Replacment Rectifier for MIG Welder

Discussion in 'Electrical Systems' started by MikeJohns, Sep 18, 2011.

  1. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    I have a 250 Amp single pahase Mig Welder which had a diode short circuit in the rectifier.

    The replacement rectifier was available but expensive and I didn’t really trust it already having had one die.

    Stud mount power diodes were very expensive but after a little research I simply bought a few 50Amp diode rectifier blocks in alloy cases (CM5006) added a few for a large FOS and wired them all in parallel.. I mounted them on a common heasink for close thermal coupling and kept the lead lengths the same and it works very well.
    The professional welder who was using my MIG machine thought it had a smoother arc than before it failed.

    If anyone wants to produce a high current cheap rectifier for any purpose the diode blocks cost me around $8 each for 10 and I used 8 for good measure because I didn’t know how well it would work.

    That gives a 100% duty cycle of 400A and they don’t even get warm at 180 Amps continuous , if I did it again I’d just divide the current by 50 and round up ( 5 blocks for 250Amps ) the rectifiers are already overrated and handle surge current an order of magnitude higher for short periods.

    I'd also use crimp connectors too rather than soldering.
     

    Attached Files:

  2. lasersailpete
    Joined: Feb 2013
    Posts: 3
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: newzealand

    lasersailpete New Member

    Hi am considering building one these for my 200amp mig how reliable has proved to be? Sounds like an economical fix genuine rectifiers are so expensive......cheers pete
     
  3. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Pete

    I researched this well, before I built the replacement.

    It has been a great success. The welder has been in intermittent use for some quite large jobs without any problem at max current.

    I did mount a fan on the heatsink. Just a 12v inline plugpak and a computer cooling fan because I mounted it away from the main units fan where the old rectifier had been.

    I'd simplify the construction. Mount the devices the same way on the heatsink with thermal paste. It's important that they are coupled thermally so they must all mount on the same heatsink.

    But I'd use multistrand wire and electrical spade crimp fittings ( the diode blocks have the matching lug already). Make all the tails the same length, connect the required ends together and solder the ends of those into copper pipe with one end flattened to make a connection lug.

    The rigid construction shown was not easy to solder, I had to use a small gas torch to get enough heat.


    Using this type of setup anyone can make a cheap very high capacity rectifier for any purpose, the current limit is set by the number of devices that can be physically mounted on one heatsink. But only for silicon power diodes , not schottky or other types.
     
  4. lasersailpete
    Joined: Feb 2013
    Posts: 3
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: newzealand

    lasersailpete New Member

    hi thanks for that have sourced & purchased 6 x rectifier blocks any chance of posting a wiring diagram ie what are those capacitors etc shown in the pic etc? had a lot of experiance with rc cars planes etc and am confident i can build this just dont want to get it wrong cheers pete..
     
  5. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Pete
    Here, drawn with 4 rectifier blocks just keep stacking them in parallel the same way. The components across the DC output are rf interference and spike suppression you can buy the RC as one component called a snubber. Probably fine without them. I just reused those components off the old rectifier.

    Cheers
     

    Attached Files:

  6. Mark Cat
    Joined: Jun 2011
    Posts: 115
    Likes: 4, Points: 0, Legacy Rep: 66
    Location: Michigan

    Mark Cat Senior Member

    In paralleling semicondcutors, it always interesting to ask, how will the devices share current? Which are subject to the greatest electrical stress, either forward or reverse? Which are dissipating the most heat? Why?

    Not easily answered.

    Overall to me the behavior (modelling) of the design with only 4 devices for the bridge far outweighs the cost penalty.
     
  7. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Mark


    The balancing act with silicon power diodes is simply temp, the hotter device has more resistance and quickly 'protects' itself. Providing the devices are evenly matched for their curves it's a very reliable technique. You'll find it in many high power applications. Mounted on a common heatsink or immersed in a common cooling fluid all the devices are close enough in temp and characteristics to achieve very close current sharing.

    As I posted before I opted for overkill and with forced cooling and 25A per diode (200A) the heatsink doesn't even get warm to the touch. I'd be very happy pushing that to 400 A continuous.
     
  8. jonr
    Joined: Sep 2008
    Posts: 721
    Likes: 11, Points: 0, Legacy Rep: 57
    Location: Great Lakes

    jonr Senior Member

    As I recall, most mosfets go the other way - the sharing gets worse until one of them burns up. But you can actively rotate among them to distribute heat.
     
  9. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Yes, most semiconductors that are actively controlled get positive feedback from temp rise. Same with the silicon transistor, but the bog standard silicon rectifier is a gift in this regard. Other devices like fast recovery diodes don't have the same characteristics and will fry if you try the same config.
     
  10. Mark Cat
    Joined: Jun 2011
    Posts: 115
    Likes: 4, Points: 0, Legacy Rep: 66
    Location: Michigan

    Mark Cat Senior Member

    MikeJohns,

    First of all I would like to say keep up the good work. The best way to learn power electronic design is to build and experiment. Then come up with ways to test and abuse the circuit to see what smokes first. Obviously, wearing the required protective gear of course.

    What is the input frequency?

    In modelling the circuit we would need to include all of the stray parasitic devices (lead inductance for example). For a 4 device bridge the task is easier. Even more so if the design includes a Printed Circuit Board (repeatable parasitics), thus taking out more variability in manufacture.

    As you said the temp coefficient helps. But we should dig a little deeper into the actual switching behavior over the operational current and input frequency. This type of analysis would perhaps illustrate the need of a snubber.

    All for now,

    Mark Cat
     
  11. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Mark

    Full wave rectifier 50 or 60 Hz around 50 Volts RMS Current and voltage depends on the welders setting. But say 200A and 50V

    Also consider the silicon power diode has a gentle switching region. 'Soft knee' in the forward conduction curve. That makes them quite forgiving in parallel applications.
     
  12. lasersailpete
    Joined: Feb 2013
    Posts: 3
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: newzealand

    lasersailpete New Member

    rectifier

    hi many thanks for all the info. built my rectifier over the last couple of days using your suggestions, made the large terminals out of 15mm copper pipe . used six rectifier blocks mounted to a large heat sink with thermal paste in between, made up four leads consisting of six 150mm lengths of 12g super worm silicon wire, i use this for lipo battery leads on my rc race cars nice and flexible, soldered on spade terminals to the end of each lead plugged them all in as to your diagram ended up secring them in place with solder, mounted the unit into the mig directly behind the fan bolted up the four leads switched it on and wow it works perfectly! welded continuously for fifteen minutes at full power the rectifier barely got warm and it welds so smoothly so impressed !! cost well under $100 nz dollars to build purchased the mig on trade me . the description was stopped running may be a fuse who knows grab a bargain !! got it home had a look and found a home made rectifier in there with some fried leads on a couple of the rectifiers the farmer i bought it of was the original owner from new so he knew all about the rectifier... at $600 for a new rectifier mm again many thanks cheers pete.. PS the pic of the unit mounted in the mig is quite deceiving the four large lead ends are at least 50mm away from each other and securely held in place...
     

    Attached Files:

  13. TerryKing
    Joined: Feb 2007
    Posts: 595
    Likes: 25, Points: 28, Legacy Rep: 289
    Location: Topsham, Vermont

    TerryKing On The Water SOON

    Looks like a good approach. Usually cost VS capability is best with commonly-used devices.

    Highly important that the exact same devices are paralleled, and best if bought in a batch from the same supplier, same date code...

    For the Specifications for the CM5006 see:
    http://www.farnell.com/datasheets/1662309.pdf
     
  14. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Terry

    Not much of an issue apparently, I got different batch codes with my diodes. I was told just the same type pref from the same manufacturer but not to worry if they were from different manufacturers so long as they had the same current rating and a similar fwd V drop.
    Apparently these devices ( Silicon power diodes) are very forgiving in parallel as they have a positive temp coefficient a slow turn on conduction knee and providing they have similar characteristics and share a heatsink they will share current quite closely and no single device can ‘run away’.

    The identical device strategy is important for different devices with faster switching or negative temp coefficients.
     

  15. sdowney717
    Joined: Nov 2010
    Posts: 1,175
    Likes: 85, Points: 58, Legacy Rep: 274
    Location: Newport News VA

    sdowney717 Senior Member

    I have a 200 amp AC stick arc welder.
    Could this be used to make a DC arc welder?

    How many rectifier diodes would you use?
    I can get 50 amp rectifier diodes pretty cheap.

    Will open circuit voltage be an issue for voltage rating the diodes?
     
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.