Author Topic: Ultra quiet step down switching converter topology? 28v ->10V, 5A  (Read 1685 times)

0 Members and 1 Guest are viewing this topic.

Offline max_torqueTopic starter

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
I have an existing power supply device that drops around 28Vdc to 10Vdc for a Defense application that currently only requires an average current of 500mA, enabling that psu to be a simple linear dropper, that obviously dissipates an average of 9W, but that is very quiet (no switching or high frequency content).

A new application i'm looking at looks to require around 3A continuous, which is getting difficult to do linearly as the power required to be dissipated is now around 54W.

As such i'm looking at options for moving to a switching buck converter, but i need to be very careful of the EMC aspect, as the MIL RFI/RFC standards are draconian for this application .

So, some questions:

1) What topology should i start looking at?   Overall efficiency is not really important, the current device necessarily handles 9W of loss, which would be a hugely in-efficient switcher for just a 3A transfer.   Only voltage reduction is required as there is always at least 8 volts of voltage headroom.

2) is there any merit in following the switching stage with a low voltage drop linear one?  My suspicion is that at the frequencies that are important (harmonics of the base frequency) that high frequency content will probably pass right though any linear stage anyway?  One advantage of a switcher with a linear follower is that the basic switcher can be very slow and not that responsive, relying on the linear bit to provide the low impedance output across a decent bandwidth.

3) If i significantly oversized the switching elements (mosfets), then i could accept higher switching loses, and hence have slow rise times, is this practical? There is obviously a trade off between switching speeds and switcher size and cost, with lower speeds meaning larger inductors and larger capacitors to get the same current ripple.

4) Any merit in a parallel dual path architecture to interleave the current pulses from the switching elements

5) Off the shelf switching controller, or DIY approach with custom logic?



Space/size is not really an issue, nor is cost particularly, and the existing linear PSU does not include any current limiting beyond some output resistance and some PTC fusing

« Last Edit: July 19, 2019, 06:26:43 pm by max_torque »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #1 on: July 19, 2019, 07:05:47 pm »
LT's Silent Switcher line should be of use.

I don't think I'd suggest a custom design: even with the quirks that a lot of integrated controllers have, the fact is they offer so much more functionality than a modest sized discrete circuit can.  You'll either need hundreds of components to pull it off (the last discrete design I did, used about 100 components per channel), or as many hours programming the MCU/FPGA to provide equivalent functionality in a smaller package.

I don't think you can get away from noise in general: even resonant architectures still have spiky dV or dI/dt in places, stuff like diode recovery and all that, and that will be reflected in the emissions spectrum at some level.  So, expect to need filtering and shielding regardless.  What you're really investigating is, how much filtering is ultimately needed, and at what frequencies?

There may be some interest in a phase interleaved design for example, which acts to reduce or null the fundamental ripple (in effect, multiplying the ripple frequency by N for N phases).  That, with a Silent Switcher style control, could give quite handy improvements.  That could make the difference between, say, board-level shields, versus a full shielded enclosure, or board-level filtering versus feed-thrus.

The aim is to push the ripple up to a higher frequency, so that filtering the fundamental is easier, while also bringing down the worst case harmonics, so that your filter bandwidth does not need to be insane.

Higher Fsw also helps reduce the size of components, but at the expense of higher switching losses.

It doesn't need compressed into so little spectrum as to be a pure tone; like, you could make a class E amplifier at 13.56MHz and keep everything very soft and free from harmonics, and filtering would basically amount to 13.56MHz traps (which could be a bit of space savings actually, over a lowpass of equal attenuation).  Controlling something like that would really be the bigger issue though, and also the control bandwidth would fall outside of the notch filter, which really means that you don't have any filtering from load to source (i.e., the control must respond almost instantaneously to load changes).

I don't know if that kind of filtering is necessarily intended in your application (i.e., in effect, load variations being reflected as source emissions), but it may also be equivalent to the specifications you're given.

If you have to meet dropout/holdup requirements (avionics often have this requirement; yours may too?), you may need a lot of energy storage anyway, and turning that into a highly effective filter may not be much additional cost (and, cost in the most general sense, not just direct BOM cost but cost to the size or weight as well).

If size/weight is priority, it's interesting to note that the attenuation-per-energy-storage ratio (and therefore, the attenuation-per-volume, more or less) is maximized with a modest number of stages (the number depending on how much attenuation is required).  That is, for a given attenuation, a 3rd order (CLC) filter would need a fairly low cutoff frequency and therefore relatively large components, but a 5th or 7th order uses more but smaller components; while a still higher order filter uses simply too many parts with size of diminishing returns (because number of components keeps rising, while cutoff frequency approaches Fstop).  So there's a minima between extremes.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 7521
  • Country: ca
  • Non-expert
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #2 on: July 20, 2019, 12:40:30 am »
5) Off the shelf switching controller, or DIY approach with custom logic?
Space/size is not really an issue, nor is cost particularly, and the existing linear PSU does not include any current limiting beyond some output resistance and some PTC fusing

If space and cost are not an issue, are there no off the shelf modules available that are certified to those specific MIL standards, have you looked?
Maybe certified to a similar standard perhaps. Downside of this is these modules are usually potted so any modifications you do would likely just be additional input or output filtering.

3) If i significantly oversized the switching elements (mosfets), then i could accept higher switching loses, and hence have slow rise times, is this practical? There is obviously a trade off between switching speeds and switcher size and cost, with lower speeds meaning larger inductors and larger capacitors to get the same current ripple.

Yes slowing rise time with series gate resistor or similar will cut down EMI.

Look at the standard for emissions graphs, usually they will have some higher allowed dB rating in the lower frequency range. If your switching frequency and second harmonic is below the point where they drop the maximum emissions down, that would probably help. Do you have a copy of the standard?
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 

Offline max_torqueTopic starter

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #3 on: July 20, 2019, 01:51:12 pm »
Thanks for the replies!

Those LT parts look interesting, as they package most of the components into what is persumably a better RFI optimised package than i could manage!

My biggest choice is what switching frequency to use?  There are higher limits at lower frequencies, so that could help. Limits are currently expected to be at the least Land Class C (DEF STD 59-411), but i'd like to get a Class B to give future headroom for the device.




Doing a 10Khz "switcher"  (is it even classed a switcher at that frequency!  :-DD )  looks to have some merits in terms of having the highest limits, but of course, it'll be a 'mare to filter due to that low frequency.


Some of those LT silent switcher parts do have a frequency control pin, so i can at least set and hold a fixed switching frequency. I wonder if a fundamental switching frequency of say 100 to 200Khz has some merit?


 

Offline IDEngineer

  • Super Contributor
  • ***
  • Posts: 1950
  • Country: us
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #4 on: July 20, 2019, 08:37:39 pm »
Doing a 10Khz "switcher"  (is it even classed a switcher at that frequency!
I am embarrassed to admit that back when I was first working on what we then called "DC-DC power supplies", a 10KHz frequency wasn't considered low at all.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #5 on: July 21, 2019, 01:24:12 am »
And hey, you can use cheap, reliable BJTs for it too!... :P

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #6 on: July 21, 2019, 02:04:23 am »
does the topology involved with low noise decrease transient regulation etc
 

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #7 on: July 21, 2019, 03:26:56 am »
Have you worked on a switcher under these restrictions before?  Or maybe someone else in your company?  It's always nice to learn from other people's mistakes.

Do the relevant standards for conducted emissions only consider the input leads? Do you have requirements for the quality of power output?

Also keep an eye out for other MIL requirements like input over voltage conditions (spikes/surges/RF noise immunity).

In terms of 'just get on with it...'
  • Seems like a good place for a buck converter.  Multi-phase could be useful, but the power level isn't super high
  • Using a linear post regulator may not be much use unless you want low power standby.  Low power standby often involves running the converter in pulse skipping mode, which results in something of a sawtooth output voltage waveform
  • Oversizing the MOSFETs and using large age resistors might be helpful - but using larger components will mean larger leakage inductances and stray capacitances which could cause trouble
  • Fancy control schemes might not be particularly useful, and there's a lot to be said for making things simple.  However, getting something with spread-spectrum switching could help spread out the fundamental switching energy a bit.
  • A buck converter in Discontinuous Conduction mode can form an LC resonator (main inductor oscillates with diode and switch capacitance), maybe in the 5 - 30 MHz range. Fitting a small snubber here could help damp that out.

I agree with T3sl4co1l that LT Silent Switchers may be useful.  In general, I would suggest looking at integrated power controllers (i.e. controller + MOSFETs in the same package) because they shrink the most critical switching loops which experience the highest dV/dt.

I've never used one, but Vicor claim to have converter modules with fancy soft switching inside (reduced dV/dt and dI/dt), which could be quite good for knocking down higher frequency EMI (i.e. frequencies much higher than switching frequency).  The ripple from the basic converter switching frequency won't change much .

Spread spectrum operation may help, too.

EDIT: fixed list formatting
 

Offline max_torqueTopic starter

  • Super Contributor
  • ***
  • Posts: 1327
  • Country: gb
    • bitdynamics
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #8 on: July 21, 2019, 11:11:11 am »
I have experience of designing to EMC limits, including high power 3 phase motor controllers, but not directly with DC/DC to those draconian limits!

The concern i have with using an COTS unit such as the Silent switcher is that the development targets for that device may be different to mine.  They will have had some trade off between efficiency and noise, simply in order to meet thermal requirements, but also to support the very necessary "high efficiency" line in the marketing bumph......

In my case, efficiency really doesn't matter that much. The current linear design is "Low efficiency" by design after all, at max continuous output current of 1A, it delivers 12W to the load, and pulls 28W from the source, loosing 16w of heat on the way

For the new switching design, i am aiming for a 3A contiunous output at 12v (36W), so for the same "Loss" as the linear device at it's max continous rating, that would be 52W of input power (36+16 = 52), equivalent to a conversion efficiency of "just" 69%.

I terms of dV/dT, i'm switching 28V, and the question becomes "how slowly can i turn on the mosfet and it survive?"  If we took my mythical 10KHz fundamental, that's a cycle time of 100us, and given that my device pretty much just chops the voltage in half (ie 50% duty cycle) i could immagine switch times of as much as 10us, for an average dV/dT of 2.8 MV/sec

One option i did think of to avoid having to run the switcher in discontinuous mode was to conbine the linear converter with the switcher, with the linear path having an output voltage just a bit higher than the switching path, and the linear path having suitable output imedance to cause the voltage to sag into the range that enables the switcher at higher loads.  That would of course need careull control and sequencing from some form of custom control system.....

I suspect i will have to perhaps build a couple of prototype supplies, one with a COTS device and one with a more custom solution, and pre-test both devices to see which offers the best solution??
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #9 on: July 21, 2019, 11:11:45 am »
1) What topology should i start looking at?   Overall efficiency is not really important, the current device necessarily handles 9W of loss, which would be a hugely in-efficient switcher for just a 3A transfer.   Only voltage reduction is required as there is always at least 8 volts of voltage headroom.

I would start by avoiding topologies which have the highest discontinuous voltages and currents like flyback in favor of topologies which have the lowest input and output ripple current requirements.  The Cuk converter might be the best in this respect.

Or use an inverter for a majority of the step-down but this will require linear regulation to produce a regulated output.  Note that controllers like the LT1533 mentioned below can be used to drive an inverter making it even lower noise.

Quote
2) is there any merit in following the switching stage with a low voltage drop linear one?  My suspicion is that at the frequencies that are important (harmonics of the base frequency) that high frequency content will probably pass right though any linear stage anyway?  One advantage of a switcher with a linear follower is that the basic switcher can be very slow and not that responsive, relying on the linear bit to provide the low impedance output across a decent bandwidth.

In my experience there is not.  Linear regulators have poor rejection at the frequencies of interest compared to passive filtering and frequency response is limited by the output filtering anyway.

Quote
3) If i significantly oversized the switching elements (mosfets), then i could accept higher switching loses, and hence have slow rise times, is this practical? There is obviously a trade off between switching speeds and switcher size and cost, with lower speeds meaning larger inductors and larger capacitors to get the same current ripple.

Sure that is practical and some switching controllers are specifically designed to take advantage of it by explicitly controlling dI/dT and dV/dT.  Linear Technology makes some like this and they would be my first choice no matter what topologies is used.  Linear Technology application note 70 is the best place to start for these parts.  Note that only small losses in efficiency are required to make a big difference.

Quote
4) Any merit in a parallel dual path architecture to interleave the current pulses from the switching elements

Multiple phase designs are more appropriate at higher power levels where parasitic inductance and ripple current become significant problems limiting maximum power from a single phase.

Quote
5) Off the shelf switching controller, or DIY approach with custom logic?

With parts like the Linear Technology LT1533 and similar, there is no reason to do a fully custom design.

Quote
Space/size is not really an issue, nor is cost particularly, and the existing linear PSU does not include any current limiting beyond some output resistance and some PTC fusing

I remember military requirements including a lot of requirements to handle overloads at the input without failing.
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: Ultra quiet step down switching converter topology? 28v ->10V, 5A
« Reply #10 on: July 21, 2019, 10:56:24 pm »
I have an existing power supply device that drops around 28Vdc to 10Vdc for a Defense application that currently only requires an average current of 500mA, enabling that psu to be a simple linear dropper, that obviously dissipates an average of 9W, but that is very quiet (no switching or high frequency content).

A new application i'm looking at looks to require around 3A continuous, which is getting difficult to do linearly as the power required to be dissipated is now around 54W.

As such i'm looking at options for moving to a switching buck converter, but i need to be very careful of the EMC aspect, as the MIL RFI/RFC standards are draconian for this application .

So, some questions:

1) What topology should i start looking at?   Overall efficiency is not really important, the current device necessarily handles 9W of loss, which would be a hugely in-efficient switcher for just a 3A transfer.   Only voltage reduction is required as there is always at least 8 volts of voltage headroom.

2) is there any merit in following the switching stage with a low voltage drop linear one?  My suspicion is that at the frequencies that are important (harmonics of the base frequency) that high frequency content will probably pass right though any linear stage anyway?  One advantage of a switcher with a linear follower is that the basic switcher can be very slow and not that responsive, relying on the linear bit to provide the low impedance output across a decent bandwidth.

3) If i significantly oversized the switching elements (mosfets), then i could accept higher switching loses, and hence have slow rise times, is this practical? There is obviously a trade off between switching speeds and switcher size and cost, with lower speeds meaning larger inductors and larger capacitors to get the same current ripple.

4) Any merit in a parallel dual path architecture to interleave the current pulses from the switching elements

5) Off the shelf switching controller, or DIY approach with custom logic?



Space/size is not really an issue, nor is cost particularly, and the existing linear PSU does not include any current limiting beyond some output resistance and some PTC fusing

Hi,

whats the PARD level allowed in mV ? I think well below a mV over 0-10MHz is possible, if ...

My approach would be:

- a good input filter, cutoff well below switching frequency
- Low switching frequency
- risetime limited switching signals to cut harmonics down (inefficient, but thats not the main issue).
- low fT bipolar transistors
- LC output filter for switcher
- linear postregulator with some preload
- very good shielding between switcher and postregulator

a resonant Royer could do, maybe look up some Jim Williams App notes.

regards
  Wolfgang


 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf