Has anyone done any recent work with hybrid switching / linear power supplies, especially at low voltages?
I was thinking a simple linear power supply (15-18V) could feed a DC-DC buck converter that would provide a regulated <=13V and an linear LDO would have to drop the last few hundred mV to get to 12V.
This way you get something simple, compact, and easy to assemble from off the shelf modules. There are a few DC-DC buck converters on Amazon for a few dollars (LM2596 and MP1584EN) and LT3045 modules are available from ldovr.com.
I was thinking a simple linear power supply (15-18V) could feed a DC-DC buck converter that would provide a regulated <=13V and an linear LDO would have to drop the last few hundred mV to get to 12V.
This way you get something simple, compact, and easy to assemble from off the shelf modules. There are a few DC-DC buck converters on Amazon for a few dollars (LM2596 and MP1584EN) and LT3045 modules are available from ldovr.com.
It depends on power and efficiency needs. If you're feeding a preamp and the unregulated voltage is close to the desired voltage, then an LDO is the easy approach.
If you have large differences between Vin and Vout or large currents, then a switching regulator would work better.
I've successfully combined the two in an isolated SMPS where I would have separate windings for supplying preamps or crossovers, that's where I usually add linear regulators.
If you have large differences between Vin and Vout or large currents, then a switching regulator would work better.
I've successfully combined the two in an isolated SMPS where I would have separate windings for supplying preamps or crossovers, that's where I usually add linear regulators.
Do away with the linear supply completely and use an isolated smps, like a laptop brick, to feed the buck regulator. This way you are free from the mains earth and any potential ground loops. You don't need to connect the chassis ground to earth, for safety, with the laptop brick
Most laptop bricks I've worked with (probably the new ones) have the secondary ground tied to earth.
Hibrid schemes are ok, but you say about input voltage 15-18 Volts, and output 12 V - the real efficiency will be about the same as for just linear scheme, but much more difficult. (Yes, it depends of value of a current).
Low voltage dc-dc modules are ok - their noise isn't "bed" for filtering. High voltage dc-dc suppies are bed - they have too big amplitude of high frequency ripple, so they are very powerful source of "bed" commonn noise.
Low voltage dc-dc modules are ok - their noise isn't "bed" for filtering. High voltage dc-dc suppies are bed - they have too big amplitude of high frequency ripple, so they are very powerful source of "bed" commonn noise.
Doesn't have to be the case as the late Jim Williams demonstrated:
YouTube
See this application note:
http://www.analog.com/media/en/technical-documentation/application-notes/AN118fb.pdf
You can also get a very low noise switching supply by taking the gate drive and filtering it so that the transition or slew rate is reduced. This drives the device through the linear portion of its curve -- the tradeoff is that you lose a lot of efficiency but at the same time shed a lot of noise. Linear (now owned by ADI) has a bunch of driver chips which do this.
Last edited:
Yes, you are right, there is low noise dc-dc types, but, unfortunately, they are not in very common use.
There are also resonant mode controller ICs that are getting more sophisticated, thus helping to make low noise switching supplies easier to build. PWM Controllers and Resonant Controllers | Products | Power ICs | TI.com
Jan Didden made a switch-mode supply followed by low-noise series regulators some time ago. The whole thing can be supplied by a USB charger.
The SilentSwitcher | Linear Audio NL
The SilentSwitcher | Linear Audio NL
The problem with a laptop brick is wanting to have an adjustable output voltage to feed the linear regulator.
Looking at spec sheets for transformers, the primary is usually rated for 115VAC and my house is consistently 120V. Factor in the over-voltage some manufacturers hint at to combat the huge amount of regulation in smaller toroids, and I was afraid of having a wide range or secondary voltages to account for. That would mean a potentially large voltage drop for an LDO, and I thought that feeding the LDO via a closely dialed in SMPT voltage might be a good idea.
Maybe, maybe not? I just have not found any such products.
A simple option is to give up on an SMPS or DC-DC and use a higher current linear (pre)regulator and just heatsink it well, and feed that into an LDO.
Looking at spec sheets for transformers, the primary is usually rated for 115VAC and my house is consistently 120V. Factor in the over-voltage some manufacturers hint at to combat the huge amount of regulation in smaller toroids, and I was afraid of having a wide range or secondary voltages to account for. That would mean a potentially large voltage drop for an LDO, and I thought that feeding the LDO via a closely dialed in SMPT voltage might be a good idea.
Maybe, maybe not? I just have not found any such products.
A simple option is to give up on an SMPS or DC-DC and use a higher current linear (pre)regulator and just heatsink it well, and feed that into an LDO.
- Status
- Not open for further replies.