ZIG-ZAG TRANSFORMER / REACTORS IN MULTIPLE ROLES
THE POWER CONDITIONING OVERVIEW
“Independent studies indicate that voltage fluctuations are an ever increasing problem with varying internal and external loads, as well as demand for power.” (Dr. M. Mherdad, Columbia University 1978)
The utility providers generate excellent quality power. However, there are factors which can change what is finally delivered to the customer. Weather, load behavior, primary transformer back wash, and other accidents outside of utility provider control may cause spikes, sags, dips, and other noise that have the potential that can literally destroy or severely damage sensitive electronic equipment that we find in almost all businesses today. These small fluctuations actually are responsible for more real damage than even massive power failures from extreme inrush conditions. We know of few solutions to this problem other than what a good heavy zig-zag multirole transformer in harmony with what tuned power capacitor banks can provide. By further explanation, there is such a thing as a ferroresonant transformer, and transformers that are regulators, but each is usually a stand alone device and we will talk about them all here. Why that matters is because the transformer we built is all three of those instruments at the same time and not just a zig-zag. It is making this device all of those things together that was the difficulty both to make and explain.
The purpose of this paper is to defend why we make such a device. We have heard many who argue and ask why we do it when no one else will. Here are the reasons and the logic behind it.
In some ways the transformers we make change the entire approach to Power Quality Control and the look at Power Conditioners in general. The last reason is that by doing this we now can deal with not typically or successfully mitigated multiple problems which were assumed to be impossible to reduce together. We had great difficulty even finding a manufacturer who dared to make what we asked for, and argued strongly against doing it. Now that we have successfully made them we have no idea as to what else could possibly replace what they do.
It just isn’t good enough to protect a power customer with Power Factor capacitors here and harmonic filters there and think that we have done a satisfactory job to protect the end user. Sufficient Power Conditioning is far more than that. Voltage regulation, ride through in short power breaks, line noise, ground loops, ongoing surge suppression, generator backwash from slowing down motors, inrush current bringing down voltage stability during motor startup, contactor sine wave breakup, and real power sags take a toll on equipment and usually have no mitigation when only standing power caps are present. Worse, notch filters left alone may cut back on some larger than other harmonics but they nearly always create harmonic increase around the frequency mitigated or targeted. The result is that stand alone notch filters reduce the dominant harmonic but only spread the damage around instead of actually reducing the THD in the circuit. (Total Harmonic Distortion)
We insist that while we are improving power quality just by knocking out a few dominant harmonic frequencies and reducing Power Factor; that is not really doing a good enough job for the client. Except for total power failure, a good power conditioner is of far more value to the power customer to provide clean, consistent power under most conditions for far less cost than any UPS system. End users with sophisticated and sensitive equipment will spend a fortune on an expensive UPS system fearful of a shutdown while ignoring far more damage from the constant attack not dealt with at all by the best of UPS devices. But then will buy not integrated equipment to help these problems which often react adversely between themselves and create more problems than they help. Still, here is a list of four ways to keep voltage in check:
1. Voltage Regulators. Very expensive electronic systems which react to voltage changes in an effort to keep voltage steady. One flaw is that they do not have longevity and need replacing often.
2. Isolation Transformers. This is really common answer when harmonics become unbearable. Not knowing why computers are locking up, printers fail or become inaccurate, data is altered and other nightmares become apparent when too many devices share the same circuit. The answer has been to put in a dry transformer to isolate one load in an office or other area from interference between systems on the same line. However, because little is understood about why the errors occur, variations in voltage is the only problem so Isolation Transformers are installed which often seems to solve the problem. But what really happens is that the isolation is now only keeping the distortions from overloading the working area.
3. Line Conditioners. These often sophisticated and expensive electronic devices address both line isolation and integrated circuits when internal interference is common between data and other accuracy sensitive devices which will sometimes exchange information between each other. Especially when the only link between them may only be the power plug.
4. Ferroresonant Filters. These are not common, and good ones are almost impossible to find but they are transformers which are detuned between devices. Actually they can work well but if all systems are not considered, they can become horrendous to control. They also act as traps to hold against small power interruptions which show up for no reason at all and don’t even originate from the utility power source. By the way, one needs to understand that these transformers are zig- zag transformers. The difference is that these transformers stop at voltage stability and do not go into multirole properties which is what this paper is about. My question then is why stop there when you are in the neighborhood? Making other transformers to perform the other needs in the same circuit creates automatic inductive interference and the potential difference creates a ground loop phenomenon and flowing current between them. So why create more headaches when the idea is to solve them?
What we often see is desperate clients that have purchased all four and sometimes even more devices in an effort to handle all their problems. There is an unknown but common mistake for doing that. It is that they are never integrated and will fight each other to get the job done. And because of that, usually end up creating more problems than they are helping. Almost every time we see these together, the client finally just gives up and will use only one, typically and most often isolation transformers. We often go into a building and see them installed everywhere and know why they were bought. But then often see one or more of the other three sitting there turned off and gathering dust.
PRINCIPLES OF OPERATION
Power Quality Regulation done better is not to ignore any of the headaches but to recognize that they are all there and need to be addressed. But there are many as we have listed. Here is a list and what can be done to deal with them. Specifically why many of them are best dealt with by using a single transformer which is not made anywhere until we made them. How and why we do it is reported here. First we will list the two greatest power loss causes and explain them:
1. Power Factor Can one install Power Caps to reduce Power Factor Error against inductive reactance and lagging current against on time voltage? Sure. But we remind everyone that equipment comes on and off constantly. Failure to keep the capacitor offset changing as fast as the load changes is a mistake. The only way to do that is to have an intelligent control system that will bring in KVARS as fast as they are needed then turn them off as fast as they are not. But after 1984, power caps were burning out so fast that what they saved in Power Factor was off set by the installation costs of constant replacement. This fact brought about a near loss of knowledge of PF itself and what can be done to correct it. Not only that, when they were used, static caps never kept PF better than .85 even when they weren’t burning up. Our solution is that the mitigation must be as active as the changing condition if a .98 PF is even possible. We declare that it is and in fact our norm for correction.
2. Harmonic Distortion This did not exist in enough quantities to even pay attention to before the mid 80s’. After that, the whole condition has grown to become a bigger monster than even Power Factor. What was not recognized along the way was that harmonic distortion at the higher frequencies was the cause for burned out capacitors. Believed to be just inferior manufacturing of caps, what was really burning them out was the harmonics in the 65 KHz to 75Khz range which was cutting through the insulation inside the power capacitor. Eliminate that range, and the caps stop burning up. But that was not all. Mid range HD was heating up wire, iron cores, and even burning up semiconductors as well. Recognized as the ongoing problem, attacking HD alone without recognizing PF and treating it at the same time only resulted in less than satisfactory results anyway. The whole HD problem is a subject for more than a few words here and we suggest reading white papers that deal with this to create an understanding.
Earlier in this paper we brought up far more about power quality than just Power Factor and Harmonic Distortion. Our third area is that the two big concerns without recognizing the rest is a serious error. We will deal with them now and explain that, all of the areas of Power Conditioning MUST be looked at together to ever really arrive at satisfactory answers to the Power Quality problem. This paper is mostly about the role of transformers that did not exist until we forced them to be made. Long ago we asked why all the problems inherent with power quality are not integrated? That list of four in the first page had cause and need. Why not put them all together into one system? That seems so simple the very question answers itself. Our reply is that we need to. More information as to how.
Some systems will install a transformer tuned with a capacitor such that they are called regulators because they dampen capacitor inrush and fluctuation. In electronic circuits they are called chokes for that reason. But in a full power operation, choking a circuit can lead to an explosion. A fully tuned regulator with an on frequency capacitor gives us this formula:
Cr = Ir = Infinity ohms [or]
Capacitive Reactance equal to Inductive Reactance results in a circuit dead short.
You see, the heart of a regulator is the constant voltage transformer. Why? Because it is the calming factor when capacitors are flying around charging and discharging. And to have the best dampening effect upon the capacitor, it needs to be close to the opposite tuning. But the exchange also forces inductors to fluxuate as well instead of remaining constant. And If it does not remain constant then it might accidently wander into pure tune with the capacitor, and we get big trouble. So the result is that most regulators are never designed to be close to a tuned condition because of the danger. For decades that has been the caution, and why most regulators are only marginal in their dampening effect. The answer should be obvious but has been ignored forever. Why does not matter. So, let’s look at it.
A true constant voltage transformer has a magnetic core structure different form a conventional transformer. Most transformers operate by creating a magnetic shunt with a fixed gap to exchange magnetic flux between coils. The primary coil will induce (why we call all coils inductors) current into a secondary coil. If the primary coil has so many windings, then if we only have half as many windings in the secondary then the voltage or power pressure will be half. That way a high voltage line needed for long distant transmission and wire size reduction can be reduced down to a low safe voltage where it can be used. In most industrial locations the voltage arrives at the site at about 13,000 Volts and is stepped down by a transformer into a usable 480 volts. Simple enough ok? But back to making a constant voltage transformer that we can keep close to the capacitor tune without danger.
All transformers one way or another have primary and secondary coils even when the coils are electrically tied. This fact is why a simple coil of one wire treats part of itself as primary as well as secondary inducing current and voltage inside of the very single coil that it is. This induction is what forces current to arrive late behind voltage in a circuit and creates the condition of Power Factor Error. Multiple motors and transformers will combine to create what Power Factor can be seen in a circuit. This is the origin of the condition.
Inside of regulators operating with capacitors, the internal secondary voltage is shunted by the fixed capacitor. Upon application of current upon the transformer, the primary part of the coil causing secondary voltage increases to a point at which the core saturates. As soon as that happens, whatever dampening that may be needed becomes maxed out. In nearly every application we have seen in power quality effort where regulators are present, they are all usually too small. But there is a limit to size and some saturate faster than others but all will saturate when used as a single coil due to simple design. It just happens.
Therefore we think a double design is needed. Why not make a regulator as a zig-zag transformer as well? Every engineer I know of is afraid of that, and most insist that doing it just makes the whole thing doomed to explode anyway. But let us look at it and why we choose to do it and still keep it safe.
Now let us tune a three phase zig-zag transformer and use it as a regulator. We will assemble the core as a three part, two hole doughnut with all three cores coupled as one piece of iron. Let us name one of the three cores as phase A part of the transformer and call it the primary core and coil. Then call the other two parts the B and C cores and coils secondary while the A core is powered. But we will wind the A core as a primary with multiple windings in one direction then with the same wire in the circuit, reverse the direction of the coils with few windings when calling B & C coils secondary. Then as the A phase dies out, and B phase powers up, then the B core of many windings becomes primary and wire from the A circuit now has few winding in an opposite direction on top of the A and C core. Then we repeat the same circuit configuration when the C core is powering up. That is what defines a Zig-Zag transformer and it forces voltages beween phases to imitate each other.
But most Zig-Zag transformers have a typical problem. As the primary core normally approaches saturation, it cannot carry additional magnetic flux and will saturate just like it always does. But the increase in secondary voltage with smaller turns on the other two phases is always less than the primary one no matter what the proportional increase in the primary might ever be. This creates a condition of relative stability in both of the secondary coils and cores, since both B and C cores are charged equally they must therefore be identical. Keep in mind that we are not only going to wind this transformer as a Zig-Zag, we are also going to tune the coils to the capacitor as close as possible without allowing them to resonate. Why would we do this? And also because of the need for a good tank circuit, why not make both the zig-zag and reactor large so it has holding capacity? Well, we do.
Remember that all the cores and wires in all three are connected. By reverse winding the two secondary cores with the primary, our doughnut core has two polarities inside itself. At the A phase core the polarity is one way and inside the B and C cores are reversed in polarity. Like it or not, the iron core cannot saturate with opposite polarities inside of itself. No matter how high you drive the A phase, B and C will not allow A to go wherever it wants because the magnetic flux will not ever be one sided. Over the range of the secondary windings both may become saturated but that is not what matters in this circuit. It is the A phase that is dampening the A side of the capacitors, the B and C phases are not active. With both of the secondary cores active, this produces voltage stability between all three phases so long as the A phase is charged. Any attempted change in voltage between the three phases will be resisted because one phase is dictating a flat voltage between all of them whatever it is at that moment. But again, due to the magnetic shunt between the primary and secondary windings, that part of the core under the primary is not saturated and provides all the dampening available.
Now let us look at the fact that this is a three phase alternating circuit. Alternating current is sinusoidal and must change direction and return to the same place every 1/60th of a second. A phase is going to reverse momentarily and B then C phase will follow. As A phase is dying away B is coming up. As they pass each other, the B part of the transformer becomes the primary coil and A and C become secondary. Remember that all three phases are not connected electrically, but the iron cores are. Now as the phases pass they take control of their part of the core and flip polarity in the other two parts. This transition repeats every 360th of a second because each phase is running at 60 Hz but the reversal in magnetics is multiplied by three, and reversing just as fast on the other side of the sine wave adding three more changes to the process. This whole transition is now so fast that any variation in voltage between phases must find a common ground and remain steady. No matter what change in voltage each phase my want to move, the other two phases which are magnetically linked now will not allow it.
As a stable reactor we are safe to keep the tune close enough to give us the dampening need we are looking for, and the interrelationship between the reactor and the capacitors now will not and cannot wander. Any danger of a direct tune between the reactor and the capacitors is now impossible and can tune it very close to where maximum dampening effect can be enjoyed. In reality this is not complicated even if it seems so. The fact is, we have shown a very simple set of transformer principles working together. let us continue dealing with the rest of the Power Quality issues and needs:
3. Voltage Stability Excellent voltage regulation and stability now also forces current to fall in line. And a steady current will also reciprocate and would force voltage to remain flat. Steady current and voltage are very comforting to circuits of all kinds especially three phase motors. The result is increased efficiency and prolonged life. Other circuits that may not be three phase operate much better when the voltage and current are steady, especially sophisticated equipment that must remain accurate at all times. We have seen a lot of end users seek voltage stability with tap switchers. That method is better than nothing, but results in sudden and not smooth change transitions which are hard on equipment and seldom accurate. That idea reminds me of trying to thread a needle with snowmobile gloves on. Rough to accomplish well. The zig-zag keeps it clean and steady.
4. Capacitor Dampening Here is a reactor with a heavy core that cannot saturate and takes capacitor inrush and download in stride. With a safely held tuned reactor close to the caps, alterations in circuit fluctuations are held in check because the transformer refuses to allow offsetting voltages to fluctuate. If the Wattage draw is constant, and voltage is held in check, current rise and fall typical of charging and discharging caps is vastly reduced as well. Therefore the typical inrush and out charge spikes seen when caps must do that are heavily suppressed by a powerful and closely tuned reactor.
5. Surge Suppression For the very same reason, any sudden surge is resisted by the transformer as the magnetic mains resist all and any change in stability. We mentioned that our transformers are large. Now it matters. A heavy core coupled with large power capacitors creates a tank that has sufficient holding power to keep any short surge in check. Another fact is that small surge suppressors that are at most desks burned out many surges ago and are essentially now just power strips. The tiny circuits inside them are not capable of surviving multiple hits. We talked about long term surge suppression. The heavy combination of power caps and transformers automatically do it at the source so even small devices which burn out are themselves now protected against damage. We do guarantee voltage spike attenuation at a ratio of 250 to 1. Note is taken that all surges and sags can come from changes in either voltage OR current changes. One affects the other. By not allowing either to move, whatever the cause of the sag or surge is irrelevant. That is why voltage only surge suppression systems burn out. By dealing with current as well, no surge can pass this circuit –except one. ( See paragraph #12).
6. Ground Loop Correction This problem is created when too many grounds are so separated that they have different resistance between them. This differential can create a flow of power between them. This power flow becomes a totally isolated circuit where no circuit should be present as well as change the entire purpose for having a ground at all. By creating swift magnetic changes inside the zig-zag, the power flow in the normal circuit is no longer looking at changing grounds that oppose it. Therefore ground loops that might be created by variations between phases, now disappear.
7. Motor Slowdown Backwash Motors and generators are essentially the same thing. Motors convert power into kinetic energy and work. Generators made the same way use kinetic force to create power. When all motors are turned off, and their inertia keeps them going for a short time until they stop, that slowing down turns them into short time generators and makes power instead of using it. This backwash is never in harmony with the existing power in the line and always creates odd power which is disruptive in the circuit. This disruptive current and voltage can be wiped out within the induction coils of a zig-zag transformer by forcing the fast changing magnetic fields within it to force non compliance power to join in with harmony or be converted into magnetic flux energy while engaged in fast polarity exchanges .
8. Sags We are a bit interested in why so many worry about surges and buy voltage surge suppressors. When we talk about surges we are referring to voltage spikes. But voltage does not usually damage. It is current in amps that is destructive. We do not have breakers that trip when voltage changes. We do have breakers which go when current exceeds a set amount because that is what burns out wire. Any utility company will tell you that for every surge there are between six and more sags. Remember what steady power draw does? It means that watts remain the same. A 60 watt light bulb will draw 60 watts without regard as to what the voltage or the current may be. If voltage sags, then it is current that will be forced up to maintain the constant power draw. If the voltage sags enough, then the resulting current spike will burn out that light bulb or other power device - not the sag in voltage. But the reality is that current surge devices that are self correcting are far too expensive to put in a desk size power strip. Breakers and/or fuses take care of that problem but must be reset or replaced. And as we mentioned, that power strip surge suppressor burned out long ago because of the sag in voltage that spiked the current and destroyed it. So, along with voltage change resistance, this big circuit will also force current sags to remain very constant and within reason. Really big current changes are dealt with in other ways as we will discuss.
9. Harmonic Distortion What do zig-zag systems have to do with that? They don’t. But we saw something common and took advantage of it. Frequency notch filters are made from delta loaded capacitors and coils attached and tuned to the three terminal connections. Frequency notch filters are usually small enough to hold in your hand. We reasoned that we were already using capacitors to suppress Power Factor Error and dampening them with reactors. The filter parts were already there. By making the power capacitors in a delta loaded form, and tuning each reactor phase to the offending harmonic, now have a giant notch filter for a specific frequency. As we have several stages needed in keeping Power Factor steady, it is logical to merely make each stage a notch filter while also providing KVARS to correct Power Factor. Now we have made our caps multidimensional against the dominant harmonics which motors naturally create. (And only three of them make up 90% of that list) If we take a bite out of the dominant harmonics, we will reduce total distortion (THD) along the way. And it works amazingly well. But the biggest advantage comes in common integration. Remember the problem with multiple devices that were not designed to work together? By forcing systems to do multiple tasks, they internally do not react against themselves the way stand alone mitigators do which were not intended or designed to operate together.
10. Noise Suppression. Some time back we looked at where line noise came from. Most of it is created by a combination of dominant harmonic interaction, voltage and current variations, motor and capacitor inrush and outdrain, and switching that goes on in all busy circuits. By not directly attacking noise problems, if we eliminate most of what causes it, by natural selection much of it is reduced. The system we have when under test revealed a steady noise suppression of better than 60 DB. As many buildings may not have more than that anyway, that much suppression can eliminate all of it. Circuit noise is the major reason or cause of most data accuracy errors. Where precision is a must, noise can make that precision undependable. Some noise comes from harmonic disturbance. And some noise IS harmonic distortion. We will look at more of that in another area.
11. Line Interruptions and Power Breaks. Ride through is a major problem when everything in a building must work in harmony such as a refinery or assembly line. Any interruption in one place makes the whole plant go through a startup procedure to get back on line. We talked about the reason for heavy reactors and over size capacitors. Now here is the other reason for it. The power tank that forces voltage and current to sit still will also delay an interruption for 25 milliseconds or 1.5 cycles. The average power switchover that commonly occurs every three days in most utility systems is .71 cycles or only about 10 milliseconds. That is enough to force a refinery or other critical line integration conditions to go through a startup once every couple of months, or so. Some assembly lines may have to go though startup more often but 25 m/s is long enough that our circuit tank should keep even that from happening. We do not have an answer when the line break is longer. That is a full utility error and nothing much can be done to avoid that. However, we will say that the tank circuit will maintain 40% of the power for a full three cycles which is a full 1/5th of a second. In some cases that will keep a non critical plant on line without a startup. If voltage is critical, we guarantee that the tank will hold voltage to 99.7% for 3 to 8 milliseconds. This ability to clamp under line loss is what makes the transformer a ferroresonant type in holding on to current over time and not just a zig-zag, frequency choke, or even only a reactor.
12. Excessive Sag or Spike Suppression. Lightning strikes or substation power rebound however, will overwhelm our system. That is why we do analyze max power conditions and install MOVs large enough to save the plant. What will happen is the entire Power Conditioner will give its’ all instead. The strike will be taken internally but the critical equipment in the plant should survive. We have experienced that twice now with no damage at all to the end power user. Lightning rods are only marginal, and often run damaging power back through the ground to destroy sensitive equipment anyway. The only answer is to clamp the ground at the incoming location where our system is attached to the main panel. When power overwhelms the tank circuit, the MOVs then will shunt all the excess current internally to the tank circuit where it is delayed until the condition has passed. The same ability to allow for some ride through will contain damaging power long enough to only hurt the protection circuit. When the quick lightning strike is over, the containment circuit is gone but held it back long enough to save whatever it is protecting. Where extremely important equipment is at stake, this can be a very significant matter. If power is so severe that some damage is unavoidable, it is better to sacrifice lesser equipment instead of high value.
13. Short Circuit Protection Outside of the MOV, if output voltage drops to zero, we will limit output current to 200% of rated power. That is enough delay of sufficient back wash that any average breaker will trip. And even if they do not, this is enough load headroom to assure that no power transformer damage will occur. We have documented this claim many times in past installations.
14. Efficiency: Average to date for all we make is between 94% and 99% so long as all the components are in working order. Neglect for years will let anything fail. We have had equipment go without any replacement or service for seven years so far. But to ask a contactor that may have two million or more operations behind it to keep running on is not a reasonable expectation. It should be replaced. And for the money it saved the user, $300 in a few years is a bargain. Our intelligent switcher will keep track of exactly how many operations each part has had to endure. There is a way to ask it to tell you what that is. We built that feature into the computer it contains so that answer is available to you. Maintenance is a good idea for everything. Nobody makes things that last indefinitely, and taking care of it will extend the life of anything. It is all about economy. Does it pay for itself several times? If so then allow it to continue to do that job with good upkeep. Remember that in enough time and under enough stress every part will eventually fail. Keeping parts in good order as needed will give the same return indefinitely.
15. Volt/Amp Ratings: From 140 VA to 75 VA
16. Load reduction or Headroom This is an area which really does not need to be included but is worth calling attention to if not understood. All of the categories here are about power waste and loss. It is hoped that by reducing so much of what is inefficient, then the load has to go down unrelated to what is called for. A common example is when an operation is at maximum power availability. The transformer or main panel just cannot give out another amp. Instead of obtaining another larger transformer or installing a bigger panel, If we reduce the consumption by 30% through various means, the user now has 30% more room to add on more power equipment without having to upgrade or replace their present transformer or panel. I mention 30% for a reason. In nearly all factories or manufacturing facilities, a 30% headroom expansion is common for us to give to the customer who installs our systems.
17. Broadband Harmonics. When THD is beyond the ability of a few notch filters in the H-2 to H-12 range, then more harmonic reduction is needed. When sufficient amount of Harmonic Distortion is detected between 3 KHz and 100 KHz, we make an Intelligent Harmonic Filter to deal with any frequency in that category. When high frequencies exist in enough power, damage is inevitable. Look for an entire white paper dealing with just this problem and how we deal with it.
Written by John Jackman, Principle Engineer Power Control Company November 2009
THE POWER CONDITIONING OVERVIEW
“Independent studies indicate that voltage fluctuations are an ever increasing problem with varying internal and external loads, as well as demand for power.” (Dr. M. Mherdad, Columbia University 1978)
The utility providers generate excellent quality power. However, there are factors which can change what is finally delivered to the customer. Weather, load behavior, primary transformer back wash, and other accidents outside of utility provider control may cause spikes, sags, dips, and other noise that have the potential that can literally destroy or severely damage sensitive electronic equipment that we find in almost all businesses today. These small fluctuations actually are responsible for more real damage than even massive power failures from extreme inrush conditions. We know of few solutions to this problem other than what a good heavy zig-zag multirole transformer in harmony with what tuned power capacitor banks can provide. By further explanation, there is such a thing as a ferroresonant transformer, and transformers that are regulators, but each is usually a stand alone device and we will talk about them all here. Why that matters is because the transformer we built is all three of those instruments at the same time and not just a zig-zag. It is making this device all of those things together that was the difficulty both to make and explain.
The purpose of this paper is to defend why we make such a device. We have heard many who argue and ask why we do it when no one else will. Here are the reasons and the logic behind it.
In some ways the transformers we make change the entire approach to Power Quality Control and the look at Power Conditioners in general. The last reason is that by doing this we now can deal with not typically or successfully mitigated multiple problems which were assumed to be impossible to reduce together. We had great difficulty even finding a manufacturer who dared to make what we asked for, and argued strongly against doing it. Now that we have successfully made them we have no idea as to what else could possibly replace what they do.
It just isn’t good enough to protect a power customer with Power Factor capacitors here and harmonic filters there and think that we have done a satisfactory job to protect the end user. Sufficient Power Conditioning is far more than that. Voltage regulation, ride through in short power breaks, line noise, ground loops, ongoing surge suppression, generator backwash from slowing down motors, inrush current bringing down voltage stability during motor startup, contactor sine wave breakup, and real power sags take a toll on equipment and usually have no mitigation when only standing power caps are present. Worse, notch filters left alone may cut back on some larger than other harmonics but they nearly always create harmonic increase around the frequency mitigated or targeted. The result is that stand alone notch filters reduce the dominant harmonic but only spread the damage around instead of actually reducing the THD in the circuit. (Total Harmonic Distortion)
We insist that while we are improving power quality just by knocking out a few dominant harmonic frequencies and reducing Power Factor; that is not really doing a good enough job for the client. Except for total power failure, a good power conditioner is of far more value to the power customer to provide clean, consistent power under most conditions for far less cost than any UPS system. End users with sophisticated and sensitive equipment will spend a fortune on an expensive UPS system fearful of a shutdown while ignoring far more damage from the constant attack not dealt with at all by the best of UPS devices. But then will buy not integrated equipment to help these problems which often react adversely between themselves and create more problems than they help. Still, here is a list of four ways to keep voltage in check:
1. Voltage Regulators. Very expensive electronic systems which react to voltage changes in an effort to keep voltage steady. One flaw is that they do not have longevity and need replacing often.
2. Isolation Transformers. This is really common answer when harmonics become unbearable. Not knowing why computers are locking up, printers fail or become inaccurate, data is altered and other nightmares become apparent when too many devices share the same circuit. The answer has been to put in a dry transformer to isolate one load in an office or other area from interference between systems on the same line. However, because little is understood about why the errors occur, variations in voltage is the only problem so Isolation Transformers are installed which often seems to solve the problem. But what really happens is that the isolation is now only keeping the distortions from overloading the working area.
3. Line Conditioners. These often sophisticated and expensive electronic devices address both line isolation and integrated circuits when internal interference is common between data and other accuracy sensitive devices which will sometimes exchange information between each other. Especially when the only link between them may only be the power plug.
4. Ferroresonant Filters. These are not common, and good ones are almost impossible to find but they are transformers which are detuned between devices. Actually they can work well but if all systems are not considered, they can become horrendous to control. They also act as traps to hold against small power interruptions which show up for no reason at all and don’t even originate from the utility power source. By the way, one needs to understand that these transformers are zig- zag transformers. The difference is that these transformers stop at voltage stability and do not go into multirole properties which is what this paper is about. My question then is why stop there when you are in the neighborhood? Making other transformers to perform the other needs in the same circuit creates automatic inductive interference and the potential difference creates a ground loop phenomenon and flowing current between them. So why create more headaches when the idea is to solve them?
What we often see is desperate clients that have purchased all four and sometimes even more devices in an effort to handle all their problems. There is an unknown but common mistake for doing that. It is that they are never integrated and will fight each other to get the job done. And because of that, usually end up creating more problems than they are helping. Almost every time we see these together, the client finally just gives up and will use only one, typically and most often isolation transformers. We often go into a building and see them installed everywhere and know why they were bought. But then often see one or more of the other three sitting there turned off and gathering dust.
PRINCIPLES OF OPERATION
Power Quality Regulation done better is not to ignore any of the headaches but to recognize that they are all there and need to be addressed. But there are many as we have listed. Here is a list and what can be done to deal with them. Specifically why many of them are best dealt with by using a single transformer which is not made anywhere until we made them. How and why we do it is reported here. First we will list the two greatest power loss causes and explain them:
1. Power Factor Can one install Power Caps to reduce Power Factor Error against inductive reactance and lagging current against on time voltage? Sure. But we remind everyone that equipment comes on and off constantly. Failure to keep the capacitor offset changing as fast as the load changes is a mistake. The only way to do that is to have an intelligent control system that will bring in KVARS as fast as they are needed then turn them off as fast as they are not. But after 1984, power caps were burning out so fast that what they saved in Power Factor was off set by the installation costs of constant replacement. This fact brought about a near loss of knowledge of PF itself and what can be done to correct it. Not only that, when they were used, static caps never kept PF better than .85 even when they weren’t burning up. Our solution is that the mitigation must be as active as the changing condition if a .98 PF is even possible. We declare that it is and in fact our norm for correction.
2. Harmonic Distortion This did not exist in enough quantities to even pay attention to before the mid 80s’. After that, the whole condition has grown to become a bigger monster than even Power Factor. What was not recognized along the way was that harmonic distortion at the higher frequencies was the cause for burned out capacitors. Believed to be just inferior manufacturing of caps, what was really burning them out was the harmonics in the 65 KHz to 75Khz range which was cutting through the insulation inside the power capacitor. Eliminate that range, and the caps stop burning up. But that was not all. Mid range HD was heating up wire, iron cores, and even burning up semiconductors as well. Recognized as the ongoing problem, attacking HD alone without recognizing PF and treating it at the same time only resulted in less than satisfactory results anyway. The whole HD problem is a subject for more than a few words here and we suggest reading white papers that deal with this to create an understanding.
Earlier in this paper we brought up far more about power quality than just Power Factor and Harmonic Distortion. Our third area is that the two big concerns without recognizing the rest is a serious error. We will deal with them now and explain that, all of the areas of Power Conditioning MUST be looked at together to ever really arrive at satisfactory answers to the Power Quality problem. This paper is mostly about the role of transformers that did not exist until we forced them to be made. Long ago we asked why all the problems inherent with power quality are not integrated? That list of four in the first page had cause and need. Why not put them all together into one system? That seems so simple the very question answers itself. Our reply is that we need to. More information as to how.
Some systems will install a transformer tuned with a capacitor such that they are called regulators because they dampen capacitor inrush and fluctuation. In electronic circuits they are called chokes for that reason. But in a full power operation, choking a circuit can lead to an explosion. A fully tuned regulator with an on frequency capacitor gives us this formula:
Cr = Ir = Infinity ohms [or]
Capacitive Reactance equal to Inductive Reactance results in a circuit dead short.
You see, the heart of a regulator is the constant voltage transformer. Why? Because it is the calming factor when capacitors are flying around charging and discharging. And to have the best dampening effect upon the capacitor, it needs to be close to the opposite tuning. But the exchange also forces inductors to fluxuate as well instead of remaining constant. And If it does not remain constant then it might accidently wander into pure tune with the capacitor, and we get big trouble. So the result is that most regulators are never designed to be close to a tuned condition because of the danger. For decades that has been the caution, and why most regulators are only marginal in their dampening effect. The answer should be obvious but has been ignored forever. Why does not matter. So, let’s look at it.
A true constant voltage transformer has a magnetic core structure different form a conventional transformer. Most transformers operate by creating a magnetic shunt with a fixed gap to exchange magnetic flux between coils. The primary coil will induce (why we call all coils inductors) current into a secondary coil. If the primary coil has so many windings, then if we only have half as many windings in the secondary then the voltage or power pressure will be half. That way a high voltage line needed for long distant transmission and wire size reduction can be reduced down to a low safe voltage where it can be used. In most industrial locations the voltage arrives at the site at about 13,000 Volts and is stepped down by a transformer into a usable 480 volts. Simple enough ok? But back to making a constant voltage transformer that we can keep close to the capacitor tune without danger.
All transformers one way or another have primary and secondary coils even when the coils are electrically tied. This fact is why a simple coil of one wire treats part of itself as primary as well as secondary inducing current and voltage inside of the very single coil that it is. This induction is what forces current to arrive late behind voltage in a circuit and creates the condition of Power Factor Error. Multiple motors and transformers will combine to create what Power Factor can be seen in a circuit. This is the origin of the condition.
Inside of regulators operating with capacitors, the internal secondary voltage is shunted by the fixed capacitor. Upon application of current upon the transformer, the primary part of the coil causing secondary voltage increases to a point at which the core saturates. As soon as that happens, whatever dampening that may be needed becomes maxed out. In nearly every application we have seen in power quality effort where regulators are present, they are all usually too small. But there is a limit to size and some saturate faster than others but all will saturate when used as a single coil due to simple design. It just happens.
Therefore we think a double design is needed. Why not make a regulator as a zig-zag transformer as well? Every engineer I know of is afraid of that, and most insist that doing it just makes the whole thing doomed to explode anyway. But let us look at it and why we choose to do it and still keep it safe.
Now let us tune a three phase zig-zag transformer and use it as a regulator. We will assemble the core as a three part, two hole doughnut with all three cores coupled as one piece of iron. Let us name one of the three cores as phase A part of the transformer and call it the primary core and coil. Then call the other two parts the B and C cores and coils secondary while the A core is powered. But we will wind the A core as a primary with multiple windings in one direction then with the same wire in the circuit, reverse the direction of the coils with few windings when calling B & C coils secondary. Then as the A phase dies out, and B phase powers up, then the B core of many windings becomes primary and wire from the A circuit now has few winding in an opposite direction on top of the A and C core. Then we repeat the same circuit configuration when the C core is powering up. That is what defines a Zig-Zag transformer and it forces voltages beween phases to imitate each other.
But most Zig-Zag transformers have a typical problem. As the primary core normally approaches saturation, it cannot carry additional magnetic flux and will saturate just like it always does. But the increase in secondary voltage with smaller turns on the other two phases is always less than the primary one no matter what the proportional increase in the primary might ever be. This creates a condition of relative stability in both of the secondary coils and cores, since both B and C cores are charged equally they must therefore be identical. Keep in mind that we are not only going to wind this transformer as a Zig-Zag, we are also going to tune the coils to the capacitor as close as possible without allowing them to resonate. Why would we do this? And also because of the need for a good tank circuit, why not make both the zig-zag and reactor large so it has holding capacity? Well, we do.
Remember that all the cores and wires in all three are connected. By reverse winding the two secondary cores with the primary, our doughnut core has two polarities inside itself. At the A phase core the polarity is one way and inside the B and C cores are reversed in polarity. Like it or not, the iron core cannot saturate with opposite polarities inside of itself. No matter how high you drive the A phase, B and C will not allow A to go wherever it wants because the magnetic flux will not ever be one sided. Over the range of the secondary windings both may become saturated but that is not what matters in this circuit. It is the A phase that is dampening the A side of the capacitors, the B and C phases are not active. With both of the secondary cores active, this produces voltage stability between all three phases so long as the A phase is charged. Any attempted change in voltage between the three phases will be resisted because one phase is dictating a flat voltage between all of them whatever it is at that moment. But again, due to the magnetic shunt between the primary and secondary windings, that part of the core under the primary is not saturated and provides all the dampening available.
Now let us look at the fact that this is a three phase alternating circuit. Alternating current is sinusoidal and must change direction and return to the same place every 1/60th of a second. A phase is going to reverse momentarily and B then C phase will follow. As A phase is dying away B is coming up. As they pass each other, the B part of the transformer becomes the primary coil and A and C become secondary. Remember that all three phases are not connected electrically, but the iron cores are. Now as the phases pass they take control of their part of the core and flip polarity in the other two parts. This transition repeats every 360th of a second because each phase is running at 60 Hz but the reversal in magnetics is multiplied by three, and reversing just as fast on the other side of the sine wave adding three more changes to the process. This whole transition is now so fast that any variation in voltage between phases must find a common ground and remain steady. No matter what change in voltage each phase my want to move, the other two phases which are magnetically linked now will not allow it.
As a stable reactor we are safe to keep the tune close enough to give us the dampening need we are looking for, and the interrelationship between the reactor and the capacitors now will not and cannot wander. Any danger of a direct tune between the reactor and the capacitors is now impossible and can tune it very close to where maximum dampening effect can be enjoyed. In reality this is not complicated even if it seems so. The fact is, we have shown a very simple set of transformer principles working together. let us continue dealing with the rest of the Power Quality issues and needs:
3. Voltage Stability Excellent voltage regulation and stability now also forces current to fall in line. And a steady current will also reciprocate and would force voltage to remain flat. Steady current and voltage are very comforting to circuits of all kinds especially three phase motors. The result is increased efficiency and prolonged life. Other circuits that may not be three phase operate much better when the voltage and current are steady, especially sophisticated equipment that must remain accurate at all times. We have seen a lot of end users seek voltage stability with tap switchers. That method is better than nothing, but results in sudden and not smooth change transitions which are hard on equipment and seldom accurate. That idea reminds me of trying to thread a needle with snowmobile gloves on. Rough to accomplish well. The zig-zag keeps it clean and steady.
4. Capacitor Dampening Here is a reactor with a heavy core that cannot saturate and takes capacitor inrush and download in stride. With a safely held tuned reactor close to the caps, alterations in circuit fluctuations are held in check because the transformer refuses to allow offsetting voltages to fluctuate. If the Wattage draw is constant, and voltage is held in check, current rise and fall typical of charging and discharging caps is vastly reduced as well. Therefore the typical inrush and out charge spikes seen when caps must do that are heavily suppressed by a powerful and closely tuned reactor.
5. Surge Suppression For the very same reason, any sudden surge is resisted by the transformer as the magnetic mains resist all and any change in stability. We mentioned that our transformers are large. Now it matters. A heavy core coupled with large power capacitors creates a tank that has sufficient holding power to keep any short surge in check. Another fact is that small surge suppressors that are at most desks burned out many surges ago and are essentially now just power strips. The tiny circuits inside them are not capable of surviving multiple hits. We talked about long term surge suppression. The heavy combination of power caps and transformers automatically do it at the source so even small devices which burn out are themselves now protected against damage. We do guarantee voltage spike attenuation at a ratio of 250 to 1. Note is taken that all surges and sags can come from changes in either voltage OR current changes. One affects the other. By not allowing either to move, whatever the cause of the sag or surge is irrelevant. That is why voltage only surge suppression systems burn out. By dealing with current as well, no surge can pass this circuit –except one. ( See paragraph #12).
6. Ground Loop Correction This problem is created when too many grounds are so separated that they have different resistance between them. This differential can create a flow of power between them. This power flow becomes a totally isolated circuit where no circuit should be present as well as change the entire purpose for having a ground at all. By creating swift magnetic changes inside the zig-zag, the power flow in the normal circuit is no longer looking at changing grounds that oppose it. Therefore ground loops that might be created by variations between phases, now disappear.
7. Motor Slowdown Backwash Motors and generators are essentially the same thing. Motors convert power into kinetic energy and work. Generators made the same way use kinetic force to create power. When all motors are turned off, and their inertia keeps them going for a short time until they stop, that slowing down turns them into short time generators and makes power instead of using it. This backwash is never in harmony with the existing power in the line and always creates odd power which is disruptive in the circuit. This disruptive current and voltage can be wiped out within the induction coils of a zig-zag transformer by forcing the fast changing magnetic fields within it to force non compliance power to join in with harmony or be converted into magnetic flux energy while engaged in fast polarity exchanges .
8. Sags We are a bit interested in why so many worry about surges and buy voltage surge suppressors. When we talk about surges we are referring to voltage spikes. But voltage does not usually damage. It is current in amps that is destructive. We do not have breakers that trip when voltage changes. We do have breakers which go when current exceeds a set amount because that is what burns out wire. Any utility company will tell you that for every surge there are between six and more sags. Remember what steady power draw does? It means that watts remain the same. A 60 watt light bulb will draw 60 watts without regard as to what the voltage or the current may be. If voltage sags, then it is current that will be forced up to maintain the constant power draw. If the voltage sags enough, then the resulting current spike will burn out that light bulb or other power device - not the sag in voltage. But the reality is that current surge devices that are self correcting are far too expensive to put in a desk size power strip. Breakers and/or fuses take care of that problem but must be reset or replaced. And as we mentioned, that power strip surge suppressor burned out long ago because of the sag in voltage that spiked the current and destroyed it. So, along with voltage change resistance, this big circuit will also force current sags to remain very constant and within reason. Really big current changes are dealt with in other ways as we will discuss.
9. Harmonic Distortion What do zig-zag systems have to do with that? They don’t. But we saw something common and took advantage of it. Frequency notch filters are made from delta loaded capacitors and coils attached and tuned to the three terminal connections. Frequency notch filters are usually small enough to hold in your hand. We reasoned that we were already using capacitors to suppress Power Factor Error and dampening them with reactors. The filter parts were already there. By making the power capacitors in a delta loaded form, and tuning each reactor phase to the offending harmonic, now have a giant notch filter for a specific frequency. As we have several stages needed in keeping Power Factor steady, it is logical to merely make each stage a notch filter while also providing KVARS to correct Power Factor. Now we have made our caps multidimensional against the dominant harmonics which motors naturally create. (And only three of them make up 90% of that list) If we take a bite out of the dominant harmonics, we will reduce total distortion (THD) along the way. And it works amazingly well. But the biggest advantage comes in common integration. Remember the problem with multiple devices that were not designed to work together? By forcing systems to do multiple tasks, they internally do not react against themselves the way stand alone mitigators do which were not intended or designed to operate together.
10. Noise Suppression. Some time back we looked at where line noise came from. Most of it is created by a combination of dominant harmonic interaction, voltage and current variations, motor and capacitor inrush and outdrain, and switching that goes on in all busy circuits. By not directly attacking noise problems, if we eliminate most of what causes it, by natural selection much of it is reduced. The system we have when under test revealed a steady noise suppression of better than 60 DB. As many buildings may not have more than that anyway, that much suppression can eliminate all of it. Circuit noise is the major reason or cause of most data accuracy errors. Where precision is a must, noise can make that precision undependable. Some noise comes from harmonic disturbance. And some noise IS harmonic distortion. We will look at more of that in another area.
11. Line Interruptions and Power Breaks. Ride through is a major problem when everything in a building must work in harmony such as a refinery or assembly line. Any interruption in one place makes the whole plant go through a startup procedure to get back on line. We talked about the reason for heavy reactors and over size capacitors. Now here is the other reason for it. The power tank that forces voltage and current to sit still will also delay an interruption for 25 milliseconds or 1.5 cycles. The average power switchover that commonly occurs every three days in most utility systems is .71 cycles or only about 10 milliseconds. That is enough to force a refinery or other critical line integration conditions to go through a startup once every couple of months, or so. Some assembly lines may have to go though startup more often but 25 m/s is long enough that our circuit tank should keep even that from happening. We do not have an answer when the line break is longer. That is a full utility error and nothing much can be done to avoid that. However, we will say that the tank circuit will maintain 40% of the power for a full three cycles which is a full 1/5th of a second. In some cases that will keep a non critical plant on line without a startup. If voltage is critical, we guarantee that the tank will hold voltage to 99.7% for 3 to 8 milliseconds. This ability to clamp under line loss is what makes the transformer a ferroresonant type in holding on to current over time and not just a zig-zag, frequency choke, or even only a reactor.
12. Excessive Sag or Spike Suppression. Lightning strikes or substation power rebound however, will overwhelm our system. That is why we do analyze max power conditions and install MOVs large enough to save the plant. What will happen is the entire Power Conditioner will give its’ all instead. The strike will be taken internally but the critical equipment in the plant should survive. We have experienced that twice now with no damage at all to the end power user. Lightning rods are only marginal, and often run damaging power back through the ground to destroy sensitive equipment anyway. The only answer is to clamp the ground at the incoming location where our system is attached to the main panel. When power overwhelms the tank circuit, the MOVs then will shunt all the excess current internally to the tank circuit where it is delayed until the condition has passed. The same ability to allow for some ride through will contain damaging power long enough to only hurt the protection circuit. When the quick lightning strike is over, the containment circuit is gone but held it back long enough to save whatever it is protecting. Where extremely important equipment is at stake, this can be a very significant matter. If power is so severe that some damage is unavoidable, it is better to sacrifice lesser equipment instead of high value.
13. Short Circuit Protection Outside of the MOV, if output voltage drops to zero, we will limit output current to 200% of rated power. That is enough delay of sufficient back wash that any average breaker will trip. And even if they do not, this is enough load headroom to assure that no power transformer damage will occur. We have documented this claim many times in past installations.
14. Efficiency: Average to date for all we make is between 94% and 99% so long as all the components are in working order. Neglect for years will let anything fail. We have had equipment go without any replacement or service for seven years so far. But to ask a contactor that may have two million or more operations behind it to keep running on is not a reasonable expectation. It should be replaced. And for the money it saved the user, $300 in a few years is a bargain. Our intelligent switcher will keep track of exactly how many operations each part has had to endure. There is a way to ask it to tell you what that is. We built that feature into the computer it contains so that answer is available to you. Maintenance is a good idea for everything. Nobody makes things that last indefinitely, and taking care of it will extend the life of anything. It is all about economy. Does it pay for itself several times? If so then allow it to continue to do that job with good upkeep. Remember that in enough time and under enough stress every part will eventually fail. Keeping parts in good order as needed will give the same return indefinitely.
15. Volt/Amp Ratings: From 140 VA to 75 VA
16. Load reduction or Headroom This is an area which really does not need to be included but is worth calling attention to if not understood. All of the categories here are about power waste and loss. It is hoped that by reducing so much of what is inefficient, then the load has to go down unrelated to what is called for. A common example is when an operation is at maximum power availability. The transformer or main panel just cannot give out another amp. Instead of obtaining another larger transformer or installing a bigger panel, If we reduce the consumption by 30% through various means, the user now has 30% more room to add on more power equipment without having to upgrade or replace their present transformer or panel. I mention 30% for a reason. In nearly all factories or manufacturing facilities, a 30% headroom expansion is common for us to give to the customer who installs our systems.
17. Broadband Harmonics. When THD is beyond the ability of a few notch filters in the H-2 to H-12 range, then more harmonic reduction is needed. When sufficient amount of Harmonic Distortion is detected between 3 KHz and 100 KHz, we make an Intelligent Harmonic Filter to deal with any frequency in that category. When high frequencies exist in enough power, damage is inevitable. Look for an entire white paper dealing with just this problem and how we deal with it.
Written by John Jackman, Principle Engineer Power Control Company November 2009