Effecient 3-Phase Buck PFC Rectification - Diodes out... Fets in?

[bentcrafter]

Bluntly, not my area of expertise.

I'm in-the-hunt for the most efficient rectification. The alternator is the leece Neville 110-555 - 24Vac @ 100A 3-phase, 400Hz max.  What's available? bolt-on, kit, DIY. This PDF got me to thinking about it:

http://www.pgembeddedsystems.com/controlpanel/upload/project/IEEE/Power%20Electronics/PG2012PE59/towards_a_99_efcient_three-phase_buck-type_pfc_rectier_for_400-v_dc_distribution_systems.pdf

[mobile_bob]

the following is my opinion only

the 110-555 alternator is native 12volt machine
if you are running at 24volts, that ok, but figure on closer to 500hz

the efficiency of the 555 running at the higher voltage (24) will be about 80%
the standard rectifiers will have about 5% losses attributed to them. so lets say a perfect rectifier would raise the alternators efficiency to about 85%.

there likely is no off the shelf three phase controller rectifier such as the one
in the technical paper, leaving you to design and build your own.

in my opinion the gains in efficiency afforded by the controlled rectifier while significant probably would not be worth chasing after, at best i would be very surprised if the alternators efficiency would increase to 83% with its use.

what they show for a schematic is a simplified and idealized circuit, using mosfets
but does not show the driver circuitry, that little omission in details constitutes probably 90% of the work required to build the thing and have it operate as designed at the hoped for efficiency and reliability.

in other words this is a very sophisticated piece of equipment , on the surface of which appears to be quite simple, but the reality is the development of which is going to be very expensive.

this is not to mention issues with emi and the fcc reg's that either would or should be complied with.

bottom line, while i see this as very interesting, i am not overly optimistic of the success of getting one designed, built and operating.

maybe there is something already available off the shelf from the aircraft sector? they run 400hz stuff, and 24volts is not uncommon either?  however the paper was written in 2010 so the technology might not be in the field yet, and if it is it probably is going to be very costly.

what are your goals?

bob g

[bentcrafter]

Thanks Bob,

So, are you implying that I have no other choices? Let's initially leave cost out of the equation.

I have the mechanics pretty much completed, and the focus is now on rectification - what's the options here? 

[mobile_bob]

remove the oem rectifier bridges
remove the oem diodes from the heat sinks
replace them with three of the #8772 and three of the #8773 diodes

those two numbers are for aftermarket 70amp 200piv rated replacement diodes
they are standard fast recovery diodes and not avalanche type.

i don't know who's numbers they are, however any supplier ought to be able to cross reference the numbers and get you what you need.

alternatively you could go to a rebuilder and have them supply you replacement rectifiers fitted with the upgrade diodes.

the diodes list for 12 bucks each, but you should be able to get them for much less if you tell them what you are doing with them, most alternator shops are fascinated by this sort of re-purposing. they don't get to do this sort of thing normally, so they sometime will be interested and give you a break on the price...i used to get them for a bit over 5 bucks each.

bob g

[mobile_bob]

it looks like those are "Transpo" numbers

bob g

[SteveU.]

Roger that on those being Transpo supplier part numbers. They ARE the auto electrical Rebuilders source for wholesale diodes for all auto applications and some voltage regulators. Jimco is the wholesale manufacturer/supplier heavy on the voltages regulators and testing.

Yes. Doing the same old thing day after day, year after year does get wearing. Be very polite and realistic and honest when you approach and ask though. These fellows get asked weekly to energy wise and mechanically make a purse out of a sows ear. Good ones will smell a free energy Con a mile away.

Hey Mobile Bob if you have an interest, here is the Rebuilders Association library link:
http://www.electricalrebuilders.org/techlib/tech_lib.php
You should be able to at least read the articles listings. Have to join to be able to download.
Halfway down they list a manual for a stator rewinding manual for hand use.

Regards
Steve Unruh

[BruceM]

Hi Brentcrafter,
What you are looking for is more commonly called "synchronous rectification", and has become somewhat more popular in more advanced designs, especially for low voltage power.  It uses Mosfets in lieu of diodes, with synchronous (to the AC waveform) switching of the Mosfets at zero cross time.

As Mobile_Bob suggests, its not for someone not skilled in power engineering. At the voltages you're looking at, it is not worth the added complexity and lack of reliability. 

[bentcrafter]

Hi Brentcrafter,
What you are looking for is more commonly called "synchronous rectification", and has become somewhat more popular in more advanced designs, especially for low voltage power. 

I do have a mosfet gem in my grubby little hand,.. but sadly, it's rated for only 50 amps. Would it be feasible to parallel two of these? - they tap directly off the stator.

[mobile_bob]

using mosfets is not cut and paste proposition, again the schematic they posted
on the paper was an idealized and very simplified diagram.

there are all sorts of things you have to have in order to use a mosfet successfully
for this purpose, as a simple switch not so much, but as a controlled rectifier very much so... capacitors, resistors to name a few, and of very specific values of generally very tight tolerances.

there are driver chips needed to force a hard on and hard off condition for the mosfet
otherwise they will go into linear operation and make heat and die.

there is just a lot of other bits and pieces to get the job done, and then you have things like part positioning, ground planes, stray capacitance and all sorts of other crap to attend to.

put simply, if you can successfully build a controlled rectifier or as BruceM properly calls it a syncronous rectifier, for this project, you in my opinion are wasting your gift!  you would make yourself a very wealthy man making mppt controllers, inverters and all sorts of other power electronic control units.

i don't mean to be a doom boomer on this project, i am just not overly optimistic of success. 

but, don't let me stop you from giving it a go!  it would be wonderful to have a member do such a project because it has such wide application.

bob g

[bentcrafter]

using mosfets is not cut and paste proposition, again the schematic they posted
on the paper was an idealized and very simplified diagram.

there are all sorts of things you have to have in order to use a mosfet successfully
for this purpose, as a simple switch not so much, but as a controlled rectifier very much so... capacitors, resistors to name a few, and of very specific values of generally very tight tolerances.

there are driver chips needed to force a hard on and hard off condition for the mosfet
otherwise they will go into linear operation and make heat and die.

there is just a lot of other bits and pieces to get the job done, and then you have things like part positioning, ground planes, stray capacitance and all sorts of other crap to attend to.

put simply, if you can successfully build a controlled rectifier or as BruceM properly calls it a syncronous rectifier, for this project, you in my opinion are wasting your gift!  you would make yourself a very wealthy man making mppt controllers, inverters and all sorts of other power electronic control units.

i don't mean to be a doom boomer on this project, i am just not overly optimistic of success. 

but, don't let me stop you from giving it a go!  it would be wonderful to have a member do such a project because it has such wide application.

bob g

My apologies. I should've been more clear... I have the assembled and fully functioning unit, not a single component. It functions fine, but I'm convinced that I can easily toast it - considering its 50a rating.

[mobile_bob]

oh i see

cool, got any pics or a schematic?

you might well be able to parallel connect them to increase the capacity, don't know without taking a look at what you have there.

bob g

[bentcrafter]

[BruceM]

The block diagram indicates that they've integrated a low side buck regulator with the high side half of the 3 phase bridge, eliminating half of the diodes.  They are still using diodes for the high side, so this is not true synchronous regulation.  But it is still potentially an efficient design.

I would not put two of these in parallel without manufacturer's approval, as we don't have enough of schematic to consider stability issues.

Buck converters provide no isolation, so you must be aware of this.  For your application the alternator output should float (no grounding) if the battery has one side grounded.

Who's product is this? 

[mobile_bob]

i wholeheartedly concur with Bruce
don't parallel these until you have talked directly to the tech dept of the manufacture
at the very least, and preferably the engineering department.

surprisingly if you are polite and to the point, you can generally talk to the engineering department and get answers to all sorts of questions. if you don't succeed with the first call, wait an hour or a day and try again,,, sooner or later someone will put you through

sometimes i have had really good look starting with the sales dept, who connects me to tech dept who then connects me to the engineer that knows the answer.

that is a good way to get past the gate keeper.

bob g

[bentcrafter]

... until you have talked directly to the tech dept of the manufacture at the very least, and preferably the engineering department.

Won't happen.

First, I don't speak Japanese,
And second, I'd rather not waste a professional's time by needling him/her with product usage... for equipment it was never designed for. I have a Fluke and an accurate digital DC watt & Wh meter, so it's more then usable as-is as long as i don't get abusive. Besides, I don't yet have a second unit so paralleling isn't possible right now.