Topics: Voltage Multiplier Boost Converter
on Power Supply
Voltage Multiplier Boost Converter
Most of the case we go with full bridge topology with HF transformer to achieve high voltage ratio (~16 times) where as with simple boost converter we can achieve 6 to 7 times (~85% of Duty cycle) ,
So , Is there any alternative way to achieve high voltage ratio like 16~18 times of input voltage? I know Voltage Multiplier Boost Converter, but never implemented in any of my design. So wanted to know how much high voltage ratio can be able to achieve by this topology? Why this is not broadly addressed in the industries like phase shifted full bridge topology? If any one can write few line about this topology's advantage and disadvantages will be greatly appreciate...Also Looking for experts comments to know best suitable way to achieve such a higher voltage ratio with power rating of let say 1kW to 5kW range...
Thanks to all in advance.
09-09-2013 04:59 AM
Traditional voltage multipliers need a low impedance high-side drive and low-side drive to make the capacitors pump their current. A flyback only provides a high-die low impedance.
I had designed a high turns-ratio flyback, and it worked, but the output leakage inductance was a big problem. The output spikes and the very poor output regulation were the major problems. I had to resort to an interleaved transformer to increase the coupling and cross-regulation feedback to make it work. That was only a 5:1 increase in output voltage.
I would personally recommend to stay with the wisdom of your forefather's hard work on this. A forward converter with diode or synchronous output rectifiers and the diode/capacitor classic multiplier. There are many examples on the web. Electropherisis (misspelled) is a classic example (used for extracting DNA plots).
good luck and wear rubber soles on your shoes.
09-09-2013 05:00 AM
For high output voltages and high output power, try using a buck current-fed push-pull or buck current-fed full bridge. Very interesting topologies, and no high current output inductor...it's at the input.
09-09-2013 05:01 AM
I understand that you do not need isolation in your design and you are willing to build a converter without a transformer. In some applications these converters can be cheaper and more efficient than isolated converters.
As you said, the boost converter have a limited voltage gain. But there are many other topologies that can achieve higher voltage gain than the boost converter. I did a comparison of some of the topologies I found in the following paper:
=> Comparison of non-insulated, high-gain, high-power, step-up DC-DC converters
One that I found specially interesting for high power applications is the Interleaved Double Dual Boost, as shown in the following paper. It gives you about twice the voltage gain of the boost converter. It can reach the 16X voltage gain you need, but keep in mind that this voltage gain is very high and I am not sure if you will have some advantage over using the FB with a transformer to step-up the voltage.
=> Modeling and Control Design of the Interleaved Double Dual Boost Converter
Please take a look at them. If you do not have access to IEEE Xplore, just drop me an email at firstname.lastname@example.org and I will send you a copy of these papers.
09-09-2013 05:02 AM
I am very much happy to see ur comments.
Its nothing like I do not need isolation. I am doing some literature survey to know what are the topologies possible to do get such 16X voltage boosting.I can find few papers published on IEEE about high voltage gain such as interleaved voltage multiplier. But I am not sure why those are not popular yet in the industries like FB with a transformer to step up.
One of the above comment from Sir Marty about "traditional voltage multipliers need a low impedance high-side drive and low-side drive to make the capacitors pump their current"
I tried to get some feeling on PSpice open loop simulation, but really I could not able to do any conclusions......:(:(
Anyway I have very limited access of IEEE. Please send me few papers on
=>Modeling and Control Design of the Interleaved Double Dual Boost Converter at email@example.com . It will be a great help Fellipe.
I am still waiting for others comments.....I am ok with idea and papers what other said. But I am more interested to know practical aspect of those topologies.
Finally.......what should be best topology to get 16X voltage gain???? Is this the only one with HF step up transformer with FB? a cheap solution and fewer components.......OR.......(_________)
09-09-2013 05:02 AM
What about a resonant converter (like a parallel resonant converter)? You are wanting to pump kWs at high voltage...this sounds like one of those specialized applications where a load resonant converter would come in handy...use those parasitics to your advantage.
BTW, what's your application? Are you charging capacitors? If so, use a series resonant converter. It works like magic, and you can constant current charge caps to very high voltages very quickly. You can use that leakage inductance in your resonant tank if your turns ratio is really high and your transformer is leaky. Use IGBTs with soft recovery anti-parallel diodes.
09-09-2013 05:03 AM
As Jay said, it would help if tell us the application of your converter. What is the input and output voltages? Do you need isolation?
If you are using going to use any transformer-isolated converter, the voltage gain is not a problem (ouput voltage / input votage). Of course you may have many issues to worry about if your output voltage is too high, but the voltage gain by itself is not limited.
On the other hand, if you wish to use a converter without a transformer, the voltage gain is limited. As you said, it is not practical to design a boost converter with a voltage gain of more than, say, 7, especially when the power is high.
If your application allows you to consider a non-isolated converter, there are some topologies to consider in order to reach the voltage gain of 16.
I sent you some papers by email. Some interesting ones are:
=> A New DC-DC Converter Circuit with Larger Step-up/down Ratio
=> New high-power high-ratio non isolated DC-DC boost converter for fuel
=> Current-Mode Control for a Quadratic Boost Converter with a Single Switch
=> Transformerless DC-to-DC converters with Large Convertion Ratios
One time I participated in a project where we had to power a small AC motor (I think it was about 200W) from a 12V (automotive) battery. The motor required about 200V at the inverter DC link. So we needed a voltage gain of about 17 and isolation was not necessary. The most simple solution we found at the time was to use two stages (two boost converters in series, a.k.a. a quadratic boost converter). So the first one stepped up the voltage to around 50V and the second one to 200V.
Surely, it is not a beautiful solution because the energy is processed twice. But I think is faster to design than a transformer-isolated converter with the same specs and it is competitive on price, size, and efficiency.
09-09-2013 05:04 AM
Thanks a lot for those paper. It will help me a great to complete my analysis on high voltage gain boost converter...
I will let you know the final conclusion and will discuss to best suitable one.
it is not so special application like what u r thinking. it is part of my analysis and literature survey to know the possible topologies in the market to get such 16X boosting.
If u see in solar, fuel cell application this is a very good topic where u have to boost up the voltage nearly 16X to get ~230VAC from 24VDC. Lot of people use HF topologies for ~1kW range. For higher power application people use with a simple boost of ~4X or 5X time boosting than a LF transformer which is much bulky solution. In some of the case people simply use only LF transformer at the output of inverter to boost up the voltage with out any intermediate boost for a bit higher voltage input like 48VDC~96VDC.
If u see and complete a literature survey It looks
1) For higher voltage VDC input people use non-isolated topologies. (a 1:1 transformer will solves isolation issue if application demands for that)
2)For medium voltage people use non-isolated boost with LF step up transformer
3)For lower voltage input people use HF transformer with FB topologies.
But point number 3, where it is limited by power. Most of the case people use it for 1kW~1.5kW. We can still do it upto 5kW but complexity of magnetic properties increases.
A voltage doubler could be a possible solutions but once again problem with boost trap capacitor charging issue with low and high side gate drive.
Anyway a comment by Marty "I would personally recommend to stay with the wisdom of your forefather's hard work on this"....:):)
Seems like we have to listen to them. Otherwise it will take another century ....just kidding.
09-09-2013 05:05 AM
Current fed topologies can push serious power (20kW) at high output voltages. Try searching for "Voltage Fed and Current Fed Full Bridge Converter for the Use in Three Phase Grid Connected Fuel Cell Systems."
Switching losses with the current fed topologies are lower compared to the ordinary hard switched bridge, which is nice.
You have the current fed push-pull, the current fed full bridge, and then the buck current fed FB, and the buck current fed PP. The buck current fed topologies are much easier to analyze and understand in my opinion.
As far as to where a topology is limited by power, that depends upon switching losses, conduction losses, device stresses (snubbers), snubber dissipation, and how big of a heatsink you are willing to use.
09-09-2013 05:05 AM
It is important to know the absolute value of the output voltage and an estimate of the parasitic capacitance in order to determine whether to adopt a hard-switched or resonant circuit strategy.
If for example, the output voltage is say, 3 kV, transformer secondary parasitic capacitance Cp is say 100 pf, and switching frequency fsw is 50 kHz, then the capacitance transitions will cause an additional loss of .5*Cp*V^2*fsw = 22.6W in the switches. Then depending on efficiency and output power requirements, one can make appropriate choices with regards to switching frequency, hard-switched or resonant strategy as such like.
09-09-2013 05:06 AM
Since I used to work in high voltage (>4kV), one could use a flyback transformer/inductor in the inverter stage and it would be suitable for driving a voltage multiplier on the output side. It is sometimes referred to as a Cockroft Walton multiplier. This would cut down tremendously on the losses/parasitics associated with transformers with extremely high turns ratios. These multipliers are stable due to their characteristically high impedance and the ripple is virtually nonexistent considering the output voltages. Your only concern may be finding high voltage picofarad caps :). Also, one has to deburr all sharp edges on conductors and your solder joints must be globular (ball solder joints) to minimize corona effect.
09-09-2013 05:07 AM
Thanks for the comments Girish and Smith.
@Fellipe and Jay....I think we can conclude to go with push-pull or half bridge, as this will be more suitable for such a high gain voltage boosting with ~1kW power rating? Resonant will help us to increase efficiency .What is ur idea or thoughts on this two topology? I don't want to go with full bridge converter because I am not looking for such a high power like ~20kW or so...I guess half bridge or push-pull can use upto 1kW~1.5kW rating?? and may be a optimal selection....
It will be great if others also comments on this conclusion.
Non-isolated topology is not a best choice for such a high gain voltage boosting. Disadvantages are more as compare to isolated topologies.
09-09-2013 05:07 AM
Personally, I think it's possible to push that power with a current-fed PP, but transformer leakage inductance will be an issue for the PP stage...you will need a turn off clamp snub. The buck current fed is easier to implemet: all you need is a "50%" square wave oscillator for the PP switches with brief overlap to give the buck inductor a current path during PP switch commutation. All of the control is at the buck stage, since the PP stage simply free-runs. The inductor is at the input to the PP stage, so no HV inductor design, and multiple outputs are cake since you are designing a single low voltage inductor. Also, switching losses in the PP stage are much less than a hard-switched PP due to the quasi-ZVS and the splitting of inductor current during commutation.
I think you can tell I love this topology. I'm sure there are other (and possibly better) options.