the opamp in the last solution solves the problem of imbalance for small currents. Just to be on the safe side it is really useful to use a push-pull current amplifier configuration so the opamp doesnt get so stressed and the ground is more stable for a wider range of currents :)
Ever since I opend and old DVD player, I was curious about the -5 volts and -12 volts in the power supply. And thanks to you now I know why it was needed.
When joining 2 secondary coils of the same transformer, one must double check the correct sides are connected, or else you will end up with 2 separate positive outputs and a common ground. Further on combining coils of transformers... When combining 2 outputs of a transformer on both ends to increase the amount of current, you must do that after the bridge rectifier you should use for each... Learned that the hard way ;)
Why would you need to "When combining 2 outputs of a transformer on both ends to increase the amount of current, you must do that after the bridge rectifier you should use for each... Learned that the hard way ;)" If you don't swap the polarity of the windings it should just act as a primary with more thicker windings
@@alexstone691 Those windings are never exactly the same, so you get some strange currents flowing. When combining it from 2 separate transformers you must for sure match polarity and still may see the same 'strange currents'. When I started learning about electronics (begin '90s), it took me quite a few fuses to find out why they blew so quickly. Glad I used fuses in my designs back then ;)
actually, it will still work. It will even work off of one winding. The problem with this one-winding setup is that if the current consumptions on the two rails differ, that difference current will appear as dc across the winding, causing core saturation. That makes the transformer roar and heat up unreasonably.
You can also make a dual +/- power supply from a single output transformer. You just need to use a single diode for each side. The disadvantage of this solution is that each side will be effectively half-wave rectivied, so you need to use bigger smoothing capacitors, but the advantage is that single output transformers are much easier to find.
What about using two identical single output transformers and connecting their primaries in parallel, then connect the secondaries in series (essentially making a center tap) and connecting that to ground?
You stated that you want to teach how to build 'a dual rail power supply so that your students won't be clueless when such a requirement arises ... ' Your statement has helped me better understand what I most need and, most benefit from, in your excellent programmes. It is this: I most need some of the vocabulary used in areas of technology outside of my present knowledge. This ignorance excludes me from very much and from many areas of interest. An example from another area of interest to me is medical science. Thank you for the important work you do in bringing into my purview whole areas of facts that would be excluded to me but for your superb tutorials and the high level of language that you explain away. David Lixenberg
I'm not an electrical engineer, I'm a heating engineer actually, but I love watching your videos even though it's something I'm never going to need to do..👍
I am just a hobbyist. My passion for learning & how the individual components work together has increased from watching your Channel. I only do it in my spare time & have been at it for about 3 yrs now. I am starting to understand more and more as time goes by. I really appreciate you taking the time to explain to us how things work! I have learned, the knowledge of electronics is never ending! The other day I went to a local well-known television/ computer & phone repair shop, to see if I could buy a couple of individual electronic components? I was working on a project & I didn't want to wait 3 days on my order from Mouser. The technician looked at me like I was crazy! He said we don't replace individual Parts like that anymore. We just replace the entire board nowadays. I thought to myself.....ok, just about anyone can do that. That's not really repairing Electronics. That is replacing Electronics! Apparently, if I ever need to work at a job like that, I shouldn't have any problem qualifying
You'll probably find that even repair places that actually repair the boards probably wont have what you need for you project since most modern electronics use SMD components, they may have some through hole lying around for fixing older electronics but I wouldnt count on them magically having everything you need and I very much doubt they would have components unrelated to TVs, computers or phones, so probably would not have variable resistors, usable displays (they would have ones for phones but they are pretty much unusable), motor drivers, leds, anything like that. If they wanted to sell components to hobbyists then they would open a shop not a repair business, it is probably a bit of hassle for them to sell to you even if they have what they need. The employee would probably have to ask their boss, and then find out how much it costs them and then have their boss decide how much they will charge you for it, essentially it could take a while and be very inconvienient to them because they arent set up as a shop to sell components to hobbyists. It isnt cost effective for a lot of these repair shops to actually fix the board, it is cheaper and faster to just replace the broken parts and it means they dont need to hire someone who can replace SMD and BGA components. Also repairing the boards would take time to track down the exact cause of the problem, its a lot easier to discover the problem is with this board, rather than this exact component on the board. They would have to first identify the exact component on the board that is damaged and then replace it, which not that many people could do anyway, especially if it is very small SMD components or BGA. Then it may not be the only damaged part. They also need to have access to datasheets and schematics to fix all but the most basic problems, which the manufacturer wont publically release. Then there is the problem of sourcing the components, since some might not be publically available, or the ICs dont have any markings or dont show up when you google them. It is by far easier just to order the boards to replace them with. It also means they dont need to hire an expert to fix the boards. An expert would still take quite a long time to fix a phone for example, and in that time, a lower paid employee could replace the board and have multiple phones fixed. These repair workers that just replace the board do still have to have knowledge and experience of what to replace and why and have knowledge of lots of different devices, so just because you are a hobbyist doesnt make you automatically better than them. In a lot of cases mechanics are the exact same, they just replace parts, whether or not that is because they cant fix them themselves or not is irrelevant but most kinds of repair work is just replacing parts as quickly and cheaply as possible.
This simple charge pump circuit I used when I wanted to get an old scrap LCD matrix display from an old phone to work. With a polish datasheet and too much free time I was able to make a library for it, to use it with an Arduino. Only thing left was the negative voltage. So I googled and found this charge pump design which I put together with scrap parts from old circuitbords and it worked :D
The ICL7662 has been around forever and is an easy way to get a negative voltage. With just two added capacitors it will convert up to 20 volts to a like negative voltage. I have used it for many different circuits and it always works well. There is also an ICL7660 but it has a max 5 volt input. The output is not regulated so it will drop with load but it is good for low power circuits in the 50mA range.
If you could get the right boost converter you could probably invert the polarity of the output by reversing the input voltage connection. I've tested it in circuit sim and it works there
Xl semi provides a few ic's for positiv and negativ output. The xl6008 and the xl6019 are good enough, both have a higher output current and both are more stable. I have makes a breadboard power supply with the xl6008 ic, work pretty well.
My very first power supply I built back in 1983 (13)used a 555 for the oscillator and charge pump for negative rail. Built from striping down a car battery charger which the bridge rectifier had failed and been thrown out. I have used the charge pump for several negative rails. The idea came from an everyday electronics magazine.
Most switching power adapters are isolated in reference to AC power ground, and we can cascade such outputs and achieve positive/negative power rail. If you want a micro volt potential difference between the “center ground” of your DC power supply and the “earth ground”, connect them with a 1 M-ohm resistor.
Hahaha, this is exactly the kind of video I was looking for a year ago when I was working on my MIDI piano keyboard!! Bit late for my needs, but I'm glad it's out there for the next guy!
I build stringed instruments in my retirement, which led to preamplifiers for transducers such as piezo pickups hot glued to the soundboard. To build a preamp I selected opamps as the basis for the preamp and started experimenting with an lm741. For several days I could not figure out why I could only get the opamp to amplify the signal as an inverted signal. Finally it dawned on me that the ground they referenced in the data sheet was not a typical chassis ground and was actually as some call it a virtual ground in a dual supply power supply. To test my new assumption I placed two 6 volt battery packs in series and used for lack of better words the junction of the two packs as a reference for ground in my circuit. Wala, now the lm741 amplified both in inverted and non inverted modes. I see the dual battery packs in series as the simplest way to make a dual power supply that has no real down side I can see unless somehow you draw power from just one battery pack which then has the issue associated with a voltage divider when placing an extra current demand on just one rail. I love the scheme and can't help think that using two voltage regulators such as that output 5 volts in parallel and then placing their outputs in series would also be very effective and likely without the voltage divider issue because ultimately drawing from just one rail is drawing from the source for each voltage regulator.
One common solution is to add a push pull pair of transistors to the output of the opamp virtual ground. The transistors can boos up the amperage capability of the opamp and maintain the virtual ground under a higher load. Also they make small form negative voltage regulator chips now as well. I use one that takes in +5v and spits out -5v and they make them for other specific voltages. They are also very simple to use. Good job, just thought I would mention a few other options available as well
unbelievable, I'm working with amps and this useful video pops up. Love your work, so useful to me over the years and so clear and explanatory. I'm just sad that in my opinion the old intro was much more catchy!
Be careful on virtual ground approach. Never share the same ground with other circuits that uses the true 0V grounding. If you need a signal route to external circuit, use optocoupler or bypass capacitor to isolate the signal grounding.
The last one is probably the most useful to me. I’ll normally have access to a bench power supply or can wait for a mailed chip to do that power converting for me, but sometimes I just want to test something that involves an op-amp needing + and - rails using components I have just laying around. Knowing that an op amp in the middle can help stabilize is pretty useful.
you're truly great, scott! this indeed is a brilliant 0.10$ hack! I checked it on oscilloscope it works, and the ripple is no different than the regular positive output(20-50mV). didnt test yet on load.
So, if you have 3 output pins at 0V, 1V and 2Vs But decide to label them -1V, 0V and 1V, is that all a negative voltage is? You hook up your probe common to what you choose to be ground (0V) and everything else is measured relative to that point?
I was recently thinking of making a negative voltage supply by using a reversed flyback converter. I wonder if this would be able to provide a higher output current. Great video. I never knew of the charge pump hack.
Yeah flyback converters can be used, you only need to flip the secondary transformer to get a negative output There's a lot of flyback converters that include both a positive and negative branches, with 2 secondaries windings on the transformer and essentially the same circuit as the output
Fantastic video with fantastic information! I really loved using a highlighter to show the functional blocks of the circuit. Something that I could do myself when needed.
I use two buck converters in series. - Input negative become the target negative voltage - First stage output positive becomes the target positive voltage - Second stage positive becomes the target ground voltage There are two buck converters to adjust, which is a disadvantage and, for reasons that I did not investigate, the first buck converter heats-up like crazy and ends-up dying after a while.
Great info, I've got a waveform generator I built for a Uni project and have never gotten around to building a dual rail supply for it, of course at uni we just powered it off the bench supplies, wasn't until I took it home I realised I couldn't use it. Yet.
Can you just use 2 boost converters, set one to 12v and one to 24, then use the 12v output as virtual ground, the original ground as -12v and the 24v output as +12v? I guess you can't but why not?
Another option you didn't mention: Just use _two power supplies._ You need to make sure that the supplies you use can function without using a common output ground between them, but many power supplies have effectively isolated input and output grounds, so many actually can do this. Then you just connect the positive output of supply B to the ground from supply A, and B's "ground" output then becomes a negative voltage (relative to A's ground). This can even allow you to tune the size/type of the different power supplies to match the current (or noise, etc) requirements of each rail. You can also use complementary linear voltage regulators (i.e. a 7812 and a 7912) to maintain stable positive and negative voltages relative to a reference VGND...
Two supplies is an elegant and simple solution, although it's not based on a single circuit. I used this option when I was studying op amps and I had no time to build a dual-rail-supply, neither wanted to buy one.
for even higher current, you can add a pair of npn/pnp transistors (e.g. 2n3055 mj2555) on the output in half bridge (bases to op amp out , emitters together connected to op amp - and virtual ground. collector of pnp to V- collector of npn to V+) add a couple of smoothing caps between both rails and virtual ground. note that since you're basically burning the imbalance current, the transistors will heat up a lot in some cases (you're getting I*rail voltage, so better make sure there's adequate cooling).
I have just completed my school project a 350 W bluetooth speaker and the biggest problem for me was to make simetrical power supply(for supplying preamp) as i have only dc source and making the inverter for transformer is not efficient. I had found the same page that you have shown in the video and i made a divider with two transistors and everything works fine. Its interesting that you have uploaded this video now when I had already gone trough this all..
Of course, you could also add another power supply to add the necessary negative voltage with the needed current; if it's a lot of current and also loaded uniquely to the positive supply.
Such a good video that I needed some weeks ago when i designed a positive and negative voltage and now I have more info thanks. Another way is to use a Buck converter, if you see the schematics of a inverter converter and a buck convertee you will see the same image just flip so the lm2596 can generate negative voltage if you put GNd to output and Voltage to.input, it bootstrap the voltage so if you get -12V from 5V the input will see 17V that is the disavantage the limit would be the max input voltage of the IC but it is pretty straight forward to get negative voltage like that without soldering anything
You can actually use an LM2596 buck module as a negative voltage generator, and you can draw loads of current from it since it's not a charge pump! The great thing is that practically no hardware mods. Search for LM2596 negative voltage, and you'll find a Wordpress blog post that explains how to do it.
You can also use almost all Buck Converter Modules. Connect the + Supply with Vin of the Module, GND to Vout of the Module and You will get the negative Voltage at the GND-Pin of the Module. The Summ of Input Voltage and negative Output Voltage has to stay below the maximum allowed Input-Voltage of the Buck Converter Module.
@@juergenschimmer960 True, but I'm not sure if it can be accomplished with a synchronous rectifier where the "diode" is instead a MOSFET, and built into the chip. The other downsides are the large inrush current and the fact that the output tends to be noisier compared to when not generating a negative voltage. I've solved this in the past by adding another inductor in the series and an additional output capacitor, a pi filter.
Another approach is to use an adjustable buck boost converter with the positive output lead connected to ground. I used this setup for delivering negative voltage to a large LCD. Of course the converter has to have isolated potentials in order not to create short circuit.
This circuit is the Cockcroft-Walton circuit. It is slightly different from the generally said active charge pump circuit.Although it is a convenient circuit, the current carrying capacity is small.
Eigenkrafts!! Opamp Vgnd toll!!! I used two similar boost converter to create negative voltage in one of my skiffs, it uses the 12v out of one to ground input of the other, the inverted out of the latter gives neg rail. Plenty of noise anyway, not the best approach, but with some filters gets the job done very efficiently. Never thought of the pump solution, quite interesting!
1. Get a boost converter. 2. Replace its inductor with a center tab transformer.(use boost con as flyback) 3. connect the center tab to gnd.. 4. Do a positive and negative rectification with the other two ends of the tran's secondary winding. I've done it so many times. The negative output can do as the positive. Current limiting also works with negative output since most current sensers are placed in gnd.. DVD power supplies does it except step down( but only for referencing purposes)..
I use two diodes and two caps on the secondary of a single secondary transformer. It uses the upper half-wave to charge the positive al cap and the negative half-wave to charge the lower voltage rail cap. No need of center tap this way
There's also the Cuk converter which can achieve good efficiency and low noise on both the input and output. It's a good choice for noise sensitive applications like audio.
Fun fact. Telecom carriers use -48volt DC because it is easy to store in deep cycle batteries and can backup equipment during a power outages for much more time than a regular UPS.
7805’s output stage is not bipolar, so I’m not sure if you can sink significant current into its output without it misbehaving. If you want to use linear regulation, you’d be better off with a power opamp as shown in the video since these devices are designed to work correctly for both sourcing and sinking current.
Can we add another isolated center tapped coil with the inductor? Another solution is to boost the 12v in to 24 v and using the input positive (12V) as virtual gnd
Can it also be done with a voltage multiplier? a capacitor and diode arrangement to tap ground from between the two capacitors, a bit like the one you did with the resistors.
7660 charge pump plus a couple electrolytic capacitors is a simple & inexpensive way that can run on +1.5vdc to +10vdc (like +5vdc from USB) to generate corresponding negative DC voltage supply.
How about using a H-Bridge switching fast, filter output with diodes and smooth with Caps to create Pos and Neg supply? Would be able to get much more current out of something like that...
H bridges are good for reversing the direction of current like through a motor but they dont actually have a ground point that you could use and the voltage difference is only relative and between the two output terminals. It changes the direction of the potential difference and therefore current flow but it doesnt create negative voltage. If you measured the voltage between either output and the circuit ground, one would be at the supply voltage and the other at ground, there isnt any negative voltage there.
I tried making my own DC/DC supply once, taking ~14V and using a buck converter and a buck-boost converter to output +V and -V. I ran into issues with total current supply and in-rush current.
If you want to use the transformer alternative without using mains voltage you can use an oscillator from 0v to vsupply to convert the dc voltage to ac and then supply that to the primary of the transformer.
The easiest way to get a dual rail power supply is using one buck converter. supply 24v to it, set it to 12v positive. Now use the buck converter's positive (V+) terminal as your GND. Between V+ and the input (VIN+) of your converter there is 12v and -12v between V+ and GND of buck converter. You can supply as much current as your BC can handle, assuming your power supply can too.
The moment you said "Check out the links in the description" and I saw that manila background and the dual 2N3904 with the two 1N4148 forming the voltage divider I was like.. Yep, read that page already! That's a great reference for creating virtual grounds using essentially linear components and no switching or charge pumps.
That page is fantastic indeed :-)
. ....
haha my exact thoughts
@@greatscottlab......Dave Plumber from Dave's garage could use your help......he is the retired Microsoft software engineer and tons more cooler stuff.....he's got a problem with negative voltages and addressable leds
I am an electrician. Ive made my own way in my company by proving I could reverse engineer and repair the power supplies in our specialized computers. The main thing that breaks is -12v. Its been interesting seeing the different ways people get there as there are several generations of power supplies and each generation is completely different from the last
Another source of negative voltage that comes to mind is an inverting buck boost circuit commonly used in PC power supplies for the -12V rail.
though that seems only gd for pre amp...power amp....just to low power
Another way to create negative voltage is also to:
- use one diode as a voltage dropper, okay for low power IC that require small biasing to allow output voltage go to GND
- use isolated power supply and connect them in series, for example use isolated 5V-5V module and connect it +out pin to -in pin to create +5V and -5V
Old serial ports used that voltages, but in ATX power suppy its has easy obtained from a inverted diode on the 12V line right out the transformer...
@@servantsochrist the -12v rail is pretty low power (usually under 2A)
@@vincentschumann937 yes...only good for line level stuff...op amp logic.... but enough to run headphones i guess...nothing more....or just a ref voltage for whatever your use case
This is the only video in my feed that I don’t understand 90% of the content but like watching
When I was a teenager my mind was blown by a power supply I was working with and figuring out if you tie +12 and -12 rails together you got 24v
Yeah, your intuition would tell you +12 + -12 = 0.
24v rms peak to peak, but still
That's why you shouldn't become part of the circuit.
Minor correction here... It's not when you "tie +12 and -12 rails together" that you get 24v, because if you did that then you would short out the power supply, possibly causing damage to it... Instead, it is rather when you measure across or put a load across the -12 and +12v output terminals of the power supply, that THAT'S when you get 24 volts.
@@JoeJ-8282 Cmon man, that's just being pedantic. Everyone knows what I meant.
such a great deal. i sold my wife and my 6g72 engine and bought the altium designer. no brainer to me.
Which market did you use to sell your wife? I want to get as much as possible.
important to note that the opamp in the last solution solves the problem of imbalance for small currents. Just to be on the safe side it is really useful to use a push-pull current amplifier configuration so the opamp doesnt get so stressed and the ground is more stable for a wider range of currents :)
Ever since I opend and old DVD player, I was curious about the -5 volts and -12 volts in the power supply. And thanks to you now I know why it was needed.
Glad I could help :-)
Why was it? lol.
Also some old HP modems uses -78 volts, always wondered...
When joining 2 secondary coils of the same transformer, one must double check the correct sides are connected, or else you will end up with 2 separate positive outputs and a common ground.
Further on combining coils of transformers...
When combining 2 outputs of a transformer on both ends to increase the amount of current, you must do that after the bridge rectifier you should use for each... Learned that the hard way ;)
Thanks for the feedback :-)
Why would you need to "When combining 2 outputs of a transformer on both ends to increase the amount of current, you must do that after the bridge rectifier you should use for each... Learned that the hard way ;)"
If you don't swap the polarity of the windings it should just act as a primary with more thicker windings
@@alexstone691 Yeah but on some transformers the two center taps on the secondary side are connected to each other
@@alexstone691 Those windings are never exactly the same, so you get some strange currents flowing. When combining it from 2 separate transformers you must for sure match polarity and still may see the same 'strange currents'.
When I started learning about electronics (begin '90s), it took me quite a few fuses to find out why they blew so quickly. Glad I used fuses in my designs back then ;)
actually, it will still work. It will even work off of one winding. The problem with this one-winding setup is that if the current consumptions on the two rails differ, that difference current will appear as dc across the winding, causing core saturation. That makes the transformer roar and heat up unreasonably.
I was taking apart a humidifier, and I realized it had a negative supply. I had no idea how to make my project work, but your video helped me a lot.
Genau das richtige Video zum richtigen Zeitpunkt, wir nehmen im Studium gerade Op-Amps durch und die brauchen ja eine positive und negative Versorgung
Glad I could help :-)
You can also make a dual +/- power supply from a single output transformer. You just need to use a single diode for each side. The disadvantage of this solution is that each side will be effectively half-wave rectivied, so you need to use bigger smoothing capacitors, but the advantage is that single output transformers are much easier to find.
nice tip... gonna use that thanks
What about using two identical single output transformers and connecting their primaries in parallel, then connect the secondaries in series (essentially making a center tap) and connecting that to ground?
@@MuhammedGemci, I think that ought to work, but be careful with the polarity.
You stated that you want to teach how to build 'a dual rail power supply so that your students won't be clueless when such a requirement arises ... '
Your statement has helped me better understand what I most need and, most benefit from, in your excellent programmes. It is this: I most need some of the vocabulary used in areas of technology outside of my present knowledge. This ignorance excludes me from very much and from many areas of interest. An example from another area of interest to me is medical science.
Thank you for the important work you do in bringing into my purview whole areas of facts that would be excluded to me but for your superb tutorials and the high level of language that you explain away.
David Lixenberg
I'm not an electrical engineer, I'm a heating engineer actually, but I love watching your videos even though it's something I'm never going to need to do..👍
I am just a hobbyist.
My passion for learning & how the individual components work together has increased from watching your Channel.
I only do it in my spare time & have been at it for about 3 yrs now. I am starting to understand more and more as time goes by.
I really appreciate you taking the time to explain to us how things work!
I have learned, the knowledge of electronics is never ending!
The other day I went to a local well-known television/ computer & phone repair shop, to see if I could buy a couple of individual electronic components?
I was working on a project & I didn't want to wait 3 days on my order from Mouser. The technician looked at me like I was crazy! He said we don't replace individual Parts like that anymore. We just replace the entire board nowadays.
I thought to myself.....ok, just about anyone can do that. That's not really repairing Electronics. That is replacing Electronics!
Apparently, if I ever need to work at a job like that, I shouldn't have any problem qualifying
You'll probably find that even repair places that actually repair the boards probably wont have what you need for you project since most modern electronics use SMD components, they may have some through hole lying around for fixing older electronics but I wouldnt count on them magically having everything you need and I very much doubt they would have components unrelated to TVs, computers or phones, so probably would not have variable resistors, usable displays (they would have ones for phones but they are pretty much unusable), motor drivers, leds, anything like that. If they wanted to sell components to hobbyists then they would open a shop not a repair business, it is probably a bit of hassle for them to sell to you even if they have what they need. The employee would probably have to ask their boss, and then find out how much it costs them and then have their boss decide how much they will charge you for it, essentially it could take a while and be very inconvienient to them because they arent set up as a shop to sell components to hobbyists.
It isnt cost effective for a lot of these repair shops to actually fix the board, it is cheaper and faster to just replace the broken parts and it means they dont need to hire someone who can replace SMD and BGA components. Also repairing the boards would take time to track down the exact cause of the problem, its a lot easier to discover the problem is with this board, rather than this exact component on the board. They would have to first identify the exact component on the board that is damaged and then replace it, which not that many people could do anyway, especially if it is very small SMD components or BGA. Then it may not be the only damaged part. They also need to have access to datasheets and schematics to fix all but the most basic problems, which the manufacturer wont publically release. Then there is the problem of sourcing the components, since some might not be publically available, or the ICs dont have any markings or dont show up when you google them. It is by far easier just to order the boards to replace them with. It also means they dont need to hire an expert to fix the boards. An expert would still take quite a long time to fix a phone for example, and in that time, a lower paid employee could replace the board and have multiple phones fixed. These repair workers that just replace the board do still have to have knowledge and experience of what to replace and why and have knowledge of lots of different devices, so just because you are a hobbyist doesnt make you automatically better than them. In a lot of cases mechanics are the exact same, they just replace parts, whether or not that is because they cant fix them themselves or not is irrelevant but most kinds of repair work is just replacing parts as quickly and cheaply as possible.
Yep, that old adage 'Time is money'. No job satisfaction with just replacing boards.
>No job satisfaction with just replacing boards.
a job usually isn't fun? imagine my shock!
I just started working with Altium last week in college, kinda cool seeing them sponsor this video, now I'm more motivated to learn it
i was JUST THINKING about negative voltages! this video is incredibly well-timed!
This simple charge pump circuit I used when I wanted to get an old scrap LCD matrix display from an old phone to work.
With a polish datasheet and too much free time I was able to make a library for it, to use it with an Arduino.
Only thing left was the negative voltage. So I googled and found this charge pump design which I put together with scrap parts from old circuitbords and it worked :D
Great :-)
The ICL7662 has been around forever and is an easy way to get a negative voltage. With just two added capacitors it will convert up to 20 volts to a like negative voltage. I have used it for many different circuits and it always works well. There is also an ICL7660 but it has a max 5 volt input. The output is not regulated so it will drop with load but it is good for low power circuits in the 50mA range.
If you could get the right boost converter you could probably invert the polarity of the output by reversing the input voltage connection. I've tested it in circuit sim and it works there
Xl semi provides a few ic's for positiv and negativ output. The xl6008 and the xl6019 are good enough, both have a higher output current and both are more stable. I have makes a breadboard power supply with the xl6008 ic, work pretty well.
My very first power supply I built back in 1983 (13)used a 555 for the oscillator and charge pump for negative rail. Built from striping down a car battery charger which the bridge rectifier had failed and been thrown out.
I have used the charge pump for several negative rails. The idea came from an everyday electronics magazine.
Perfect timing for this video - I was just starting to think about designing something that will need dual rails for an opamp!
Most switching power adapters are isolated in reference to AC power ground, and we can cascade such outputs and achieve positive/negative power rail. If you want a micro volt potential difference between the “center ground” of your DC power supply and the “earth ground”, connect them with a 1 M-ohm resistor.
Hahaha, this is exactly the kind of video I was looking for a year ago when I was working on my MIDI piano keyboard!! Bit late for my needs, but I'm glad it's out there for the next guy!
This video arrived some weeks later, but it delivered... it saved my thesis proyect
Wouldn't it be much easier to just swap the Positive and Negative wires? 😁
Well......
😂😂😂👌
Not allways. For example if you need a Positive *and* a negative voltage.
But if you want a rail to rail output from a simple dc you will need his video 🤩
@@flos251 I expected you to take it as a joke.
Thank you so much for your great content I already order 2.5v - 5v circuit from your link
I build stringed instruments in my retirement, which led to preamplifiers for transducers such as piezo pickups hot glued to the soundboard.
To build a preamp I selected opamps as the basis for the preamp and started experimenting with an lm741. For several days I could not figure out why I could only get the opamp to amplify the signal as an inverted signal. Finally it dawned on me that the ground they referenced in the data sheet was not a typical chassis ground and was actually as some call it a virtual ground in a dual supply power supply.
To test my new assumption I placed two 6 volt battery packs in series and used for lack of better words the junction of the two packs as a reference for ground in my circuit. Wala, now the lm741 amplified both in inverted and non inverted modes.
I see the dual battery packs in series as the simplest way to make a dual power supply that has no real down side I can see unless somehow you draw power from just one battery pack which then has the issue associated with a voltage divider when placing an extra current demand on just one rail.
I love the scheme and can't help think that using two voltage regulators such as that output 5 volts in parallel and then placing their outputs in series would also be very effective and likely without the voltage divider issue because ultimately drawing from just one rail is drawing from the source for each voltage regulator.
One common solution is to add a push pull pair of transistors to the output of the opamp virtual ground. The transistors can boos up the amperage capability of the opamp and maintain the virtual ground under a higher load. Also they make small form negative voltage regulator chips now as well. I use one that takes in +5v and spits out -5v and they make them for other specific voltages. They are also very simple to use. Good job, just thought I would mention a few other options available as well
Thanks, i needed this for a small audio project
Glad I could help!
Hey! You are amazing!!!!! You solved my last puzzle in fixing my LCD monitor. Big manufacturer also uses the similar $0.1 hack.
a charge pump. that was so obvious when you think of it.. BRILLIANT LITTLE HACK. THANK YOU BUDDY, THIS IS GOING IN MY SAVED FILES... : )
unbelievable, I'm working with amps and this useful video pops up. Love your work, so useful to me over the years and so clear and explanatory. I'm just sad that in my opinion the old intro was much more catchy!
Be careful on virtual ground approach. Never share the same ground with other circuits that uses the true 0V grounding. If you need a signal route to external circuit, use optocoupler or bypass capacitor to isolate the signal grounding.
yeah, super short circuit current from doing that, things puff up the smoke real fast
I think the circuit started without a true ground signal to begin with, only the first 3 out transformer had true ground.
But what is "true ground"? Are you guys talking about the Earth pin in your wall mains?
@@Nintendjyes
The last one is probably the most useful to me. I’ll normally have access to a bench power supply or can wait for a mailed chip to do that power converting for me, but sometimes I just want to test something that involves an op-amp needing + and - rails using components I have just laying around. Knowing that an op amp in the middle can help stabilize is pretty useful.
you're truly great, scott! this indeed is a brilliant 0.10$ hack! I checked it on oscilloscope it works, and the ripple is no different than the regular positive output(20-50mV). didnt test yet on load.
This is useful for very very old PC that requires -5v output for something like Soundblaster
Good to know :-)
many pc power supplies have a - 5v rail and a -12 rail generally the white and a blue wire
@@ShahZahid yup that’s why I save my old atx’s lol.
@@greatscottlab Was hast du dir bei deinem Kanalnamen gedacht?
So, if you have 3 output pins at 0V, 1V and 2Vs
But decide to label them -1V, 0V and 1V, is that all a negative voltage is?
You hook up your probe common to what you choose to be ground (0V) and everything else is measured relative to that point?
the best videos you make are electronic basics that is so simple and informative please make more about it
I have wanted a video like this for so long. Thank you!
Awesome vid Scott ty.
I was recently thinking of making a negative voltage supply by using a reversed flyback converter. I wonder if this would be able to provide a higher output current. Great video. I never knew of the charge pump hack.
Yeah flyback converters can be used, you only need to flip the secondary transformer to get a negative output
There's a lot of flyback converters that include both a positive and negative branches, with 2 secondaries windings on the transformer and essentially the same circuit as the output
Fantastic video with fantastic information! I really loved using a highlighter to show the functional blocks of the circuit. Something that I could do myself when needed.
I use two buck converters in series.
- Input negative become the target negative voltage
- First stage output positive becomes the target positive voltage
- Second stage positive becomes the target ground voltage
There are two buck converters to adjust, which is a disadvantage and, for reasons that I did not investigate, the first buck converter heats-up like crazy and ends-up dying after a while.
I agree 100%. Negative voltages are very important !
This would be pretty useful with op-amps requiring dual rail power supply too..Thanks for this informative video
Playing with buffered rail splitters is a rite of passage for anyone building opamp-based headphone amps.
True👍
Exactly the information I was looking for today.
Brilliant :-)
Great info, I've got a waveform generator I built for a Uni project and have never gotten around to building a dual rail supply for it, of course at uni we just powered it off the bench supplies, wasn't until I took it home I realised I couldn't use it. Yet.
I waited for this video for far too long, THANKS ALOT
i was waiting for same video , thanks greatscott
Me too brother.
Hope you enjoyed it!
Can you just use 2 boost converters, set one to 12v and one to 24, then use the 12v output as virtual ground, the original ground as -12v and the 24v output as +12v? I guess you can't but why not?
Another option you didn't mention: Just use _two power supplies._ You need to make sure that the supplies you use can function without using a common output ground between them, but many power supplies have effectively isolated input and output grounds, so many actually can do this. Then you just connect the positive output of supply B to the ground from supply A, and B's "ground" output then becomes a negative voltage (relative to A's ground). This can even allow you to tune the size/type of the different power supplies to match the current (or noise, etc) requirements of each rail.
You can also use complementary linear voltage regulators (i.e. a 7812 and a 7912) to maintain stable positive and negative voltages relative to a reference VGND...
Two supplies is an elegant and simple solution, although it's not based on a single circuit. I used this option when I was studying op amps and I had no time to build a dual-rail-supply, neither wanted to buy one.
I actually made a large dual rail power supply from 12 PC power supplies. The main use I have from it is a large LCD project I'm working on. (:
Man this is one of those videos I wish had existed 1 year prior. lol
for even higher current, you can add a pair of npn/pnp transistors (e.g. 2n3055 mj2555) on the output in half bridge (bases to op amp out , emitters together connected to op amp - and virtual ground. collector of pnp to V- collector of npn to V+) add a couple of smoothing caps between both rails and virtual ground.
note that since you're basically burning the imbalance current, the transistors will heat up a lot in some cases (you're getting I*rail voltage, so better make sure there's adequate cooling).
The "remove every component from the board so we can trace it" is a true double Picard facepalm one. I can't even.
If the boost converter uses a small toroidal inductor, you can easily add a second winding to derive the required separate (negative) rail.
But then what happens when you load down one of the rails but the feedback resistors are on the other?
I have just completed my school project a 350 W bluetooth speaker and the biggest problem for me was to make simetrical power supply(for supplying preamp) as i have only dc source and making the inverter for transformer is not efficient. I had found the same page that you have shown in the video and i made a divider with two transistors and everything works fine. Its interesting that you have uploaded this video now when I had already gone trough this all..
Great and timely video! I have recently been considering negative voltages and this video provides lots of useful info for me. Thanks
Of course, you could also add another power supply to add the necessary negative voltage with the needed current; if it's a lot of current and also loaded uniquely to the positive supply.
I made the discrete transistor version for a 60w amplifier and it works perfectly fine
Thanks for the feedback :-)
Such a good video that I needed some weeks ago when i designed a positive and negative voltage and now I have more info thanks.
Another way is to use a Buck converter, if you see the schematics of a inverter converter and a buck convertee you will see the same image just flip so the lm2596 can generate negative voltage if you put GNd to output and Voltage to.input, it bootstrap the voltage so if you get -12V from 5V the input will see 17V that is the disavantage the limit would be the max input voltage of the IC but it is pretty straight forward to get negative voltage like that without soldering anything
You can actually use an LM2596 buck module as a negative voltage generator, and you can draw loads of current from it since it's not a charge pump! The great thing is that practically no hardware mods. Search for LM2596 negative voltage, and you'll find a Wordpress blog post that explains how to do it.
You can also use almost all Buck Converter Modules. Connect the + Supply with Vin of the Module, GND to Vout of the Module and You will get the negative Voltage at the GND-Pin of the Module. The Summ of Input Voltage and negative Output Voltage has to stay below the maximum allowed Input-Voltage of the Buck Converter Module.
@@juergenschimmer960 True, but I'm not sure if it can be accomplished with a synchronous rectifier where the "diode" is instead a MOSFET, and built into the chip. The other downsides are the large inrush current and the fact that the output tends to be noisier compared to when not generating a negative voltage. I've solved this in the past by adding another inductor in the series and an additional output capacitor, a pi filter.
Another approach is to use an adjustable buck boost converter with the positive output lead connected to ground. I used this setup for delivering negative voltage to a large LCD. Of course the converter has to have isolated potentials in order not to create short circuit.
Nice iron burn! We match! Twinning!
I literally sent an email to my professor, asking for help with dual power supplies, half an hour ago... If only your released this yesterday.
This circuit is the Cockcroft-Walton circuit. It is slightly different from the generally said active charge pump circuit.Although it is a convenient circuit, the current carrying capacity is small.
This was another excellent video. I really enjoyed all the content. I learned quite a bit. Thank you for sharing this with the world.
Eigenkrafts!! Opamp Vgnd toll!!! I used two similar boost converter to create negative voltage in one of my skiffs, it uses the 12v out of one to ground input of the other, the inverted out of the latter gives neg rail. Plenty of noise anyway, not the best approach, but with some filters gets the job done very efficiently. Never thought of the pump solution, quite interesting!
1. Get a boost converter.
2. Replace its inductor with a center tab transformer.(use boost con as flyback)
3. connect the center tab to gnd..
4. Do a positive and negative rectification with the other two ends of the tran's secondary winding.
I've done it so many times. The negative output can do as the positive.
Current limiting also works with negative output since most current sensers are placed in gnd..
DVD power supplies does it except step down( but only for referencing purposes)..
Pretty interesting indeed! Thanks a bunch, dude! 😃
Stay safe there with your family! 🖖😊
Nice video mate. Keep going. 👌😊
I use two diodes and two caps on the secondary of a single secondary transformer. It uses the upper half-wave to charge the positive al cap and the negative half-wave to charge the lower voltage rail cap. No need of center tap this way
There's also the Cuk converter which can achieve good efficiency and low noise on both the input and output. It's a good choice for noise sensitive applications like audio.
Thanks for the feedback :-)
Fun fact. Telecom carriers use -48volt DC because it is easy to store in deep cycle batteries and can backup equipment during a power outages for much more time than a regular UPS.
Another fun fact, it's not really a negative 48V it's just 48V with the positive terminal connected to Earth
@LabRat Knatz You're correct negative supply was used to reduce verdigris
6:12 with a FULL BRIDGE RECTIFIER.... ElectroBOOM^^
Thanks for another Great video, Scott!!!
much like your voltage divider circuit, I used an LM7805 between +10VDC and GND, this made a +-5V with good current.
7805’s output stage is not bipolar, so I’m not sure if you can sink significant current into its output without it misbehaving. If you want to use linear regulation, you’d be better off with a power opamp as shown in the video since these devices are designed to work correctly for both sourcing and sinking current.
Oh no? What do you think is inside 7805’s?
Very well explained
Thanks :-)
Great Great Scott 👌
Thanks 👍
@@greatscottlab Welcome 😘
a voltage double circuit also can give dual supply from a regular single supply AC transformer
I didn't really get the "Why"-Part.
Ok, there are ICs out there that want negative voltages and audio stuff wants it too. But why do they need it?
Every voltage is both positive and negative, depends of how you are connecting meter.
Can we add another isolated center tapped coil with the inductor?
Another solution is to boost the 12v in to 24 v and using the input positive (12V) as virtual gnd
thx mate
240watt Fast charging how its work
Oohhhhhhh hi
but if i run 2 in parallel? one at 12 and another one at 24? that gonna work?
Many Thanks for always explain and present useful topics
Her zamanki gibi kaliteli bir video olmuş. Teşekkür ederim
Another way is using a buck-boost converter. It inverts the output voltage naturally so the Vout+ is the input ground.
Can it also be done with a voltage multiplier? a capacitor and diode arrangement to tap ground from between the two capacitors, a bit like the one you did with the resistors.
Watched it again = Helpful again.
Very cool video! Always feel like I learn something
Glad to hear it!
Sir please make video on Thevins theroem and max power dissipation
I can have a look :-)
7660 charge pump plus a couple electrolytic capacitors is a simple & inexpensive way that can run on +1.5vdc to +10vdc (like +5vdc from USB) to generate corresponding negative DC voltage supply.
How about using a H-Bridge switching fast, filter output with diodes and smooth with Caps to create Pos and Neg supply? Would be able to get much more current out of something like that...
H bridges are good for reversing the direction of current like through a motor but they dont actually have a ground point that you could use and the voltage difference is only relative and between the two output terminals. It changes the direction of the potential difference and therefore current flow but it doesnt create negative voltage. If you measured the voltage between either output and the circuit ground, one would be at the supply voltage and the other at ground, there isnt any negative voltage there.
Thank you so much Scott !!
I tried making my own DC/DC supply once, taking ~14V and using a buck converter and a buck-boost converter to output +V and -V. I ran into issues with total current supply and in-rush current.
How come you didn't mention the simplest solution: just put two isolated power supplies in series?
I think it is not an elegant solution. Sorry.
@@greatscottlab Could you elaborate? I thought using two batteries in series would work
If you want to use the transformer alternative without using mains voltage you can use an oscillator from 0v to vsupply to convert the dc voltage to ac and then supply that to the primary of the transformer.
The easiest way to get a dual rail power supply is using one buck converter. supply 24v to it, set it to 12v positive. Now use the buck converter's positive (V+) terminal as your GND. Between V+ and the input (VIN+) of your converter there is 12v and -12v between V+ and GND of buck converter.
You can supply as much current as your BC can handle, assuming your power supply can too.