High amp (250A +), low foward voltage loss, Rectifier build - SCR or IGBT build?

Thread Starter

RogueRose

Joined Oct 10, 2014
375
I've been looking for some high amperage, low voltage rectifiers for a simple project (Rectifying various modified transformers). The voltage has been as low as 2.8v and I'm suspecting it is capable of about 350-430A (at that voltage as the transformers are rated at 1000 and 1,200 watts). I have found one rectifier that seems nice for the $$ but may not work well with such low a voltage.
https://www.ebay.com/itm/New-MDQ-40...441466&hash=item2837d8d1b4:g:fsYAAOSwpONZRPbw

I've read that full wave bridge rectifiers can be made by using mosfets, specifically IGBT's and SCR, but I'm not certain if it is possible to make one that can handle this high an amperage (though the notes I've read said they were capable of very high amperages with almost no forward voltage loss).

I'd like to build one if possible and I need to order the parts necessary for this, so if anyone has any experience with this, I'd really appreciate suggestions on where to go with this and if what I am doing is possible. I'm still kind of green when it comes to electronics so I may not have used the correct terminology to describe the usage of the mosfets. I do know that the setup required 4 of them to function properly (at least maybe in a full bridge..?)

Finally, if I am looking for parts, what feature of the IGBT or SCR determines the forward voltage loss and amperage (is the standard amperage rating what I look at or would it be calculated differently when being used in this application?).


So if I found IGBT's rated at 40A, would I need 40 of them to make this work (that's if they can be paralleled) so 40A * 10 pieces, then each bridge needs 4 pieces?


I'm not sure if these would work but they seem to be priced attractively and rated at 200A.
GT30J127 30J127
Spec sheet
 

Attachments

cmartinez

Joined Jan 17, 2007
8,220
The voltage has been as low as 2.8v and I'm suspecting it is capable of about 350-430A
Can you physically verify that the amps are as high as you're expecting them to be? ... on the other hand, IGBT's are quite capable of very high currents compared to mosfets (although a bit slower, so switching frequency could be an issue) and seem ideal for what you're aiming for. You're going to need a special circuit to drive them, though. Just like mosfets do.

If I were you, I'd look for IGBT's capable of handling at least 50% more current than what you're expecting. And I'd also protect them quite well with proper snubbers.

Digikey has a very useful interface that lets you filter whatever component you want to specific parameters:

https://www.digikey.com/product-detail/en/ixys/IXXX300N60B3/IXXX300N60B3-ND/4321707

https://www.digikey.com/product-detail/en/ixys/IXGX320N60B3/IXGX320N60B3-ND/3586374
 

tcmtech

Joined Nov 4, 2013
2,867
So if I found IGBT's rated at 40A, would I need 40 of them to make this work (that's if they can be paralleled) so 40A * 10 pieces, then each bridge needs 4 pieces?
Unfortunately IGBTs have a forward voltage drop similar to a common silicon diode so they won't help much in a low voltage active rectifier application. Low voltage mosfets with very low on resistance values would be what you want. That or high current germanium diodes.

As for reducing losses change your design over from a full bridge to a center tapped winding and half bridge will cut your parts count and rectification losses by 50%.

Also for a ~ 400 amp circuit you would want to have at least 600 - 800 amps of rectifier capacity on each current carrying leg so a full wave bridge would require some 2400 - 3200 amps of diodes and center tapped half bridge would require 1200 - 1600 which in both cases are going to require a heavy buss bar type power conductor system of at least 1/2 a square inch (.5" x 1" or .25" x 2") cross section for copper or 3/4 square inch (.75" x 1" or .375" x 2") cross section for aluminum.
 

Thread Starter

RogueRose

Joined Oct 10, 2014
375
Unfortunately IGBTs have a forward voltage drop similar to a common silicon diode so they won't help much in a low voltage active rectifier application. Low voltage mosfets with very low on resistance values would be what you want. That or high current germanium diodes.

As for reducing losses change your design over from a full bridge to a center tapped winding and half bridge will cut your parts count and rectification losses by 50%.

Also for a ~ 400 amp circuit you would want to have at least 600 - 800 amps of rectifier capacity on each current carrying leg so a full wave bridge would require some 2400 - 3200 amps of diodes and center tapped half bridge would require 1200 - 1600 which in both cases are going to require a heavy buss bar type power conductor system of at least 1/2 a square inch (.5" x 1" or .25" x 2") cross section for copper or 3/4 square inch (.75" x 1" or .375" x 2") cross section for aluminum.

Thanks for the informative reply! the few articles I was reading mentioned a rectifier that was capable of incredibly low forward v loss, maybe on the order or .05v or .1v and they stated they were capable of enormous voltages. It mentioned that it was derived from the need of a low loss solution in computers that run at 1.1 - 1.2v at high amperages. I only state this to make sure we are on the same page with what I was discussing.

I understand what you mean about the center tapping and that is a neat solution! Thank you.

Now that I recall, they may have mentioned SCR's before IGBT's, but said both were capable. Maybe it was the SCR that could produce such a low V drop.

this whole project is pushing the boundaries of my ability a slight bit, not construction, but design so maybe I'll have to shelve it for the meantime, but I am still interested in learning the concept.

As a note on the amp rating, if there are 4 diodes, wouldn't they basically be switching back and forth, so each carries the full load 50% of the time? IDK if that counts as pulsed amperage that some diodes are rated for, where the ratings are .5 - 3x the continuous amp rating.

I still need to come up with a solution for dealing with 120-220A at from 8-24 (three identical 8v transformers in series) V. I ordered 10 50 rectifiers.
50A 1000v rectifier
I have some really great heat sinks & fan I can use and thought I'd try 6 in parallel and see how long it would last.

I was planning on using these rec's for 8-24v power tools with the transformers as the bench supply. then select proper voltage tap, connect to plug (with rec inside batt housing) and run the tool. I think these rec's will be fine for that (suggestions welcome) as they are standard 12-24v power tools that run on NiCd or Li-ion.

Now just need to figure out the best method for low V drop becuase at such a low V that will be A LOT of heat that will need to be dissipated - correct? If it doesn't produce heat = A * Vdrop, and something different, then I can make the V higher to deal with loss, but current will be the same. (it's for large electrolysis cell and running plates in series isn't option for a gas seperation setup, easiest is with minimal V and lots of current)
 

tcmtech

Joined Nov 4, 2013
2,867
Now that I recall, they may have mentioned SCR's before IGBT's, but said both were capable. Maybe it was the SCR that could produce such a low V drop.
Unfortunately SCR's are a based around silicon junctions just as most other switching devices are they too have teh same ~.6 or more more volt forward drop issue as any IGBT or common power diode has. Germanium or actively switched Mosfet are teh only two things that I Know of that can get under that forward drop voltage level.

As for current each diode in the current flow path carries 100% of the current plus if its a sine wave that current has a peak well above the average hence the need for at least a 2X overhead.

Regarding high current capacity diodes, they make them in devices that go into the 1000's of amps ratings as well and the cost per amp gets a lot cheaper at that point with having one large device rather than trying to work with multiple or many smaller one in parallel.

Realistically if it was me I would be looking for some used diodes from a larger arc welder and just modify the transformers to give a bit more voltage to compensate for the loss. Overall the added power loss is inconsequential to the point that if you can't afford to loose an added 100 - 200 watts you probably cant afford the components and materials to do the work you're wanting to do anyway.

And BTW, if you're winding custom transformers HID lighting ballasts work way better than old MOT's do. ;)
 

Thread Starter

RogueRose

Joined Oct 10, 2014
375
Unfortunately SCR's are a based around silicon junctions just as most other switching devices are they too have teh same ~.6 or more more volt forward drop issue as any IGBT or common power diode has. Germanium or actively switched Mosfet are teh only two things that I Know of that can get under that forward drop voltage level.

As for current each diode in the current flow path carries 100% of the current plus if its a sine wave that current has a peak well above the average hence the need for at least a 2X overhead.

Regarding high current capacity diodes, they make them in devices that go into the 1000's of amps ratings as well and the cost per amp gets a lot cheaper at that point with having one large device rather than trying to work with multiple or many smaller one in parallel.

Realistically if it was me I would be looking for some used diodes from a larger arc welder and just modify the transformers to give a bit more voltage to compensate for the loss. Overall the added power loss is inconsequential to the point that if you can't afford to loose an added 100 - 200 watts you probably cant afford the components and materials to do the work you're wanting to do anyway.

And BTW, if you're winding custom transformers HID lighting ballasts work way better than old MOT's do. ;)

Thanks again for the very comprehensive and informative supply! I will say in the articles I read mentioning rectifiers that listed SCR/IGBT it specifically mentioned MOSFETs ( and stated that the SCR/IGBT's WERE mosfets, so I thought that mention was with relation to the SCR/IGBT). I'm embarassed to admit that i'm not good at understanding how Mosfets or amplifiers work but I know what they are used for. I have some strange learning issues where I can "paint an entire picture - or parts - but I can't explain what is going on or what it really means". It was the same when I learned IT/system/network admin. I knew all the parts, could build and do all the work, but didn't know how they interacted (other than things like major components - HD/RAM/Mobo/CPU/Video/Buses/etc) until one day when I had an epiphany and all the terms/acronyms all made sense and I completely understood the relationships and everything at a deeper level. Sorry for the long explanation, but I guess it could be a learning disability of some kind but IDK what type.. The same thing is happening now with Electronics and chemistry as i'm studying both. I only explained b/c I think you had some "issues" about why I was asking dumb questions a while back.

I said before I have 8 250w HID ballasts (Magnetek 1110-246-c-tc - $90-120 each - 120/277v, ~15.5L x 3.25 W x 2.75" H" ~ 19lbs But got them for an AMAZING deal!) but they have this nasty tar that I would need to remove somehow if I wanted to rewind them - any suggestions on how to handle this would be GREATLY appreciated! The units I have are
http://www.ballastshop.com/1110-246c-tc-universal-250w-mh-fcan.html
The problem is that they are new and I just can't bring myself to tearing them apart. I'm trying to figure out where I may be able to find old ballasts from lights that are being torn down.
 

tcmtech

Joined Nov 4, 2013
2,867
I really don't mind long replies that give decent explanations of details. I do them all the time given I rather enjoy writing even if I am not the best at it.

Like you, I have the same odd learning curve. The stuff all backs up until one day it explained in my head and all of a sudden it all just makes sense down to be a near subconscious ease of understanding and application.

As for finding used ballasts most any electrical contractor will likely have old takeout they will sell cheap, scrap yards and recycling centers get them all the time too given the amount copper in them.

Odds are if you took you new units to a electrical contractor who does commercial lighting work they would likely appily trade you a number of old and large HID ballasts for 1 - 2 of your new ones given a good quality posted HID ballast in the size range of yours are worth $75 - $150+ new and old open frame HID ballasts are worth about $3 - $10 depending on how much copper they have in them.

As for learning the details of what an electronic component is and how it actually works, that's what internet searches (or the forums own educational links above) and a few hours of educational reading is for. ;)
 

Thread Starter

RogueRose

Joined Oct 10, 2014
375
I really don't mind long replies that give decent explanations of details. I do them all the time given I rather enjoy writing even if I am not the best at it.

Like you, I have the same odd learning curve. The stuff all backs up until one day it explained in my head and all of a sudden it all just makes sense down to be a near subconscious ease of understanding and application.

As for finding used ballasts most any electrical contractor will likely have old takeout they will sell cheap, scrap yards and recycling centers get them all the time too given the amount copper in them.

Odds are if you took you new units to a electrical contractor who does commercial lighting work they would likely appily trade you a number of old and large HID ballasts for 1 - 2 of your new ones given a good quality posted HID ballast in the size range of yours are worth $75 - $150+ new and old open frame HID ballasts are worth about $3 - $10 depending on how much copper they have in them.

As for learning the details of what an electronic component is and how it actually works, that's what internet searches (or the forums own educational links above) and a few hours of educational reading is for. ;)
thanks for the reply! I've tried to think of ways to create learning tutorials that would bring together the "sticking points" that held me up - but IDK if those are usually unique to the individual or what. I find that analogies are often the vest in learning things, as long as someone knows another topic well. In electronics, comparing air or water flow has made learning electronics easier until I understood the basics of electronics better.

I picked up a couple 250 and 400w HID transformers (thanks for the suggestion on where to get them!! Great idea!). I found something that I didn't expect in that there is only one connector on the primary which looks like it connects to the secondary (after the entire winding), and then there are 5-6 output wires. All wires have continuity with each other, touch any of them, including the primary and they connect to all wires in the secondary. Very different from a MOT. I see why these may be better as they have much more wire than MOT's but they are rated at much lower wattage for basically the same size - though I can run more of them in parallel or series (series more likely) if needed.
 

tcmtech

Joined Nov 4, 2013
2,867
I picked up a couple 250 and 400w HID transformers (thanks for the suggestion on where to get them!! Great idea!). I found something that I didn't expect in that there is only one connector on the primary which looks like it connects to the secondary (after the entire winding), and then there are 5-6 output wires. All wires have continuity with each other, touch any of them, including the primary and they connect to all wires in the secondary. Very different from a MOT. I see why these may be better as they have much more wire than MOT's but they are rated at much lower wattage for basically the same size - though I can run more of them in parallel or series (series more likely) if needed.
Double check your taps and what windings they belong to. With HID ballasts the bigger most often untapped or single tapped winding is the secondary not the primary.

By far the multi tap primary Constant Wattage Autotransformer (CWA) type shown on page 14 are the most used and likely what you have.

And yes the primary and secondary are all interconnected but that's easily changed by unhooking the one lead which with the HPS, Metal halide and Mercury Vapor type ballasts that connection point is on the 120 VAC primary tap.

If you have not seen this before it's a great reference booklet/PDF for working with the various types of ballasts and will give you a pretty good idea of what type and configuration of ballast you are working with.

http://media.hydroponics.net/productdocs/HID_Pocket_Guide.pdf

Also for non constant wattage output use you can take the two laminated shuts out of the core that are between the primary and secondary. Just use a punch of some sort to drive them out if they are press fit in or if they are part of the core sections themselves grind them off when you have them apart.
 

Thread Starter

RogueRose

Joined Oct 10, 2014
375
Double check your taps and what windings they belong to. With HID ballasts the bigger most often untapped or single tapped winding is the secondary not the primary.

By far the multi tap primary Constant Wattage Autotransformer (CWA) type shown on page 14 are the most used and likely what you have.

And yes the primary and secondary are all interconnected but that's easily changed by unhooking the one lead which with the HPS, Metal halide and Mercury Vapor type ballasts that connection point is on the 120 VAC primary tap.

If you have not seen this before it's a great reference booklet/PDF for working with the various types of ballasts and will give you a pretty good idea of what type and configuration of ballast you are working with.

http://media.hydroponics.net/productdocs/HID_Pocket_Guide.pdf

Also for non constant wattage output use you can take the two laminated shuts out of the core that are between the primary and secondary. Just use a punch of some sort to drive them out if they are press fit in or if they are part of the core sections themselves grind them off when you have them apart.
I haven't had a chance to check the PDF you posted but I'll do that soon. I did take some pics of the transformers I took out of the casings, and I double checked all the wires, and every wire has continuity with every other wire on the transformer - I haven’t' checked if they are connected to the case.

The first 2 are of the 400w transformers (I got 2 of them) and the wiring diagram. Both came with a 24uf 400vac cap that looks like the ones that come out of microwave oven.

The last 3 are of the 2 transformers, it looks like the one has a burn400w.jpg 400w label.jpg 250 - good.jpg 250w-burnt.jpg 250w label.jpg 400w.jpg 400w label.jpg 250 - good.jpg 250w-burnt.jpg 400w.jpg 400w label.jpg 250 - good.jpg 250w-burnt.jpg 250w label.jpg 250w-burnt.jpg 250w label.jpg t secondary b/c it's black, but nothing flakes off the wire, so it kind of looks like it was painted black for some reason...? The label is the last image. They came with a 15MFD 400vac cap.

I have to say that the ballast I took apart, the 250w one that is 15.5" x 3.5" x 2.75" has A LOT more copper wire in it, maybe 3-4x as much. I think I had about 9-10lbs of copper from the unit and I can guarantee that these 250w's I found have no were near that much copper. The thing is that the ones I just picked up are A LOT easier to work with and don't have 4 lbs of tar coating the windings, steel case and cap 48uf @ 330vac


Also, does anyone know what the highest wattage rating is for one of these ballast's? I've seen 1000 and 1500 advertised but not sure if they go higher. It'd be really nice to find some old units that were to be decommissioned to pick up or trade for my good 250's or even 400's I've just found.

SORRY FOR ALL THE IMAGE PROBLEMS. I Always have problems when selecting Thumbnails as it always adds them 2x and full image....
 
Last edited:

tcmtech

Joined Nov 4, 2013
2,867
The one is burned and 1000 and 1500 watt are as big as they go.

Also, being they are designed to run continuous duty in extremely hot environments, in normal working conditions like you will be using them, they can be substantially over driven for long periods without harm. I've seen many with oversized capacitors on them that had them running at 120+% of rated power that were that way for years that never burned up.

The winding insulation is typically class H or H+ which will easily take 150c (300 F) operating temperatures with peaks as high as 220C (425 F) without damage. ;)

https://www.hammondpowersolutions.com/faq/what-is-meant-by-class-in-insulation/

As for finding larger units it s just a matter of asking around at the local electrical contractors and with whomever services your local streetlights and large signs plus local scrap yards. They're easy to find once you know the local sources.
 
Top