 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 : F&P@PE
  Page 3 of 12   |
|||||
| Author | Message | ||||
GWatPE Senior Member  Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
The carbon fibre was a flop. Current was transferred from the coils to the carbon, and effectively created shorted turns. I decided to do some testing at 100, 200 & 300rpm, of some different configurations. The VFDrive was set to torque mode. increasing torque with increasing drive frequency. I tested no rotor, to see a baseline for the motor and bearings. Up to 86W was needed to spin at 300rpm, and the response was quite linear with increasing rpm as expected. I added a 56 magnet rotor. and std 80series stator. this config required 168W @ 300rpm. The difference was 80W. I swapped for a 48 magnet rotor. This typical 7phase arrangement required 189W @ 300rpm. The difference was 103W. This is not too much a penalty for the reduced hum and cogging. I substituted a twisted pole stator and there was a 3W increase at 300rpm over the std stator unloaded with the 56 magnet rotor. This is probably a better/cheaper/simpler option than to go 7phase. I finally tested a with and without caps loaded into a 24V Nom battery. I could find no significant difference in performance up to the 70W max test power level. The without caps produced 0.1A more in 2.6A Approx 4% more current at the max test power level. I was unable to test to any higher power levels as the test rig consumed 650W to produce 72W output. Not very efficient use of energy. My observations on the mill with capacitors at higher power levels indicate that as the alternator output approaches double the load voltage, that additional current is obtained. The power level on my mill where the double voltage occurs is approx 130W. The test rig is unable to step up to the required power levels. There is no indication at low power levels that caps would improve power output at higher power levels, but they do. I am still looking for a better test rig, mechanical gearing may still be the best way. Gordon. become more energy aware |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Interesting. And should be a cause for some concern to those who use F&Ps... 72/650 = 11% conversion efficiency of the total chain, from electrical to mechanical to electrical energy again. Have you (or anyone else) actually measured the conversion efficiency of a F&P generator, using a De Prony brake to measure mechanical input power ? I think this would be a *very* interesting experiment to perform. Gordon, I assume you used a 3-phase motor in your experiment (with VFD; though there exist single-phase VFDs). Assuming the motor runs at 80% efficiency and the VFD at 95%, that means the 650W electrical input gets converted to 500W mechanical power. The F&P converts this 500W mechanical power into 72W electrical power... that indicates a conversion efficiency of about 14%. Which is low. Very low. And seems to confirm Flux's opinion of the uselessness of the F&P as windgenerators. Again, the above is under some assumptions I've made. It would be very interesting if someone could set up an experiment with a De Prony brake (as I've done here for my 500W motorconversion: http://www.thebackshed.com/Windmill/FORUM1/forum_posts.asp?T ID=1246&PN=4) for the F&P to give some hard, measured results of the F&P's efficiency as a generator. For comparison, that motorconversion yielded efficiencies from 80-50% over its operating range. Efficiency of an axial flux would be even higher. If the F&P has efficiencies around 15% (as your experiment seems to show) then that would be a big cause for concern for me, if I were using F&Ps. Peter. |
||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Nah something not right there. At only 14% efficient the F&P would make a good room heater! There is an article somewhere here written by a Dr Tom Chalko. The Doc did a bunch of testing on the F&P, and used a pony break from memory. His measured efficiencies were over 80% for the F&P, and I think he managed to get it over 95% with some fiddling. See if I can find the article. Flux, as clever as he is, has never seen a F&P in real life and "has no interrest in them", so I wouldnt take his comments too seriously. There are hundreds, if not more, F&P based windmills out there. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Found it, was uploaded by Bryan some time ago. Its a zipped up PDF. Link to Highlanders file Tom Chalko managed 97% efficiency from a F&P at 100 watts. The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
oztules Guru  Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Glen, I'm not sure I would rely on the good Docs report with respect to F&P and real world battery charging with a mill. In fact the relationship to mill/battery charging and his work (half a stator into a resistor @700+rpm) is fanciful. He used a stator with half the fingers cut out, ran it at over 700 rpm to deliver only 100 watts........ and all that into a resistive load of some 7-8 ohms. My crummy chainsaw mill put out over 5kw into a 4 ohm load and could have done it all day from a heat perspective, (stator only dissipates some 600W driving a 4 ohm load @5000W.....) but into batteries it would be a flaming ball of fire mill, which would have burnt up in minutes. The stator would have to dissapate at least 5kw in the coils to deliver that kind of power into batteries. As I recall, no-one else has emulated the good doctors 3uf experiments where he gets 200% power increase..... because no-one else runs at 700rpm+ into a relatively high impedance load..... real people seem to need considerably higher values due to the lower rpm, and lower impedances incurred. He didn't use a pony brake or any other quantifiable means to justify his power measurements. He seems to have extrapolated from "similar motors" power curves....A big ask really, and well.... I don't think he got to the truth of how much power was actually delivered to the F&P. He claims 1% inaccuracy.... whilst measuring 5% deviation in the belts efficiency alone......I remain unconvinced.. That all said, the F&P seems to be capable of very useful performance as Gordon has shown, and I will await Gordons final power measurements .... rather than relying on the good doctors efforts. I think Gordons results will better reflect the F&P under normal windmill application speeds, and battery charging conditions. Those two pre-requisites alter the whole equation from what the good Doc was playing with, versus what F&P windmillers are experiencing. I expect when all is done, that the F&P will not be too shabby when the numbers come in. It will be a little less than the Axials... but even then....only if they match their load too. Thats what it is mostly all about. I suspect that Dinges first order assumptions on the efficiency of the motor and VFD will be wrong because of the mismatch between the motor torque/speed and the F&P power curve. I suspect it was laboring, and was not within the normal operating curve of the motor or the VFD ... like driving at 10kph up a steep hill in top gear. .........oztules Village idiot...or... just another hack out of his depth |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi dinges, The point of this exercise was to compare the relative power lost in the stator for twisted/std poles and the rotor for 7phase conversions. It was good to get some comments re the efficiency. In these examples the efficiency is of no concern. The tests were identical except for a single change of component. The question of efficiency and bagging of the F&P as a generator gets a bit tiring. In an ideal world I would not consider a motor conversion either. I made my axial flux motor a while before flux and others promoted them as an alternative generator. My design is quite different as well. The F&P as a motor in a washing machine is rated at a few 100 Watts. I have recorded 605W from this motor as a windmill generator. I had said that my test rig was poor. The alternator does perform well in a windmill. I have the luxury of a high efficiency axial flux mill in close proximity, so I can directly compare real windmill outputs in similar conditions. IRL the axial flux is more efficient. The rotor on the F&P has 50% greater area than the axial flux, yet only produces 30% more power. The axial flux starts with approx half the windspeed, and produces approx 15W before any output is recorded from the F&P mill. I will not be pulling the F&P mill down and replacing it with an axial flux unit just yet. I would suspect that the 3phase motor operating at 15Hz would be a lot less than 80% efficient. I thought the 300rpm might have been a bit of a giveaway, and my mention of mechanical gearing being a better way. Gordon. become more energy aware |
||||
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344 |
If I can add my 2 bob in this thread  my intentions of 'Jerry Rigging' will either prove or disprove without any prony brake. Real wind is what matters and with my own genny situtated in a quiet zone it still draws 20 amps when the winds blowing with caps so if any real gains are there I'll need to get a bigger amp guage  Yes I need to put my genny ontop of the hill and watch it fail  Then I'll need to make a 5hp motor conversion to show just how easy it is really is when ya got the drive.
Cheers Bryan |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Hi Oz I agree, the link to Doc Chalko findings was more of a "lets look closer at this before making assumptions about the F&P efficency". 97% or 14% are both just plane wrong. The 97%, if achieved, was not a practical setup, and 14% would mean the F&P would burn out in the first wind gust. My single stator ceramic hub windmill has made 340 watts without caps, so at 14% efficiency, the stator had to be disipating over 2000 watts in heat! It would have turned very black very quickly. I expect the F&P to be somewhere in the 70% to 80% range, depending on the operating conditions. With caps and the right tuning, I would expect over 80% I agree with Dinges, we need to do a pony test on a F&P. His technique is easy and accurate, unfortunately I dont have a drive big enough to do the test. My baby lathe is only 500 watts. I can access a bigger lathe, but it will have to wait for a couple of weeks before I can do the tests. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344 |
OK Glenn, Ya got me hooked now for this prony brake setup our good Dutch mate will need to give some decent details as my work lathe is 10Kw. So that will give enough grunt and yes my work is interested in my idea's so if I can show the boss's the benefits, I might be able to take this one step further.
Cheers Bryan |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
As always Bryan, you show up with the goods just when we need them. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
imsmooth Senior Member Joined: 07/02/2008 Location: United StatesPosts: 214 |
Sorry for coming in late on this discussion. I was looking at your F&P 100S. I will post pictures, but I just rewired my 100S stator and increased the number of turns by a factor of 1.5. Total turns per coil are now 70. I will post pictures of my interrogation where I tested a standard and modified 100S. Then, I tested it with the neo rotor. I got over 1200W at 500 rpm with the neo rotor and the rotation was smooth and noiseless. I decogged my rotor not with twisting, but with filing the laminations. I plan on getting it up in the air this weekend. |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Glenn, thanks for the link to Dr. Chalko's tests. Is he the same person who is behind the Mt.Best conversion from freezers to highly efficient refrigerators ? I agree with Oztules, the 97% efficiency is plain silly and not credible. To give an example, if the F&P has two bearings (of good quality) which loses about 2% of power, it'd mean that only 1% of the losses would be iron and copper losses. Not believable (nor is 3% iron & copper loss, for that matter). When I measure results like those I usually wake up soon afterwards in bed and realize it was all just a dream...  (if I'd measure efficiencies over 100% it would have been a nightmare though and I'd wake up covered in sweat...) 
I know what you were doing here Gordon, I've read your previous post. Yet to me, the most interesting part was where you (implicitly) claimed a very low efficiency for the F&P, hence my remark. I know testing efficiency of the various configurations was not the core of your tests. Bagging of F&Ps was not the intention of my previous reply. The F&P isn't perfect. Neither is the motorconversion. The axial flux is probably (from an efficiency point of view) the best alternative we have at the moment. That doesn't stop me from preferring conversions though, as they have other advantages which outweigh their lower efficiency. Flux has also said on numerous occasions that he wouldn't be building motorconversions himself, as he considers the effort required too large for the result. I take careful note of Flux's remarks and objections, and then do my own thing and build conversions anyway. The best we can do is educate and inform ourselves of the pros and cons of various gennies, and then choose the type that best fits our needs and abilities. If for some that would happen to be the F&P, great for them. I have no grudge against the F&P (or AF). In fact, I'd love to have one to play around with a bit. But, if the efficiency of the F&P would *really* be 15%, then that should be told as well. I do the same (rigorous testing and criticizing) with my own gennies. It's the only way we can hope to improve things in the long run. No fairy tales/hearsay/superstition/wishful thinking etc. for me, please. (or, to quote Kant: "Sapere aude!") By using VFDs which vary both voltage and frequency (V/f control) one can shift the torque curve in such a way that maximum rated (and more!) torque becomes available at much lower speeds than the motor's synchronous RPM. The efficiency curve shifts too. So I think the guess of 80% efficiency was a reasonable one, given the lack of real hard data (measured mechanical input power). See also this link for some measured results of motor efficiency with VFDs: http://www.itrc.org/reports/vfd/r06004.pdf Check out fig.11 (p.14) and fig.12 (p.15) and notice that efficiency is only slightly lower as when the motors are working at synchronous RPM from sinusoïdal 50/60 Hz grid power, and in practically all cases above 80%. Gordon, don't get me wrong, I'm not 'picking' on you; I applaud the fact that you take the time and effort to measure these things and then take the trouble of reporting back in this forum. Too much work in windturbine construction is hearsay and 'we always did it like this' and 'I *think* setup A works better than setup B, but I haven't really measured'. There's a Dutch saying: "meten is weten" (measuring is/equals knowing). Anything else is [insert pejorative remark of your choice] and a waste of time. So let me finish by thanking you, as my reply may sound too negative otherwise. Bryan, real wind results are what matters in the end, yes. But its much better to measure generator performance on the bench (or lathe) and NOT on a windtower, as the environment is much more controllable and repeatable. By 'testing' generators on a windturbine we're suddenly faced with a *lot* more variables that can mess things up ('does the genny really have higher efficiency, or is it by chance perhaps better matched to *these* particular blades in *this* particular situation ?'). There's nothing as simple and clean as actually measuring efficiency, determining the power curve charging a battery, etc. than on the bench. And yes, I realize this doesn't give the complete picture of its performance as a windgenny. But it provides a lot of information of the generator, isolated from other messy factors (which we already have too much of in wind. Or, how many people can *accurately* (i.e. calibratedly) measure windspeed... ?) Oztules, see my remark above to Gordon. Check out fig. 12 (p.15) in the pdf file I linked to, and notice that it shows efficiency at 40% synchronous RPM (Gordon did his tests at 30%). Without knowing more of Gordon's setup (size of the motor, for example) I think 80% efficiency was a reasonable guess. But, only a truely designed setup (De Prony brake measurements) can take the guessing and assuming out of the above. Glad you reconsidered from your earlier remark (see my reply above). I think most of the info on how to set up the experiment can be found here (
http://www.thebackshed.com/Windmill/FORUM1/forum_posts.asp?T ID=1246&PN=4). The hope that others would repeat this test was the main reason for going into fairly large detail of it. You'll need a lathe of suitable size, a weighing scale (or 'spring' scale, as used for weighing fishies) and a way to measure RPM. And of course some batteries for the generator to charge. All fairly simple stuff, really. Let me know if/how you need more help in conducting the experiment. With a 10 kW lathe you could actually test a F&P to destruction, to reveal its maximum power... (ignoring the cooling effect of wind, or the generally low duty-cycle of wind); I bet that sounds like music to your ears, letting the majic smoke escape from things.
Peter. |
||||
| KiwiJohn Guru Joined: 01/12/2005 Location: New ZealandPosts: 691 |
Hmmm.... I dont think the efficiency of the F&P needs to be be all that bad. Unfortunately I dont have a nice lathe or motorised rig to do any tests but I did build a rather simple pedal generator on an old exercise cycle. When I started on that project I found the F&P almost impossible for me to pedal at even a few hundred RPM even without a load connected. So I started changing things. My final configuration retains the heavy metal flywheel of the exercise bike and I added another chain stage to drive the F&P which now has about a 6:1 ratio with the pedal crank. According to the cycles at the gym I can pedal for extended periods of time at about 100-110 RPM. What is the output of a F&P at 600RPM into a battery load? A few hundred watts I imagine. Around about now readers of this post will be thinking I am away with the fairies as there is no way a 60+ overweight office worker can put out that sort of power, and they would be right especially if I recall correctly a Tour De France rider puts out about 500 watts! What is different though is that I modified the F&P by cutting away about 60-70% of the stator! Thats right, with my trusty saw I cut around the stator removing all the coils for about 2/3 circumferance. This of course also reduced the magnetic losses and 'cogging' drag by the same amount. Gentle relaxed pedalling produces about 60 watts into the battery and just a little more pressure on the pedals, I dont have to get out of the seat or anything like that, produces in the region of 120 watts. From this I conclude that the F&P has very high losses at low RPM and that the efficiency is higher at higher RPM though no doubt there is a peak above which efficiency falls off with rising RPM and maybe rather steeply. I think that if one built two windmills with the same blades one driving an F&P directly and the other geared up driving an F&P with half (or so) of the stator cut away there would actually be more power produced by the later. 
|
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi dinges et al, The drive motor is the questionable unit in this equation. The motor I was given from the tip is a 2kW 415V star 3phase SCI type. Unfortunately the VFD produces only 240VAC 3phase. To indicate how much slip was involved in it, the drive was set to 15Hz, to produce the loaded 300rpm on the F&P. The correct speed for the motor would have been 855rpm. The slip was 0.35 instead of the usual 0.95. This is where most of the power is going in the rudimentary test. The tests were short in duration, but yes there was a lot of heat in the motor considering the short run times. I could spend many months writing up tests in a fully scientific way. I try and keep the info presented short and to the point, keeping the changes to a minimum, so other aspects do not cloud the picture. I hope someone else is able to provide approximate efficiency data of a F&P for a range of rpm and loadings. This will possibly remove this aspect from future discussions, where the intention was to compare something else. My testing has indicated that loading is the most significant problem with windmills. The fact that caps can double power output, with all other factors being the same is a strong case to support the loading aspect. Unfortunately with only a small number of components to blame losses to, the efficiency problem gets lumped to the alternator. Efficiency does not always manifest itself as heat. In a windmill this could be the opposite. The losses probably stem from a rejection of power capture. The rotor blade could be the real culprit, under and over loading both lower blade performance. Gordon. become more energy aware |
||||
herbnz Senior Member Joined: 18/02/2007 Location: New ZealandPosts: 258 |
hi note posting doubling fisher paykel output my posting 19 nov page 4 |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Understood. Far from an optimal situation. I'd even go so far as calling the input power figures practically useless for most applications (perhaps the only exception being what you used them for, a relative comparison of different stator/rotor configurations) as you are mostly measuring power dissipated in the drive motor, which is distorted by the extremely high slip of it, rendering your results very dubious. I'm not sure I'd even trust the input power figures for what you used them for, a relative comparison. I know it sounds a bit harsh. I may start to sound like a stuck record but the only reliable way to perform the test you intended to do was (is) by measuring mechanical shaft input power. Period. Any other method is just introducing extra errors. Large errors, in this case. BTW, 3-phase motors over here can usually be connected in either star (for connection to 3-phase power, 400V) or delta (3-phase 230V), as each phase of the motor has both ends available in the connection box, i.e. 6 connections (2 per phase). This allows quick and easy reconnection from star (99.9% of the applications) to delta. I've personally used a delta connected motor to a 3-phase (American) VFD, the VFD being supplied by single-phase 230V grid. It was the simplest solution for me to power a 3-phase motor from single phase grid power. slip = (synchronous speed - shaft speed)/synchronous speed. In your case, synchronous speed = 900 RPM (it's apparently a 2-pole motor, 3000 RPM @ 50 Hz) and shaft speed = 300 RPM, so in that case slip would be s=(900-300)/900 = 0.67, not 0.35; of course, a slip of 0.67 is still insanely bad (0.01-0.05 being normal) and an indication of the large losses in the motor. The good news (for F&P users) is that apparently much of the loss in your test setup is caused by the motor and not the F&P genny. How much exactly... ? Hopefully Brian can tell us soon. I for one would be interested in knowing, and so should anyone else be, I think, who wants to properly match his F&P generator to a set of blades.
Peter. (sorry if the above sounds a bit negative; I'll end my thread hi-jack here). |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi herb, that clears things up. efficiency around 80%. Gordon. become more energy aware |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
There you go Dinges, Herb has done the legwork already. Good job Herb. Also Dinges, Re:Fig 12 P15, pdf. Note the last graph at 40% speed. The slope is getting very steep, and even then...only around the .2 load mark. Mentally visualise going to half the speed again and imagine what the slope will be, and where you will end up if you try and push the output.... won't be pretty. .... and then there is also the voltage problem.......
Herb has some definative figures in the article he points to. Very much more sensible in the 70-80% range. Perhaps Glen should have pointed to this test rather than the good quack. 
The other problem is that the efficiency is dependent on the output. At low outputs my axial is very efficient(>90%)..... but as it gets up to 2kw or so, the efficiency has dropped to about 50-60%, and over 2kw we drop below the 50% mark.... and things rapidly fall to bits from there.... as the stator heats up and thermal runaway takes hold, the power absorbed in the stator will be truly ugly  ...... so the efficiency rating for a genny I figure must include at what power into the batteries we are talking about....... and it all comes down to stator resistance. For F@P the iron losses/bearing losses etc, are linear, and if?? the caps in series of the F@P negate the leakage reactances, then only R (plus the mechanical losses) are left. This will mostly determine the efficiency at X power levels I think.
............oztules Village idiot...or... just another hack out of his depth |
||||
| KiwiJohn Guru Joined: 01/12/2005 Location: New ZealandPosts: 691 |
[quote] For F@P the iron losses/bearing losses etc, are linear,[/quote] Is that true Oztules? So I dont understand why chopping out 2/3 of the stator makes an F&P apparently more efficient though with a lower output for same RMP.  |
||||
| Dinges Senior Member Joined: 04/01/2008 Location: AlbaniaPosts: 510 |
Anyone has a direct link to Herbnz's story ? I can find it neither in the index nor by googling the board. Oztules, I agree with your remark about efficiency and slope. All this could be circumvented by measuring mechanical input power. My previous remarks and links to the pdf file were pretty useless in hindsight as Gordon's setup was/is so far off from what's normal in driving motors with a VFD. So, by now I agree with his remark that his motor setup was lousy...
As far as efficiency varying with output goes: yes, exactly, and that is why I've made an efficiency *curve* for my 500W conversion, not just a stated efficiency figure. |
||||
| Page 3 of 12 |
|||||
 Print this page |
