Why there are no Schottky diode bridge rectifiers?

[LinuxHata]

Actually there are some, but quite few and in very limited specs.

I really don't understand, why you should have 1.2 volts drop on say 10A bridge and waste 12W, when you can have 0.6V drop and waste only 6 watts?

Of course, I can assemble a bridge by myself, but I'm just curious...

[Siwastaja]

There are some.

Low Vf schottkys have the following issues:
* Quite limited reverse voltage ranges,
* Significantly higher leakage current than standard PN diodes,
* Leakage current which shoots exponentially through the roof with increasing die temperature, easily causing thermal runaway.
* Leakage current also increases with increasing reverse voltage, although not as aggressively as with temperature.

There are some very low-Vf schottky diodes available, but the lower the Vf, the higher the leakage current usually is, so you need to be careful in design. Unlike with normal diodes, you can't assume reverse power dissipation to zero; you actually have to calculate it in worst-case conditions, and even if it does not lead to thermal runaway and self-destruction, it could offset the benefit to zero.

[Kleinstein]

If very low voltage loss is needed there are ideal diode circuits that use MOSFET switches as active rectifier.

There are quite some dual schottky diode diodes that can be used to build a bridge rectifier, though one would need 2 types with common anode / kathode.

With higher reverse voltage rating the Vf is no longer that low, so chances are one would end up with more than 0.6 V lost for 2 diodes.

[LinuxHata]

Well, currently I need to rectify 20V AC at about 6A current.
Tried "standard" rectifier bridges, like GBU806, KBPC2510 and so on - they got hot as hell.

Assembled the bridge from the Schottky diodes sourced from old ATX power supplies - (these are rated 30A @ 45V) - difference is stunning, they're only a bit warm.

So this is why I asked.

[MrAl]

Actually there are some, but quite few and in very limited specs.

I really don't understand, why you should have 1.2 volts drop on say 10A bridge and waste 12W, when you can have 0.6V drop and waste only 6 watts?

Of course, I can assemble a bridge by myself, but I'm just curious...

There are many back-to-back dual Schottky diodes even 45 amperes, or at least there used to be.
Get one common cathode and one common anode and make a bridge rectifier with those two.

[Siwastaja]

Tried "standard" rectifier bridges, like GBU806, KBPC2510 and so on - they got hot as hell.

Assembled the bridge from the Schottky diodes sourced from old ATX power supplies - (these are rated 30A @ 45V) - difference is stunning, they're only a bit warm.

Not only does it dissipate less power, but being as big as they are, you have a lot more surface area for cooling, too. The leads of these large diodes are thick and also act as heatsinks, so it's a good idea to solder some large wires to them, or even some copper sheet, to improve cooling further. Maybe in your tests, you already use some thick wire which gets the heat out? Compare to integrated bridges which do not cool through the legs as much but require mounting to a heatsink.

[LinuxHata]

I know how to make bridge rectifier
I'm asking why aren't they are readily available, using the Schottky diodes

[voltsandjolts]

Search for "ideal bridge rectifier"

e.g.
https://www.nxp.com/products/power-management/ac-dc-solutions/ac-dc-controllers/active-bridge-rectifier-controller-demo-board:TEA2206DB1583
https://www.digikey.co.uk/en/products/detail/nxp-usa-inc/TEA2206DB1583/14301834

Only £150 :-)

[Siwastaja]

Well, the reason for something not existing in a more integrated form is usually there is no large enough market for it. If the market is too small, and it can be constructed out of individual parts, why bother?

Linear low-voltage DC power supplies are rare today, switch mode converters have been mainstream for over 20 years. Switch mode converters need full bridge rectifiers too sure, but that's for mains voltage, so diodes need to be rated for 600V minimum. And you can't have a standard schottky at such high reverse voltage, that's why they use normal diodes. Output stages may use schottkys or dual schottkys and sure enough, dual schottkys in one package are a thing.

[SeanB]

Your standard high power schottky is most common as a common cathode half bridge, but for a bridge rectifier you need the common anode version, or the series type with 2 diodes, and one with anode connected to cathode of the other. Not common other than in tiny little SO packages, where they are common, and sadly limited to 40V 100mA instead.

[MrAl]

Your standard high power schottky is most common as a common cathode half bridge, but for a bridge rectifier you need the common anode version, or the series type with 2 diodes, and one with anode connected to cathode of the other. Not common other than in tiny little SO packages, where they are common, and sadly limited to 40V 100mA instead.

One common anode and one common cathode.
See Reply #4.

[LinuxHata]

Folks, I'm not asking HOW to create a bridge rectifier from standalone diodes, I'm asking WHY they are not commercially available with Schottky diodes

[TimFox]

A quick search on DigiKey found 198 Schottky bridges, mostly SMT.
https://www.digikey.com/en/products/filter/diodes/bridge-rectifiers/299?s=N4IgjCBcoJwGxVAYygMwIYBsDOBTANCAPZQDaIALAAwDMVFATCALqEAOALlCAMocBOASwB2AcxABfQgFYA7FUQgUkDDgLEylKrVkAOFuy6ReAkeIkWgA
The three large (chassis-mount) Microsemi products (2 at 60 A and 1 at 90 A) are listed as obsolete and out of stock.
There are a few TH parts rated at 1 or 3 A, but not very popular.
For some reason (probably bad searching), I found only 2 on Mouser.

[dobsonr741]

The 4 diode bridge is an old legacy IMO, it has 4 diode dies assembled inside, usually in a through hole package, with no standard heatsink capable configuration. I saw SMD version only for low power applications.

Compare it with the dual anode/cathode Schottky - they are more frequently used as it servers more use cases, has well established footprints for SMD. Space requirements of dual Schottkys are not more than the 4 diode bridge.

So at the end dual Schottky is a practical choice for designers: a potential BOM reduction, space saving. For manufacturers it might be unwillingness to carry over though hole legacy into the SMD "world". At the same the the legacy still lives on for the old, tried and tested classic.

[bdunham7]

I really don't understand, why you should have 1.2 volts drop on say 10A bridge and waste 12W, when you can have 0.6V drop and waste only 6 watts?

Here are some thoughts:

What applications do you think would benefit from this and at what cost?

If you are building an unregulated battery charger or something like that, then you'd save a little power at the cost of increased price--but would it matter?

If you are building a regulated linear supply, you end up dissipating power in the regulator anyway...

If you want a more efficient supply you would use switching as opposed to linear. 

Also. comparing an 8A bridge rectifier with Schottkys rated for 5X more current is apples to oranges--a larger bridge rectifer will have somewhat lower V_f at currents that are a fraction of what it is rated for. 

Producing a whole line of Schottky bridges would result in a huge number of part numbers because of the variety of reverse voltages.  You can't just sell a 600PIV part that covers almost all applications like you can with silicon diodes.

Lastly,  you can easily build up a bridge out of two parts--a CA and a CC dual packaged Schottky--at a reasonable cost and with better thermal characteristics than a packaged bridge.  In your case, find the mate for this one and I think you have a good solution for $4 (sqp) in parts.  So the answer here is that there likely isn't really much demand for large or TH Schottky bridges. 

https://nz.mouser.com/datasheet/2/395/SR1620PT_SERIES_J2103-2509572.pdf

[Vovk_Z]

Folks, I'm not asking HOW to create a bridge rectifier from standalone diodes, I'm asking WHY they are not commercially available with Schottky diodes

Low-voltage bridge rectifiers are not used typically in most actual PS designs.
We use either two-phase rectifiers or something more complicated , e.g. syncronous rectification (with active mosfet rectifiers) etc.

[Terry Bites]

As ever, if there was a market - they'd exist.

For most recitifers the total Vf is/ was usually small compared with the output voltage.
By the time power Schottky's appeared, if you were worried about heat and effciency an smps was the way to go.
Put another way, the industry doesn't give a solitary about low voltage high current linear supplies.
SMPS can out-perform linear supplies in every respect.
They are no longer a money maker. In fact, quite niche.
Its assumed that you'd bolt your FBR to a chunk of heavy steel or Ali. So no fun in a plastic enclosure.

I wonder how well AD/ LTs ideal MOSFET thingy is selling.  AD ?? BTW, how do you get this kind of info?

[mariush]

Even those "mosfet thingy" aren't always suitable ... for example that LT4320 (controller to which you add 4 mosfets) only works when voltage goes above 9v
You can see that in datasheet  https://www.analog.com/media/en/technical-documentation/data-sheets/4320fb.pdf  and SDG Electronics demonstrated it in his video about bridge rectifier alternatives

See video from around 9:00

https://youtu.be/GJctTneVeFk?t=554

[MrAl]

Folks, I'm not asking HOW to create a bridge rectifier from standalone diodes, I'm asking WHY they are not commercially available with Schottky diodes

Hello again,

Ok, but why are you asking why a part is not available when the part is available?

Efficiency is an important issue for a switching converter in this day and age, and if it needs to rectify the output from a transformer, which is fairly common, then the more efficient way is to use a dual winding and just two Schottky diodes in a full wave rectifier configuration, not a four diode bridge rectifier, because that way you only have to deal with one diode voltage drop instead of two diode drops.  Only one diode conducts per half cycle instead of two.  The secondary would then be wound bifilar. 
I've done a full wave Schottky diode bridge rectifier circuit in the past for an automotive application where i needed a galvanically isolated power source using the car battery as the power source.  I did not want to have to wind a transformer with two secondary windings that's the only reason for that.  The efficiency was not that important because the power level was so low, much less than 1 watt.

[LinuxHata]

Dual windings are less effective in terms of utilizing transformer powers - for the same amount of coil area taken, you receive less copper, because you waste two times on insulation.

[MrAl]

Dual windings are less effective in terms of utilizing transformer powers - for the same amount of coil area taken, you receive less copper, because you waste two times on insulation.

Hi,

Many, many, many modern wall warts use the bifilar (two winding) method.  The old ones used a full wave bridge rectifier.

[TimNJ]

These days, low voltage AC rectification (i.e. <100V) is rather uncommon because most AC/DC power supplies are switch-mode, not linear.  For low-voltage rectification, Schottkys are a fine solution, and are frequently used in secondary-side rectifier circuits of switch-mode supplies. But most bridge rectifiers are used to rectify 100-240Vac, so rather high breakdown voltages are needed, usually 600 - 1000V to account for line transients. Silicon Schottkys essentially do not exist at these breakdown voltage due to issues of manufacturing and performance trade-offs (as far as I understand). SiC Schottky is readily available now at 600V+, but it does not necessarily have better forward voltage performance, and always has worse forward surge performance compared to PN (which is often important for passing lightning surge and so on).

[Zero999]

Folks, I'm not asking HOW to create a bridge rectifier from standalone diodes, I'm asking WHY they are not commercially available with Schottky diodes

Low-voltage bridge rectifiers are not used typically in most actual PS designs.
We use either two-phase rectifiers or something more complicated , e.g. syncronous rectification (with active mosfet rectifiers) etc.

True. Switched mode power supplies have a transformer anyway, so it makes sense to have a centre tapped secondary and use two diodes, to give half the voltage drop.

[jwet]

There are a couple of other reasons, but mainly because it just doesn't make that much sense in most cases- economics, process and manufacturing wise.  They do exist but probably in the right rare quantities where their use makes sense.  The recovery time is their main attribute which isn't required in LF/LV rectifier circuits.  With their lower Vf , they come with some baggage- they tend to be borderline thermally unstable in high current apps, poor PIV ratings and poor reverse leakage.  Where they would make sense, a MOSFET might be cheaper and much better for Vf if not tr.  Shottky's are majority carrier only devices unlike a PN junction which conducts with minority and majority carriers- this is what makes their recovery so fast.  They have to be made on a more complex almost IC process basically vs. just very basic processes used for a PN diode- this makes them more expensive and not so easily integrated especially in opposite polarities- this is why you'll see duals but not quads (bridges).  The other main takeaway if they make sense in your app is not to run them hot- underrate them to the point where they'll run around 100C max- they won't take the 200 C peaks of standard diodes without letting you down.

[ejeffrey]

Well, currently I need to rectify 20V AC at about 6A current.

That's a relatively uncommon thing to do, doubly so if you are concerned about efficiency. That's why there aren't a lot of parts.

In addition, the temperature dependence of leakage current is a reason to avoid packaging a bridge in a single package.  That makes each diode subject to the heat dissipation of the other 3.  Obviously that doesn't completely preclude it: you can design a bridge that would have enough cooling, but you need to use a bigger package with thicker leads and that may limit the advantage of integration for something that is already a niche product.

[james_s]

I suspect it's because most of the applications for bridge rectifiers are things where the characteristics of Schottkey diodes are not really needed. Typically they are used to rectify either line voltage or the secondary voltage of a 50/60Hz power transformer. The latter application is rapidly vanishing as linear power supplies go away and when they were in more widespread use nobody cared about such small gains in efficiency and in the former case you have plenty of voltage available and the current is usually relatively low. Since it's possible to make a Schottkey rectifier yourself fairly easy out of a pair of dual diodes, there simply isn't a large market for a full bridge type.