My 18 foot Dual Rotor eats rectifiers

[mitcamp]

When I built my first 18 footer it was a 12 volt machine. I had the wrong shunt on my digital amp meter and it indicated 200 amps, so after I got the proper shunt it showed 100 amps in a 20 m.p.h wind, cut in speed at 80 r.p.m.

Recently I built a new 24 volt stator for this machine. I also built another set of blades. The old blades were built with red pine 2 x 8, mainly because I could find nothing else that was dry. The new blades are built with clear pine boards 1 in thick and 12 in. wide, 4 boards laminated together allowed me more pitch at the root and a wider blade.

 



My problem is this,  I must have my rectifiers at the top, because I use sliprings and for some reason, this machine eats a couple rectifiers every couple weeks. I have been using 2 35 amp bridge rectifiers, paralleled on each phase for a total of 6 rectifiers. I have them on a sizeable heat sink and use heat sink compound. Yesterday the machine again acted like the brake was on and I had 2 rectifiers shorted again.





This time I used the 50 amp bridge rectifiers. I only had 3 of them, but will parallel them as soon as I can get more replacements. Will this cure my problem and if not what should I be useing.

I have thought of using the bridge rectifiers out of Leese Neville truck alternators, the ones that are 3 diodes on a heatsink and use one package on each phase. Would this be more robust than the 50 amp bridge rectifiers that I just installed or do I have to build 3 bridge rectifier using Stud mounted Diodes. I would like to resolve this problem before the cold weather comes. My 18 footer is up 100 feet. Any information would be helpful.          Thanks  Mitcamp

[SamoaPower]

Personally, I wouldn't use these block bridges on a bet, either the 35 or 50A. I guess you've noticed that they're fairly inexpensive and is probably why many use them. There's been a number of reported failures.

Don't know the specs on the alternator diodes, so can't comment.

My suggestion would be to build a 3 phase, 6 diode bridge using schottky diodes to reduce losses and heat.

Your heat sink looks adequate but would be better mounted with the fins vertical.

Another thing noticed from your photo is wire nut connections. At these current levels, I don't think that's a good idea. Better to use a proper terminal block and soldered ring connectors on the wire.

I assume your slip rings only have two contacts. Three, of course, would allow the rectifier at the bottom.

[Nothing40]

I second that.

Those bridges aren't up to it,and sometimes they can do "funny" things when paralleled (current sharing issues,the one with the lower voltage drop wins and ends up taking all/most of the current).

Schottky diodes would be a good idea for everyone!

The lower voltage drop is a good thing,a little less heat to get rid of also.

Some "Beefy" diodes I've used for varous things are:

MUR20030CT

MUR10020CT

Both "obsolete",but there are 'new' versions available,I'm unsure of the new numbers though..a crossrefrence should find them.

Another good one, 122NQ030 Schottky 120A/30V (kinda low voltage rating,but okay for general 12V stuff.) 123NQ080 is 120A/80V and 123NQ100 is a 120A/100V part.

Or some big Stud-mount diodes,or whatever. Just be sure that the diodes are able to handle the voltage and current.

A Google search for "122NQ030 Datasheet" (or whatever part-number) will usually land you a datasheet on almost any diode you can dig up.

[Ungrounded Lightning Rod]

Be sure you have some heatsink compound between the rectifiers and the heat sink.

I can't make out how you're wired.  Is that a 3-phase output with 3 AC leads coming from the alternator, running a foot of wire from the wire nut to each of the bridges' AC terminals?  Or are you jerry-rigging it?

It looks like you're running about a foot of wire from a big wirenut to each of the AC terminals for each phase.  That's nice - it will serve as a set of balancing resistors.  EXCEPT that I'd SOLDER them, rather than using a wirenut.  Wirenuts can end up with some internal resistance to the connection that doesn't balance well.  (I'd also take care in positioning the wires in the twisted-together bundle before soldering so they have similar current paths within the joint.)

[tecker]

Looks like you a little light on the wire size anything over 30 amps needs a buss bar to connect to .The schottkys are a good idea bigger than you need . Every good power supply deserves a fan to kick in when the temp gets critical.

[tecker]

Belive me 50 amp bridges should say 50 amp momentary 25 amp continuous. they are too hot at 30 amps.

[jimjjnn]

00 amps through 35 amp bridges is really pushing it. You can drive them for short periods that way but sooner or later, they will short out or open which is worse as your turbine will overspeed.

Also Peak Inverse Voltage rating can be a problem too.If the PIV is too low, you are going to lose lots of bridges.

Myself, I would use stud mount Schottky individual diodes with at least a high rating for your 100 amp input to them.

[DanOpto]

Buy bigger rectifiers

Here is a selection From Ebay on Tuesday Oct 3

YASKAWA 200A 3PHASE 3 PHASE DIODE RECTIFIER BRIDGE GUAR

Fits Yaskawa PT-4090 or you can use for your purposes Item number: 320033812283

MBR7030WT, 70A 30V, Schottky Power Rectifier, Qty 5 Item number: 320033838331

200CNQ030 30V 200A schottky center tap rectifier, diode Item number: 160035785539

DD171N14K Eupec Dual Diode Power Module- 1400V, 171A Item number: 260037362065 $2.50

SKKD100/14 Dual Diode Power Module- 1400V, 100A Item number: 260037360692 $2.99

Lot of 5 New 1N4045 Power Rectifier Diodes 275A 100V Item number: 200032298196 $27.00

Go Big!!!!!!!!!

[harrie]

Hey Danopto, Thanks for the list. Ive been haveing the cheep 35 amp problem also.

[DanB]

I agree with most of the advice in all the comments here - except for the suggestion of using a fan.  Not much sense in that since the rectifier appears to be mounted to the wind turbine.  Even if it's on the ground though - I think an over sized heat sink is more reliable, simpler, and quieter than involving a fan.

You can search it yourself on ebay - but I often find fairly good prices on 3 phase rectifiers there, it's like the bridge rectifiers you're using now - except a 3 phase unit instead of a single phase bridge.  They make things nice/simple and clean - 3 leads in, two out and I often see them rated up to 150 amps for $50ish.

Ive had problems with using smaller single phase rectifiers too - although in most cases if we overbuilt it enough it's been fine.  I used 9 on my old 14' 12V machine and it was always fine.  I think if you use those it can be fine... you just need more of them, and things start to get messy.  Easier just to buy a 3 phase rectifier I think.

[Flux]

I agree with all that has been said but I want to throw in another possible problem.

If you are rectifying before the slip rings and the rings open circuit momentarily you may be killing the bridges on over volts rather than over current.

If that is the case then schottkys will be in even more trouble.

Large metal oxide varistors may offer some protection, but if the break is long enough they may not be adequate.

If you had 100A at 12v, you should not have much over 50 at 24V. 6 bridges should handle that if your heat sink has any cooling. If it is boxed in it will not be effective.

I would look at 400v piv min and preferably 600V with slip rings.

Flux

[mitcamp]

When I built this machine I never thought about the rectification. Because it is a workhorse at low r.p.m always showing 5 - 10 amps in low winds when my other generators are turning and not producing anything, I have to go the distance and build a rectifier system to accomidate it.

I have looked at the Ebay rectifiers,thanks Danopto and I agree with Dan B that 9 35 amp bridges makes a maze of wires to contend with.

I am leaning towards building with 6 Schottky diodes as suggested or building with the large stud mount rectifiers. I imagine the 6 Schottky diodes are the most expensive and probably the best bet over all at $21.00 ea. Would thier be a noticeable differience in preformance between the Schottky set up and the large stud mount setup.

I am considering ordering a 3 phase bridge like the YASKAWA bridge to do the job until I get another rectifier together.

I have not built such a rectifier with individual rectifiers on a heatsink,so if you guys got a picture,point me to a link.

Flux, my sliprings are quite robust and have 2 brushes on each brass ring 180 degrees apart. I can tell by the wear marks and the amount of tension that they are doing the job well. Thanks that certainly could have been one of the things contributing, but I believe my main problem was the bridge rectifiers.  Thank you all and a picture would help.     Mitcamp

[tecker]

Ac down the pole would be the best mod here . Exposing the rectifiers to the weather with the wirenuts as connectors is probably the problem .As far as a temp senseing it's hard to know if the cooling is good enough if your not monitoring it . Turning on a fan will give some indication of how close your comming to rated temp ( given that your working in large range of possibillities ). you should be 50 percent or less of rated temp to keep the bridges from a partial short or open. Really nice work on the blades I wish I have a better hand at that stuff.

[Flux]

You will not see any difference with the shottkys at 24v, you probably would at 12v.

The stud mounted diodes are much tougher than the potted bridges, but they do need at least a split heatsink, use 3 stud anode and 3 stud cathode diodes.

Even the press fit type diodes are quite good but not as easy to mount as the stud ones.

In view of the slip rings, the cost and the fact that you will see no difference I personally wouldn't use the shottkys in this case unless your heat sink has no cooling at all and you need lower loss.

Flux

[mitcamp]

Good advise Flux and Tecker, A.C. down the pole is out of the question on this machine. It would have to come down to revamp the slipring setup. I will use the large stud diodes on it and make it do for the winter. I will get rid of the wire nuts and make a positive connection. The machine has a cover to keep most of the weather out and the fins of the heatsink are exposed to the outside air.        

I am going to build another similar machine this winter and I will include a 3 slipring modification to it, in order to place the rectifiers at the bottom.  I will be  using the 3 x 1 1/2 x 3/4 magnets on the next one.





Flux, I do recall seeing a Bridge rectifier set up, sometime ago using those big stud diodes. I spent a couple hours looking for it today with no luck, but I will find it. The guy was using aluminum channel and isolated them with a non conductive material, I think wood. I will isolate mine with a p.v.c sheets or squares. Thanks again you saved me a few dollars, as I was going to order Schottkys.

[mitcamp]

I,ll try to repost the picture of the weather housing around the machine.   Mitcamp

[willib]

mitcamp ,not sure if anyone sugested this yet ...

but how did you wire the alt with 2 bridges in parallel per phase?

http://www.windstuffnow.com/main/3_phase_basics.htm

as you can see on Eds page you only need two bridges per alt to hook it up normally...

i dont know how you used six bridges on one alternator , ??

because to double it up you only need four

just trying to help

[scottsAI]

Hello mitcamp,

Many years with power electronics has shown,

Never get close to the rating.

Years ago I started out with a 10a circuit, the 10a power transistor lasted less than 1 hour.

Have found to operate at around half current and/or voltage seemed to be a safe number.

Voltage rating of Schottky is low, any chance to go over, could fry them.

Best not to exceed half the reverse voltage rating of a diode.

Digging into the power rating I found it dependent on temperature. The hotter it gets the less power it can handle, details are not always included in the data sheet. Some will have a derating curve based on temperature.

With 18' blades at 20mph, the power is 3727w at 24v that should be 155 amps.

This is assuming 36% efficient. (60 * 80 * 75). limit * gen * blade.

Since your getting something must less than that the internal voltage of the generator is much higher.

The 1kw out your getting is leaving 2-3kw as heat in your stator.

Cut in at 5mph = 24 * 4 = 96v peak voltage is 137v Not many Schottky diode can go here.

Peak current is 38.8amps.

Cut in at 6 mph 24 * 3.x=80v peak 114v. Need 200PIV rated diode as a minimum.

Peak current is 46.6amps. Stator heat is 2600 watts. Usable power 1118watts.

Must be where your operating. Efficiency is 30% of available power goes into battery.

Generator heat and voltage will go up 24 volts for each cut in incremental MPH.

Example cut in is 5mph, at 25mph the 96+24 would be output voltage.

This Simple explanation is not including diode drop, so voltage is likely higher.

The generator has some inductance, if the slip rings open, the voltage spike can be MUCH higher. The blades will speed up very fast adding to the output voltage (not to the spike).

All things considered, 400piv would be the minimum I would consider.

50 amp current, then 100amp min. Paralleling then should be ok.

Three things will kill a diode.

Temperature - for 50 amps the heat sink looks ok should be about 100w.

Current - you should be able to measure. Then determine the heat generated.

Reverse voltage - If the first two are ok, then this is the reason.

Heat in watts is 2 * 1 * 50 = 100w (two diodes in series, 1 volt drop under heavy load) I know most will use 0.7v, check data sheet.

I did not know there was so much to write about diodes! Such simple things!

And there is more, not needed here:-)

Hope this helps.

Have fun,

Scott.

[mitcamp]

Ed, maybe I screwed up on the wireing, but here goes. Take a look at the closeup picture of the 6 rectifiers mounted on the heatsink. First off the stator is wired star with 3 wires coming out.

This picture is the first set up of 6 35 amp rectifiers back in March when I installed it. One output wire from the stator goes to the 4 ac terminals of the two bridge rectifiers on the top, the wires are #10 orange. I used a red wire nut to make the connection.

The 2 neg. #10 black wires from the 2 top bridge rectifiers go to bottom slipring which is neg.

The 2 pos. doubled #12 red wires from the 2 top bridge rectifiers go to the top slipring which is positive.

The remaining 2 sets of bridges are wired the same, with the same colours. All the neg. black wires from 6 bridges are connected together and go to the bottom slipring which is neg.

All the pos.red wires from the 6 bridges are connected together and go to the top slipring which is pos.

After passing through the sliprings, I use 2 pieces of 2/0 welding cable to go to my batterys, one is pos and the other neg.

Every couple of weeks, I would have to change a bridge rectifier, so I changed my system from 12 volt to 24 volt. I was trying to keep my wire losses down and lower the incoming amperage. As I said earlier, I was contending with over 100 amps with 12 volt and now I am seeing half of that. Actually last night in a 20 m.p.h.front I saw 59 amps. I may reset my machine to furl at 20 m.p.h instead of 25 m.p.h  At the moment I have 3 50 amp bridges on that same heatsink. These are temporary, as I am building a bridge with 6 D08 Stud Diodes, similar to the one electronbaby built a while back.

 I am anxious to upgrade this machine because of its low wind output and would not have got this far without the help from this site and the members who helped me along the way. I learn by making mistakes, my next machine will be better.

                       Mitcamp

[mitcamp]

Scott, I welcome the education on diodes, I am in the works of building a better rectifier. Don,t understand how you come out with the calculations of stator heat and useable power to the batterys though. I am building a similar machine this winter. Any suggestions ? What can I do to make my next machine more efficient, bigger mags, bigger wire, add another slipring and bring 3 a.c wires to the bottom. What about higher voltage and using a transformer at the bottom. This is certainly addictive,and fun.  Thanks Mitcamp

[DanG]

thats good info - I lost my hard drive days ago and couldn't find the data sheets to back up a my posting a reply last evening; Scott just nailed it... I got 40 of these in the basement at the moment http://tinyurl.com/mgfse

Some other factoids here...

Also working RMS voltages absolute maximum rating is 70% of reverse voltage rating - so to play it safe AC RMS voltages should be 25-35% or reverse rating.

Also for these small bridges the design spec sheets are from tests using a 8" X 8" .25" thickness aluminum heatsink FOR EACH BRIDGE, equivalent to 2.4 pounds of heatsink w/ alot of surface area!

The photo of the bridge heatsink MIT shows would have to be 6 times larger to meet design specs! Ooo, those fins should be mounted vertically (shown horizontally in picture) to get better smoke stack effect, alot of dead air space if they "work" as shown"!!

And also published specs often are 60Hz single phase half-wave, so pulsed DC w/ sine waveform, not full wave - and at that, the 25 amps the 1000V 50A briges in spec sheet above are acting as heaters of 50 watts each, so a rating of 50A at 50° C plummet to 25A at 100°C, etc..

Anyhow just more things to consider, believe that serious heatsinks are required!!

[fungus]

Also they say on the dataseet to derate the rating by 20% if used on capacitive loads.

[scottsAI]

Hello mitcamp,

This is certainly addictive,and fun.

Yes it is.

My view is not to use PM gen on the larger machines.

Due to the problems outlined above and others.

Remember anything can be made to work.

Look around here you will see some of the things people have done to get better efficiency out of the PM gen. They add complexity, reliability issues and cost.

PM gen is great for the smaller wind generators. Not sure where the cross over is, depends on your goals.

Web link showing how to calculate the power in the wind:

http://www.windmission.dk/workshop/BasicBladeDesign/bladedesignleft.html

I made a spread sheet with the calculations, reports power in watts and HP in 1 mph increments.

Work sheet shows the RPM of TSR 1-15, so is big.

I can change any number... I use open office.

The power in the wind is NOT the power you get from the generator.

The RPM of a wind gen is directly proportional to wind speed over a good range.

Voltage output from PM gen is directly proportional to RPM.

If the PM gen is generating 24v at 5mph, then at 10mph it is outputting 48vdc.

The batteries impedance is so low it yanks the voltage down to what it needs. The impedance in the gen is higher so the power is dissipated in the stator. At 10mph, half the power is output into the battery, the other half is heat in the stator. You can see why low cut in is not good for PM gen.

Larger PM gen people have fried their stators, sometimes not even in huge winds.

Good air cooling can help. I want as much of the power as I can get, not satisfied with just getting a little more in higher winds, I want it all. (if free:-)

I am working on a 10kw design. Considerations:

Grid connected. Too large to store this much power in a battery.

Not interested in power in the winds above 20MPH, storm, costly.

Design LOW cost. 5 year payoff. 10 cents KwHr.

Use as many off the shelf parts as possible.

Recover all the power 5-20MPH. Helps with payoff.

3 ph power at 240vac.

If your needs are not the same then reconsider.

Generator is an unmodified 4 pole 1800rpm 20HP 3ph 240v induction motor.

24' Blades spin at 80RPM, using speedup gears to spin motor at 1855RPM full power.

To use induction motor as generator you need to spin it as fast as it slows for full motor load.

Gears are from a vehicle power train. Nice quiet gears, junk yard stuff. 97% efficient.

The downer, why this solution is not used more often:

Requires active pitch control, must keep RPM nearly fixed, Needs high T/D blades.

Pitch changes TSR from 14 down to zero. Challenging.

Furling is done by pitch control.

Normal looking Unusual Tail allowing yaw rate change control.

If interested, click on my user name look at my post, most of this is covered in them.

For the slight cost and increase in complexity I expect to recover 3 times the power vs simple PM gen.

Looking at the cost of the complex PM gen with equal performance, $12k. Induction $6k.

Which fits in your wallet?

For me the grid tie was the deciding factor, induction gen connects directly to it. Huge inverter would be needed for PM gen. Became simple decision to select induction.

Questions email me. Or post.

Have fun,

Scott.

[willib]

in the picture

on the top left bridge rectifier , it looks like BOTH of the AC inputs are comming  from the same phase ?

if so that is not correct.

a single bridge  should have two different phase wires going to it , with the second bridge having just one phase it may sound strange but thats the way it works..

and of course the plus outputs should be tied together like you have them, and the minus outputs tied together.

[mitcamp]

Willib, You are right about the wiring. I had the same phase, split at the black wire nut and had them going to the same A.C terminals of the same rectifier. Don't know where I come up with this arrangement. Too busy I guess. You would think after changing rectifiers every 2 to 3 weeks it would raise a flag.

This continued until August when I installed the 50 amp bridge rectifiers and only at that time did I run a single A.C wire to each bridge. Yesterday I removed the wire nuts and replaced them with mechanical nuts that you tighten with a screw. I havn't had a failure since, but still plan to build a heavy duty bridge with the DO8 stud diodes.

Another thing that slipped my mind was, it seemed that my rectifier failures showed up after I had the machine braked for either a storm front approaching or when we were going out of town for a few days. When I returned and threw the knife switch from brake on to brake off, this is when the machine acted like the brake was still on. I would replace the 1 or 2 bad rectifiers and away it would go again. Still learning. Thanks for the keen eye you have.       Mitcamp

[ghurd]

It is using 2 of the 4 diodes.

I believe that is a good design, just not for so many amps.

The diode lots match, I hope, as do the Vf as close as can be expected.

I expect it would be fine for 15~20A. Usually.  This is a big windmill!

G-

[jmk]

 What happens if they are both hooked up to the same bridge ac terminals? Is it just an over load? I think I hooked mine up that way too. I'm going to have to pull it off the wall and check.      

[jmk]

 So if you have three bridge rectifiers do you have two with two differant phases of ac and one with the third phase going to the same bridges ac terminals?

[mitcamp]

jmk, not sure if the question was directed to me, If so, I have 3 50 amp bridges, I have one a.c line going to each bridge, meaning there is a unused a.c terminal on each of the 3 rectifiers. this will have to do until I get something bigger put together.       Mitcamp

[paulpic]

was this it?

http://www.otherpower.com/images/scimages/351/charge_controller_box_3

[paulpic]

was this it?

http://www.otherpower.com/images/scimages/351/charge_controller_box_3

[mitcamp]

Paulpic, where do I get more information on this rectifier setup. I am assuming there is 1 stud anode and 1 stud cathode used on each of the 3 phases. Neat control box.       Mitcamp