What's new
What's new

Bridge rectifier

rj1939

Stainless
Joined
Jan 31, 2008
Location
southern il
Is there any reason why I couldn't use a 1000v bridge rectifier for 120v useage? I have one on hand. Amp rating is plenty high for the application too.
 
One more question: Should I put a capacitor across the two DC terminals, I have read that bridge rectifiers have some spikes on the DC side and capacitors are used to smooth it out a bit?

And thanks for the replies.
 
One more question: Should I put a capacitor across the two DC terminals, I have read that bridge rectifiers have some spikes on the DC side and capacitors are used to smooth it out a bit?

And thanks for the replies.

Too general a question. Depends on what you intend to power. An improperly applied Capacitor can have "most unwanted" consequences.

Filters can be non-existent, simple, or complex, are meant to be selected for the operating environment and needs of the load.

"Classical" LC low-pass filter involves a fairly heavy Iron & Copper lumped-inductance AKA "choke" and a capacitor. Sometimes a resistor is substituted for the inductor.

SOME applications a Capacitor ONLY or an Inductor ONLY will do the do.

That's one stage. More serious filters have more stages.

Capacitor-input ELSE Choke-input act differently even with the same values of L and C.

All this stuff is online, calculators included. Simplest is to find a similar need, copy it, as the Maths are a tad grotty if they aren't your "Day Job".

Audiophiles publish tons of info on reducing primary line-Frequency "hum", for example. Then there are "harmonics".

What are your needs and intentions? And budget...
 
PIV = PRV = the do not exceed voltage in the reverse direction on the diodes. Go more than that and the diode junction breaks down and it becomes a dead short, with dire consequences. This is one reason to always take into account any errant voltage spikes that might kill it, and use an over-rated component for inverse voltage. Semiconductor stuff like this is pretty cheap.

Current ratings are what really cost, again it's wise to overspec this by a factor of two or so. And use a generous heat sink if one is called for.
 
One more question: Should I put a capacitor across the two DC terminals, I have read that bridge rectifiers have some spikes on the DC side and capacitors are used to smooth it out a bit?

And thanks for the replies.

the output will have ripple current, only way to size it is having an oscilliscope to test the output ripple with different sized capacitors. Caps should be rated double the voltage of line to line.
 
A bridge rectifier in simple terms cuts a sinewave down the middle and folds 're other half around.

A 60 cycle some wave turns into what looks a saw blade with 120 teeth in the same space where there was 60 before.

Each high spot drops to zero between so filtering is required to convert from pulsed DC to filtered DC.

Sent from my SAMSUNG-SM-G930A using Tapatalk
 
At least a small capacitor is a good idea if there is anything sensitive to spikes on the DC side. When the thing is turned off, there will be a spike.

Can be on the AC side of the rectifier (0.1 to 0.33 microfarad) or 1 to 10 uF on the DC side. Rate for more than the actual voltage, as mentioned above.
 
Is there any reason why I couldn't use a 1000v bridge rectifier for 120v useage? I have one on hand. Amp rating is plenty high for the application too.

Will the part last longer at much less stress than it can withstand? Any opinions from the in-house PhD's about difference in reliability? Pronounced as rrrrreeeeellllliiiiiaaaaabbbbbiiiiillllliiiiitttttyyyyy.

For the scientific folks:
Does repeated cycles impart less damage to a high rated part. More reliable system.

For the bean counters:
Is it really worth the extra cost to make something more reliable. Ahh Soooooo. No way, not when spoiled silly Americans are the customer.
 
Last edited:
Will the part last longer at much less stress than it can withstand? Any opinions from the in-house PhD's about difference in reliability? Pronounced as rrrrreeeeellllliiiiiaaaaabbbbbiiiiillllliiiiitttttyyyyy.

LOL! Dunno about the Piled higher and Dry'er crowd.. but the bean-counter here sez the FWB were about three bucks each ordered ten - or fifty - at a go to meet minimum orders or keep the POSTAGE from eating up yer lunch. Thereafter, just grab one and go.

Moved up to 1600 PIV when the loads started to include DC motors.

I buy MOV's and bleeder resistors in bulk and over-rated as well.
Economics again. They don't each much space, stashed.
 
MOV will work also, of course.

Spec'ing one is actually a little more "interesting" than it seems, since MOVs have resistance and they do not really have a hard limit on voltage the way a zener does for instance (and zeners also have the same effect, just less so).

If you want to use an MOV, pick one somewhat higher (say 30%) than the voltage it will normally see, and no more than (if possible) half the voltage limit on the thing to be protected.

MOVs also have the fun feature that their voltage decreases a bit every time they get "hit" with a transient. So the buffer amount over the normal voltage basically buys time.
 
MOV will work also, of course.

Spec'ing one is actually a little more "interesting" than it seems, since MOVs have resistance and they do not really have a hard limit on voltage the way a zener does for instance (and zeners also have the same effect, just less so).

If you want to use an MOV, pick one somewhat higher (say 30%) than the voltage it will normally see, and no more than (if possible) half the voltage limit on the thing to be protected.

MOVs also have the fun feature that their voltage decreases a bit every time they get "hit" with a transient. So the buffer amount over the normal voltage basically buys time.

Simpler to double the working Voltage. 1200 or 1600 PIV bridge or SCR matrix doesn't much give a damn about transients clipped at 480 or 600 V. So long as it isn't tasked with any significant energy left at 2,000 Volts.

Which is is not, a(ny) DC motor, contactor reversed, typically kicking only five times its working DC Voltage - a tad under 280 VDC, here, so under 1500.

If-even I DO use "contactors" vs sanely controlled "4Q" Solid State.

Read: "only at the equivalent of gunpoint".

:D
 
The load, is a winding for an electric clutch.

Among the nastier of "inductive kick back" spike sources for its otherwise very modest size and load, actually.

A purpose-made packaged "snubber" in parallel should do yah. Major maker's of such clutches might even recommend which one. Or even have built it in. Let yer keyboard do the lookin', it's a common situation.

Couple of my Dee Cee motors here have a Dings and a Warner as brakes. Same critter, electrically, as a clutch.

Wisely, the motors are Dee Cee but their brakes are AC. Easier to manage cheaply.
"Zero Cross" and probability thing.
 
A single diode reverse biased across the clutch winding will solve the inductive kickback. However, it can add up to a couple hundred of milliseconds to the decouple time of the clutch, as it allows the magnetizing current to continue circulating with minimal resistance, slowing the collapse of the field. If timing is critical, that's when to look into snubber networks or zener diodes (my personal favorite for fast coil demagnetization).

The coil inductance itself will smooth out the rectified voltage just fine, and only some small filter capacitors as mentioned are needed. It is good practice to put a large value resistor across the capacitors to they cannot hold a charge long enough to shock someone who goes to work on the machinery after shutting it off. Less important with small filter caps, but a serious consideration with ripple smoothing capacitors.

A 1000V rectifier will drop a bit more voltage than a 600V or less rectifier (2-2.4 volts versus 1.6-2 volts), and will get a bit warmer than the lower voltage model, but insignificantly so in this application. There is no reliability concern with putting too little power through a semiconductor.
 
BS....... "However, it can add up to a couple hundred of milliseconds to the decouple time of the clutch, as it allows the magnetizing current to continue circulating with minimal resistance, slowing the collapse of the field"........ The free wheeling diode will shunt the rev flow and drop the voltage like a rock, the drop out time will be decreased , same circuit I used for big scrap magnets for drop out....Phil
 
BS....... "However, it can add up to a couple hundred of milliseconds to the decouple time of the clutch, as it allows the magnetizing current to continue circulating with minimal resistance, slowing the collapse of the field"........ The free wheeling diode will shunt the rev flow and drop the voltage like a rock, the drop out time will be decreased , same circuit I used for big scrap magnets for drop out....Phil

LOL! Yazz.. what was that Signetics "Product Engineer"'s name? Robert Weber?

The device was the "three terminal regulator". He introduced it, there was but one page of reference circuits.

By the time I got on to blowing them up driving 42U racks full of broadcast studio endless loop DC tape decks for weather, traffic, and such telephone dial-in...and busting his chips?

We had added another whole page of fast-switching avalanche-mode diode "reference circuits" to let 'em survive the abuse!

ISTR the time involved was in integral nanoseconds. Low ones, too. Mebbe even fractional?

Telephone calls "back in the day". Compuserve was still a novelty. "Internet" was still mostly plotting, scheming, and dreaming, US Gov "ARPA" contracts.
 








 
Back
Top