Precision Rectifiers: Get Perfect DC from Sinusoidal Wave

  • Thread starter Thread starter smruti
  • Start date Start date
  • Tags Tags
    Precision
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
Sign up Log in
11 replies · 2K views
smruti
Messages
20
Reaction score
0
If you are given a pure sinusoidal wave and a perfect DC is required from it. then what is the best way to get it? should we go for full bridge rectifiers? or any better way is there?
 
Engineering news on Phys.org
smruti said:
If you are given a pure sinusoidal wave and a perfect DC is required from it. then what is the best way to get it? should we go for full bridge rectifiers? or any better way is there?
Are you , for your purposes, content to regard a full-wave rectified sinewave as "perfect DC"?
 
Pedantically it is impossible to convert a perfect sine wave to a perfect DC level. There will always be some level of distortion (If only at the switchover point; such as with the motor generator pair -- and a pedantic, noiseless DC generator.)

The question is, how close to perfect do you want? What types of distortion can you live with and how much of each can you live with? Rectifiers are really pretty good, with some filtering. Their big problem is ripple voltage which is caused when the load starts to exceed the capacitor's storage capacity. Technically any load will do this, but practically you will design for some acceptable level of ripple. Bigger caps can help. If your load is constant, you might even throw in an inductor building a low pass filter well below the sine wave's frequency. (Though that is as rare as hen's teeth. Usually there's some variation in the load, plus large power inductors are expensive.)

If the wave is 60Hz (or 50Hz), try following your rectifier with a linear voltage regulator. Many of these have extra circuitry to eliminate ripple at that frequency. Even at other frequencies, they may help a lot.

In the end, you should be able to eliminate distortion for most purposes.
 
Also you need to consider frequency response. When the AC signal goes away. how quickly must the DC signal track it. This is a balancing act with how much ripple you can accept in your output.

I"m assuming you are familiar with the standard precision rectifier circuits. What about these circuits do not meet your needs?
 
a long time ago I learned that you should use slo diodes for ac rectification as fast diodes can introduce rf noise and need snubbers, as long as you use the right capacitance you ripple will be low, then you need an additional small value cap also to bypass because of internal resistance in the big cap at high freq, linear technology makes some better replacement 3 terminal regulators with better specs in every way than the lm317 they also come in negative, remember to use a 10uf cap after the regulator also
 
thankz said:
a long time ago I learned that you should use slo diodes for ac rectification as fast diodes can introduce rf noise and need snubbers, as long as you use the right capacitance you ripple will be low, then you need an additional small value cap also to bypass because of internal resistance in the big cap at high freq, linear technology makes some better replacement 3 terminal regulators with better specs in every way than the lm317 they also come in negative, remember to use a 10uf cap after the regulator also
Most high value capacitors are electrolytic. Electrolytic capacitors have a phenomena called effective equivalent series resistance (ESR) which is frequency dependent. The purpose of the small "bypass" capacitor is to cancel this effect. It is not needed if you pony up for an expensive ceramic capacitor as your main capacitor. Pairing a cheap electrolytic and a cheap small (perhaps mica?) capacitor is often the most economical choice for medium to large production runs.

There is a danger in slow rectifiers. Rectifiers which are on draw little power. Rectifiers which are off draw little power. But rectifiers which are switching do draw power which needs to be dissipated (and is an efficiency loss to boot). This needs to be calculated lest you let out the magic smoke.

After a short search I wasn't able to find any support for RF noise being eliminated by slow rectifiers. I don't doubt you and it sounds reasonable, so could you explain the theory behind this?
 
I don't remember exactly the characteristics, it must have been at least 15 years since I read about it.
 
thankz said:
a long time ago I learned that you should use slo diodes for ac rectification as fast diodes can introduce rf noise and need snubbers, as long as you use the right capacitance you ripple will be low, then you need an additional small value cap also to bypass because of internal resistance in the big cap at high freq, linear technology makes some better replacement 3 terminal regulators with better specs in every way than the lm317 they also come in negative, remember to use a 10uf cap after the regulator also
It seems this type of advice goes for many things in electronics. For example we use large electrolytics in parallel with small micas. This seems stupid to the untrained/unexperienced. There is no one component that covers from DC to GHz.
 
and just for good measure remember to bypass that regulator with a 1n4001 just in case of back emf.
 
To echo J.R. and not make any excuses about it; "perfect" ether needs to be defined technically, or does not exist.