Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
HeadsUp Regular Member Joined: 06/12/2009 Location: AustraliaPosts: 43
Posted: 10:09am 31 Dec 2009
Copy link to clipboard
Print this post
Hi Bob , thanks for sharing
just wondering if someone can enlighten me why the coils are longer than the magnet coverage ?
wouldnt there be output losses this way ?
if there is an article here on coil design just point me to it rather than spending your time answering if that saves you some time
also , are your magnets direct mounted to the steel backing plate or insulated ?
i read a study where eddy current harmonics in Perm-magnet motors was reduced by using a thin epoxy layer between magnet and stator.
also , has anyone experimented with using aluminium heat sink material bonded into the hollow coil center to dissipate heat ?
thanks for sharing your controller circuit by the way , mucho appreciato ( i think thats a real word )
bobshau Newbie Joined: 22/11/2009 Location: United StatesPosts: 27
Posted: 02:08pm 31 Dec 2009
Copy link to clipboard
Print this post
Hi HeadsUp,
You had some questions as follows:
"just wondering if someone can enlighten me why the coils are longer than the magnet coverage ? "
A well designed coil/magnet system is one where most, if not all, of the magnetic flux (lines of force) cut through coil conductors as each magnet N/S pole approaches a coil and then again as it leaves the coil. For this to occur, it is most efficient if the coil is slightly longer than the magnet coverage.
you asked:
"wouldnt there be output losses this way ?"
Yes, the longer the coil, the higher the I squared R losses. That is why most coils are long enough to be cut by the lines of force but no longer than necessairy.
You asked:
"if there is an article here on coil design just point me to it rather than spending your time answering if that saves you some time"
I'm new to this site, but I suspect if you search, you will find what you need.
You asked:
"also , are your magnets direct mounted to the steel backing plate or insulated ?"
The magnets were coated with a non conducting "rust proof" coating at the factory. I then used "superglue" to fasten them to the disc. Then, I inserted the pin through the hole in each magnet and the steel backing plate. I saw no need to electrically insulate the magnets from the backing plate. In fact, the thickness of any non-magnetic material inserted between the backing plate and the magnet increases the air gap and hurts performance.
You said:
"i read a study where eddy current harmonics in Perm-magnet motors was reduced by using a thin epoxy layer between magnet and stator."
The alternating current in the coils will induce some eddy currents in the magnets and plate. This "counter EMF" is a second order effect and may or maynot cause excessive heating or drag/cogging under higher load conditions. Perhaps insulating the magnets from the plate would reduce eddy currents and possibly improve performance. I haven't seen any conclusive data on this. Maybe someone else could enlighten us?
You asked:
"also , has anyone experimented with using aluminium heat sink material bonded into the hollow coil center to dissipate heat ? "
I don't think it is a good idea to put electrically conducting material into a coil. The magnetic flux will induce eddy currents in the aluminum and create much heat.
Thanks for your questions,
BobBobS
Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333
Posted: 03:23pm 31 Dec 2009
Copy link to clipboard
Print this post
Well answered Bob and enjoyed the simple direct Q and A approach .
Pete.Sometimes it just works
HeadsUp Regular Member Joined: 06/12/2009 Location: AustraliaPosts: 43
Posted: 03:43pm 31 Dec 2009
Copy link to clipboard
Print this post
thanks Bob
Eddy currents are a complex science in themselves from what i have read , and may only occur at certain RPM that are not experienced by your turbine design , the flux harmonics / eddy currents can be analysed with software like Opera 3D but it may not exist as an issue at all at your design speed.
i will PM a page link to you on that later , my computer is a bit crashy due to having about 35 web pages open right now.
i love the acetone heat-pipe cooler . love your work , looking forwards to more detail if you get time one day
i see your point regarding casting metallic cooling elements into the center of the coil , just looking for ways to improve cooling , i am sure you will find something better.... ;)
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 06:01pm 31 Dec 2009
Copy link to clipboard
Print this post
Headsup,
I think eddy currents due to flux perturbations will be present in all rpms after cutin. They will be proportional to the current drawn by the batteries, and will have the signatures of all the noise of the rectifiers encapsulated within.
Whilst the back MMF of the stator will never be able to penetrate to the rotor steel, the back MMF (which will be a reflection of the currents flowing) will modulate the flux from the magnets, and effect their respective return paths.
I put the magnitude of the problem into the who cares anyway basket.
However, galvanically isolating the magnets from the steel rotor is a must in marine environments like I experience here. I mounted my magnets similarly to Bob, but the salt air got between the magnets and the steel, and produced a wonderful corrosion mess...... and I have the rotten magnets to prove it too
This time (with no interest in eddy currents in the rotor) I used epoxy paint to coat the steel first. So I whole heartedly support the use of epoxy barriers between the rotor and magnets.... but for corrosion reasons... not eddy current reasons.....
Even after 2kw runs, there is no discernible plate heating, so the problem is acedemic rather than practical.
Eddy currents that count are when you place any conductive material in the stator.... that was bad, and many orders of magnitude worse than harmonic flux interactions with the rotors.
Stator heating is a product of poor design.... and not a weakness in the axfx system. Simply put, get the furling adjusted at the correct step off point for the stator/magnet design you chose.
...............oztulesVillage idiot...or... just another hack out of his depth
Altenergy Newbie Joined: 19/09/2010 Location: South AfricaPosts: 7
Posted: 12:00pm 19 Sep 2010
Copy link to clipboard
Print this post
Hi Bob and others. After reading this forum for some time i have finally joined up. I know it is extremely late after the event, but I am building Bobs analog MPPT for my turbine. One question i have has to do with what some members suggested, to go to single rail Op Amps, eg LM358 or LM324 etc. I ask then how do i get the negative voltage (-2.5V) out from OP2 by adjusting P3. Surely this would only go almost to Gnd and not minus? Or am i missing something? I am trying the LM324 1st. My idea is to stay away from mains power completely.
Has someone tried a small switchmode +/-12V converter off the 48V batteries?
Otherwise all works and i can set it up, am using a 4050 as a buffer into the FETs via 50 Ohms Rs. I still have to couple it to the turbine (500W) home made Hugh Piggot inspired with slight blade mods for very low speed wind in Durban most of the year. I have tried the voltage doubler and at idle i now get 50W instead of 0W !
I have seen over 500W at times..lovely.
Thanks fot a great forum
LeonHappy Solarwinds
imsmooth Senior Member Joined: 07/02/2008 Location: United StatesPosts: 214
Posted: 07:59pm 19 Sep 2010
Copy link to clipboard
Print this post
Bob,
Excellent work. Great presentation as others have said.
Dave (Marcwolf) commented on converting the device to code. I have worked on a 10kw induction heater that tracks resonance. I found that using an ATMEGA328 uP was cleaner and more accurate than just digital circuitry. You won't have to worry about drift as the uP will track exactly as you define it. It also comes with a pwm output.
Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190
Posted: 10:54pm 19 Sep 2010
Copy link to clipboard
Print this post
Hi Altenergy, and welcome to the forum.
With regard to your question about how you get -2.5V out of a 324 or similar, the answer is that -2.5V is measured relative to the "ground" rail, not to the potential of the minus rail of the + and - supply rails shown in the original circuit.
A 324 can be used with a single supply, inasmuch as it can bring its output down to the value of its negative rail, which is often all that you need. At the upper end of the output scale though, it can only bring its output up to about 1.5V less than its positive rail voltage, even though this can be as high as 32V.
This makes the 324 much more useful that a 741, which can't get its output any closer to its negative rail than to its positive rail, and, consequently, is almost invariably used with split supply rails if its output must swing to ground (or below). Additionally, with the 324 and the like, you get 4 op amps in the one package at a very low cost, plus a number of other improvements.
If you want to use a 324 with split supply rails, then its maximum supply voltage becomes + and - 16 Volts (which is, I suspect, something less than the DC value of the supply rails in the original MPPT circuit). With +/- 16V on the supply rails, the 324 output can come down to -16V relative to "ground", and also up to about +14.5V.
It is not clear from your post whether you have followed the power supply part of the original circuit. This involved what was effectively a centre tapped transformer which gave + and - power supplies and a "ground" that was mid way between the supply rail values.
If you want to use a single supply rail with the lower side at 0V potential, then, to get a -2.5V swing, one alternative is to set the "ground" rail in the original circuit at, say about 3V above the 0V rail. This is often easier said than done in most circuits, as there are usually some following components that need ground referenced to 0V.
Alternatively, it is possible to add active circuitry involving a diodes and a capacitor to a single rail power supply to produce a negative supply rail for the 324 relative to ground that would let you swing the 324 output below ground to permit the -2.5V output that you need.
If you post your circuit then maybe it will be possible to suggest some ways in which a single supply rail could be used for the MPPT.
Regards
Don B
Altenergy Newbie Joined: 19/09/2010 Location: South AfricaPosts: 7
Posted: 05:54am 20 Sep 2010
Copy link to clipboard
Print this post
Hi Don, thank you for the reply. I built the cct excactly as the original except that I used a 324 and coupled the Gnd and -12V lines together to form a single supply rail. I did not use the transformer part, but 48V DC dropped to 12V in 2 stages (LM317 and 7812). I built it with enthusiasm until I hooked up the meter, ‘scope and power supply and wanted to adjust the output to -2.5V, and then… oh OH! Hello! Am I having a Senior moment here? (again!) Then I realised that it would not be possible unless I had a negative supply. I like the idea of using diodes and a capacitor, probably 4 diodes??. Would battery negative still be “GND” ? I have this thing about staying away from utility power and their ripp-off so I don’t want to use mains as a power source, even for this minute current drawn here. I could of course run it off the inverter, but that would be like carrying a bottle of Oxygen for air to run with. Then when the inverter is off I cannot run the MPPT for the turbine, soon to be replaced by a bigger unit.
DON I don’t have facilities at home to scan a drawing of the CCT diagram, I will have to work on this one, but it is exactly like the original analog MPPT CCT posted by Bob except of course using a LM324 instead of the 741 and a 0V to 12V single supply. I am still using the LF351 and 555 of course and then a 4050 to drive the FETs via 50 Ohm resistors as suggested in this forum. I would dearly like to run this off the 48V batteries. A power supply sketch would be handy. Another question, if I sketch a diagram on paper, perhaps have it scanned as jpg or pdf. could it then be posted without too much difficulty? Remember the Senior moments!! (But still tinkering)
Briefly my system, 500W Turbine, 2KW Solar (now need MPPT for this too, HELP!) 2 x banks of 4x12V 102Ah deep cycle Silver Calcium batteries in 48V setup and a 7Kw Sinewave inverter. Not spactacular, but getting there. At least when the whole neighbourhood is dark, we run as per normal.
So, thanks once again. Hope i have not left out anything.
Leon
Happy Solarwinds
Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190
Posted: 10:00pm 20 Sep 2010
Copy link to clipboard
Print this post
Hi again Leon
With regard to your first post, I overlooked your comment about using a switchmode converter to give you plus and minus rails to power the electronics. This is certainly possible and, as the electronics have such a small power requirement, you could use a really small converter.
Actually, all that you need is a converter to give you the below ground negative rail. Have a look at small DC-DC regulated converters that give you an output that is isolated from the input. An example is the SPRO1L-09, which has a 24V input and a 5V output. A -5V negative rail should be more than adequate for this circuit. Because the power draw is so small, you could even use a small battery instead. This should last for many months.
Actually, there is no reason why this type of circuit can't be modified to run with only a positive rail and ground, but it would need to be worked through carefully to see if any changes are necessary.
A quick scan of the thread reminds me that there were some comments on Bobshau's original circuit and wiring that might be worth repeating. The first was that, although a 4050 driver and 50 Ohm resistors might square up the Mosfet driver pulses and reduce their losses (and hence the heating), it might also increase the radiated interference - particularly in view of the open wiring that he used and the air cored inductor.
If interference does prove to be a problem with your circuit, you can slow down the switching of the Mosfets by increasing the value of the 50 Ohm gate resistors. Don't be tempted to replicate Bob's wiring methods though, as long wires with coiled terminations are just what a switching circuit does not need - especially for the Mosfet connections.
Another thing that puzzled me is that his original transformer power supply arrangement for the electronics showed 18V (rms value) secondary windings. With minimal load on these, you would expect that the supply rails would have been about plus and minus 25 Volts, unless he had connected the transformer primaries in series to halve this value. Plus and minus 25 Volts is way too high.
The other thing was that it appeared as though the Mosfet gate driver circuit would swing negative with respect to ground when it turned off. This is not helpful to the Mosfet, which should have only as much positive gate voltage as it needs to turn it fully on, and zero volts to turn it off.
A common precaution is to use a zener diode with a voltage rating just above the (12V?) gate voltage connected hard across the Mosfet gate to source terminals. This is intended to protect the gate by swallowing both positive and negative spikes. Also, make sure that your 4050 driver is powered from the positive rail and ground, not from the minus rail.
Good luck with your efforts, and please post your results on the forum.
RegardsDon B
Altenergy Newbie Joined: 19/09/2010 Location: South AfricaPosts: 7
Posted: 05:58am 21 Sep 2010
Copy link to clipboard
Print this post
Thanks Don
Yes the small switchmode converter was the 1st thing to come to my mind, but I like your idea of a battery as I have a few 6V 3Ah gel cells lying around looking for work to do. I realise that they would probably not last very long doing a little more that nothing, but at least I could get going like this until I can build (not buy) a small switching converter. Like you say, the emphasis on “small”. Any idesa Don?? I will have to try and measure the current drawn by Bob’s cct, hope my meter will cope with the overload! As you say I only really need -5V or so, the +V is already there.
For the interest of all who tinkers at 48V and above: Probably like others here on the forum, I have been searching for some time for 48V diagrams for both solar and turbine controllers but can only find 12V and perhaps 24V ccts on the ‘Net. So I have dreamed up the scheme of using a LM317 to drop the 48V (actually up to about 60V) down to about 30V or so and then a LM7812 to get to 12V. This is to overcome the maximum differential or input voltages for these devices. Then I found another interesting regulator called a TL783 at the local emporium. This is a 1.25V to 125V regulator able to source 700mA in a TO220 package. This is GREAT! It works in a similar fashion to the LM317, LM350 LM338 and others, but can handle input up to 125V DC. with all the short circuit protection, thermal shut-down and stuff like that. The 48V guy’s prayers answered. So now it is one step from 48V to 12V, just requires a small heatsink and all’s well. Of course we could use resistor dividers……Don I thought I saw somewhere that Bob had some voltage regulators in there? This would have brought the 18-25V down to 12V.
I have used the Zener trick on the Gate of a FET in the past, so will do it here again, thanks for the reminder. And yes the 4050 is powered off the +12V rail.
Coming from an RF background, when I saw the wiring on Bob’s prototype, I also thought, “My, what lovely antennas” I am sure it also adds a bit of inductance to the cct, but as he says “ If it ain’t broke, don’t fix it.” I mean, how many time have we built something, got it working and then one of the child’s lunchboxes disappears? I had a similar thing happen years ago when building my Amateur Radio antennas and I made up a balun and put it inside (what I found out later to be my wife’s favourite ) plastic soap box. Boy oh boy! They just don’t understand, do they? Buying a new one did not help either. Back to the question of scanning and posting diagrams and photo’s. Any answer for me as I would like to post my final pictures at least and so keep everybody informed of my progress. I am new to this forum thing, so don’t know all the in’s and out’s.
So ever on we go. If anyone else out there has any idea on this or other 48V power supplies relating to alternative energy, please let me/forum know as it is not that easy to find.
Thanks again
Leon
Happy Solarwinds
Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190
Posted: 07:57am 21 Sep 2010
Copy link to clipboard
Print this post
Hi Leon,
This is a very quick partial response to your post.
The only problem with linear regulators like the 7812 and 317 is that they are essentially dropping resistors, and their losses and heating becomes significant if you try to pass too much current through them. While you are only powering your control electronics through them they are probably the best solution, but if you want to do any serious voltage dropping, you need to use switchmode techniques. Have a look at buck converter circuits on the net as they are, in essence, really simple. There are also good driver IC packages available that do just about all of the work for you
When dealing with 48V and up, you need to realize that you are getting near the limit of the typical Mosfet voltage capability. There are higher voltage ones available, but usually with a significant increase in on resistance.
I doubt if your control electronics will draw more than a few 10s of milliamps, so even a gell cell battery is overkill. A small dry cell battery would probably be quite adequate.
Regards
Don B
SparWeb Senior Member Joined: 17/04/2008 Location: CanadaPosts: 196
Posted: 05:41am 02 Oct 2010
Copy link to clipboard
Print this post
Hi Don,
Could you tell me which book you scanned there? It looks useful, I don't recognize it, and I like to "collect" these things.
Thanks!
Steven T. Fahey
SparWeb Senior Member Joined: 17/04/2008 Location: CanadaPosts: 196
Posted: 05:49am 02 Oct 2010
Copy link to clipboard
Print this post
Yes, it isn't harder than that. If you have your own scanner, then the software the controls it will likely ask you for the "jpg" format, or just offer it to you by default. When the file is saved, then are ready to use it in a posting on the forum. While writing a reply in the box, click on the button like this:
A box will pop up and you can go looking for the file to "upload". On some computers, you will see a bar appear first, which is checking that you intentionally want this to happen - to prevent hacking code from running (it's just a security measure). Once you're done browsing for the file, and hit "OK" to upload, it does take a moment to finish.
When the file is uploaded, you will see a new line of text with the location of your image, now copied onto the website server address. You can type before or after that image line.
Good luck.Steven T. Fahey
Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190
Posted: 02:12am 03 Oct 2010
Copy link to clipboard
Print this post
To Sparweb
Hi Steven. The book that I extracted the information from is called Wind Energy for the Eighties. Not sure of the author (my notes show BWEA???), and the publisher was Peter Peregrinus Ltd. The copy of the book that I found was in a university library.
To Altenergy,
Hi Leon. I have had a further look at Bobshau's original circuit, and can see no reason that you would need a below ground rail if you were using a 324 op amp chip. The 555 is a triangle wave generator that should have an output swinging between 4V and 8V if operated off a 12V rail. The fet op amp used as a comparator should have no problems working between these values, and it would certainly be no problem for the 324. To ensure that you get the Mosfet gate hard to ground when you want it off though, you will need a 4050 driver.
Using only a +12V and ground rail only would mean that you would need to revise Bobshau's set up voltages somewhat to get correct operation.
With a switchmode boost circuit like this, it is essential that the Mosfet never stays hard on, as this places a constant short circuit on the wind turbine output. You would therefore need to ensure that the minimum output from the 324 final op amp never goes below the triangle wave generator minimum output voltage. As I believe that you have a 'scope, you should be able to check that the fet op amp comparator driving the Mosfet gate is always switching, and hence never remains fully on, even with no output from the wind turbine.
Good luck with it.
Regards Don B
Altenergy Newbie Joined: 19/09/2010 Location: South AfricaPosts: 7
Posted: 06:14pm 04 Oct 2010
Copy link to clipboard
Print this post
Don, thanks for all the info once again.
Yes I will look at the set-up. Thanks for the tip on using the ‘scope to set up the driver. I have included a 4050 on my board, so driving the FETs on and off should be better now. The FET op amp seems to be doing it’s thing ok and drives the 4050 with no problem.
It is now our rainy season so setting it up by the mast under the turbine is not going to be easy for a while. I will also post pictures of the voltage doubler I built up and is currently running, and as I reported here before, if I flick the switch in low winds, the charging goes from zero amps to 1 amp at 48V, and keeps this one amp advantage and increases slightly till the main turbine voltage is high enough to take over. Surely 50W for nothing. OK, the cost of a few parts.
Steven, Hi and thanks for the tips on scanning and posting etc. you make it sound so easy. I will try it 1st with photos of the units, doubler and then the MPPT unit.
As a matter of interest, for an inductor, I wound a clip together type split core inductor with former until I got the correct value, about 9 turns did the trick for 26uH (close enough to 25uH), keeps size down etc, but as I haven’t checked the specs and I don’t know if it will saturate and cause problems. Maybe then create a small gap with paper in between the cores. I seem to remember ETD34 N41 250 or something to that effect. I can only try, but I have also wound the air core as Bob did to compare.
Does anyone know of a good reason, (or any reason for that matter) why this MPPT cannot be used on 48V Solar panels? I am dying to try this, perhaps with different settings, but if anyone else has tried and either succeeded or failed, let us know. Otherwise if someone’s got some PIC code and diagram for one, that’s fine too.
Great dealing with you Guys.
Regards
Leon
Happy Solarwinds
jean Newbie Joined: 09/06/2013 Location: FrancePosts: 2
Posted: 04:42am 26 Jun 2013
Copy link to clipboard
Print this post
Hi guys.
Maybe a bit late :(
I based my design on a Boost circuit chopped by mosfets and driven by a PIC. All going fine. I have a 48V chinese WM, 500 amps battery and 6 KW inverter feeding the house. I have current flowing to the battery as soon as 20v from the WM.
Today I have tested this device wich allows to set the optimum rpms/amps ratio for the turbine by adjusting(manualy for the moment)the duty-cycle.
Don't blaim me about the PCB-it is just a prototype.
The results are very good, i have to solve the automatic duty-cycle adjustment,may be @jean can help me.
Tomorow I will show the schematics.
domwild Guru Joined: 16/12/2005 Location: AustraliaPosts: 873
Posted: 02:23pm 20 Jul 2014
Copy link to clipboard
Print this post
Perhaps this deserves a separate MPPT page as it deals with a digital solution to MPPT.
Now here comes a tall order: There is much more available on MPPT via analogue means compared to MPPT via PICs. Would it be possible to see a schematic and/or code? I realize this is a fluid project and components/code changes all the time but it might just be possible to send comments/pseudocode alone (no code) to get the gist of things. And an "approximate" schematic.
Something more detailed than
Main
Read parameters, e.g., rpm, voltage into batt, amps, wind speed
Hill climbing algorithm for optimisation, like measure power into batts, increase pulse width, measure power in, increase/decrease pulse width
Goto Main
If only it were that simple!Taxation as a means of achieving prosperity is like a man standing inside a bucket trying to lift himself up.
Winston Churchill
kitestrings Senior Member Joined: 23/04/2014 Location: United StatesPosts: 102
Posted: 06:15am 22 Jul 2014
Copy link to clipboard
Print this post
I just wondered if anyone had heard anything recent from bobshau? I'd followed his design with interest described here and on OP, but his last post appears to be ~January '10. I thought he had some very creative ideas.
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
