Cooler Offline smps Bridge Rectification

My offline (nonisolated) smps bridge rectifies the 120VAC line voltage prior to DC to DC conversion. In my app, the bridge rectifier wastes about 2Watts and needs heat sinking.. What can I do to improve the efficiency? Maybe with active rectification using mosfets or IGBTs? But wouldn't that be really tricky? Are there modules for offline power bridge rectification? D from BC

Reply to
D from BC
Loading thread data ...

If the application allows this: reduced the filter capacitance, so that there is more ripple voltage across it. This will decrease the peak current into the cap and through the bridge, thereby reducing dissipation. Of course, line regulation capability of the regulator must be up to the job, and hold-up time for brown-outs may be shortened.

Running the bridge hotter also reduces its dissipation by reducing forward drop across the diodes.

Paul Mathews

Reply to
Paul Mathews

Actually, low frequency rectifier losses depend on the average current and aren't signifigantly affected by peak-to-average ratios. The filter caps will appreciate this consideration, though.

Sharing their rise with other components is a bad idea also, as these may have a possitive temperature coefficient of loss, or lower permissible operating limits.

RL.

Reply to
legg

snip

A full-wave doubler will have half the rectification losses of a full-wave rectifier at the same average line current. One of the (few?) advantages of switched input voltage range.

RMS current ratings aside, your capacitance should be calculated based on permissible voltage ripple (droop) and hold-up time, where applicable.

RL

Reply to
legg

These 'truisms' are often repeated and often true, but not always. FIrst, it's average POWER that matters, not average current. To the extent that voltage drop per rectifier is constant, average power and average current will track linearly. However, rectifiers have a resisitive component to their characteristic, which, together with their exponential diode characteristic, means that voltage drop rises with current. This becomes significant, particularly for high crest ratios and small rectifiers. Second, given a certain amount of power dissipation P in a component, the sum total of heat added to surrounding components will be equal to P. However, the rise in operating temperature of that component will be approx P*G, where G is its thermal resistance. My suggestion was to allow the temperature of the rectifiers to rise, and I meant by raising its thermal resistance. This turns out to slightly reduce P in the case of Si rectifiers, with their negative tempco of Vf. It also slightly reduces P. In many cases, eliminating a heatsink or reducing its size also diminishes the amount of heat added to adjacent components, due to reduced radiant coupling between the hot component and adjacent components. Here's a rough calculation to illustrate the point:

  1. heatsink reference case: P = 1.6 V * 1A = 1.6 W, G = 20 C/W, dT = P * G = 32 C
  2. remove heatsink case, iterative approximation: P = 1.6 V * 1A = 1.6 W, G = 40 C/W, dT = P * G = 64 C --> 32 C hotter --> Vf(hot) = Vf - dT * 4mV/C = 1.6 - 0.128 = 1.472 V

In other words, at 1 A average, operating the bridge rectifier at G =

40 C/W raises its temperature about 32 C, resulting in a 128 mW reduction in power loss, which is about 8% of the original 1.6 W. This approximation is meant for illustration, only. You can quibble about whether 128 mW is significant. In my designs, every last percent counts.

Operating Si rectifiers at higher temperatures, even low frequency rectifiers, is a common technique for improving power conversion efficiencies. Although it raises the temperature of the rectifiers themselves, it actually reduces the total amount of heat added to the surrounding components. Paul Mathews

Reply to
Paul Mathews

Cool..Kinda like fighting fire with fire.. D from BC

Reply to
D from BC

snip

Well put.

Only a few manufacturers actually publish useful ratings information that considers peak-to-average stress levels. With commodity bridge rectifiers, the published data can amount simply to plagiarized rubbish that may never be correctly proof-read while the company lasts.

Keep in mind also, that input rectifiers are often associated with commodity filter components in fan-free situations. Safety approved parts have a limited thermal environment and should be applied carefully in that case, as they are forced to inhabit the same proximity, and employ the same (usually printed) wiring.

Suggesting that a smaller heatsink will radiate less (~net) while it's surface temperature rise above ambient doubles, may be a misunderstanding of the nature of radiation. While hardly logarithmic over such a narrow range, the relationship between radiation and deltaT is worse than linear in the situation cited..

RL

Reply to
legg

Heat flows through parallel conducted and radiated paths, and models of temperature can get quite complex in a multi-component environment. Your suggestion that a hotter component can have unintended consequences is valid. My point was simply this: The total amount of heat added to the power supply environment by the rectifier(s) is proportional to their power dissipation, which actually declines when they are operated at higher temperatures. In many power supplies, the primary external determiner of asymptotic component temperature is the total heat evolved in the enclosure. There are certainly cases where the peak temperature of an adjacent component is a stronger factor. However, note (among many effects) that the removal of a heatsink often effectively moves the hot component farther away (by making it smaller). So, it is by no means a foregone conclusion that heatsink removal and the consequent rise in temperature of a bridge will elevate the temperatures of adjacent components, particularly since removal of the heatsink lowers its dissipation. Paul Mathews

Reply to
Paul Mathews

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.