Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Bridge Rectifier

  1. Jul 3, 2007 #1
    I want to make a bridge rectifier circuit to give me 6v or 9v DC power from 220v 50-60 Hz AC supply.

    I know the circuit, but don't know which diodes and capacitors to buy for it.

    This would be the first time I'm doing something like this, so any suggestions/tips would be appreciated.
     
  2. jcsd
  3. Jul 3, 2007 #2

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    You will need a transformer to reduce the input voltage to ~ 10VAC, then you will feed that to your bridge.

    Moving this to electrical engineering.
     
  4. Jul 3, 2007 #3

    ranger

    User Avatar
    Gold Member

    Any general purpose diode (that meets your requirements) will suffice for the rectifier. You can even purchase a bridge rectifier module that has all the diodes in one package.

    You'll generally want an RC network with a relatively large time constant hooked up to the output of the rectifier. This may vary as we don't know what sort voltage regulator you're using (if any).
     
  5. Jul 3, 2007 #4

    berkeman

    User Avatar

    Staff: Mentor

    It sounds like you have not worked with the high voltage AC mains before, so I want to caution you to be very careful, and be sure to design and build your projects in accordance with the safety agency requirements for your country (for me, it is Underwriters Laboratories here in the US). There are some *very* important safety practices with respect to fusing, grouding, isolation, insulation, etc., that need to be followed, or you will be in danger of hurting yourself or others, with both shock and fire hazards.

    The safest thing for you to do for this project, is to use a pre-approved, monolithic wall transformer that plugs into the AC mains socket, and gives you the 12Vac or 24Vac that you can then experiment with making the full-wave bridge and output capacitor (not RC) smoothing filter. The plug-in wall transformer will already be safety approved, because all the AC mains high voltage parts are insulated so that you cannot touch them. All you can touch is the lower output voltage, which is much safer.
     
  6. Jul 3, 2007 #5
    Thank you for pointing that out. Ill look into that, but I want to make this from the ground up. I know you can buy components that are mostly prefabricated, but I don't want to do that. I'll ask the person I'm buying the components from, hopefully he can help me...
     
  7. Jul 3, 2007 #6
    What would be a large time constant? Again, here I know the theory, but I'm kinda lacking where the practical part is concerned...
     
  8. Jul 3, 2007 #7
    If I want more than one output voltage, will I need two transformers or is there some other way I can manage that?
     
  9. Jul 3, 2007 #8

    berkeman

    User Avatar

    Staff: Mentor

    The usual way is to use a transformer that has multiple output windings, sized for the output voltages that you want. Like, if your goal is to have +5V and +12V DC regulated outputs for a power supply, then you would chose a 220Vrms AC mains transformer that had output windings that you could rectify and smooth into about 15V and 8Vdc, and then use linear regulator chips to make the final output voltages. The transformer would need to be sized to handle the output power, and the regulator chips would need to be heat-sinked correctly for the output power.

    And as I said, since you would be connedting AC mains wiring to the input of the transformer yourself, you would need to obey the applicable safety requirements with respect to wire connection techniques, fusing, grounding, and insulation/isolation. Please be careful.
     
  10. Jul 3, 2007 #9

    ranger

    User Avatar
    Gold Member

    I suggested a large time constant there would be periods when the capacitor would discharge through the load (resistor). If the time constant is small enough, the cap would discharge quickly, hence there would be considerable ripple in the output DC waveform. Having a large time constant would ensure slow discharging, which means a more smoother output waveform.

    But as I mentioned in my initial reply, the application of this would vary. You need to ask yourself if you're using voltage regulator; if so what kind. For example, if you're using a zener regulator, you'll have to choose R in such a way that current through the zener is within operation parameters. If you're using a voltage regulator IC, the IC chip would take care of everything, provided you have followed the manufactures specifications. An example of an adjustable voltage regulator IC is the LM317.

    But if you're looking for simple, a zener voltage regulator would be the way to go.
    http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/zenereg.html
    You'd definitely want a large valued cap though. Probably like 2200uF.
    A large time constant in this case would be greater than the time between two adjacent peaks.
     
    Last edited: Jul 3, 2007
  11. Jul 3, 2007 #10

    mheslep

    User Avatar
    Gold Member

    chaoseverlasting-

    There are several approaches to obtaining your 9-10 VDC from wall power AC with pros and cons to each depending on your application. You need to make a couple more decisions to select your approach.

    o Whats the power load?
    o Do you need regulated DC? That is, do you care if the DC output voltage varies when a) you vary the load on your power supply, or b) if your house voltage (supply) goes up or down when, say, the air conditioner starts?
    o How much ripple? The output DC will always have some amount of left over AC riding on top of your DC. You can specify this.

    1. AC/DC Switched Power Supply. This is the modern approach and probably nothing that you want to build, but they are inexpensive to buy. A switcher is generally more efficient and certainly smaller and lower weight than 60 cycle transformers described above. See here
    In this case your bridge is connected directly to the 120/220 producing high voltage VDC = 1.7*VAC. The high VDC is chopped at high frequency(~10KHz), much higher than the house AC (60Hz) and thus the transformer can be relatively tiny. There's also a feedback circuit IC to control the chopper which enforces the DC output voltage you desire regardless of load or input voltage. Finally, the you need the output rectifier and RC circuit, but the ratings for these are also much smaller (nano or micro farads, 20VDC rating) than the would be needed in conjunction with the 60Hz transformers mentioned above. You can buy, and for safety sake, you should buy your switcher via any big catalog electronics distro. (Newark, etc). The switcher will be a bit more expensive than #2 below.

    2. AC/DC 'Adapters', unregulated. You can also buy AC/AC adapters as described above but then you still have to build your bridge to get DC. Just buy the AC/DC which has bridge built in for you. Note this: these adapters are almost all unregulated as I described above which means they're probably not suitable for running an electronic digital device like your cell phone, but they are perfectly fine for charging your cell phone batteries which don't care if the voltage varies a great deal. There will also be a lot of 60Hz ripple left over on your DC with these adapters, maybe 5-10% of your output voltage. Again not a problem if you want to charge batteries.

    3. AC/DC Adapter plus linear regulator. This is my #2 above feeding a linear regulator chip like the LM317 described up thread. Absent a vary sophisticated switching supply, a linear will give you the cleanest (microvolt 60Hz ripple), regulated output power. Linear's work by resistively burning up the Voltage delta * out put current so the IC will get hot and for most jobs (> ~100ma) you'll have to heat sink it. Plost_to_linear = (Vout-Vin)*Iout. They typically want an input voltage 1-3 volts above the output so you'll have to buy your AC/DC adapter accordingly.

    mheslep
     
  12. Jul 4, 2007 #11
    Thank you. I will look into that. And maybe I'll just buy a prefabricated ac transformer that follows the norms... It would be simpler working with that considering its my first time doing something like this.

    Why would the transformer need to be 220v rms? The rms value would be 220/1.414 right for 220v rated AC?
     
  13. Jul 4, 2007 #12
    Actually, I just recently got interested in electronics, and built my first radio ( which isnt working for some reason, but I cant check it till I get my multimeter), and the battery supply for that is 4x1.5v batteries. If I further pursue electronics as a hobby, I need this eliminator as buying batteries for every device I make will be expensive. Most of them will be in the 3/6/9 v range, so I thought I'd build one for those voltage ratings.
     
  14. Jul 4, 2007 #13
    Ok. So a large time constant would be something like 1/50 or 1/40. If I use a zener diode, then RC needs to be something in the 1/50 to 1/40 range. Depending on the diode ratings, I can find R, and from the Time Constant, I can find C.
     
  15. Jul 4, 2007 #14

    berkeman

    User Avatar

    Staff: Mentor

    I was just emphasizing the fact that when we talk about 220V or 110V AC mains supplies, those numbers are the RMS values. The peak-to-peak values are considerably larger!
     
  16. Jul 4, 2007 #15

    ranger

    User Avatar
    Gold Member

    If you consider a regular sine wave at 60Hz, it has a period of 16.67ms. When we consider the rectified voltage out of a bridge rectifier, we will want to find the time between two charging peaks. In this case it is obviously 8.33ms. An RC time constant significantly greater than this would give an acceptable output DC waveform; remember the cap will fully discharge at 5RC. So RC = 25ms means a full discharge in approx. 125ms. I'd same aim for an RC approaching 100ms.
     
  17. Jul 4, 2007 #16

    ranger

    User Avatar
    Gold Member

    Mains is already given in RMS, so 220V mains supply will have a peak of 220V/0.707. However, when you're dealing with mathematical calculations that relate to the power supply, its a good idea to have everything in peak.
     
  18. Jul 5, 2007 #17
    I didnt know that. I thought everything given was in peak values.
     
  19. Jul 6, 2007 #18

    ranger

    User Avatar
    Gold Member

    It is rated in this way for an important reason. The rms value allows us to find the effective value of power delivered to the load. If we have something rated as 12V AC; it usually means rms (unless otherwise stated). So a household incandescent bulb if connected to a 120V AC and 120Vpk will illuminate brighter with 120V AC, because 120Vpk has less effective power over the same time interval as 120V AC.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Bridge Rectifier
  1. Bridge Rectifier (Replies: 6)

Loading...