Switch mode PSU with linear output stage

[PinheadBE]

Hi,

I vaguely remember having seen a design of a PSU combining the best of both worlds, with a first switch mode stage for efficiency, followed by a linear LDO regulation stage for better noise performances

The two stages were kind of syncronized, so that the drop out voltage on the linear stage always remains minimized.

Does this ring a bell to anyone ?  Anyone remembers having seen this ?

Thanks a lot.

[nctnico]

IIRC R&S has these kind of PSUs

[PinheadBE]

I was looking for a design so that I could add a linear low-loss stage to a noisy switch mode PSU

[Kean]

There are a bunch that work like this.  The switching section is set to be a few volts above the linear output stage.

I believe the R&S HMC804x PSU is considered a "linear switching" power supply, so high efficiency, low weight, and low-ish noise.  There are probably others they have like this.

The EEZ BB3 DCP405 module is an open design using this concept so you can check the schematics and I think they published some design notes on their website.  See https://www.envox.eu/dc-power-modules/dcp405-power-module/

[coromonadalix]

some Kikusui has this topology

phase controlled /  buck / and end with some linear section

[sonpul]

Riden has this topology with the letter P in the marking. For example Riden RD6006P. This is a great option if you need an inexpensive solution.

[robert.rozee]

see:
https://sound-au.com/articles/preregulators.htm#s5


cheers,
rob   :-)

[PinheadBE]

Riden has this topology with the letter P in the marking. For example Riden RD6006P. This is a great option if you need an inexpensive solution.

I knew Riden made good modules lately, but I wasn't aware it was *THAT* complex.  Thanks for the schematics !

see:
https://sound-au.com/articles/preregulators.htm#s5

Very interesting in my case where I already have the Switching part !   I surely will experiment something this way !   Thanks !

[sonpul]

PinheadBE

keep in mind, the Riden RD6006P regulation is made relative to"+". This means that the output of the main power supply  " - " should not connect with ground, as sometimes happens. Because then connected to the oscilloscope scheme with grounding will make a short circuit.

[PinheadBE]

Yes, thanks.   I will keep that in mind.
Part of my lab bench is equipped with an isolation transformer.   That shoud help keeping the magic smoke within my scope.

[DrGeoff]

This was usually done in SMPS that had a primary rail and several auxilliary rails. Back in the 80's we had SMPS that had a beefy 100A 5V rail and a couple of 1A +/-12V rails. The SMPS regulation was performed using feedback from the 5V rail and the transformer windings for +/-12V did not have the same load regulation as the 5V rail so were fed through a the 7812/7912 regulators to maintain good regulation. Linear regulators won't help you with noise, it passes straight through them as the frequencies are usually very high (>100kHz). Usually good ferrites and chokes were used to keep the noise down.

[tridac]

There were several of that type, even decades ago. Some even used a thyristor as a preregulator. The key thing is that the pre and post regulators track, or are in sync. Maintaining a constant voltage drop across the linear part to minimise power dissipation and improve efficiency. The main reasons for such a design approach, while providing a low noise at the output...