Low Noise Switching Regulator

I should have put in the Current required(sorry). I need >1.5A. Cost is not a limitation. I dont mind spending more on the regulator now rather than have my video all messed up later on.

The way I understand it is that the 'low noise' regulators are low noise because they smooth out the corners of the switching waveform and hence reduce harmonic energy and is not due to external component filtering. I have seen other low noise regulators on the linear site with low external component count.

Thanks.

You say you don't have room for heat sinks, implying that you're moving some hefty current. How much are we talking about?

Any switcher that produces very little noise is going to include "lots of other external components" since these will be needed to filter out the noise that is inherent in the operation of a switcher. A simple buck regulator will probably need at least two inductors and several capacitors in addition to the switching components and control logic.

What output current do you need? Is cost a serious limitation?

Rounding the corners of the switching waveform reduces the highest harmonics that tend to radiate from any antenna. But it also decreases the efficiency, making it more likely that you will need heat sinks. But no amount of rounding gets rid of the interrupted current going into the buck regulator which conducts noise back out into the 15 volt supply and ground busses. I prefer to contain the radiated noise of fast switching, high efficiency designs with good layout and grounding techniques and shielding. Then I address the conducted noise with low pass filtering components on both sides of the regulator.

If you pick a candidate design, we can discuss how to apply those noise containment techniques on that example.

How about syncing the switcher to the pixel clock in the video board so that the noise aliases to DC?

Maybe you could use the switcher to get down to 12.5 Volts and use an LDO or an LDO + NPN power transitor for the last 1.5 V?

Is there any airflow?

Also, don't forget that switchers also put noise on the power input voltage rail. So you will need to filter it, too. This increases the board space of a switcher to the point where it may not save space over a linear regulator, even for your 6+ Watt drop. Especially if there is guaranteed airflow.

--Mac

I would try to go for a linear regulator. A switcher at 80% efficiency will dissipate 3.3W. A linear regulator will dissipate 6W. If you use

I'm talking to a UK transformer company at the moment, for the LT3439 since the coiltronics supplier in the UK have a MOQ of 300, and are not at all helpful.

I've been trying to catch up on transformer design (groan)and have come across Planar E cores from Ferroxcube, who now own the Philips breed. These cores sit in a cut out in the PCB.

martin

Serious error. All shortcuts have disappeared. Screen. Mind. Both are blank.