IGBT as a shunt regulator

Discussion in 'General Electronics Chat' started by excaliber, Dec 6, 2014.

  1. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    I need some help... I havent used a IGBT before but I think its whats needed for my project, though I am open to other suggestions. First I will explain. I am off grid and like building everything myself. I built a shunt regulator for my windmill consisting of a stack of doides (150 amp ones from junked welding machines) from memory 21 or so in series mounted on aluminium plates.

    what happens is this stack starts to conduct at 13 volts dc (maybe 1 amp at this point) but by 15-15.5 volts the stack is fully turned on and sheding power away from my batts...its worked great since 2007... But this took a lot of high powered diodes and I dont know if I can get more for free.
    I am building a alternator (otherpower style) to be driven by a engine at 1800rpm.
    The alt will be 3 phase and will be rectified to DC. its at this point I would like to find a way to clip peak voltage much as my diode stack did for me with the windmill.

    I figured shunt regulation would be the way to go. I would like to have the alt supplying power up to 13.4 volts then have the Igbt clip anything above that, if it could be adjustable that would be even better.
    I am open to other and any ideas, and understand using the IGBT as a series pass element may be another way to go.

    I was thinking shunt as Its 100 percent efficient up to 13.4 with all power going into my bank up to that point.
    Ideas Please...thanks.
     
  2. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    I have just been reading some of the post on here and maybe a fet as shunt regulator might work better? its more like a variable resistor, that may be what I need. Ididnt understand that a Igbt is better for on/off control, but thats what I am gathering from the post I am reading on here.
     
  3. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    I'm guessing you need to limit the speed of the windmill with this? Otherwise if it is a standard alternator you could just control the field current to control the output voltage - much less power. How many amps is the alternator? FETs or transistors would work. Often big resistors or some other load is used with them to keep the power (hence the number) down since they are a little expensive.
     
  4. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    yes for the windmill example the stack of diodes serves as both charge control and as a load to limit windmill rpms.
    I was not clear, the alternator is not of typical construction,rather its of the otherpower style, if you go to thier web site you will see what I mean, a flat steel disk with neo mags around the outside then hand wound coils cast into fiberglass for whatever voltage/current you wish then a second disk of neo mags, thats how my mill is but for this engine it will just be a single disk with neo mags, I dont need the second disk as I have so much more RPMs to play with. I would like to construct the alternator so its capable of hammering out 40 amps tops if I want to do a eqalization charge up to 15 volts dc. But for normal use 15 amps maybe 20 is fine. dont really need more. Idealy if I could set my control so at 13.4 volts it starts to pass current that would be great I would set my engine rpms so that my alternator roughly tops out there and my regulator just clips any extra. now if I want more power for a equalization charge I would bump my rpms up a bit and would like to be able to then set my control to 14.9-15 volts and I could just let it run until the batt voltage hits this level and any extra is just clipped. The engine will be a 80 cc one set up for 3600 rpm but I am changing the rpms to around 1800 maybe less. I am building this to get the longest charging run time for a liter of fuel.
     
  5. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    Hmm, maybe a series regulator would work better? Then when there was no load (battery charged) there would be no load on the generator either. I didn't find the alternator on the site yet.
    To calculate a shunt we would need to know the no load output voltage and the resistance of the coils. That way we can calculate how much power will be in the transistors.
    Generally speaking permanent magnet alternators are used for wind because they work at low rpm, but for this a plain old car regulator would work the best because then you can regulate it with the field current.
    But I bet you want to build your own.
     
  6. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    I have thought of a car alt, and in the end maybe I will look back and wished I had gone this way, but I do suspect I will get better effeciency this way and learn something along the way. If I go series pass do you recommend a IGBT then?, having it close and apply full power to supply the loads and hammer the batts up would work for me.I have some from junked welders.I took electronics in school, but that was decades back and I dont even remember if they were invented then so I am not up on a drive cct. Your right about the application for neos being low rpm, but they can be used for higher...the alt will be 12 pole and with a rpm of 1800 (though I can go up or down from there to tune it in) I will be around 360 hz, so I also suspect I should use fast recovery diodes for the rectifier. If I go serries pass with a igbt can the device just be triggered on with a zenner voltage divider network?
     
  7. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    It's hard to say what the best way to do it would be. A lot depends on the resistance of the coils in the alternator. Do you know their resistance, or what size wire and the length would also tell us. It would be more complicated than just a zener, but maybe not to bad.
     
  8. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    I havent built a test coil yet, but I suspect each of the nine coils will be of 16 gauge wound two in hand, maybe three in hand but not likely, inside dia of the coil would be 1.25 inches, 14 wraps per coil. So lets say two // 16 gauge wires 60 inches in length per coil, 3 coils per phase, nine coils total. Thats roughly wear I am at if I had to guess
     
  9. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,800
    1,103
    When IGBTs are switched fully on they waste more power than FETs (for a given current). However, in this particular application you are actually trying to waste power rather than save it. So IMO I don't think it matters whether you use FETs or IGBTs.
     
  10. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    Unless you build a closed loop control amplifier to drive the insulated gate input (same as a MOSFET) - the VGSthr can have a fairly wide spread on value. The VBE of a bipolar transistor is far more predictable, but parallel devices to shunt that sort of current is more complex - and of course you need more current to drive the bases.

    IGBTs were pretty much an interim measure to combat on losses in SMPSUs due to RDSon - which could be pretty high in early MOSFETs.

    For a shunt regulator which should never be into saturated switching, an IGBT has no particular advantage over a MOSFET - and a MOSFET is probably cheaper.
     
  11. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    I was kind of afraid of that. It has a very low resistance so what you would be regulating with a shunt regulator would be engine speed.:D It also means the shunt has to dissipate the full 40 amps - so over 500 watts. It also means the engine is under full load even when the battery is fully charged. Not so good for the gas mileage. That would lead us to a series regulator. Now what you need is a bit more than 13.6 volts so you can use some thru the regulator. So lets say you set the engine to give you 15 volts at 40 amps. Now the regulator only drops 1.4 volts at 40 amps or 56 watts. Of course as the battery charges and the current goes down the engine will speed up, but the regulator will keep the voltage constant.
    This would still take a lot of heat sink and there is danger if the motor speed is set to high so we need to think about that. I want to look at a possible PWM, but it may take a bit.
     
  12. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    The IGBT mentioned by the OP was an interim measure to avoid the high RDSon of early MOSFETS, the IGBT has no particular advantage over a plain old MOSFET in this application. IWHT; sheer production volume would make a very robust MOSFET much cheaper.
     
  13. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    I will try to explain my application better. Lots of ways to charge a large battery bank, and hitting it with full 40 amps and bulk charging it up then tappering off and finally holding it at 13.4 is likely ideal. Nov, Dec, and January is my gennie charging season, and while I have done my charging that way its not ideal for me in my situation. I see -40 c every winter, -30 likes to visit and hang around, often brings his buddy -20 C who could stay for a month, fighting to start a gennie every day or second day, or maybe a couple times a day grows old fast, or if I am not home, others have a hard time starting the gennie.
    So this year I started playing around and found I could drop rpms on my genni down to around 2400 (its only a 80 cc engine), the ac voltage goes down to something like 54 volts ac at a freq of 42hz if I lower rpms any more I collaps the field in the alternator...now this 54 volts ac is fed into a hacked 120volt mig welder transformer, that was from the junk yard...this works very well, I am getting 20 amps charging current at 12.5 volts or so, and it tapers of to about 15 amps as the voltage rises to 14 volts, now the system averages lets say a 10 amp draw steady state, sure it spikes up when the fridge comes on or the water pump but these are more momentary things..so I have found I am getting run times of 12 hrs on 5.5 liters of fuel. I like this..I can fire up the gennie leave for the day and know the gennie will carry the average base load if the sun doesnt come out, the few extra amps the gennie puts out does not seem to harm my bank as its large at 1600 amp hrs, and no one has to fight to start the gennie when I am not there.
    Engine wear and noise are greatly reduced. its proving to be a good idea for my application...but I know if I could remove the multiple inefficiencies, first from the AC alternator, then from the step down transformer I should be able to drop rpm further and extend run times greatly.
    So if I build this three phase alternator and get it through tweaking to roughly top out at around 14 volts dc, then my thoughts were that if I could build a shunt that just clips a half volt or so off that when the batts are full, I thought that might only consume a amp or two ., now as my batts need power and/or a load comes on the voltage would drop lower than my 13.4 volt set point and current is passed to the batts as needed. the lower the volt drop the greater the out put from alternator to drive it back up.
    If i ever need lots of power like 40 amps I would just dial my set point up and bump my rpms up to get it for things like a equalization charge. Thats how I was thinking I would use this, and how the shunt would operate.
     
  14. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    If you hadn't talked of liters I would have guessed you lived in North Dakota. :)
    I'm going to try and steer you back toward a car alternator with a custom regulator. It's my turn to think out loud. Everyone can start laughing now.
    I think most alternators now days will put out 40-50 amps at 2400 rpm which is probably the sweet spot of you engine. So you could gear it for that or a little higher.
    Now the regulator takes care of everything for you. If you need 40 or 50 amps to recharge or run the fridge you can get it (within the engine power) and if they are charged it will drop the current. The only thing missing is a way to increase the voltage for the equalization. You can either do this with a charger or we can make a regulator that is adjustable. I think the only other thing needed is a switch so you can turn it off to start the engine.
    Lets see what other people think.
    Edit:
    The thought just occurred to me that maybe your gen set keeps the speed constant regardless of load??????
     
    Last edited: Dec 9, 2014
  15. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    pic00007 437.jpg
    I would love to hear what others think, just so you know though I will be building this, have started and am not adverse to building one with a typical alternator as well, in fact I think I have one sitting in the garage.I like redundancy.
    thought I should add, if everyone thinks a standard alternator is better...would they think a servo motor would be even better...I have up to 3hp 180 volt dc ones Glentek if I remember right
     
    Last edited: Dec 9, 2014
  16. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    I made an edit at the same time as your post. I'm thinking maybe your gen set keeps the speed constant with some type of speed regulator to compensate for load?
    Looks like fun. Keep in touch.
     
  17. excaliber

    Thread Starter New Member

    Dec 6, 2014
    9
    0
    yes, the rpms stay mostly where I set them...a heavy load (like 40 amps) would knock them down though. not if I was running it at 3600 rpm, but a heavy load at 1800 may, though the disk I am mounting my magnets to weighs more than the motor so this flywheel may give the engine govenor a chance to recover....I am outside my scope of knowledge so this all is a wag for me.
     
  18. InterSparky

    New Member

    Dec 21, 2014
    1
    0
    Would it be possible to vary the motor speed? As the voltage falls, speed the motor up, as the voltage rises above 13.4 slow the motor down. This would reduce the fuel consumption when the battery is fully charged.
     
Loading...