24V Vehicle PSU / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

24V Vehicle PSU

previous 1 2

Move Thread

LAN_403

Conundrum

Wed Feb 25 2009, 07:07PM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4061

Beefed up switched capacitor inverter?

Just a thought, the noise *should* be lower due to the lower operating frequency and it isn't hard to build.
-A

idea *2, build the SMPS in a heavily shielded box with feedthroughs and smoothing up the wazoo.

Mattski

Thu Feb 26 2009, 04:03AM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

The Cuk converter can be made so that it has (in the case of ideal components) zero ripple on input and output current in the steady state, if L1 and L2 are wound on the same core. This doesn't seem to be described in the Wikipedia article on Cuk converters, but it's in the book I have (Pressman, Switching Power Supply Design). This paper also treats the idea of ripple cancellation with coupled magnetics in a more generalized fashion.

I don't have any specific experience with making Cuk converters, but it seems to be a good candidate for a low-noise switching power supply. It will no doubt require more up front effort than a dual 12V battery design, but it might be worth it.

KLH

Fri Feb 27 2009, 03:43AM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

Don't worry about radiated magnetic fields if you use toroids in your converters; the magnetic field won't go anywhere. As for RFI radiating from conductors, a few ferrite beads slipped onto the input and output leads, selected for the proper frequency, would work very well (use the Amidon databook or another magnetics company's (Steward) resource to help you select the material). Alternatively, the power supply can be placed VERY close to the equipment it's powering.

A boost converter is a very good idea, as suggested by others, mainly because the components required for your power supply specifications are readily available (MOSFETs, simple control ICs such as op amps and gate drivers, and 20 amp inductors).

Push-pull converters are viable, but they come with vastly increased complexity with the design of the transformer (and it is VERY complicated to get an efficient, low loss, lightweight transformer). However, higher powers warrant this topology. The increased design effort required for this topology is simply not necessary for your power requirements; in the end, the boost converter is the best choice.

Neither of these topologies, designed well, will require electrostatic or magnetic shielding, as long as ferrite beads and toroidal inductors / transformers are used properly.

GeordieBoy

Fri Feb 27 2009, 02:49PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Although the boost converter has the simplest power circuit, there are some hidden benefits to the push-pull converter:

1. There is isolation between input and output sides, effectively disconnecting the output side from any input side noise. (You can even employ a faraday screen between pri and sec in the transformer if necessary.)

2. Power dissipation is spread between two switches and two diodes, instead of all being dissipated in a single switch and diode.

3. The closed magnetic circuit of the push-pull transformer (no air-gap) will result in less radiated RFI than the gapped boost inductor.

4. If designed carefully the push-pull converter can run at a duty ratio very close to unity. This makes input and output ripple currents low, and minimises energy storage in the output buck choke. Again reducing RFI emissions. A boost converter from 12V to 24V will always run near 0.5 duty ratio leading to greater input and output current ripple figures.

5. The fundamental switching frequency of the push-pull converter (and it's odd harmonics) cancel out in the full-wave rectifier at the output of the push-pull converter so its output has a less rich RFI spectrum.

6. Closed loop control of the buck derived push-pull converter is simpler than that of the boost converter due to the lack of right-half-plane zero in the converter's forward transfer function.

7. Forced reverse recovery of the boost diode and subsequent ringing in the boost converter is always a potential source of RFI. This is easier to manage in a push-pull converter where the leakage inductance of the transformer limits the rate of rise of secondary current.

8. Although the original post made no mention of it, output short-circuit protection (or over-current protection) is easier to implement in the push-pull converter. The boost converter always has a DC path from input to output making it unable to shut down the output in the event of a short-circuit or excessive load current.

As stated previously, almost any topology could be made to work - Those are just some comments in defense of the push-pull converter from practical design experience.

-Richie Burnett,

KLH

Fri Feb 27 2009, 06:37PM

Registered Member #1819 Joined: Thu Nov 20 2008, 04:05PM
Location:
Posts: 137

GeordieBoy wrote ...

2. Power dissipation is spread between two switches and two diodes, instead of all being dissipated in a single switch and diode.

3. The closed magnetic circuit of the push-pull transformer (no air-gap) will result in less radiated RFI than the gapped boost inductor.

7. Forced reverse recovery of the boost diode and subsequent ringing in the boost converter is always a potential source of RFI. This is easier to manage in a push-pull converter where the leakage inductance of the transformer limits the rate of rise of secondary current.

2. Power dissipation may be spread out, but with increased total losses, especially in the secondary rectifiers. The majority of energy loss in SMPS designs occurs in output rectifiers. With only one diode drop to overcome, compared with two for a full bridge output for a push-pull transformer, the boost converter has much lower losses overall, provided proper components (MOSFETs and inductors) are selected. The losses in the secondary rectifiers can be reduced to that of the boost converter's losses with a center-tapped secondary, with the added benefit that each diode sees half the total power loss, but this means poorer core and winding utilization. Again, with the required power level, components are readily available and fairly cheap for the boost converter.

3. Toroidal cores (IP for boost, ungapped ferrite for push-pull) will not radiate any significant amount of EMI. A boost converter doesn't imply an inductor with a wide-open magnetic circuit.

7. Reverse recovery will occur in both topologies. There is only ringing when the boost converter operates in discontinuous mode, which it even shouldn't be running in for this application. Continuous inductor results in no ringing, very small core loss, hysteresis loss, and AC winding loss, all of which are higher in the push-pull topology. AC winding losses are also very problematic in push-pull transformers.

One of the additional advntages of using a boost converter for this application is that no custom magnetics have to be wound.

GeordieBoy

Sat Feb 28 2009, 03:29PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

For those trying to compensate a continuous-mode boost converter there is a healthy discussion about the dreaded right-half-plane zero here:

-Richie,

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob