The Synchronous Voltage Converter: A DIY Project

I'm excited to share with you my recent project – a synchronous voltage converter that I built myself. In this article, I'll walk you through the entire process, from designing the circuit to testing its performance.

My goal was to create a device that can convert AC power to DC power with high efficiency and reliability. To achieve this, I decided to use a synchronous design, which is commonly used in commercial applications.

The concept of a synchronous voltage converter is quite simple. It consists of two MOSFET switches, one for each phase of the AC input. When one switch is on, it conducts current from the capacitor and through the inductor, while the other switch is off. This process repeats continuously, creating a sinusoidal output.

To make this design more practical, I used an IR2184 Half-Bridge MOSFET driver IC. It's a high-power device that can handle both high-side switching and bootstrapping, making it ideal for my project.

The first step in designing the circuit was to choose the components. I selected a 2-layer PCB design with all the required components, including the IC, capacitors, inductors, and resistors.

Next, I soldered the components onto the PCB. This took around an hour, but it was worth it to see the final product come together.

After completing the circuit, I connected the PWM signal generated by my Teensy board to the PCB. This was the critical step, as it allowed me to control the switching of the MOSFETs.

The next challenge was handling the discontinuous conduction mode (DCM) and continuous conduction mode (CCM). In DCM, the current flow through the inductor is zero during certain periods, which decreases efficiency. To address this issue, I used a small inductance value of 29uH, which made the negative current flow noticeable on my oscilloscope.

With the inductor in place, it was time for the first tests. I connected the device to my AC power source and measured its performance. The results were impressive – nothing blew up, and the synchronous converter behaved like a buck converter.

One of the most significant advantages of this design is that the MOSFETs stay at room temperature even during high-power operation. This is a major benefit over traditional designs, which often lead to overheating and reduced efficiency.

To further improve the performance of my device, I would consider implementing a feedback system to regulate the output voltage. This would require some additional components, but it's an area worth exploring in the future.

For now, I'm happy with the results of my project. The synchronous voltage converter has proven itself to be a reliable and efficient device that can convert AC power to DC power with high accuracy.

In conclusion, building a synchronous voltage converter is a challenging but rewarding project. With the right design and components, it's possible to create a device that exceeds commercial standards in terms of performance and reliability.

If you're interested in learning more about synchronous converters and buck converters, I recommend checking out my video description, where I linked some excellent resources on the subject.

WEBVTTKind: captionsLanguage: enThis is my Synchronous voltage converter thatI built in the past week.Now you might be asking yourself what makesit so special since you can already get switchedmode voltage converters like buck and boostconverters everywhere on the internet.And by the way for those just starting outwith electronics such converters can takea DC input voltage and convert it efficientlyinto a higher or lower DC voltage that wecan then use.But getting back to the question what makesthis synchronous one so special and to answerthat we can have a look at this beautifuloverview chart from Würth Elektronik of almostall switched mode power supply typologies.As you can see their efficiency is around85 to 90% at max which is not bad but a synchronousconverter can supposedly reach 95% efficiencywhich actually sounds to good to be true.That is why I built my own to measure itsefficiency but before I do that why don'twe go back in time together so that I canshow you how exactly I made mine and explainwhy this converter type is quite a bit morecomplicated to build than you might think.Let's get started!This video is sponsored by JLCPCB which isa PCB manufacturer that I can truly trust.I actually ordered the PCBs for this projectfrom them which was super simple to do byuploading my Gerber files, selecting my PCBpreferences, paying very little money andwaiting for a week and just like that I gotbeautiful purple PCBs that you can also getif you try out their service.First off I had to decide what kind of topologyI wanted to transform into a synchronous oneand I went with the buck converter becausein the case of a switch fault we either onlyget the full input voltage or nothing at allat the output which is great for testing.In comparison if a boost converter would actfaulty we could get a rather high destructiveDC voltage on the output which I wanted toavoid for my first tests.But anyway, next we need to understand howthe circuit works and if we simplify it asmuch as possible then we could say that theswitch creates a PWM voltage with a high frequencythat gets smoothed out and thus lowered byan LC low pass filter.It does not get simpler then that but to bemore precise and accurate let's imagine theswitch really gets controlled by a high frequencyPWM signal and right now it is closed.Now current flows through the circuit andsupplies the load while charging up the capacitorand building up a magnetic field around theinductor which by the way opposes the flowingcurrent and thus only let it rise slowly.This continues until the Switch opens at whichpoint the inductor uses it electromagneticfield energy in order to let current flowthe same way as before by now acting as ourenergy source.This is possible because our Diode which blockedthe current flow before now becomes part ofour current loop.But of course the current is this time lineardecreasing and depending on a couple of factorsthe current can either stay above zero beforethe switch once again closes or fall to zerobefore that at which point the capacitor hasto supply the load for a short time.But either way depending on the utilized dutycycle the output voltage can be fine adjustedbetween 0V and the supply voltage.So all in all the circuit itself even whenconverting to real electrical components aswell as the voltage and current waveformsare not that hard to grasp, right?I thought so as well which is why next I startedto create my own DIY version on a perfboardin order to show you exactly where energygets wasted.What I didn't expect though was that my firstcircuit idea with an N-Channel MOSFET andGate-Drive Transformer didn't feel like workingwith me for mysterious reasons.So after 3 hours of troubleshooting I gaveup and instead went with this more straightforwardP-Channel MOSFET alternative according tothis final schematic which once again provedthat everyone can have a bad day with theirprojects, no matter the experience.And after writing a bit of code for the Teensymicrocontroller in order to create an 80kHzPWM signal whose duty cycle I can change witha potentiometer, it was time to connect a100uH inductor to the circuit, attach a 4.7ohm load to the output, power everything with12V and have a look at the oscilloscope whileplaying around with the duty cycle.As you can see the previously discussed theorypretty much corresponds with the practicalapplication and thus we can fine adjust theoutput voltage by simply changing the dutycycle.So next we have to find out how this circuitcan be more energy efficient by once againhaving a closer look at the main components.During the first switching state we only gotthe MOSFET, inductor and capacitor which allhave a very low losses so there is not muchto improve.During the second switching state however,current flows through the diode whose voltagedrop we can measure in the practical circuit.As you can see it is around 200mV peak soaround 100mV on average which multiplied bythe practical average current that is around400mA equals a power loss of around 40mW whichdoes not sound too bad but wait until youhear the synchronous design power loss.Because with it you simply replace the diodewith another MOSFET switch and it is calledsynchronous because in phase one the upperMOSFET is on while the lower one is off andin phase two that simply flips around synchronously.And since the MOSFET I want to use comes witha resistance of 2.8mΩ the new power lossthrough it should be around 1% of the diodepower loss which is quite a big difference.So all in all a pretty good concept but howcan we easily turn on and off the 2 MOSFETsalternating considering that we are dealingwith high side switching and the MOSFETs arealso never allowed to turn on at the sametime.The solution is of course such an IR2184 Half-BridgeMOSFET driver that not only requires one PWMsignal in order to turn on/off both MOSFETsalternating with bootstrapping but it alsoin cooperates a dead time so that both neverturn on at the same time.And if you are completely confused right nowthen I would recommend you to watch my videoabout MOSFET drivers or basically every videoI did so far about switched mode power suppliesor voltage converters.But anyway, after I added complementary componentsaround the IC, I turned this schematic intoa proper 2 layer PCB design which I orderedat JLCPCB.After one week of waiting I was not only greetedwith my beautiful PCBs but also with all therequired components which means it was timeto solder them all in place.And after 1 hour of SMD and THT soldering,the circuit was almost done.So I connected the PWM signal generated bythe Teensy to the PCB and was about to solderin the previously used inductor when I realizedthat we might have a problem.Before we saw that with this inductor whiledrawing only a bit of power the current fellto zero during the second switching statewhich only changed when we drew more currenton the output.Those two modes are called discontinuous conductionmode aka DCM and continuous conduction modeaka CCM and they both come with their ownunique advantages and disadvantages.Now before in DCM mode the diode preventedcurrent from flowing from the capacitor andthrough the inductor.But now that we got a MOSFET here this currentflow can actually happen which will once againdecrease the efficiency.The solution is of course to turn off thelower MOSFET when the current flow reacheszero but that is easier said than being donewhich is why there are actually lots of synchronousconverter ICs out there that do this complicatedjob for you and all you have to do is simplyadd a couple of complementary components.For my DIY attempt however I simply ignoredthis fact because only by using a rather smallinductance of for example 29uH and a low currentdraw this negative current flow was noticeableon the oscilloscope but with the other 100uHinductor this problem was pretty much neglectable.So with the inductor in place, it was timefor the first tests and I was super happyto see that nothing blew up and the synchronousconverter basically acted like a buck converter.It was also fascinating to discover that whiledrawing quite a bit of power and thus heatingup the 50W load resistor, the MOSFETs prettymuch stayed at room temperature.So it was time to measure some input and outputpower values and calculate a couple of efficiencieswhich in my case peaked at almost 92% witha 9V output voltage which is of course notthe 95% I was hoping for but not bad at all.To achieve that we would have to fine adjustevery component and also hit the load currentsweet spot which as you can see in this datasheetcan be certainly possible.Now of course at this point I could implementa feedback system into my design so that Iget a stable output voltage which I mightdo in the future and meanwhile I encourageyou to check out the video description whereI linked really good articles about synchronousconverters and buck converters in general.With that being said I hope you enjoyed thisvideo and project.If so consider supporting me through Patreonso that I can continue producing more videoslike this.As always don't forget to like, share, subscribeand hit the notification bell.Stay creative and I will see you next time.