Paralleling of Bridge Rectifiers

[ATSman]

Has anyone ever connected 2 bridge rectifiers in parallel to increase the current capacity?
This will be used to momentarily energize a large solenoid in an ATS (480V @ 80ms.)
I have heard pros and cons, even if they are matched perfectly to share the load, PIV differences, etc.
Just thought I'd throw this out there for to hear any real life experiences on the subject.

 

[kwired]

Has anyone ever connected 2 bridge rectifiers in parallel to increase the current capacity?
This will be used to momentarily energize a large solenoid in an ATS (480V @ 80ms.)
I have heard pros and cons, even if they are matched perfectly to share the load, PIV differences, etc.
Just thought I'd throw this out there for to hear any real life experiences on the subject.

I don't see what it should hurt. Think about battery chargers - you have a rectifier as well as the battery output connected in parallel. Larger capacity chargers connected to a three phase supply are essentially putting at least three rectifiers in parallel for one output.

 

[Besoeker]

Has anyone ever connected 2 bridge rectifiers in parallel to increase the current capacity?
This will be used to momentarily energize a large solenoid in an ATS (480V @ 80ms.)
I have heard pros and cons, even if they are matched perfectly to share the load, PIV differences, etc.
Just thought I'd throw this out there for to hear any real life experiences on the subject.

How much current do you need?

 

[GoldDigger]

Just do not count on it safely delivering fully twice the current.

Sent from my XT1585 using Tapatalk

 

[Ingenieur]

That is sort of what is done in a 3 ph rectifier

 

[Besoeker]

That is sort of what is done in a 3 ph rectifier

Not one that I've seen.
Unless/until we know what current is required, it is difficult to reply constructively.

 

[ATSman]

Not one that I've seen.
Unless/until we know what current is required, it is difficult to reply constructively.

I can't remember the last time I tried measuring the current. The DC resistance of these
solenoids are 22 - 26 ohms. I would guess that the inrush currents are in the hundreds of amps
range. Don't forget it is an 80ms pulse, not a steady state rated device. (5 cycles?)

 

[Ingenieur]

.....

 

[Ingenieur]

Make sure the bridges are the same
same as 3 ph with 1 ph removed

 

[Besoeker]

I can't remember the last time I tried measuring the current. The DC resistance of these
solenoids are 22 - 26 ohms. I would guess that the inrush currents are in the hundreds of amps
range. Don't forget it is an 80ms pulse, not a steady state rated device. (5 cycles?)

OK. You don't need to parallel units. It isn't something I'd suggest anyway.
IXYS and most mainstream suppliers can offer a single unit to do this.

At 480Vdc and 22 ohms, steady state would be around 22A - if you ever got there.
With DC, the inrush ought to be an irrelevance. The coil inductance would limit the rate of rise of current (di/dt).

So, easily within the capability of one module.
Unless I'm missing something.

Another thing to consider is PIV. At 480Vdc you'd need devices rated at 1600V.

 

[Ingenieur]

AC 480
DC 430 full wave rectified ( ac rms x 2 x sqrt2 / Pi)
at 24 Ohm 18 A
Assume a mechanically latching mechanism (since momentary and an ats)

that is a lot of current and power 7.75 kW or so

 

[mike_kilroy]

What is an ATS?

And no, you cannot simply parallel them, they will not share close enough. More tomorrow when I am at a real keyboard. Google it yourself tonight.

Sent from my SM-G900V using Tapatalk

 

[iwire]

What is an ATS?

Automatic Transfer Switch

 

[gar]

161217-2208 EST

Two randomly selected diodes of the same part number will not have identical v-i curves and when paralled will not precisely share current. To parallel semiconductor diodes you need to add an appropriate amount of linear resistance in series with each diode before paralleling.

Just buy bigger (higher current) diodes for your small application.

.

 

[Besoeker]

AC 480
DC 430 full wave rectified ( ac rms x 2 x sqrt2 / Pi)
at 24 Ohm 18 A
Assume a mechanically latching mechanism (since momentary and an ats)

that is a lot of current and power 7.75 kW or so

Interesting, perhaps telling even, that you see it in that perspective.
I don't see it as a lot of current. A single encpasulated bridge module could comfortably handle that even on a continuous basis.

 

[Besoeker]

161217-2208 EST

Two randomly selected diodes of the same part number will not have identical v-i curves and when paralled will not precisely share current. To parallel semiconductor diodes you need to add an appropriate amount of linear resistance in series with each diode before paralleling.

Just buy bigger (higher current) diodes for your small application.

.

I agree. But I'd go for a module rather than discrete diodes.

 

[Ingenieur]

Interesting, perhaps telling even, that you see it in that perspective.
I don't see it as a lot of current. A single encpasulated bridge module could comfortably handle that even on a continuous basis.

'telling even' lol of what? Your stalking and dislike of me?
a lot of current for a solenoid
a solenoid that holds an elevator brake off against the spring draws 1 kw or so

many use 2 rectifiers in parallel for redundency
277 ac to 250 dc

 

[Besoeker]

a solenoid that holds an elevator brake off against the spring draws 1 kw or so

many use 2 rectifiers in parallel for redundency
277 ac to 250 dc

Fine. Use two fully rated encapsulated bridges.

 

[mike_kilroy]

For the OP... hopefully you googled something like "series and parallel of diodes" and learned why you must try to not do it.

The issue as many have said is that the forward voltage drop will not be the same. You can buy 10 bridge rectifiers and check drop on each of the internal 4 diodes and try to pick a better matched set, but why not just the correct size to begin with? My 7kw generator in my 1973 GMC motorhome has a 25amp continuous bridge rectifier - think I paid like $ 1.50 for it.

The issue continues to a potential snowball effect: as one diode passes more current than its parallel friend, it will heat up. Heating semiconductors causes their forward voltage drop to decrease. So it will then pass evven MORE of the uneven load. Heat up more, draw more of the load, heat up. This snowball effect can overload this diode, blowing it. Now you are back to the single diode that is not rated to pass the full so it will eventually blow.

If this is required to be done, adding L or R in series with the diodes helps match them. Otherwise, as one reply stated where they put bazillions of diodes in parallel for hi power power supplies, they no doubt put then on THE SAME HEASINK - so one did not heat more than it neighbors and the snowball effect did not happen (often probably).

Next issue is WHAT SIZE DO YOU REALLY NEED?? Do you really believe your 480v 22 ohm solenoid will draw 18 amps???? Think about that for a bit. Then go measure some other INDUCTORS and COILS like other small solenoids, any transformers... Your 22 ohms is only a SMALL PORTION OF YOUR SOLENOID's IMPEDANCE! The L in that coil will limit the current!

So before you try to pick the proper diode bridge, why not stick an ammeter in series (or a clamp on) with that solenoid and MEASURE the steady state current! I bet you find it is less than 1 amp! The 22 ohms tells me that it likely will have about a 20x inrush current - or 18 amps.

 

[kwired]

.....

Make sure the bridges are the same
same as 3 ph with 1 ph removed

Now that you drew that out I might change my mind a little.

In your drawings you don't really have any of the rectifiers completely parallel to one another, it is more of a series-parallel that changes every half cycle, and gets more complex with the three phase setup and input phase angles.