 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. |
Forum Index : Windmills : visual effect of capacitors
  Page 12 of 25   |
|||||
| Author | Message | ||||
Gill Senior Member  Joined: 11/11/2006 Location: AustraliaPosts: 669 |
Glenn, I've done some initial 12v tests. From Brian's recommendation of 1,000 to 15,000 uf, I tried 1,000uf 100V 2 amp ripple, back-to-back on a series 80 in a 2 coils in series, 3 phase, Star, times 7 in parallel. I found performance down. Then tried 470uf 63v 105degC low ESR TKR. Very slight improvement at the lower revs but they smoked up. Then tried 470uf 200v 85degC (ex Computer psu). Again very slight improvement before blowing at higher revs. I conclude the cap volts capacity is high enough but the ripple of, I believe around 2 amps on the 200v caps, is insufficient. I do not know how Gordon's caps of the same value are passing the current as applicable to the max power he claims. My budget does not run to purchasing the quantity of caps necessary to give me a range of values with a sufficient amp ripple rating. My cap tests have been suspended. I guess every bit of feedback helps be it + or -. was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
robbo Regular Member  Joined: 25/03/2007 Location: AustraliaPosts: 71 |
Glenn, Were these tests loaded, or unloaded ? Wouldn't the Batts keeps the voltage down to round about 12 volts? Wouldnt this mean you would then be able to use the caps smaller than, say, 63volts if a 12v, 24v, or 48volts system ? "the Earth was not given to us, by our fathers, rather, it is lent to us by our children". |
||||
GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Gill, I have just braved the cold and taken some DSO readings of typical voltage across the caps. There will only be voltage if current is passed by the cap. 
The output frequency was 6.4mS, approx 330rpm, provided approx 150W to the load with 10VAC ripple on the caps. On my original test set of caps, 200uF 35VDC, these did heat up at high power and high frequency. These were very small and were only used to prove the doubler concept on the series80 2s7p rewire. I now use 180WV, or 400V caps. I have not specifically purchased any caps to try yet. I measured a good sine wave across the caps, so the back to back arrangement of caps seem to handle power without visible cross over distortions as the caps change function. Is it possible to check the waveform across the caps when they blow? Is it possible the caps are wired differently? My caps have been punished without problems. I have provided photos of the caps in my system. What system have you tested on? What power levels are you testing at? Would this be a hydro system or a windmill? Hi Robbo, All of the capacitor testing I have published results for is with a load connected. You really cannot provide power measured to a load, without a load connected. The frequency is what would cause the most problems with capacitors. The voltage rating should be based on the system voltage and the max rpm expected. These are related to power transfer rates. My mill has an upper rpm of 450. If a mill has higher rpm, then higher voltage caps will be needed. Gordon. become more energy aware |
||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
I'm ordering 12 100uF 100v NP axial caps. That should give me enough to play around over the Christmas break, hope they get here in time. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Gill Senior Member Joined: 11/11/2006 Location: AustraliaPosts: 669 |
Robbo, In answer to your questions: Q1. Tests were loaded by 12v battery. Q2. Yes, indeed that is so. Q3. Here is the key point I mentioned neither you or Gordon has picked Up on. I have no issue with the cap voltage. I have an issue with the cap ripple current rating. I have noticed the current ripple rating increases on caps with a higher voltage rating even though the value of capacitance is the same. I had to use higher voltage caps not for their higher voltage but for their higher ripple current carrying capacity. Perhaps re-reading my post will have you focusing on current issues instead of voltage? For further information, the 470uf 100v caps blew at 213Watts and the 470uf 200V @ 330Watts. In these cases the currents were 15.3 and 22.3Amps respectively. As the current is carried by 3 sets of back-to-back caps, could we not say the caps of each phase of a 3 phase gen is passing 1/3 the total current? This would be 5 Amps in the 63Volt caps when they blew and 7.4Amps when the 100Volt caps blew. Not surprising when the typical current ripple rating is 775mA for 470uf 63V and under 1.99Amps on a 470uf 400v caps. You see why I point out and query what I am not able to duplicate from Gordon's claims of 500Watts ++. To a 24Volt battery his Voltage may have been double mine, so that's around 30Volts. Therefor a current of 16.5Amps or a phase current of 5.5Amps. Passing 5.5Amps through caps with ripple current capacity of < 2Amps is suspect to me. was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Gill commented: You see why I point out and query what I am not able to duplicate from Gordon's claims of 500Watts ++. Passing 5.5Amps through caps with ripple current capacity of < 2Amps is suspect to me. Hi readers, I was able to duplicate Bryans measurements, as was discussed quite a while ago now. 20+ amps into 24V battery. I have progressed further. I have no current ripple data for the caps I am using. I have been testing on a windmill, as has Bryan. I suspect that Gill is testing on a hydro or other system. I am unable to comment on the rpm of his setup during testing either. I never see constant output power from my mill. I would equate wind energy where I live to music power, with the effects of gusting at high wind speeds. This is not like an inverter supplying a constant load, or the constant power produced by a hydro system. The system installed on Boltys windmill with a doubler, contributes power to the bottom of the wind energy curve. The 12 caps only need to handle approx 150W maximum at much lower frequency to a F&P. I believe that oztules has pointed out that the capacitors seem to make the blades work better in energy harvest from the wind, by improving wind loadings by the alternator/load and not necessarily alternator efficiency. I will continue to monitor the caps on my mill and Boltys mill. My mill produced 2kWhr in a 8hr period yesterday, so this averages to approx 250W. There were peaks of only 700W during this time. This was spread over 24caps in my setup. This means approx 10W average transferred per cap, not 100W. This is at 24V and 48V, so this means approx 0.5A average per cap, probably closer to 0.3A as there are the 2 simultanous loading systems. I have shown pictures and described the layout previously. The higher power levels are handled at 48V, with 400V caps. If criticism is to be made, then I suggest readers should read the postings first, noting p17,p18 and possibly others. I have provided enough info, suitable for a windmill loading, and shown pics of my full windmill loading system. Bryan has recorded more than 500W power lavels on a windmill. I cannot speculate on suitability for continuous operation at a maximum power. Gill may need to find a different solution for his system. I have recommended a first place to start would be a doubler, to increase power at lower wind levels, to lower windmill cutin, in conjunction with normal rectifiers for the higher power levels, as was installed on Boltys chinese mill. This would be suitable for a windmill that was spinning but not providing current until higher wind levels. This would not be a stall limited windmill system, or other constant loaded system. I suspect that Gill has used 6 capacitors in the back to back series configuration of a similar size, or physically smaller than the smallest size I used and is expecting the full power to be passed to the load. I have used many more caps and distributed the work done across all of them. I expect that if I tested a similar arrangement to Gill, that the caps may fail as well. I suspect most readers would benefit first with a windmill, by allowing the mill to produce current in windspeeds where normally no current would be obtained. I first replicated the results Bryan obtained, and then investigated lowering cutin and increasing low windspeeds power. In my normal testing, I start at low power levels and work up. I guess I test in a different way to Gill. Hi Gizmo, I think you get what I am suggesting, make the doubler first and test in conjunction with existing rectifiers, to check a lowering of windmill cutin rpm. Add series caps to the main rectifiers, with provision to bypass if needed. This is how I started[pic on p17]. This will increase maximum power by improving the load matching at upper power levels. This should be a start. Gordon. become more energy aware |
||||
Bolty Regular Member Joined: 03/04/2008 Location: AustraliaPosts: 81 |
My tower is currently down, as I am increasing the size of the blades to increase the average power generated. Whilst it is down I am also increasing the area of the tail, as I am attempting to improve it's into the wind facing. (In another thread, there has been some discussion about my turbine often swinging out of the wind, and continuing to generate whilst backwards...with the blade apparently spinning in the wrong direction) Whilst working on the blade mounts, I noticed that I was able to hand spin the alternator (without blades)and cause the Latronics TC48 controller to activate the brake circuit. This means that my doubler circuit was generating at least 75 volts (the brake activation voltage) Of course the inverter connected to the output of the controller was not loading the system, as it had not yet grid synchronized. I found this totally AMAZING that hand spinning the shaft could produce such a high voltage output. Clearly this is entirely due to the voltage doubler. For me, there are only small top end benefits from capacitors. This is because my machine is much lower frequency than F&P alternators. As explained earlier I would need to have massive capacitors to allow top end benefits. However capacitors allow me to use the voltage doubler circuit to very great effect at the low end. As mentioned before, my wind location is not as good as Gordon's. However in winds around 3 m/s I am now generating small but significant power in the order of 50 to 100 watts much more of the time, when previously it would have been none! Over a 24 hour period of light winds, this equates to well over 1 extra kWh. Whilst this may not be much, it is quite significant over a year. Another great benefit of this is that my Latronics inverter remains synchronized for a much higher proportion of the time, therefore being ready to collect small bundles of energy in light conditions. Prior to the use of capacitors and the doubler, the inverter would lose input for too long, and would require re- synchronization with the grid. This takes at least 2 minutes, during which time no power is able to be passed to the grid. I am looking forward to getting my machine up again later today, and seeing how it works with my bigger diameter! |
||||
Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190 |
In response to Gill's post of 19th December, I believe that the heating experienced with capacitors is as a consequence of the ripple current acting on their Equivalent Series Resistance (ESR). ESR is a recognised shortcoming with electrolytic capacitors in particular when dealing with high ripple currents, and it is possible to buy low ESR capacitors at a slight premium to the cost of conventional capacitors. Anyone with access to a Jaycar catalogue will see an explanatory box on the topic, plus the range of low ESR and standard capacitors that they sell. The Jaycar listing also gives the RMS (ie steady state)ripple current rating of their range of electros. It is interesting to note that you need to go to 1,000 MF, 35 V low ESR caps to get 2.5A ripple rating. They do not stock a 63V 1000MF cap, but their 470 MF low ESR cap has a 1.7A ripple rating. These are listed at $1.28 each. By way of comparison, their standard 63V 470MF capacitor has a ripple rating of 0.9A, or about half that of the low ESR one, and their 450V 100MF capacitor (at $4.70 ea) has a ripple rating of only 0.5A. They also sell 4,000MF 75V 4.6A ripple chassis mounting electros at $10.90 each. Getting the capacitance, voltage, and ripple current rating required is clearly somewhat of an economic balancing act. No doubt major trade stockists such as RS Components would have a greater range of capacitors available. Don B |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Don, the ripple current rating and the temperature rating and the voltage rating usually affect physical sizing. I have some 1000uF caps of the same size. One is rated at 1000uF 85degree 160VDC. The other is rated 1000uF 105degree 100VDC. Normally, the voltage rating would increase the physical sizing. The 230uF caps that Dennis started testing with, were bags of 11uF-20uF 240VAC Mains rated, parallel combo's, of all sorts. These were a bit unwieldly, so I have reduced the size somewhat. The caps still neeed to handle power, so need to be sized accordingly. The capacitance is the important value in better windmill loading matching. Gordon. become more energy aware |
||||
| Gill Senior Member Joined: 11/11/2006 Location: AustraliaPosts: 669 |
As we search for the cap rating for our 12v mills, the amp issue becomes more vital. My 12v windmill generates 18amps and is still 1/2 furled. My testing at 22Amps is not excessive and I require they not pop caps here where I would hope caps might increase output even beyond this. The testing device is irrelevant. Caps must withstand all windmill amperages, plus some for reliability. To increase amp rating we will need to parallel caps. Parallelling caps increases capacitance accordingly, so even more caps need then be serised to counter this. Now we have an awful lot of caps. Far more than I can scrounge. If I buy them, have I chosen the best value? I must really lash out and buy a mass of caps in several values. And looking at current ripple rating as my guide and not physical size I can confirm Don's comments. On doing the figures, this is beyond my budget. Good luck to any who play here. We needed to test and establish a value to recommend for retro fitting to those many 12v mills already made and those still to come. Talking higher voltage mills just doesn't cut it. was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
Bryan1 Guru  Joined: 22/02/2006 Location: AustraliaPosts: 1344 |
Gill, As said before when I initally did the serial cap test I used 400volt 470uf caps. I've given Gordon quite a few different caps to play with just so he could replicate my results. Using low voltage caps will warm them to a point of blowing so working with a 12 volt system will require a decent voltage rating on the caps. My array is 24 volts and it's been ages since I put them on and they haven't ever heated up. I only have 1 set of 330uf caps as all the others have gone to guys for their own testing. But going on a 12 volt array I would suggest a 400 volt cap rating to be the least you should use. Cheers Bryan  |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi readers, Here is a pic of the variety of caps that Dennis put together for his original testing. 
There are all types to make up the capacitance. These can be picked up from old fluoro's, motors, etc. Bryan simplified to 6 x 400V units. I was not so enthusiastic, and expanded to 24. I have no plans to test for 12V systems. I see no difference with sizing capacitors to resistors, or inductors, or mosfets etc, for power rating. Gordon. become more energy aware |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
I guess this answers my questions re your test unit. I doubt caps would offer many befefits, with a controlled energy source. Efficiency would be most important. I doubt Dennis payed much for his cap setup. I certainly have been testing on components I had, or have been loaned, or given to test with. With windmill loading matching, the important aspect is the capacitance. The degree of loading offered by a value selected, can be calculated given the system voltage and the required operation rpm range. The power handling requirement can be estimated based on selecting the highest voltage rating available. If economics is critical to justify this type of mod, I doubt there would be sufficient funds to upgrade the regulation system and possibly battery storage as well. Gordon. become more energy aware |
||||
| Gill Senior Member Joined: 11/11/2006 Location: AustraliaPosts: 669 |
Gordon, What I hear you saying is you don't have a solution on how to find which caps best suit a 12 volt F&P. You know, specific parts to buy for the basic back-to-back arrangement that won't blow up on the original backshed windmill. Ok. That's fine, I don't know either. Let's mark this down as a yet to be resolved standard cap value? was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Gill, Capacitors are usually designed for a particular purpose, nominally energy storage. In the case of electolytic capacitors, this would be nominally DC, up to a maximum voltage rating. In the application we have, additional margins need to be used. The inherant behaviour of the capacitor to increasing frequency affects these margins. So capacitors we can purchase are rated for DC voltage with ratings for ripple current and temperature and maximum voltage, or AC voltage, predominantly 50Hz frequency with temperature and voltage maximum ratings. An AC cap is rated for connection across the source of AC current, at usually the 50Hz. We cannot normally purchase a capacitor with a rating for the 200-300Hz or more operation that can occur with a F&P as a windmill. We have to increase the one rating normally given on the device, the voltage. If we used AC rated caps, then this would be roughly 4-6x the normal 50Hz frequency, times the maximum expected AC voltage expected from the windmill. This dictates 200-300VAC ratings for my 24V system, allowing for unloaded conditions. Using DC rated caps in AC shoulod only require more caps only, but the peak voltage in AC, dictates at least a 1.4x voltage rating above that calculated from the frequency and windmill system unloaded voltage. Additional margins should be used as the chemistry inside electrolytics causes additional heating when the voltage is cycled. The can of a cap can only dissipate a certain amount of power. The best components to use if you buy them are AC rated caps, compensating for frequency, if the ratings are lower than the max operating Hz. The problem you seem to be seeing is related to increasing the capacitance, just by using a single bigger capacitor. For the 12V system, I would just parallel smaller caps of the sizes I have used. I would keep the voltage the same and just use more for the higher system currents. A 12V system may just need more [twice the number] of the caps that Bryan and I have used at 24V. Remember that this was for a nominal 280W power, increased to 600W. Gordon. become more energy aware |
||||
| Gill Senior Member Joined: 11/11/2006 Location: AustraliaPosts: 669 |
Gordon, "A 12V system may just need more" May seem a very defining statement for those who haven't a clue but is just a waste of my time. Thank you. was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
There is not much difference sizing capacitors to sizing resistors as to power dissipation. The AC passed by the capacitors causes heating. A certain amount of surface area and volume is required to dissipate the energy. double the current requires double the surface area to disssipate the energy. How readers do this while maintaining the required capacitance is up to them. I am not persuing the full AC coupling on this thread anymore. This is not going to help most people with a windmill. Full AC coupling would be the last phase in the process of windmill loading matching to the available wind energy. The doubler arrrangement will help most readers. The process of fully characterising the complex interactions if easy, would have been done already in an academic environment. I feel there is enough information for resourceful readers to make a good start towards improving their windmills output in low winds. It is unfortunate that Gill was unable to determine the component sizing and chose to undertake the most difficult task before seeing firsthand the way capacitors change the loading. Gill seems to think that this forum is a place to scientifically calculate a system that will work on his and everyone elses windmill, or for that matter any other alternator/loading system. I cannot accept his criticism as useful. There are already 3 known windmills benefitting from the series capacitor arrangements, probably many more. I can accept criticism if I have made a mistake, however I do not accept that it is my role to calculate and test and then provide the step by step instructions to make a particular arrangement that will work for all possible voltages and power levels on all types of diverse systems. As far as I am concerned, unless other readers shown an interest with some feedback, I am just wasting my time providing any further discussion. Gordon. PS: There is no standard cap value, just like there is no standard windmill blade. become more energy aware |
||||
| Bolty Regular Member Joined: 03/04/2008 Location: AustraliaPosts: 81 |
Hi Gordon! As Glenn has reminded us on another thread, it is the time of the year for all people to show good will to each other! It is also the time for festivities! I would like to take time and reflect and celebrate the great development that this thread has made in the understanding of how capacitors work with windmills. When you initiated this thread you called it "Visual effect of capacitors". I doubt that anyone could have possibly contemplated the massive amount of information and opinions that have been presented here! I believe that this thread has been fantastically successful in developing understanding of how capacitors enhance the performance of windmills. This thread has produced ideas and information that has been creative and unique. Where else in the world is there any information on this topic! Along with this success there is a significant difficulty. It is the huge amount of information and opinion in this thread that needs to be interpreted by a new person willing to try capacitors. This thread now only serves to fine tune ideas. Whilst that is valuable and needs to continue (there is always something left to learn) a new thread or sticky needs to emerge. This will allow newly interested people in this topic, to more readily and quickly "get up to pace". In recognition of the efforts of those primarily involved in developing this thread, it would be prudent for one or more of these contributors to initiate a summary of capacitors as a sticky. Merry Christmas everyone! Bolty |
||||
| Don B Senior Member Joined: 27/09/2008 Location: AustraliaPosts: 190 |
Series Capacitors for 12V Mills The standard three phase F & P motor has 14 poles (or coils) per phase, all connected in series, and the three phases are connected in star. If you call this a nominal 240V connection at the phase terminals, then with, say, a nominal current of 1A per phase, the nominal motor power would be 3 X 240 X 1 = 720W. For the same coil current, and hence total power, the effect of different winding connections is: For a standard motor (14 coils per phase)reconnected in delta, the nominal terminal voltage becomes 138V, with a phase current of 1.7A. I believe that this is the connection that Gordon is using with his rig. For an F&P motor rewired with 7 paralleled sets of 2 coils per phase connected in star, the nominal terminal voltage becomes 34.3V, with a phase current of 7A per phase. I believe that this is the usual 12V LV connection. For an F&P motor rewired with 7 paralleled sets of 2 coils connected in delta, the nominal terminal voltage becomes 19.8V, while the phase current becomes 12 A. These voltages and currents are purely nominal, but do show the relative terminal voltages and currents for the same output power and frequency with different internal winding connections. The problem for using series capacitors with the low voltage F&P connections is that you are asking the series capacitors to pass higher currents with a lower voltage across them. The formula for the AC current passing through a capacitor is current = 2 X pi X frequency X capacitance (Farads) X (RMS) voltage. As the frequency for any given shaft speed will be the same, if a 330MF capacitor works with Gordon's (delta) connection method, then, for a comparable 7 X 2 LV star connected mill, you would need a capacitor that is 138/34.5 times larger (ie 4 X larger) to account for the lower voltage, and a further 7.0/1.7 times larger (4 X again) to account for the larger current needed for comparable power to flow. The equivalent capacitor therefore needs to be 16 times larger, or about 5,280MF per phase. The factor of 16 is also about what you need to adjust for the lowered impedance of the windings when connected in 7 X 2 star. Note, however, that this does not take into account that Gordon's rig has an effective 48V out. If you are chasing the same effect at 12V only, then you probably need significantly larger (4 X ?)capacitors again. There is also the old bugbear of the losses being equal to the square of the current flowing. I think that you are getting well and truly down the law of diminishing return curve to expect any benefit from series caps with a 12V rig. Don B |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
I can now understand why a 12v system using caps is a bad idea. OK, how about approaching the problem from a different direction. A 100 series in delta for 24v with caps is a proven configuation. So how about configuring a cap voltage divider? I'm sure there are ways to half the AC voltage / double the current using caps alone, could that method be used with our windmills. Use a 100 series in delta with caps, but half the output voltage in the process so it can charge a 12v system. Man I really need to dig up my old Telecom AC theory training manuals! Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Page 12 of 25 |
|||||
 Print this page |
