The rectifiers that we will typically consider look like these:


Because they are single phase, full bridge rectifiers, they have 4 connection points/terminals:
- 2 for the AC input (polarity irrelevant)
- 2 for the DC output (polarity determined by the rectifier) ie. +(positive) and -(negative)

During a purchase, the voltage and amps need to be specified. Prices range around $2 to $6 and there is no harm in getting slightly more expensive rectifiers that have excess capacity.

I gave a preference for the leftmost rectifier with the pluggable ("spade") connectors.

What do these "things" do again??
Francis

Convert alternating current AC to direct current DC.

They sit between the transformer and the capacitor in the power supply.

http://en.wikipedia.org/wiki/Rectifier

A bridge rectifier is just four simple diodes connected together so that two diodes pass electricity on the positive part of an AC cycle and two diodes pass electricity on the negative part of an AC cycle. The way that the diodes are connected together also converts the AC positive/negative voltage cycle into a POSITIVE ONLY voltage cycle. 60hz (in America) becomes 120hz.

If you looked at the output of an un-rectified 36VAC transformer with an oscilloscope, you would notice that the voltage signal goes from positive 50V to negative 50V. Yes, that's right a 36V transformer is actually producing 36V X SQRT(2), or 50V. An AC voltage meter reads the RMS (Root Mean Square) voltage of an AC signal, not the Peak voltage. (That's why a DC power supply "magically" develops 1.4X more voltage than the AC voltage that it gets from the AC transformer.)

Anyway, the bridge rectifier converts that AC positive/negative signal into an AC positive only signal. How it does it is a little complicated and requires some understanding of how semiconductor diodes work, but a simple explanation is that a diode passes voltage when it is forward biased and blocks voltage when it is reverse biased. That means that when four diodes are connected in a diamond shape, two of the diodes will conduct electricity on each half of an AC cycle and two diodes will block electricity on each half of an AC cycle. The result is that you get half the AC amplitude and twice the frequency. When you want to build a positive voltage power supply, the bridge rectifier gives you almost exactly what you want, a positive only AC signal at 2X the line frequency.

To convert that 2X positive only AC voltage into DC voltage requires that we add the large capacitor (10,000uF to 20,000uF). The large capacitor acts like a storage battery. It fills up every half-cycle and is partially drawn down by the stepper motors between cycles. That's why it takes such a large capacitor. The formula is 80,000 X Amps / Voltage. What that formula does is to insure that the stored electricity is sufficient so that the power-hungry stepper motors won't be able to draw enough electricity from the capacitor to cause the DC voltage to "ripple" more than a small amount. The "ripple" that you can easily see with an oscilloscope, happens when a device pulls more electricity from a capacitor than the capacitor can store on each "charging" part of the AC cycle. For stepper motors, a 5% to 10% ripple is usually acceptable. A little ripple keeps the size of the capacitor reasonable.

In summary, you need three parts to make a power supply:

1. A transformer to "transform" an AC voltage from one amplitude to another amplitude, i.e. from 120V to 36V.

2. A bridge diode to "rectify" the voltage, i.e. from positive/negative to positive-only and from line frequency to 2X line frequency.

3. A "filter" capacitor to store the AC voltage so that it becomes a constant DC voltage, i.e. to "filter" out the AC ripple.

(There is one minor consideration that sometimes causes people grief when they design their first DC power supply. A semiconductor bridge rectifier requires 0.7V before the semiconductor starts to conduct, so that 0.7V is effectually lost. With the power supplies that you use for stepper motors, that 0.7V is not significant; however, if you were building a 5V power supply or a 3V power supply, you would need to be aware that semiconductor diodes "consume" 0.7 volts before they start to conduct.)

Edit: Sorry, Gerald, you posted while I was writing.

Is it possible that if the capacitor is not big enough, stepper motor will turn smoothly clockwise but uglily counter-clockwise? (or vice versa).

A small capacitor would have equal effect if a motor turned CW or CCW. The only thing that happens to change the direction of a stepper motor is the sequence that the coils are energized. The capacitor has nothing to do with that sequence.

If a motor runs rough in one direction but not the other, check mechanical things (binding, loose pinion, etc.). Also, verify that the direction connection to the stepper driver is working properly. Use a voltmeter and verify that the voltage is correct and that it is steady. If the direction voltage fluctuates, then the stepper driver may be trying to change direction.

I'm still in component selection stage.

Looking at the Mouser catalog, there are a bewildering array of bridge rectifiers available.

For sake of discussion, I'm tentatively planning on building a supply from a microwave oven transformer with the secondary wound of 14ga wire to yield 26vac. The DC voltage will thus be 37vdc. My preferred motors are rated for 5a ea, for a total of 20a max. The power supply will be 26*20=520va.

Which rectifier do I need? What's the important spec?

"Reverse voltage" is AC voltage in, (26vac) right?

Is this...

http://www.mouser.com/Search/Product...b3RYU7HT9Vg%3d

... a suitable rectifier?

Jeff,
I would go with this rectifier if you are wanting to order from mouser. This one is a lot cheaper and will work fine for you. Has a little higher max reverse voltage but that is ok.

http://ca.mouser.com/Search/ProductD...fHQL6SLXu7M%3d

a little cheaper at Digikey though. Same part.

http://search.digikey.com/scripts/Dk...me=GBPC2502-ND

Full Datasheet

http://www.fairchildsemi.com/ds/GB/GBPC12005.pdf

Thanks Heath,
Judging from your recommendation, the only really relevant spec for our purposes seems to be the forward continuous current... true?

Also, the catalog indicates that it dissipates 83w. They must get pretty hot.

I'll definitely check out digikey further. Thanks again.

Jeff,
The continuous forward current spec must be capable of sustaining your max forward current, AND the max reverse voltage spec must be able to sustain the highest ac voltage you will see on the input. In your case you think you will see 20A but in reality you will see a lot less. You will probably never see ALL of your motors running at peak load at the same time. You could go higher with the current capability of the bridge but you are just wasting your money. A 30A, 40A or 50A would work though if you have them around. :)

The rectifiers are quite inexpensive. I have never worried about the voltage spec. I have just picked the cheapest square one with a metal case and spade/blade terminals.

The local industrial electrical components store only carries 400V bridge rectifiers. They said that it didn't make business sense to carry lower voltage units when the price difference was less than a dollar per unit.

It is a good idea to buy a bridge that either has a metal case, or at least one that has exposed metal for heat-sinking. If possible, bolt the bridge rectifier to a piece of sheet aluminum using some heat-sink compound. The PMDX-135 power supply has a hefty aluminum heat sink bolted to the bridge rectifier:

Copied from another thread:

I have this rectifier. I have 33 AC volts going into it and 87 DC volts coming out... What am I doing wrong? I have a 30 volt transformer so I was expecting something like 42 DC Volts.

Blake, it's not unusual for a completely unloaded power supply to give a high reading. Of course, you don't want to risk your expensive electronic devices to prove this. You can use a dummy load (resistor) across the output to get a measurement. A low wattage light bulb (like 15 or 25 watts) can act as a dummy load.

Blake, a rectifier cannot cause a 260% increase in voltage. My first reaction is that your voltmeter is (very) faulty. Can you provide some pictures of your wiring connections and your voltmeter setup?

Blake

Did you wire the rectifier like this :

http://i890.photobucket.com/albums/a...ectifier-1.png

Yes I did Sergio. Thanks for taking the time to draw that very clear diagram. I wish I would have had it sooner because I had to do some digging to find out how to wire it.

The problem cleared itself up as I worked with it. I am not sure what happened but I am suspecting that it was a grounding problem. It is securely grounded now and reading correctly.

Hi all.

Does the PMDX 135 and PMDX 136 have a build in rectifier bridge ??.

From the PMDX website for both the 135 and 136

Utilizes heavy duty bridge rectifier mounted to a large heatsink