isolating and rectifying the individual phases...?

[makenzie71]

Would it work the way I think it would?

I've always thought of three phase PMA's as having three separate generators all working in parallel...can I break them up?  So we have L1, L2, and L3.  We normally bridge all three into one DC circuit.  Would I be able to rectify the individual circuits (L1+L2, L2+L3, L3+L1) into their own isolated DC circuits?  I am thinking the voltage would be the same, but amperage would be cut by a third like it is when one of turbines drops a phase.  Does that make sense or am I missing some critical component here?

[joestue]

most motors produce a little bit of a third harmonic which is canceled out by the common star connection of the 3 windings.

so basically you don't get a full 1.73 voltage increase when connecting in star. one example i can give you is a 10 pitch winding i did on on a 24 slot 3600 rpm 5kw generator head. I got 138 volts per phase but when connected in star i only got 230vac out of it. something like 14 volts went missing.


So when you rectify each phase you only get 1x volts instead of 1.73x volts. this basically the same as the star delta transformation. you can connect a 3 phase machine in delta, or you can rectify each phase individually and connect them to the same dc bus. its effectively the same thing but with the delta connection if the harmonics are really bad then you get some circulating current at lower loads. i don't know that anyone has ever reported that this impeded startup under low wind conditions.

but if you can pull off a set of relays to change from star to delta, you will get some improved power output from low wind conditions.

other than that, rectifying each phase separately (and you can add a capacitor after each rectifier and run all 3 rectifiers in series) you can get a 3x voltage out of the machine. it won't have much  power because you'll be relying on the capacitor to provide the current, but it will work to pull a little more power out of the turbine in really low wind.


other than that, yes you can have 3 separate rectifiers and 3 separate loads, each isolated from each other electrically. not sure what the point would be....

[Adriaan Kragten]

Three phase rectification is explained in my public report KD 340 which can be copied for free from my website: www.kdwindturbines.nl at the menu KD-reports. Star rectification is explained in chapter 3.2.1. Delta rectification is explained in chapter 3.2.2. Rectification of each phase with a 1-phase bridge rectifier is explained in chapter 3.2.3. This last chapter has been added recently to KD 340.

[SparWeb]

Rectifying the phases separarely works just fine.  Done it many times.

I learned about it from (the late) Jerry, one of the Fieldlines members who tried lots of things, back in the day.

So in my diagrams I refer to separated phases as "Jerry-rigged"   

Interestingly, it is also easy to come up with a Star-Jerry switch to have two speed ranges, if it helps matching blades to generator in different conditions.  I had this in my system between 2012 and 2017.

Throwing that switch in a storm and watching the ammeter needle jump was a lot of fun.

[makenzie71]

Neat, thanks guy!  I was sure it was something that would work...I know that when one of my turbines drops a phase it still kicks out power on the remaining phase, just not as much, so I was sure it would, btu wanted to check.

As for the reason WHY I would want to do this...I want to try to couple a high voltage turbine to something like an SMA 7.0 string inverter.  That inverter has three inputs but can only handle roughly 10 amps per input.  The turbine I'm looking at will be running roughly 350~400vac and 10~15a total.  I want to feed each phase into each input and see what happens.

[bigrockcandymountain]

I'm skeptical whether it will work, but also eager to see the experiment.  Oddly enough, i was looking at the sma inverters for solar very recently.  The 3 mppt lines each track voltage separately.  I don't think they will like the lumpy single phase rectified to dc very much without a lot of capacitance to smooth it out, but i could be wrong. 


You've done a few things that i didn't think would work and you proved me wrong, so I'll watch and see. 

[Adriaan Kragten]

The advantage of a separate 1-phase bridge rectifier for each phase is that the total DC voltage is a factor 1.155 higher than for star rectification if the voltage drop over the rectifier diodes is neglected. However, for star rectification the current flows always through two diodes and for three 1-phase bridge rectifiers, the current flows always through six diodes. So the voltage drop over the diodes is a factor three higher and about 4.2 V in stead of about 1.4 V if silicium diodes are used.

So for a low battery voltage, the diode losses are relatively high for three 1-phase bridge rectifiers. Rectification in star is therefore better for a low battery voltage but for a high battery voltage, the use of three separate 1-phase bridge rectifiers is a good alternative although you need totally twelve diodes in stead of six diodes.

[mab]

I rectify AC into an sma sunny boy 1200 for my small single phase hydro generator (up to 450w). Seems perfectly happy - although i do wonder how well it would cope with the ripple if it was pushing 1200w.

[Warpspeed]

The advantage of a separate 1-phase bridge rectifier for each phase is that the total DC voltage is a factor 1.155 higher than for star rectification if the voltage drop over the rectifier diodes is neglected. However, for star rectification the current flows always through two diodes and for three 1-phase bridge rectifiers, the current flows always through six diodes. So the voltage drop over the diodes is a factor three higher and about 4.2 V in stead of about 1.4 V if silicion diodes are used.

One configuration not yet mentioned, is to separate each phase (six wires total) feed each phase into its own bridge rectifier, then connect the whole thing in series. That gives around double the voltage produced in each individual phase, but with an extremely low final ripple voltage and current.  In fact you don't really need an output capacitor.

The final dc  output waveform looks like what you expect to see from a six diode bridge, but at a voltage about twice as high.  Diode losses will be a lot higher though, but for higher voltage systems it may still have some application for a wind turbine.

[makenzie71]

The advantage of a separate 1-phase bridge rectifier for each phase is that the total DC voltage is a factor 1.155 higher than for star rectification if the voltage drop over the rectifier diodes is neglected. However, for star rectification the current flows always through two diodes and for three 1-phase bridge rectifiers, the current flows always through six diodes. So the voltage drop over the diodes is a factor three higher and about 4.2 V in stead of about 1.4 V if silicion diodes are used.

One configuration not yet mentioned, is to separate each phase (six wires total) feed each phase into its own bridge rectifier, then connect the whole thing in series. That gives around double the voltage produced in each individual phase, but with an extremely low final ripple voltage and current.  In fact you don't really need an output capacitor.

The final dc  output waveform looks like what you expect to see from a six diode bridge, but at a voltage about twice as high.  Diode losses will be a lot higher though, but for higher voltage systems it may still have some application for a wind turbine.

You can connect the outputs of rectifiers in series?

[Warpspeed]

like this:


With star connection, at any instant there are effectively two windings in series, the rectifier disconnects the third winding, so its just sitting there unloaded for a short time.

With the above circuit, all three windings are always in series and supplying current all of the time, and you get the highest output voltage possible.  Not only that, the ripple voltages overlap and cancel out producing a low ripple dc output.

There ain't no free lunch though. 
All of the windings are conducting all of the time, and you must reduce the final output current rating accordingly.
So you end up with higher final dc voltage and less available current...

[kitestrings]

...the Fieldlines members who tried lots of things, back in the day.

No disrespect to those still here, but I do miss 'those days'... and those contributors. ~ks

[Warpspeed]

The above circuit was built and successfully tested by me a very long time ago, and I know for certain it works.
Should have some pictures of it on an old hard drive around here somewhere, but they may be lost.

Thinking about all this a bit more now, adding three electrolytics, each should charge up to peak voltage, increasing the dc output even further.  The ripple in each series section may be high but should also cancel just as it does just with the bare rectifiers.

I never did test it with the electrolytics at the time though, so its unproven but should work. An interesting experiment for someone to try ?

[makenzie71]

I LOVE it when you people talk to me like I'm as smart as y'all are! lol

I think I understand all the components you're assembling there, warpspeed, but why/how it works will elude me until I can get some more education in my head!  Knowing that I can indeed isolate the individual phases more or less answers the question I had and gives me hope that I can indeed power the inverter I'm looking at suitably.  My end goal is to keep the voltage high but current low.  I don't know what configuration the turbine is in and don't think I'll be able to modify it because I'm not 100% certain I'll be "owning" it...I may have to give it back after I figure out if it'll work the way I hope it will.

[Warpspeed]

Your situation is a bit unique in that you are trying to split the power three ways. You can certainly do that, but as has already been pointed out, the output of the rectifier from one winding only is going to pulse fairly dramatically without a big bunch of filter electrolytics.  Your mppt circuit may become very confused and  unhappy, but its probably worth a try.

The main objective of trying to achieve higher dc voltage for most people, is the reduced cut in speed, which is a very different matter.

[makenzie71]

I "think" it'll work but that's based on my experience tying rectified turbines to microinverters.  It's worked there, assuming I could keep it in the voltage range...I'd like to think that these bigger, fancier inverters can deal with it.  I think the idea of an MPPT tracker being functional is out the window but I bet it'll still handle the basic function.

[Adriaan Kragten]

The advantage of a separate 1-phase bridge rectifier for each phase is that the total DC voltage is a factor 1.155 higher than for star rectification if the voltage drop over the rectifier diodes is neglected. However, for star rectification the current flows always through two diodes and for three 1-phase bridge rectifiers, the current flows always through six diodes. So the voltage drop over the diodes is a factor three higher and about 4.2 V in stead of about 1.4 V if silicion diodes are used.

One configuration not yet mentioned, is to separate each phase (six wires total) feed each phase into its own bridge rectifier, then connect the whole thing in series. That gives around double the voltage produced in each individual phase, but with an extremely low final ripple voltage and current.  In fact you don't really need an output capacitor.

The final dc  output waveform looks like what you expect to see from a six diode bridge, but at a voltage about twice as high.  Diode losses will be a lot higher though, but for higher voltage systems it may still have some application for a wind turbine.

The derivation of the formulas for the DC voltage is given in KD 340 for both star rectification and for three 1-phase bridge rectifiers connected in series. The peak DC voltage for three 1-phase bridge rectifiers is a factor 2 / square root of 3 = 1.155 higher than for star rectification. But the ripple on the voltage and the current is the same. The ripple is only reduced if more than three fases are used. A much smaller ripple is gained for 9-phase rectification (see report KD 712).

[joestue]

the ripple from a 3 phase diode block and turbine is pretty minimal and i don't think it would affect an mppt algorithm.

i have no idea why you would want to rectify each phase separately and sent them to a separate input on a solar controller intended for multiple strings of panels. just run the turbine into all three inputs in parallel.

[makenzie71]

i have no idea why you would want to rectify each phase separately and sent them to a separate input on a solar controller intended for multiple strings of panels. just run the turbine into all three inputs in parallel.

I'm not 100% sure why I'd want to, either...but I'd like to know if I could and to play with all of the potential options to see how they all perform.  Typical string converters are a dime a dozen compared to inverters specified for wind turbines...I haven't found anyone who has really tried doing it this way and I'm not certain what all the outcomes would be.  I did think it might be a little easier on the inverter if I pull from each phase, but I'm not sure if it'll make any difference.

[MattM]

the ripple from a 3 phase diode block and turbine is pretty minimal and i don't think it would affect an mppt algorithm.

i have no idea why you would want to rectify each phase separately and sent them to a separate input on a solar controller intended for multiple strings of panels. just run the turbine into all three inputs in parallel.

Through a battery first to clamp voltage?

[joestue]

That would defeat the mppt algorithm.


Use another diode block and a dumpload.

[mab]

i have no idea why you would want to rectify each phase separately and sent them to a separate input on a solar controller intended for multiple strings of panels. just run the turbine into all three inputs in parallel.

I'm not 100% sure why I'd want to, either...but I'd like to know if I could and to play with all of the potential options to see how they all perform.  Typical string converters are a dime a dozen compared to inverters specified for wind turbines...I haven't found anyone who has really tried doing it this way and I'm not certain what all the outcomes would be.  I did think it might be a little easier on the inverter if I pull from each phase, but I'm not sure if it'll make any difference.

If they are three independent MPPT inputs then connecting them in parallel to a common power source will cause problems as each MPPT will try and maximise it's output at the expense of the other two. If however the inverter has a setting to allow the three inputs to work together as one, then that would be fine.

[Mary B]

There were some experimenting with individually rectified coils being used in light wind situations by switching and combining coils to get a decent DC output while having less mechanical force needed to get there, low winds being a lot less available mechanical force to get the turbine turning...

The experiment was abandoned due to switching complexities, poor reliability of all those relays and way to much effort for minimal return...

[makenzie71]

If they are three independent MPPT inputs then connecting them in parallel to a common power source will cause problems as each MPPT will try and maximise it's output at the expense of the other two. If however the inverter has a setting to allow the three inputs to work together as one, then that would be fine.

I don't believe they would have a setting to let them all work together from one source, but can't say for certain.  I've reached out to SMA and Enphase and a few others about the possibilities of using wind turbines with their inverters and have universally been told, more of less, "we're not going to assist you in any way with this project".

I feel like pulling from the individual phases will allow the MPPT functions to work, but electricity tends to flow like water...if I hook all the phases to all the inputs one will always be trying to draw more than the others...I think...

There were some experimenting with individually rectified coils being used in light wind situations by switching and combining coils to get a decent DC output while having less mechanical force needed to get there, low winds being a lot less available mechanical force to get the turbine turning...

The experiment was abandoned due to switching complexities, poor reliability of all those relays and way to much effort for minimal return...

It'd have to be a very high output turbine to make it worth that effort :/

[Warpspeed]

The experiment was abandoned due to switching complexities, poor reliability of all those relays and way to much effort for minimal return...

Those are problems of crappy amateurish construction, not a limitation of the concept.

[Mary B]

The experiment was abandoned due to switching complexities, poor reliability of all those relays and way to much effort for minimal return...

Those are problems of crappy amateurish construction, not a limitation of the concept.

These were top notch builders who contributed a lot of what we now consider the standard today for a home built turbine. Relays wear out, contact arcing, weather exposure, even in a water tight container the humidity builds up with temp changes, trust me, I know! I run mast mounted preamps for my ham radio gear. I build preamp boxes so they can drain for a reason! Even with my skills at building electronics for outdoor use(been doing it since late 80's...) I lose a preamp now and then when humidity manages to flood the inside of a housing when insects plug drain holes. Or supposedly sealed with dry nitrogen relays that aren't sealed and the contacts fail early...

[Warpspeed]

Yes you are quite right Mary.
Relay technology has all of those limitations which is why its arguably not the best solution for that application, especially in a very harsh environment.
But today we have alternative ways to switch power that do not require mechanical moving parts.

Anything you can do with a relay can now be done better with solid state, only down side of that is susceptibility to EMP.

At one time I was repairing electronic instrumentation used in the deep mining industry.  They are fanatical about making things flame proof against igniting an underground explosion, and totally waterproof so equipment would still operate in a completely flooded mine.  It really opened my eyes to the whole harsh environment problems and solutions.

Potting things in a soft slimy jelly seems to work. The goo conducts heat, totally keeps out water (and bugs) and can be removed for servicing, although its a rather messy job.  I have no idea what that goo actually was, my employer supplied it as required. But it would be ideal for something like a mast head preamp.

One of Murphy's laws goes something like "anything made completely waterproof only holds the water in".

[JW]

Quote from: Warpspeed
I had all kinds of problems with exploding mosfets, until I realised what was actually happening, and the cause of my problems were absolutely fascinating.

Warpspeed

Bite, no.
But if you are going to quote one sentence from a completely different thread, completely out of context, that only confuses other readers and is not helpful.

Anything you can do with a relay can now be done solid state, only down side of that is susceptibility to EMP.

I'm not sure of the switching speed on this test The coil handles a high load and its a water cooled wire. Im not going to go into more

https://www.youtube.com/watch?v=T2TJQSOHgx8

Oh ya the injector is variable lift. I have been granted 2 patents on this feature CIP

[Warpspeed]

O/k fine.
The exploding mosfet experience was on the other thread, in which the original poster was asking about the possibility of coupling together two separate battery banks of different voltage (24v and 48v).

That situation can sometimes arise when people decide to do a significant system upgrade.

On the surface, it sounds like an impossible problem,  but quite by accident while working on something rather different, I came up with a very simple way that this CAN be done, with the right approach.
Its not wild pie in the sky stuff either, switching power supply design for commercial products used to be part of my day job.

I may be a newborn here on this Forum, but I am also a retired power electronics design engineer.  I do not know it all, and never will, but I may be able to pass on some of the ideas that I have from time to time for the benefit of other Forum members.

[Mary B]

Yes you are quite right Mary.
Relay technology has all of those limitations which is why its arguably not the best solution for that application, especially in a very harsh environment.
But today we have alternative ways to switch power that do not require mechanical moving parts.

Anything you can do with a relay can now be done better with solid state, only down side of that is susceptibility to EMP.

At one time I was repairing electronic instrumentation used in the deep mining industry.  They are fanatical about making things flame proof against igniting an underground explosion, and totally waterproof so equipment would still operate in a completely flooded mine.  It really opened my eyes to the whole harsh environment problems and solutions.

Potting things in a soft slimy jelly seems to work. The goo conducts heat, totally keeps out water (and bugs) and can be removed for servicing, although its a rather messy job.  I have no idea what that goo actually was, my employer supplied it as required. But it would be ideal for something like a mast head preamp.

One of Murphy's laws goes something like "anything made completely waterproof only holds the water in".

I have removed that goo form inside antenna feed points... pain in the ... hate that stuff. The stuff these guys use is sticky so every gob you scrape out sticks everywhere. Brush an arm over it you have it stuck to your skin for a week...

To much maintenance on a preamp box for my antennas. Relay failures happen... running 500+ watts of RF is hard on relay contacts, lightning hits arc them, preamp transistors blow up... so no goo!