Tag Archives: power electronics

Do you know your Capacitor?

Took a while since my last blog post but this one about capacitor’s needed some preparation time.

The capacitor story

In many designs (even some app notes) there are huge electrolytic or polymer cap’s to filter noise. And in many cases exactly those huge capacitors fail completely. They filter out the fundamental ripple of the switched mode power supply but when it comes to higher frequencies it seems they are not there anymore. But why?

Is my capacitor always a capacitor?

This is the most important and fundamental question. If the capacitor is capable of filtering out the fundamental waveform but not the higher harmonics then this leads to the assumption that the electrolytic capacitor I use is not always a capacitor.

For this reason I made a test setup to do a kind of proof of concept measurement for this assumption. Its a very simple test basically. With a resistor (150k) in series with a polymer-cap (15µF and 0,1Rser) I have built a first order low pass filter. For testing I set my frequency generator in my Keysight Scope to an offset of 3V and an amplitude of 2V (to keep the voltage across the capacitor always positive). Then I connected one channel to the input of the filter and one channel to the output of the filter – both AC coupled. For keeping the measurement as good as possible I generated a script to automatically step through the frequency range from 1Hz to 20MHz – which is the maximum I can get out of the frequency generator. Such a script should eliminate a lots of errors out of measuring by hand.

Frequency response of the low pass filter

Measurement analysis

The result of the measurement can be seen in the plot above. Theoretically the curve of the filter should start decreasing at 1 Hz already. I guess due to some measurement uncertainties this curve starts decreasing at about 10Hz. But more important is what happens after 10kHz. The filter settles after 10kHz (except this one dip which is cause by power supply noise of my printer which is nearby the test bench) at a maximum damping value – which is basically the value you get out of the voltage divider of the resistor and the series resistor of the cap. But at 1Mhz it starts to rise again – So my capacitor starts to act like an inductor. Such behaviour is quite normal within electrolytic and polymer capacitors and leads often to problems when it comes to EMC testing.

But how should I know in advance how this capacitor behaves? Simulators for the rescue! There are capacitor manufacturers like Kemet which provide a simulator which gives you an indication where you are going to land with your polymer capacitor but its just an indication because the wire length and the layout you have on your board are influencing this behaviour a lot!

What does that mean for power supplies?

Ok, what does this mean for your power supply design? As you might know a Buck converter is producing a high amount of noise at the input and a boost converter is producing a high amount of noise on the output. This noise is normally somewhere above the 20MHz I have shown here, so the ELCO and polymer caps will not be sufficient for damping such noise. In order to handle that I would recommend to put some smaller caps (ceramics) aside to this one the get rid of the noise.

When do I know i have enough ceramics for my noise? Well i would suggest a measurement with a LISN and a spectrum analyzer to determine which configuration is the best in your design. Without this it might lead to some over or underachievement in your design and either you have wasted board space or you have won some extra design cycles because your EMC noise levels are not where they should be.

Hope this analysis helps for your daily life in electronics!

j j j

LT830x µP Flyback Converter – New Product Monday

Today I want to talk about one of my favorite topics in power electronics – Flyback converter !

Flyback Converter in General

You can see flyback topologies in many different areas, for example consumer electronics, industrial, medical and many more. The power ranges are usually from a few watts up to some 75W or so. Its very rare to see them above 100W because the transformer gets big and a flyback topology is not the most efficient one when it comes to higher power ranges. So why is it so popular in lower power ranges? Thats an easy answer – its all about simplicity and BOM – cost. With a flyback you have just two power switching semiconductors and a fairly simple transformer. This makes the design cheap and compact. Thats why its mostly the only type of power supply topology for notebook chargers, TV power supplies and so on.

The LT830x Family

LT8304 Flyback Converter Application Schematic

LT8304 Flyback Converter Application Schematic

Lets come to the LT830x family consists mainly of 3 parts. First one is the LT8300 which is not a really new one but its the first one in this series. Lots of customers of mine use this one to get galvanic isolation on their sensors. Those things don’t need much power but do rely on a stable and reliable power supply, which the LT8300 delivers. To address the needs of “higher” power demands Linear Tech introduced the LT8303 and the LT8304. Both parts have the same input voltage but with a higher switching current capability. The LT8300 and the LT8303 are also pin compatible – one footprint 2 parts.

No-Opto Flyback Converter

When looking at the datasheet of the LT8300 you will recognize that there is one part missing, which is usually in every isolated power supply – the optocoupler. The optocoupler meanwhile became one of the weakest parts in a power supply and the reliability and isolation voltage often is directly related to this part and so semiconductor companies search a way around this part. A flyback topology allows the sensing of the secondary voltage right at the drain of the power MOSFet or at the AUX – winding of the transformer during the conduction phase of the secondary diode. The voltage that can be seen on the winding there is a direct relation of the output voltage and the transfer ratio of the transformer ( minus the diode forward voltage ). The LT830x – family applies some techniques to get this sensing done.

Of course the LT830x – family is not the only one who have that feature. The LT3798, which I mentioned in my post about the power factor correction , has this feature on board as well.

Key Applications

So, where is the key application for that parts. Mainly its industrial, transportation, medical and measurement equipment where things need to operate isolated, like sensors, ADC’s (for example using it together with the LTM2893 to get an ADC working isolated) or ADC’s combined with small FPGA’s or DSP’s (with the LT8304 up to 24W power consumption!). Due to the high input voltage that parts are suitable for a wide input voltage range which makes them reusable for many designs.

Meanwhile there is also a big choice when it comes to flyback transformers from all big magnetics companies – but if you like you can also wind your own transformer.

j j j

7 Ways To Improve Your Buck Converter EMC

For electronic engineers EMC is often a mystical thing with a weird test in a metal box they have to pass with their design. Its true, EMC can be a tricky thing but with the right tricks in your design its much less tricky then before. Here I show you some of this tricks for buck converter : 

1. Get input filtering to your Buck Converter! 

Buck converter produce tons of noise on the input, that means you need that input filter! It always depends on operation mode and frequency of the buck but in the datasheets are usually recommendations for the filtering take them serious.

2. Gate resistors at the gates of your mosfets

The buck converter control IC’s you use do have mosfet drivers but that doesn’t mean that you can conntect them directly to the gate of the mosfet. Sure, the faster you switch the more efficient you are but that also means a lot of noise because of fast rising and falling edges at your mosfets – that harms the EMC a lot. There are many ways to adress this and the most simple way is just putting a resistor between the pin of the IC and the mosfet gate, for low voltages this should be enough. If you switch higher voltages it may help to get diodes within the game. With two diodes and two resistors you can control the rising and falling edge at your mosfet independently this can help you a lot with improving the EMC but not ruining your efficiency.

3. Get your boost cap a series resistor

A very common thing that is ignored. In a buck converter you need a bosst cap to get the voltage of the gate from the high side mosfet above input voltage level. The circuit to get this done is always in the IC itself there is often just the need of pluging an external capacitor to the IC and it really seems that easy but it isnt. Lots of IC vendors dont get their layout and bonding properly done. This means that you create a resonance circuit with the cap, IC bonding, IC layout , mosfet bonding and mosfet gate cap. I’ve seen mosfet drivers myself where the overshoot was about 80% of the acutal switching level which caused a lot of damage to the EMC. With the small resistor in series of the boost – cap you can damp this resonance circuit – rule of thump: values between 5 – 25 Ohms schould be enough.

4. Get a fixed frequency switcher

There are several switching and control approaches on the market and some are variabel frequency like constant on or off time and some are fixed frequency like voltage mode control and peak mode control. With constant on or off time it means that your overall switching frequency changes with the load. If you have a circuit with different load situations like a µC controlled board with different peripherals on the µC your board behaves different dependending on what your software is doing with the µC and the peripherals. That often causes a lot of headaches during the EMC testing and may cause beside layout changes also tons of software changes which adds more work. With fixed frequency converters you switch at one frequency and optimize the board for that – much easier to do! BTW: when using µCs at your buck converters output you just should use peak current controllers – much faster and stable control loop. It ensures that the supply voltage is stable when changing from sleep mode to normal operation at your µC

5. Use fixed frequency switchers with spread spectrum

When having a constant frequency switcher you can optimize your board to one frequency but there is another trick you can add to this design: spread spectrum. With constant frequency you have all the noise energy you radiate is fixed in one place at the frequency spectrum, that creates a peak at one frequency. With spread spectrum the IC modulates the switching frequency in a controlled way. This can happen for example with a sine wave modulation like FM modulated radio. This spreads then the radiated energy over a certain band of frequencies which damps the peak. Here is for example a list of switchers with that technology from LT: Spread Spectrum Switchers

6. Take the layout recommendations seriously

Every vendor of power supply IC’s which understands this topic puts layout recommendations in his datasheet – read them! It also makes lots of sense to analyse the layout if its very complex it may be that the manufacturer wasn’t paying a lot of attention to the pinout of the IC, this takes revenge at the layout and higher EMC later so they have to do a more complex layout. Linear recently introduced their silent switcher series (another big fanboy moment for me!) where they have put some extra attention to the IC, the bonding and the layout which ensures in very calm EMC even at high switching frequencies and integrated mosfets with up to 6 amps. This should also help in your design!
BTW: Some vendors still recommend 2 layer boards. Do yourself a favor and use 4 layers. This makes your life easier a lot!

7. Get control about the current flow

Often there is the rule in layouts: There is a lots of cupper on the board, leave it there! Well that helps but sometimes it can make things worse. In layout it is common to pay a lots of attention to the positive power rail but not to the negative. There is often just a ground plane left with hope that its working properly – Do not use ground planes for your power needs! Ground planes do not work for transporting your energy, the current always takes the most convenient way to flow back and this is often the wrong way for your EMC! Make proper tracks even for the ground all the way back to the connector or cable! This will help you to not create ground plane antennas.
BTW: I personally use ground planes to get analog tracks shielded from the digital tracks.

 

Happy engineering!

PS: if you can add a tipp just comment below!

j j j

Power Factor Correction A Short Story

Power factor correction is always a topic in power electronics or in electronics general, but often its quite unclear what it means. In the following I try to explain it a little bit and give some practical tips and hints for your design or future design. 

First of all what is this thing called power factor. A detailed description can be found here at Wikipedia but what does that mean in practice. When talking about power factor you always talk about AC (or mains) powered devices. For the energy provider it is highly important that the load, that is drawn from its power grid and power plants is a resistive load – that means power factor is 1 or close to 1.
What does that mean in practice? Lets draw some circuits to show …

 

two basic circuits that show in simulation the impact on power factor

The upper circuit you see is just an AC source with an resistor. Pure resistive load. The second schematic you see is a AC source, with rectifier, ideal cap and resistor. Again, just a resistor load. But, show both the same ohmic behaviour to the AC source?
The simulation plot will tell us:

sim1_plot

First plot is the resistor only schematic and the second plot is the rectifying circuit, and no there is nothing wrong with the simulation, but why is it? In the first circuit you have power factor 1 – ideal world example. The second one is a more real world like example. In your application you need a certain DC voltage, so AC is going to be rectified and then transformed in the desired DC voltage rail for your application – in this example then replaced by a resistor for simplicity. When you look at the second plot and the current waveform, this waveform isnt looking like a sine wave at all like in the first plot – so the power factor turned very poor. This comes from the cap in the circuit which keeps the voltage stable for the output but it also needs to be charged from the input and so the input current is very high at a very short amount of time to get the energy into the cap that needs to be provided to the output.

What does poor power factor for the power provider mean? In the first plot with the good power factor you see the current is very low and in the second one the peak current is very high – 22mA vs ~1.2A . The load is similar in both cases but in the second case the provider has to use much bigger cables to provide the same power which is just uneconomically and waste of material. So we’ve discovered the root cause why power factor correction is necessary.

When is power factor correction needed? Well, there are different limits in standards. For industrial electronics the standards say that everything at 75W and above needs active power factor correction and with lighting there is the limit at 25W, every LED driver at and above 25W needs active power factor correction.

So what can be done for power factor correction? There are two ways: active and passive. Passive power factor correction can be done with input filters but with 50Hz frequency and the goal of power factor of 1 the filters are just huge, in size and cost. Active power factor correction is the approach to have smaller input filters but an active circuit that keeps the power factor at a high level.
In practice these circuits are either boost or flyback converters – it depends on your application. For output powers lower than 100W a flyback solution is often a very charming solution and for higher output powers there is no way around boost converters (multiple phases, bridgless rect. and so on).
These circuits put an inductor on the input side and switch it very fast. With this switching they are modulating the input current in a way, that the current seen by the AC source is looking like a sine wave. Of course there are some input filters needed to filter harmonics but that are much smaller filters than with purely passive PFC.

Lets talk about the practical hints: There are plenty of IC’s in the market that promise a good power factor correction and at nominal load this fact is true but thats not the complete truth. When you do a power supply design you design it around a maximum load that can occur. This means you almost never achieve the nominal output power in real life for this power supply. And there comes the downside. A lot of these parts, especially the cheap ones, are optimized for a maximum point of load where the power factor is good. This makes sense for simple and cheap lighting applications without dimming but when it comes to part-loads or something they quickly loose power factor and you’ll end up with a power factor of 0.4 or lower at 50% of your nominal load which is indeed not the way a circuit should perform.

When it comes to topology it again is dependent on your application. If your application needs lower than 100W a flyback maybe is a good fit – but be aware at the output of the flyback there then a 100Hz ripple that comes from the AC side so I recommend to do an AC/DC which is isolating and then a DC/DC which then takes care of your output voltage. This is also a good approach for engineers which do want to do a plattform for their power supplies. Design it once and then change the output voltage with the DC/DC in a very easy way. Also makes a lot less headache at the EMC testing.
If you need more power then go for a boost pfc stage and then an isolating DC/DC which meets your application. It can be a push pull or fullbridge for example but be cautious with LLC topologies after your boost stage. The reason is simple, Boost-PFC-stages are in general fixed frequency and variable duty cycle circuits, LLC are fixed duty cycle and variable frequency this leads to harmonics on the charge and discharge currents of the electrolytic cap between this stages and damages it which results in a short or almost no lifetime of your power supply.

My favorite part for flyback-pfc solutions below 100W is the LT3798 and this is not because I’m a Linear Tech fanboy – its because this part shows good power factor behaviour over a wide load range. I’ve tested a lot of such parts and there was just one part that was as good as this but the manufacturer made it obsolete two years ago.

At least here is a plot I’ve made some time ago about the power factor behaviour of the LT3798. As you can see at 40% of the max. load the power factor is still at 0,9, which is pretty awesome analog engineering and makes the PFC curve almost ideal.

 

LT3798 PFC Plot

on the x-axis there is the percentage of the output power and on the y-axis there is the power factor.

Hope this article helped a little for getting some insights into PFC. If there is the need of further informations or you’ve spotted mistakes – just comment.

Happy engineering !

j j j