Use of computer power supplies instead of step down transformer possible?

[minvayu]

Dear all,

First of all thank you for this board.  I have spent countless hours reading postings and have learned a lot from those experts that can share their experiences.  Thanks to all for their time and patience!

I made an experiment to see if I could use some leftover laptop power supplies to step down a 350W 220V Hugh Piggott turbine to charge a battery bank.  I though this is not done for some reason, but it seemed interesting enough to test.  I had two options, to run the turbine and dump the power to some resistors or send the power to the battery via the power supplies.  When the turbine was connected to the resistors the blades would easily spin up with higher winds and I got around 60W of power coming through.  I would switch to send the power to the battery bank and I could measure 16V at the power supply, but the inverter/battery combo would stall the wind turbine.  We used two identical power supplies in parallel that work from 70V up to 240V AC input, 16V output.  We rectified the power to DC and fed it directly to the power supply.  The power supplies used are rated for 6amps, 16V.  We connected them to a 12V battery bank.

This was done as a test to see if this setup would work instead of buying a 1kVA 3phase step down transformer (Higher rating to compensate for low frequency operation to avoid saturation).  The reason we chose the higher 220V voltage was to save on cable since the turbine is about 60 meters from the battery bank.  We also need to use the system on some small scale waterpumping systems (wind-electric) so we want to connect it directly, without batteries or inverter.  WHen we connect a motor to the turbine we have capacitors helping keep the V-I curve proportional and in this way operate the pump in "variable speed" operation.

If I can figure out a way to step down the voltage and actually not stall the turbine it would be great.  Our project teaches village mechanics how to build turbines, and they invariably end up using thinner wire than they should, so I am looking for low-cost alternatives for voltage step-down that can be implemented in rural India.

My questions are:
If I get a lower voltage, such as a 14.5V SMPS would it work better (Currently it is 16V)

I have seen several designs of "home-made" MPPT controllers out there, but they tend to be up to 70v input/output.  I need something that can work at 220V and step down to 12V.  I am not an expert in power supply engineering, therefore we can use guidance on sizing of components that would work.

We also connected a PWM solar charger hoping that that may help the system to work better, but that also did not work properly.  Of course, a solar charger does not follow the power production of a wind turbine, but it was fun to try anyway!

I will be installing our newly build arduino data logger to get windspeed/rpm and power measurements and can post more operational data soon. 

Jorge
MinVayu.org
India

[SparWeb]

Three 500kVA single-phase transformers cost less than one 1kVA 3-phase transformer.

Frequency is a problem, yes, but starting with european 50 Hz models helps.  How many poles in the WT generator?  At low wind speeds who cares about saturation when there is very little power to capture anyway?  It sounds like you have an AC pump.

I'm not qualified to help you select the right transformer rating, I can only suggest that the primary winding DC resistance may or may not stall your turbine the way your other jury-rigged system did.

I see very little reliability or overload capacity in the computer power supply arrangement you are trying to build, but I'm not opposed to the act of trying for its own sake.  I just think it's likely to stall the turbine no matter what you do.  It's kind of like making a variable frequency drive but backwards - variable frequency input + DC link + fixed frequency inverter.  There will be lots of losses and complications.

Can we discuss your reasons for avoiding transformers so far?
Can you elaborate upon the decision to build the wind turbine for 220V output?
Was this the balance of system expected when winding the stator wire?
Or was it something completely different, and you're already on "plan B"?

[OperaHouse]

It could be a good idea if you were a designer.  The 12V out is a pretty poor design and is not suitable for an all the time load.  The 12V isn't actually regulated, the 5V is and the 12V just tags along.  You could move that over to sample the 12V.  A UNO could easily drive the opto isolator that controls the voltage.  The UNO could monitor the input voltage as well as the output.  The regulator needs to be able to shut off to control stall.  Still need to add a load dump controlled by UNO.  There are lots of mod PC PS info on youtube but never the one you have

[minvayu]

I got a quote from a manufacturer or transformers, but 3 single phase transfo cost more than one 3-phase transfo.  Maybe in CA it is the other way around, but here it is not.  The 1kVA 3ph autotransformer is costing 10,000 InR, about 200 USD.  I am able to get laptop or 500W Desktop SMPS for 10 USD.  So you can say that price is the main driver in this alternative setup.  Also, the possibility of hacking the Power supply with the Arduino and connecting a dump load would be perfect.  If anybody knows of a hack that could do the trick let me know!  I will start looking at youtube.

The reason I built a 220V stator was 1) to test a HV system, 2) Test some wind-electric waterpumping for a Aquaponics setup.  THe pump is AC and rated at 220V. 3) Can use excess power for a Dehydrator (The system is in a farm).  Ideally we would love to implement a battery-charging option, but here in India affording a 200 USD transfo is difficult for a farmer, therefore our interest in testing the PC power supply alternative. 

The turbine is the 1.8m Turbine designed by Piggott.  It uses 8 mags, 6 coils.  Startup is around 240 rpm.

We originally planned on just running the pump and the dehydrator, no battery charging.  In India we need to figure out ways of generating an income.  Both these applications are important in this area.  THe battery charging would be great, but only if we can build something on the cheap to step the volts down. 

I also looked at a EV bike charger, but they come ready to charge 36 or 48V battery banks, not 12V.  We would have the same problem as the Laptop chargers I assume, of stalling the turbine.

Our other turbines are mostly 24 and 48V systems charging batteries and they are plain and simple. I have 8 turbines running in the area but this little baby is my first HV.  Of course, if I could get my hands on a MPPT Classic then my problems would be solved and I would be a happy camper, but 800 USD is not justifiable unless I am building a higher rated turbine (say 3.5m+ blade diam).

Jorge

[OperaHouse]

I pick up a used PC power supply whenever I can for 50 cents but the transformers aren't useful for much.  What is the 12V current rating?  under 10A and greatly exaggerated at that.  I like the 200V turbine idea but I would much rather use a couple transformers backwards from an old 12 or 24V inverter.

[minvayu]

I wish I could get a power supply for 50 cents...!! 

The use of a used inverter transformer is a good idea.  THere are plenty of inverters here so finding a second hand transfo is possible.  Will need to see the cost of these parts.  To make sure I am not making a mistake, the 3 single phase transformers are connected phase - to phase, or do I need to connect them Neutral to phase?  If N-Phase then the voltage output is 110V instead of 220V.  Would need to modify the output tap in order to get 14.5V+0.7 diode drop.

[SparWeb]

The economics you face are different from mine - sorry for making an assumption like that.

Still, since you're still in the experimental phase, find or scavenge old used stuff.  For example, I once obtained a dozen 500VA transformers, for 12 USD each as "surplus".  If I had a project like yours, I would be keen to try them first.  Your start up frequency could be 30Hz (just off the top of my head) so already entering the effective range for a transformer even if it's not optimal.  Getting it working, though, gives you the experience to later buy the right thing.  Or avoid buying the wrong thing!

You have many connection options depending on the wiring available from the WT alternator (3 wires Wye, 4 wires Wye, 3 wires Delta, 6 wires either one??) and then do the calculations to find the optimal combination.

[Flux]

As always there are compromises, your requirement for direct pumping conflicts with the ideal for battery charging.

I have always considered transformers an expensive option but if you increase the frequency they become more cost effective. Your 8 pole machine at 1.8m is likely to be in the region of 40Hz at full output or a little more so you may have to de-rate a bit on 50Hz transformers. If you went up to a 16 or 20 pole alternator you could use much smaller transformers and reduce the cost but you couldn't then use the ac motor for direct pumping but you may be able to rectify and use a dc pump motor.

In theory the computer power supply idea could work if you could find an electronic genius to modify tthe feedback circuit to a current control and you will need to include a speed ( frequency) component for best results. I don't know enough about these supplies to know if this is really viable and the long term reliability may not be good.

The idea is the same as a mppt converter but there is a slight difference in the quality of engineering between a cheap consumer product and a specialised wind mppt controller. If you are going to get the help of an electronics genius then it may be better to design from scratch.

You don't need to bother with the neutral with transformer connection but it will do no harm to include it if it is avaiable on the alternator and transformer primary.

It usually works out cheaper to use single phase transformers as they are mass produced and cheap. If you have a local producer then it ought to be cheaper to build a 3 phase transformer as the core will use less material than 3 single phase ones.

Possibly you could compromise with a 115v system with 115v pump motor. You could then use 240v Hz transformers without de-rating. At 350W I don't think you will have much trouble with cables even at 115v over your fairly short run. I can see it being an issue at 12v but I wouldn't have thought it was an issue even at 48v battery charging and with direct connection I suspect you will need some cable loss to avoid stall.

Flux

[Flux]

Having re read your previous question perhaps I ought to explain a little more about transformers.

A 3 phase transformer with 230v primary would connect to the 3 alternator lines and if you used a standard 50 Hz device you would need something near 350 v nominal primary to keep the core flux down as your full speed frequency is a bit below 50 Hz.

You only need to consider frequency at one speed as a transformer works in volts/cycle mode. At half frequency it will only have half volts. If the frequency is ok for full load it will still be ok at cut in speed. The only factor here is that if you use the high flux densities of modern commercial transformers the iron loss may be a drag and delay start up. Keeping the flux a bit lower than for mains supplies is a good idea but increases cost.

With single phase transformers you have the option of connecting primaries line to line and you use 230v primaries for a 230v alternator ( at 50 Hz) or you connect the primaries line to neutral so you now have 110v on your 230v primary. Ths comes down to you star connecting the primaries and in fact there is no real need for the neutral ( it has minor benefits) but not enough to justify another 60m of wire.

If you use 230v single phase transformers star connected you will have a 415 V effective 3 phase transformer and you will be underfluxed at 40 Hz and not near saturation.

You will actually see a derating in that you have changed from delta to star and your primary current rating has been reduced by root 3 but you will now be able to use standard units which are usually cheap. If you start looking for higher voltage units to work at lower frequency you may have trouble finding them and cost may be high.

You have the options of connecting your secondaries in star or delta so this give a fair choice of voltages so you should find suitable standard devices. Ideally if you use a star primary with a rectifier you ought to use a delta secondary if there is no neutral but for small machines like this it really isn't that important and star/stra seems to work ok.

Flux

[oztules]

Simple is best with wind.
Modification of the computer power supply is fairly simple, but only  if you are handy with a soldering iron, and understand more than just the basics.

I do it a fair bit when I want battery chargers for fixed current or fixed voltage use or float charging.

The best alternative if you don't want to throw good money after bad, is to get three microwave transformers from the dump, split off the secondaries, and rewind them to what you want. Leave the primaries alone to keep the insulation in tact, unless your confident of your skill set .

In the bigger ones, there is enough steel laminate to do about 1.5 turns per volt..... but experiments will tell you exactly what you need to do. Plenty of room to put big wire in there too, as your removing a mountain of finer wire.

If you want still more power you can use two E to E to make up one ( not E to I )... but you'll be up in the kilo watts to want to do that.

They are big, tough, and generally free, or even you'll be paid to take them away.

As a mains transformer, they are not too good without a primary rewind, as they are designed to run in saturation, so may need a choke in series, but for your little mill, the 240v primaries will do fine as is..... I use them for  lots of things actually....

You may need to use a freq/voltage sensor to keep the transformers turned off until up near to cut in to keep the stall at bay at very low rpm. Some bigger solid state relays are good for this application ( what we use here on our 240v mill).... In truth I think we could probably connect directly to the transformers, but we didn't for the last 8 years, and see no reason to change away from the solid state relays (3.7Meters AWP)

...................oztules