# Converting AC to DC

#### iONic

Joined Nov 16, 2007
1,650
Ok,

I have a snowblower that has a light on it that is powered by the blower motor. The voltage at the light is 15VAC. The light is terrible and I have a great replacement for it. It is an LED array that is powered by 12V-24V DC and consumes 27W. So my task is to use a Bridge rectifier to convert the 15VAC to DC.
But to insure I get the maximum DC out I want to smooth the signal with a capacitor. Some of the math escapes me and some of the info I would need for the math may not be available. I suspect a simple 25V or 50V electrolytic cap will do, but I am not sure what a good capacitance I would need.

Any clue?

#### dl324

Joined Mar 30, 2015
11,251
What is the wattage of the lamp you're replacing? Is the AC fused? 15VAC will give about 21V DC peak, minus a diode drop. Cap may not be critical because a good LED driver will be constant current and the lamp you're talking about will operate down to 12V. I'd start at 470uF and maybe try 1000uF.

#### Brownout

Joined Jan 10, 2012
2,390
Decide on what the maximum ripple you require = deltaV:

C = I/(f*deltaV), where;

I is the current you will supply to the load
f is the frequency of the rectified voltage (use 2x input frequency for full-wave rectification)
C is the capacitor.

ex: I = 1A, f=120Hz, deltaV=100mV

C = 1/(120*.1) = .083F.

#### Dodgydave

Joined Jun 22, 2012
9,313
Stick a bridge rectifier and a 35V or 50V capacitor 220 470uf will do for a led.

#### iONic

Joined Nov 16, 2007
1,650
Stick a bridge rectifier and a 35V or 50V capacitor 220 470uf will do for a led.
That's basically all I need, a ballpark figure. The LED's will do fine without the cap, but the effective DC voltage will be slightly less.

Thanks
I'll try this out!

#### iONic

Joined Nov 16, 2007
1,650
Decide on what the maximum ripple you require = deltaV:

C = I/(f*deltaV), where;

I is the current you will supply to the load
f is the frequency of the rectified voltage (use 2x input frequency for full-wave rectification)
C is the capacitor.

ex: I = 1A, f=120Hz, deltaV=100mV

C = 1/(120*.1) = .083F.
The problem here is I have no idea what the frequency is, it being governed by the blower motor. and I don't have an oscilloscope.

#### Brownout

Joined Jan 10, 2012
2,390
The problem here is I have no idea what the frequency is.
Neither do I, so I couldn't possibly suggest a value. Guess it's just trial and error time.

#### Brownout

Joined Jan 10, 2012
2,390
It occurred to me that the reason you're getting terrible light with the supplied fixture is your voltage might be a low duty cycle pulse coming out of the show blower. You can't actually verify it's a nice, sine AC wave. In that case, you might not be able to get the LED to work any better.

#### iONic

Joined Nov 16, 2007
1,650
It occurred to me that the reason you're getting terrible light with the supplied fixture is your voltage might be a low duty cycle pulse coming out of the show blower. You can't actually verify it's a nice, sine AC wave. In that case, you might not be able to get the LED to work any better.
That may be true. we'll see if I get anything out of the rectifier. But the multi-meter read zero on DC and 15V on AC. Would the pulse be both negative and positive??

The bulb is a CEC1295 (37.5W 12.5VDC) Incandescent. A 37.5W incandescent is saying a lot about it minimal output.
http://www.allbatterysalesandservice.com/browse.cfm/4,8207.html

Last edited:

#### crutschow

Joined Mar 14, 2008
25,260
I would use a bridge to rectify the voltage and, trial and error to determine the best capacitor size.

#### #12

Joined Nov 30, 2010
18,210
Just as a wild guess, I'd say you will start having diode recovery speed problems above 100,000 RPM.
If you're using a jet engine, you will need to change to a faster diode than a 1N4000 series or gear the alternator down to a slower speed. (There is no way your BS meter didn't peg on that one.)

Most, dead stupid, lawn engines run around 2500 t0 3000 RPM so your frequency number would be about 833 Hz for the worst case capacitor. Plug that into Brownout's post #3 and I see that as about 850 uf.

So there you go. Two or three people taking a rough hack at it and they all guess about 220uf to 1000 uf.
I just can't buy Brownout's ripple voltage of 0.1 volts. Light bulbs and human eyes just aren't that accurate.

#### Lestraveled

Joined May 19, 2014
1,946
I agree with the bridge rectifier. I would go with one that is rated for 200 volts (minimum), because alternator outputs are generally nasty. Rated 5 amps or more. I don't think the cap value is that critical. Once you get over 470 uF you won't be able to tell the difference.

Have fun, sounds like a good project.

#### wayneh

Joined Sep 9, 2010
16,399
I agree with the bridge rectifier. I would go with one that is rated for 200 volts (minimum), because alternator outputs are generally nasty. Rated 5 amps or more. I don't think the cap value is that critical. Once you get over 470 uF you won't be able to tell the difference.

Have fun, sounds like a good project.
Ditto all around. One more thing - check the voltage on the cap BEFORE you hook up the LEDs. It's remotely possible it could exceed 24V if the input to the rectifier is really sharp spikes instead of a nice sine wave.

#### Lestraveled

Joined May 19, 2014
1,946
Ditto all around. One more thing - check the voltage on the cap BEFORE you hook up the LEDs. It's remotely possible it could exceed 24V if the input to the rectifier is really sharp spikes instead of a nice sine wave.
Good point. The original load was 37.5 watts. If the led load is less than that, the voltage will be higher. You may need a load resistor to keep the voltage from going too high.

#### Brownout

Joined Jan 10, 2012
2,390
I just can't buy Brownout's ripple voltage of 0.1 volts. Light bulbs and human eyes just aren't that accurate.
Huh? You think made up numbers used to show how to use the equations are actual recommendations for the project?

Last edited:

#### iONic

Joined Nov 16, 2007
1,650
I took a shot in the dark and connected a three LED array (series), LED voltages of aprox. 3.7V directly to the "15V" from the blower. Yes, a bad thing, no limiting resistor. Of course I did this for about a few seconds only.
The LED's lit up nicely with no flicker. I then tried the 9 LED array rated for 12V-24V and 27W. It also lit up nicely.
Obviously, not the way to keep it. I couldn't find a ready-made bridge rectifier so I'll have to build one. I did find a 470uF cap rated for 50V, but should have one for either 16V or 24V...16V preferable given the last couple of comments.

My dang electronics are packed away in boxes pending a move that did not happen yet so I am rummaging to find parts.

Ha! just removed a GBU606 Rectifier from a broken video projector.
Rated - 800V, I=8A, Ifms200A
Overkill, but easier to work with.

The closest caps I have are 330uF @ 16V
I could parallel two of them for 660uF @16V
Think it's time to get some solder flowing!

UPDATE: Soldered up the wires to the Bridge Rectifier with the two caps. The voltage I am reading is no longer AC or pulsed, but reads as DC ~20V.

Any explanation for the 20V??

Last edited:

#### BR-549

Joined Sep 22, 2013
4,938
Is the blower running at constant speed?

#### SLK001

Joined Nov 29, 2011
1,548
Your generator is probably two magnets (one N and one S) embedded in the flywheel and a pickup coil, giving you a positive and a negative pulse each revolution. At 3500 RPM, this would give you a frequency of ~58 Hz. The capacitor banks needed to smooth this 2.5 amp output whould be huge, so I would just skip the caps. You probably will only notice the flicker at low RPMs anyway.

Build the bridge using 4 amp diodes (not 1N4000 series), or find a bridge with capacity 4 amps or higher. You might want to heatsink the bridge

#### BR-549

Joined Sep 22, 2013
4,938
Last edited:

#### #12

Joined Nov 30, 2010
18,210
Caution: 1N4000 series is rated for 1 amp.