Home |
Search |
Today's Posts |
|
Metalworking (rec.crafts.metalworking) Discuss various aspects of working with metal, such as machining, welding, metal joining, screwing, casting, hardening/tempering, blacksmithing/forging, spinning and hammer work, sheet metal work. |
Reply |
|
LinkBack | Thread Tools | Display Modes |
#121
|
|||
|
|||
I agree. Rotor inertia is necessary to drive the mechanical load in a
singlephase motor just as it is the third leg in an idler. I don't assert that this is the dominant source of energy exchange, only that it is a necessary contributor for at least part of each cycle. Not that it matters in any practical way, but I'm still curious about how much. I'm not curious enough (yet) to instrument it properly to measure it. wrote in message ... On 11 Sep 2004 16:40:13 -0700, jim rozen wrote: * In article , Don Foreman says... This doesn't prove (or disprove) whether or not energy is withdrawn from rotational energy during part of each cycle and replenished during a different part of the same cycle. There is certainly countertorque produced when the rotor field is generating the third leg while excitation voltage (hence input excitation power) goes thru zero. I would agree that "the .... energy built up during starting remains constant" as an average value, but that does not preclude significant alternating energy exhange during each cycle that nets to zero over each cycle under steady-state condx. I can certainly believe that moment of inertia is already so high that adding more produces no noticable improvement or difference. From a practical standpoint, that's probably all that matters. * Once again I would suggest the thought experiment of having a converter idler with a *zero* moment of inertia rotor. If such an instrument would not function, because the rotor would stop at the instant it is required to supply energy into the third leg, then you are most likely correct - there is *some* flywheel required. But no more so than, say, the rotor in single phase motor supplys to keep its rotor turning during the entire cycle. Jim It probably helps to look at limiting cases. If there's zero mechanical inertia, because the torque drops to zero twice per cycle, continuous rotation is not possible. If there is only JUST enough inertia to enable rotation to continue through the low torque parts of the supply waveform, the rotational velocity is not constant and drops towards zero during this low torque period. In the period that it remains below its long term average speed, as the rotor slows, it tries to take more power from the supply. In the period that it is above the long term average it returns the excess power to the supply. Long term (ignoring pesky second order effects) its input/output power balance is the same as as a motor with very large inertia. A second order effect that cannot be ignored is the effect on the back EMF waveform. Because of the velocity variation this is no longer sinusoidal and this results in a large third harmonic content in the current drawn from the supply. For the same reason, if this very low inertia motor is used as a rotary converter, the phantom phase voltage waveform will be distorted and contain a large third harmonic component. Very fortunately, for the sort of motors that we we use, rotor mechanical inertias are so large that the effects fade into insignificance. The torque fluctuations are so well averaged by the rotor inertia that the only effect we observe is the slight vibration of the frame of a single phase driven motor as it tries to accelerate and decelerate the rotor . The effect on phantom phase waveform is equally small and masked by the normal harmonic content of a commercial supply. Jim |
#122
|
|||
|
|||
On Sat, 11 Sep 2004 16:11:10 -0400, Gary Coffman
wrote: * On Sat, 11 Sep 2004 15:58:26 +0000 (UTC), wrote: The output pattern of these three phase winding can be directly measured and show that the stator generates a magnetic field pattern in a squirrel cage rotor that rotates at slip frequency in the same direction as the mechanical rotation. Because, in our real motor, this rotation of the magnetic field orientation is additional to and adds to the mechanical rotation it exactly cancels the slip speed and returns the phantom phase output to the correct frequency. Both the mechanical rotation and the magnetic correction are resisting the same drag torque from ouput loading so the energy contributions are in proportion to their speed - about 95% from the motor plus generator mechanical rotation and 5% for slowly dragging the stored field round the rotor. * This is where I have a problem. In a steady state system, the output energy equals the input energy (minus system losses). The mechanically stored rotational energy built up during motor starting remains constant (if it didn't, the rotor would slow to a stop). So the mechanical rotation contributes *zero* to the output energy. It is only the energy coupled via the slipping field which is transferred through the system. The mechanical rotation only establishes *phasing*. It isn't the source of energy for the output. This may appear to be a fine semantic distinction, but it explains why adding flywheel mass to a rotary converter doesn't help. Gary I think this is basically semantics and the usual problems of trying to keep posts reasonably short. It's really the result of trying to use single step and two step models to describe the same process. I prefer the two step model because, by conceptually separating the portion of the input current that produces torque, from the portion that rotates the orientation of the magnetic field in the rotor, it produces an easily assimilated picture of what's happening. Of course, as you say, there's only one real current and this produces the slipping field that you describe. The two step model does NOT require the rotational energy stored in the rotor INERTIA to make any long term contribution to the output energy but it makes clear that mechanical ROTATION of the rotor is an essential part of the phantom phase generation. If the rotor stops there is no energy transfer. The same two step model covers operation both with the input power power present or removed and it neatly and exactly explains the surprisingly small drop in phantom phase output power and frequency for the first few revs after the power input completely removed. The single step model can also explain the same process but it is much more difficult. Once the supply is removed the slipping field transfer is no longer operative and it is necessary to now take into account the effect of the still remaining rotor magnetic field that is the core of the two step model. It's clear that I overdid trying to keep the post short and simple. The following extended penultimate paragraph perhaps better explains what I was trying to say. Both the mechanical rotation RESULTING FROM THE STATOR INPUT CURRENT PROVIDING DRIVING TORQUE TO THE ROTOR and the magnetic correction are resisting the same drag torque from ouput loading. The energy contributions are in proportion to their speed - about 95% from the motor plus generator mechanical rotation and 5% for slowly dragging the stored field round the rotor. Jim |
#123
|
|||
|
|||
|
#124
|
|||
|
|||
All this talk got me to reading my audels electric motor book.
They state that a synchronous motor is almost identical to an alternator. Since I have a 3 PH alternator, I was wondering how hard it would be to make it run as a sync motor and once it is running to use it as a rotary phase converter. This would be more efficient than driving it with a single phase motor. My limited research seems to indicate that I need to get it very close to sync speed with a pony motor and then apply power to two phases and DC to the field. Then it should motor just like a motor and hopefully the third leg will be generating. Yes or NO? chuck |
#126
|
|||
|
|||
|
#127
|
|||
|
|||
|
#128
|
|||
|
|||
This sounds neat. Why would it be hard to configure a regulator to look at
the third leg? "Gary Coffman" wrote in message ... Should work, but there won't be any automatic voltage regulation of the phantom leg with varying load (slip does that for you in an ordinary squirrel cage RPC). You'll need to dynamically vary the field current to regulate the 3rd leg voltage with changing loads. Alternator regulators normally do this, but configuring one to only look at the 3rd leg voltage might cause a bit of head scratching. Gary |
#130
|
|||
|
|||
I think this is intriguing, might be worth a bit of head scratching. It's
intriguing because it offers the possibility of a "self-tuning" RPC that adapts to varying loads. Electronics could be quite simple because they need not produce significant AC power (as in a VFD) but only control DC excitation current. Assuming that a neutral point is not available, a neutral-equivalent voltage reference could be synthesized with opamps. It would be half the line-to-line excitation plus a quadrature component produced by an RC phaseshift circuit. Perhaps the third-leg phase voltage relative to this quasi-neutral could then be compared to the other phases relative to the same point, and excitation then self-adjusted to minimize the difference between generated third-leg magnitude and the other phase magnitudes relative to the quasi-neutral. Generated third-leg would be whatever it turns out to be, since we have only one controlled variable -- excitation. However, if the magnitude is right then I think the phase would be pretty close to right due to the geometry of the machine. Such a controller could be accomplished digitally with a microcomputer of course, but I'd find it easier to see and understand what's going on in a controller realized with a few opamps. Comments? I could be up for helping to design and even build such a gadget just to see if and how it works. "Gary Coffman" wrote in message ... Should work, but there won't be any automatic voltage regulation of the phantom leg with varying load (slip does that for you in an ordinary squirrel cage RPC). You'll need to dynamically vary the field current to regulate the 3rd leg voltage with changing loads. Alternator regulators normally do this, but configuring one to only look at the 3rd leg voltage might cause a bit of head scratching. Gary |
#131
|
|||
|
|||
On Tue, 14 Sep 2004 09:11:24 +0000 (UTC), "Don Foreman"
wrote: I think this is intriguing, might be worth a bit of head scratching. It's intriguing because it offers the possibility of a "self-tuning" RPC that adapts to varying loads. Electronics could be quite simple because they need not produce significant AC power (as in a VFD) but only control DC excitation current. Assuming that a neutral point is not available, a neutral-equivalent voltage reference could be synthesized with opamps. It would be half the line-to-line excitation plus a quadrature component produced by an RC phaseshift circuit. Perhaps the third-leg phase voltage relative to this quasi-neutral could then be compared to the other phases relative to the same point, and excitation then self-adjusted to minimize the difference between generated third-leg magnitude and the other phase magnitudes relative to the quasi-neutral. Generated third-leg would be whatever it turns out to be, since we have only one controlled variable -- excitation. However, if the magnitude is right then I think the phase would be pretty close to right due to the geometry of the machine. Such a controller could be accomplished digitally with a microcomputer of course, but I'd find it easier to see and understand what's going on in a controller realized with a few opamps. Comments? I could be up for helping to design and even build such a gadget just to see if and how it works. Should be an interesting project. A possible alternative method of synthesising the signal is a couple of low power transformers in Scott connection. The first is a centre tapped auto across the supply. The second is of any convenient ratio with primary connected from centre tap to phantom phase. Transformers are a nuisance but they have the big advantage that they fully isolate the control circuitry from the high power bits. If isolation isn't important, an opamp looking at the difference between a three equal resistor artificial neutral and the phantom phase is yet another way. The really interesting measurement will be to see what happens to the phantom phase angle when the regulator corrects the voltage drop resulting from a heavy load. If the model really describes the way it works we should see the effect of a small shift in the angular relation between the rotating field magnet and the rotating field component of the supply. There's a good chance that it will be a really useful improvement on the basic rotary converter but it's pity that 3 phase alternators are a lot rarer than motors. Jim |
#132
|
|||
|
|||
Isolation transformers would definitely be a good thing.
wrote in message news On Tue, 14 Sep 2004 09:11:24 +0000 (UTC), "Don Foreman" wrote: I think this is intriguing, might be worth a bit of head scratching. It's intriguing because it offers the possibility of a "self-tuning" RPC that adapts to varying loads. Electronics could be quite simple because they need not produce significant AC power (as in a VFD) but only control DC excitation current. Assuming that a neutral point is not available, a neutral-equivalent voltage reference could be synthesized with opamps. It would be half the line-to-line excitation plus a quadrature component produced by an RC phaseshift circuit. Perhaps the third-leg phase voltage relative to this quasi-neutral could then be compared to the other phases relative to the same point, and excitation then self-adjusted to minimize the difference between generated third-leg magnitude and the other phase magnitudes relative to the quasi-neutral. Generated third-leg would be whatever it turns out to be, since we have only one controlled variable -- excitation. However, if the magnitude is right then I think the phase would be pretty close to right due to the geometry of the machine. Such a controller could be accomplished digitally with a microcomputer of course, but I'd find it easier to see and understand what's going on in a controller realized with a few opamps. Comments? I could be up for helping to design and even build such a gadget just to see if and how it works. Should be an interesting project. A possible alternative method of synthesising the signal is a couple of low power transformers in Scott connection. The first is a centre tapped auto across the supply. The second is of any convenient ratio with primary connected from centre tap to phantom phase. Transformers are a nuisance but they have the big advantage that they fully isolate the control circuitry from the high power bits. If isolation isn't important, an opamp looking at the difference between a three equal resistor artificial neutral and the phantom phase is yet another way. The really interesting measurement will be to see what happens to the phantom phase angle when the regulator corrects the voltage drop resulting from a heavy load. If the model really describes the way it works we should see the effect of a small shift in the angular relation between the rotating field magnet and the rotating field component of the supply. There's a good chance that it will be a really useful improvement on the basic rotary converter but it's pity that 3 phase alternators are a lot rarer than motors. Jim |
#133
|
|||
|
|||
Assuming that a neutral point is not available, a neutral-equivalent voltage
The alternator is a 12 wire, so it can be connected as delta or wye. If connected as wye, a netural point is available. IF connected as delta, the voltage from either leg to the wild leg indicates the wild leg voltage. Seems straight forward? Also keep in mind it is a brushless alternator which means there are really two alternators wrapped up in one. The excitor has a field that is stationary. The rotor on the excitor is a 3 phase alternator that produces AC (at 120Hz because it has 6 poles) which is rectified by diodes on the rotor. The DC from the diodes powers the main rotating field. It only takes 30 volts at about 1 amp on the excitor field to drive the alternator to 15kw. When used as an alternator driven by a engine, the electronic voltage regulator is powered by 2 legs and drives the excitor field. It is really only regulating line 1 to line 2 and line 3 gives what it gives. Seems like the sense point could be moved to line 1 to wild leg. Anyway my point here is there is a huge time constant involved with regulating the output going through all inductance. |
#134
|
|||
|
|||
|
#135
|
|||
|
|||
On Tue, 14 Sep 2004 08:04:08 +0000 (UTC), "Don Foreman" wrote:
This sounds neat. Why would it be hard to configure a regulator to look at the third leg? Because a standard 3 ph alternator regulator doesn't work that way. You don't want it trying to regulate the two legs supplied by the utility, and it may be non-trivial to convince a standard regulator not to try to do so. Of courseyou can scratch build a custom regulator. Don Foreman had some good ideas on that. Gary |
#136
|
|||
|
|||
Probably would be a good idea to characterize it with a few experiments, as:
give excitation a step change and observe the time response of the generated voltage. Think the dynamics would change with load? wrote in message ... On 14 Sep 2004 22:20:14 GMT, (Charles A. Sherwood) wrote: Assuming that a neutral point is not available, a neutral-equivalent voltage The alternator is a 12 wire, so it can be connected as delta or wye. If connected as wye, a netural point is available. IF connected as delta, the voltage from either leg to the wild leg indicates the wild leg voltage. Seems straight forward? Also keep in mind it is a brushless alternator which means there are really two alternators wrapped up in one. The excitor has a field that is stationary. The rotor on the excitor is a 3 phase alternator that produces AC (at 120Hz because it has 6 poles) which is rectified by diodes on the rotor. The DC from the diodes powers the main rotating field. It only takes 30 volts at about 1 amp on the excitor field to drive the alternator to 15kw. When used as an alternator driven by a engine, the electronic voltage regulator is powered by 2 legs and drives the excitor field. It is really only regulating line 1 to line 2 and line 3 gives what it gives. Seems like the sense point could be moved to line 1 to wild leg. Anyway my point here is there is a huge time constant involved with regulating the output going through all inductance. Sounds logical and well worth a try. The lags involved in the time constants of a brushless alternator set up make it a bit more difficult to stabilise a regulator loop. However, if we've guessed right, the loop gain and lags of the revised regulator loop should be sufficiently similar to the previous connection for it remain stable. Moving the sense connection to wild leg has the major advantage of minimum change to the regulator loop. Theoretically it would be better to sense from a real or synthetic neutral but I don't think it's different enough to matter. What is difficult to predict is the transient behaviour as it first tries to lock into synchronous rotation. Probably better to check it out first with fixed excitation before allowing the regulator to take charge. Jim |
#137
|
|||
|
|||
On Wed, 15 Sep 2004 09:40:37 +0000 (UTC), "Don Foreman"
wrote: Probably would be a good idea to characterize it with a few experiments, as: give excitation a step change and observe the time response of the generated voltage. Think the dynamics would change with load? Would be a good move - what would be really nice is to look at the short term changes with a storage scope. Not sure how the dynamics will change with load - hopefully not a lot. There are so many second order effects! One interesting one is the effect of hysteresis in the soft iron field circuit of the exciter alternator - sort of minor electronic backlash within the servo loop. However my present feeling is that, whoever designed the existing regulator loop will have already solved most of these problems for us. Jim |
#138
|
|||
|
|||
wrote in message ... Probably would be a good idea to characterize it with a few experiments, as: give excitation a step change and observe the time response of the generated voltage. Think the dynamics would change with load? Would be a good move - what would be really nice is to look at the short term changes with a storage scope. Yup. That or a PC-based data-acq setup. I have a scope and a LabJack DAQ but no synchronous motor. DATAQ offers a simple DAQ for $24.95. http://www.dataq.com/products/startkit/di194rs.htm It's kinda slow at 240 samples/sec. That might be fast enough, given the large inductances and rotational inertias involved. Not sure how the dynamics will change with load - hopefully not a lot. There are so many second order effects! One interesting one is the effect of hysteresis in the soft iron field circuit of the exciter alternator - sort of minor electronic backlash within the servo loop. However my present feeling is that, whoever designed the existing regulator loop will have already solved most of these problems for us. Solved them, or just ignored them. Many control systems just do the best they can with the system at hand, warts 'n all. Consider, for example, an automotive cruise control. They certainly aren't quick and they're not always precise, but they usually work acceptably well. |
#139
|
|||
|
|||
On Wed, 15 Sep 2004 15:38:18 +0000 (UTC), "Don Foreman"
wrote: wrote in message .. . snip Would be a good move - what would be really nice is to look at the short term changes with a storage scope. Yup. That or a PC-based data-acq setup. I have a scope and a LabJack DAQ but no synchronous motor. DATAQ offers a simple DAQ for $24.95. http://www.dataq.com/products/startkit/di194rs.htm It's kinda slow at 240 samples/sec. That might be fast enough, given the large inductances and rotational inertias involved. snip The trouble is that suitable machines are pretty thin on the ground.I've got a pretty ancient storage scope but it's the wrong side of the atlantic! I don't know if advice from our photographic friends would help. I believe some of the better digital cameras have a 5 second maximum shutter time - would one of these do the necessary with a standard scope? Jim |
#140
|
|||
|
|||
I've done that, using an Olympus C2500L camera. It works OK if the scope
screen is hooded to cut out reflections and glare from the environment. However, for a 5 second period, I think the $24.95 DATAQ might work better. wrote in message ... On Wed, 15 Sep 2004 15:38:18 +0000 (UTC), "Don Foreman" Yup. That or a PC-based data-acq setup. I have a scope and a LabJack DAQ but no synchronous motor. DATAQ offers a simple DAQ for $24.95. http://www.dataq.com/products/startkit/di194rs.htm It's kinda slow at 240 samples/sec. That might be fast enough, given the large inductances and rotational inertias involved. snip The trouble is that suitable machines are pretty thin on the ground.I've got a pretty ancient storage scope but it's the wrong side of the atlantic! I don't know if advice from our photographic friends would help. I believe some of the better digital cameras have a 5 second maximum shutter time - would one of these do the necessary with a standard scope? Jim |
#141
|
|||
|
|||
Maybe I've got a crazy idea or have I had too many, but let's say that I buy a
Bridgeport or some old lathe, boring mill, or surface grinder that has a 3ph motor. The idea is to get this machine running on my 1ph service as cheaply as possible. Why not buy some big clunky 1 or 3 ph motor from the boneyard, use the apropiate circuit to make it run and couple it up to an automotive alternator with the diodes bypassed to get 3ph AC.Most automotive alternators will put out 100a at 12v but 10a at 120v will still handle a decent sized motor. I'd think that it would be a good idea to have the motor running the alternator to have the same speed as the motor being run from it. This would eliminate the need for pony motors although the alternator would require a 12v DC field power supply. This would also be an advantage because low voltage wiring to a switch could allow you to "freewheel" your alternator when not running your machine. Engineman1 |
#142
|
|||
|
|||
|
Reply |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Rotary Phase Converter | Metalworking | |||
Rotary phase converters | Metalworking | |||
General stuff on phase converters | Metalworking | |||
Phase Converters vs. VFDs | Metalworking | |||
Phase converter balancing | Metalworking |