Instrumentation for Home Power Systems Richard Perez When we make our own electricity we are our own power company. We are our own production crew, our own energy auditor, and our own trouble shooter. Instruments are our eyes into the electric world of our power systems. Without accurate instruments we are flying blind. While instrumentation is not necessary for the system to work, it greatly helps us operate our systems. And when things don't go right, instrumentation is essential for finding out what and where the problem is. System Measurement If we don't know how our system is performing, then we cannot effectively use the system. We are roughly in the same position as a U Boat captain, we must make operational decisions based on the state of charge of our batteries. If the batteries are full and the power source producing, then we are wasting power in the only way possible in an RE systemŠ by not using it. If the batteries are empty, then we need to ease off on power consumption, for example, wait a while before running the washing machine. So battery state of charge is the first and most important bit of info we need. Monitoring critical system points gives us an at a glance check of major component performance. We don't really need to continually know this data everywhere in our system, just at critical points. The best places to make these mostly voltage and current measurements are on power producers (PV, hydro, wind, or ?), power storage (batteries) and power processing devices (inverters and controls). Before we measure anything, we need a meter. It could be an analog meter (you know, the older types with a dial and pointer) or a more modern type that displays numbers on a digital display. The instrument may be set up to perform many types of measurements or it may be optimized to perform only one type of measurement. How accurate does the instrument need to be? As accurate as you can afford. In measurement, accuracy is the name of the game. A Good DMM vs. Discrete Instruments A Digital MultiMeter (DMM) can make a variety of measurements. The DMM will measure voltage, current, and resistance. Some DMMs will also measure frequency, duty-cycle, capacitance, test semiconductors, and record data in their memory. There are hundreds of types of these meters on the market. We use two Fluke types at Home Power. The Fluke 77 is rugged, accurate (0.1% on DC), and inexpensive (Å$140). The Fluke 87 has all the features mentioned above for about $280. These are highly accurate, capable, and reliable instruments. If you are seriously interested in instrumentation, a DMM of this caliber should be your first purchase. It will be the standard with which you will build other specialized, dedicated instruments. There are less expensive DMMs than the Flukes mentioned here. You will get what you pay for. The higher quality DMMs are more accurate, have many more features, last many times longer, and are very rugged. See HP#15, page 41 for a technical report on the Fluke 87 DMM Discrete analog meters are inexpensive and usually optimized to perform a single function, like being a battery voltmeter. The accuracy of analog meters can vary from "strictly ballpark" (as bad as ±25%) to very accurate (Å1%). Analog meters are inexpensive ($3 to $40) and easily available in the surplus market. They are powered by the circuit under test and generally require no on board batteries. They are extremely easy to tweak into accurate, dedicated meters for virtually any measurement. So the choice of instruments is up to you. Let your inclination and bank account be your guide. Battery Measurements The battery is the heart of the system. The battery is the head of the hit parade. The battery is the Numero Uno, first, last, and most essential component subjected to continuous scrutiny. The best single instrument for operating batteries is an dedicated battery Ampere- hour meter. Period. If you don't really care about fully instrumenting your system and want only a single instrument, then get a battery Ampere-hour meter. Other very useful battery instruments include a dedicated battery voltmeter and a bi-directional ammeter that measures net current into and out of the battery. We use both and find them very informative for at a glance checks on system performance. Battery Ampere-hour measurement The constant question in any battery based system is, "How full is the battery?" The easiest to understand and most accurate method uses a digital Ampere-hour meter. It's a "gas gauge" for all types of batteries, both lead-acid and nickel-cadmium. These instruments not only work well, but their information is direct and understandable by even the most nontechnical battery user. Ampere-hour Measurement There are many ways to measure a battery's State of Charge (SOC). In lead-acid cells, you can measure the specific gravity of the electrolyte with a hydrometer. But this is inaccurate as it depends on temperature, and risks contamination of the cell. In nicads, specific gravity of the electrolyte is meaningless for determination of SOC as it doesn't change with the cells' state of charge. We can use a voltmeter to determine SOC. But this is not very accurate, and varies with the battery's temperature. Measuring SOC by voltage is also dependent on the current flow through the battery. If the battery is under charge, then the voltage is higher. If the battery is under discharge, then the voltage is lower. And after you have compensated the voltage measurement for current and temperature, then you must still consult a SOC versus Voltage chart to accurately determine the battery's State of Charge. Sound confusing? Well, it is. And all this confusion is cleared up by an Ampere-hour meter. Ampere-hour measurement is the best way to determine battery state of charge. The measurement doesn't depend on temperature, cell type, and doesn't care whether the battery is being charged or discharged. The Ampere-hour meter provides a digital readout of exactly how many Ampere-hours have been withdrawn from the batteries. Ampere-hour meters come in two types. Some are totalizing types that continually add up the Ampere-hours flowing in a single direction, say the yearly current production of a PV array. Other Ampere-hour meters are optimized as Battery SOC meters, and are bi-directional, net reading meters. They measure and count current flow to and from the battery. The Ampere-hour Meter as a "Gas Gauge" for Batteries The battery Ampere-hour meter is installed on a fully charged battery. At this point the digital display will read zero (0). This makes sense since the battery is full and we haven't yet withdrawn any power from it. As the battery is discharged, the digital display counts the Ampere- hours withdrawn from the battery. For example, say our battery is full in the afternoon and during the night we withdraw 40 Ampere-hours. In the morning, the Ampere-hour meter will read -40 (that's MINUS 40) to indicate that we've withdrawn 40 Ampere-hours from the full battery. As the Sun comes up and the PV array (or any other power source) starts recharging the battery, the Ampere-hour meter begins counting up (from -40, it counts to -39, -38, -37, etc.) to zero as the battery refills. When the battery is full, the meter again reads zero. At that point any additional recharging of the battery is read as positive numbers on the display. For example, after the battery is full, if we put 20 Ampere-hours more through it, then the display will read 20 as "overcharge Ampere-hours". After charging stops, the meter resets itself to zero regardless of the number of overcharge Ampere-hours. This makes sense since overcharge Ampere-hours can not stored by the battery because it is already full. For "Things that Work!" tests of two battery Ampere-hour meters see HP#16, page 40 for a review of Cruising Equipment's meter. See HP#20, page 40 for a "Things that Work!" review of the Ample Power's Ampere-hour meter. The Cruising Equipment model is a straight Amp-hour meter for about $200, while the Ample Power version is also a battery voltmeter and battery ammeter for about $300. Battery Voltage Let face it, although battery Ampere-hour meters are the best tool for the job, they are also expensive. Many of us still use, and have used for years, battery voltmeters to aid us in determining battery state of charge. If you own a DMM, then you can use this instrument to measure battery voltage. Most of us, even those with several DMMs, still like to have a dedicated battery voltmeter on line all the time. You can buy these as digital meters for around $50, as analog meters for about $20, or you can make a very accurate analog model as follows. The Expanded Scale Analog Battery Voltmeter The idea here is to use an analog dc milliammeter in a circuit that will accurately measure the batteries voltage. This circuit produces an expanded scale voltmeter. Most analog voltmeters start reading a 0 volts. This is really a waste for battery systems as a lead acid battery will have about 10 to 11 volts (20 to 22 VDC in a 24 VDC system) even when just about empty. So the portion of the meter's scale between 0 and 10 volts is never used. Wasting this portion of the meter's scale decreases its resolution and thereby the accuracy of the meter. This circuit allows the meter to start reading at 11 volts and to display full scale at 16 volts (a very fully charged 12 Volt battery while still under charge). The 24 VDC version starts reading at 22 VDC and displays full scale at 32 VDC. This is called an expanded scale, and makes the meter much more accurate to use. All the components for this metering project are available at most Radio Shack stores, or from just about any electronics supply house. Cost of the parts should be between $15. and $40., depending on your hardware sources. Construction time is about 1 hour for an experienced assembler. This circuit is powered by the battery under measurement, and never requires the use of small batteries to power the meter. We don't have space here to give an electronics primer for those not familiar with electronic construction. What I do offer is the schematic for the circuit. If you can't figure out how to build this meter from the schematic, then please seek out an electronics person who can aid you. INSERT VOLTMETER SCHEMATIC Electronic Nitty-Gritty This circuit uses a 1 mA. DC Ammeter as an expanded scale voltmeter. The meter has its ground elevated to 11 Volts (22 Volts in a 24 VDC system) by the use of an LM 723 voltage regulator in shunt mode. This makes the meter very accurate as there are no series semiconductors in the measurement circuit. Full scale reading and the 11 Volt (or 22 Volt) ground level are both adjustable by using the potentiometers in the circuit. R1 is the adjustment for the shunt regulator. Adjust R1 until Test Point 1 (TP1) is at 11 Volts (22 Volts in a 24 Volt system). Then adjust R2 until the meter reads the battery's voltage at the time. Use an accurate DMM to calibrate this circuit. Average power consumption of this meter is about 5 milliWatts. When on line 24 hours a day, power consumption is less than 0.1 Watt- hours per day. This meter is superŠefficient and can be left on line all the time with a minimum of power consumption. We've had one on line since 1976. Battery Current A bi-directional Ammeter is a great instrument to have on a battery. The instrument measures the flow of current either into or out of the battery. Since the ammeter is in series with the battery, the ammeter must have low insertion loss. Every amperage measurement scheme has some electrical resistance. At the high currents commonly found in battery systems, the insertion resistance must be low (less than 0.001 ½). For example, an inverter starting a big electric motor may require over 800 Amperes of current from the battery. All this current must pass through the ammeter. Shunts In high current situations, use a shunt for measuring battery current. Shunts are very low resistance, precision resistors designed specifically for current measurement. Shunts are relatively inexpensive ($10 to $40), accurate (0.1%), and can handle large currents (10 A. to >1000 A.). Shunts are used for current measurement by every Ampere-hour meter and most ammeters. If you can live with a accuracy losses of less 10%, then you can even use the copper wiring in your system as shunts. It all works by the magic of Ohm's Law. In Theory Ohm's law informs us that any electrical current flowing through a material (like a piece of wire or a shunt) suffers a loss in voltage. This voltage drop across the material is due to its resistance and the movement of the electrons (current) through that material. The amount of current flowing through the material can be determined if we know two things. One, the voltage loss across the material, and Two, the resistance of the material. Or in algebraic terms using Ohm's Law: I=E/R (Equation1) where I= the amount of current in Amperes E= the voltage drop in Volts R= the material's resistance in Ohms Well, every appliance, power converter, power source and whatever is wired into the system with copper wire. The wiring in necessary to move electrical current from place to place, from source to load, etc. If we consider these bits of wire as resistors, then we can use the amount of voltage loss across a wire to determine the amount of current flowing through the wire. Wire used in such a fashion is called a "shunt" in electronics jargon. How it Works All we need to perform current measurements on our batteries, PV panels, inverters, refrigerators, or any other device that consumes, stores, produces, or converts electricity is a Digital MultiMeter (DMM) and the already existing wire in our systems. And a little help from Ohm's Law. The DMM is used to measure the voltage drop across a piece of wire carrying current. The DMM should be capable of making measurements in the millivolt DC range. Such resolution is necessary as this technique involves using lengths of wire with resistances from 0.1½ to 0.0001½. The resultant voltage drops across such small resistances will be very low, and we'll need a DMM that can make accurate measurements in the milliVolt range. We also need to know, as accurately as possible, the resistance of the piece of copper wire we are using. To find this resistance first determine the wire's size or gauge. Most wire has its gauge number printed on its insulation. Or the wire's gauge can be determined by using a wire gauge measuring tool. Once the gauge number is known, then measure the length of the wire. Copper wire has its resistance, in Ohms per foot, specified by gauge number. Once we know the gauge, we can look up the resistance (½/ft) on a Copper Wire Table. This value is multiplied by the number of feet of wire we are using to make the measurement. And the result is the resistance of that particular piece of copper wire or shunt. This technique can be used on wire of any size, and of any length. There are certain resistance values for shunts that have distinct advantages. Consider the following resistances: 0.1½, 0.01½, 0.001½, and 0.0001½. If these values are used for R in Equation 1, then we are performing division by a decimal fraction of 1. This means that the measurement taken by the DMM can be read directly and a calculator is not needed to perform the math. Only the decimal point of the reading of the DMM need be shifted to obtain the amperage measurement. What follows below is a Copper Wire Table that is optimized to display the lengths of various gauges that have resistances from 0.1½ to 0.0001½. Find the wire gauge size of the wire you are using, and the lengths necessary to produce the shunts are shown across the table. Measure the indicated length along your wire and you have a shunt with a resistance that is a decimal fraction of 1. Attach the leads of the DMM across this length and you're ready to make current measurements. At the head of each shunt column on the table, there is a reminder to shift the decimal point on the mV. reading taken from the DMM. For example, let's consider a 12 VDC light hooked up with 12 gauge wire. From the shunt table, we see that 0.63 feet of this 12 gauge wire will give us a shunt of 0.001½. The heading of the column tells us that the milliVolt (mV.) reading on the meter will equal the amperes of current through the shunt. If we measure 4.2 mV. across this 0.001½ shunt, then the current flowing the shunt (and the light) the light is 4.2 Amperes. If the shunt had a resistance of 0.01 ½ (as in 6.3 feet of 12 ga.), the the milliVolt reading on the DMM would be 42.0 mV. and would have to be divided by 10 to produce the correct amperage measurement of 4.2 Amperes. INSERT COPPER WIRE SHUNT TABLE The schematic shown below shows the electrical setup for using shunts to measure current. The measurement can be taken in the positive or negative wire, it doesn't make any difference. I've made switch panels to measure current in different places by soldering small (20 gauge) "sense" wires to the shunts (either commercially made high accuracy shunts, or just plain ole' copper wire ones) and running these smaller wires to a panel with a rotary switch. The DMM is connected to the output of the switch which selects the different shunts. We don't have to cut the wire to make a shunt. Simply make the length measurement, strip back the insulation at the shunt's length, and solder on the sense wires. In places where you don't need to make measurements often, use needle probes on the DMM to pierce the insulation without stripping. A piece of string is useful to transfer length measurements from a tape to stiff pieces of nonstraight wire and cable. INSERT SHUNT SCHEMATIC Where to Use Copper Wire Shunts Use this technique any place you wish to measure current. Here are some suggestions. On the main wires delivering current from PV arrays to the batteries. On the wires that supply current to an inverter (this is a great place for a 0.0001½ shunt made out of 2.04 feet of 0000 gauge copper cable). On the wires that connect the battery pack to the bus. And on any appliance whose current consumption needs to be measured. Advantages There are all kinds of advantages in using this technique. The wiring that we are using to make the measurement already exists to move the power to or from the device. The measurement process doesn't introduce any new losses as the shunt wiring is already there. The wiring need not be cut as in the insertion of an in-line meter. Shunts can be made with very low resistances, thus enabling high current measurements with minimum loss. The technique can be used with minimum trouble and no expense for occasional measurements than don't require a dedicated in-line ammeter. Disadvantages The big disadvantage is inaccuracy due to the copper wire changing resistance as it heats or cools. The information on the Copper Wire Shunt Table is correct for copper wire at 68”F. (20”C.). For copper wire at 32”F. (0”C.), this method will yield amperage measurements that are low by about 10%. At a wire temperature of 122”F. (50”C.), this method yields amperage measurements that are high by 10%. If you compensate for the temperature of the wire, this technique can be made more accurate. For high accuracy, use a factory made shunt instead of the copper wire shunt. Commercial shunts are made from materials whose resistance varies only slightly with temperature. Commercial shunts are also calibrated and tested for accuracy. Nerd Stuff-- Equation City The data on the Shunt Table was calculated from an equation written by the Wizard. While browsing through the Copper Wire Table one afternoon, he noticed this simple exponential relationship between wire gauge size number and the resistance of that sized wire. What follows here is a generalized equation that yields amperage through a shunt of any length and gauge of copper wire. This equation is also compensated for temperature. INSERT EQUATION where: I= current through the copper wire shunt in Amperes (A.) Lm= length of the shunt in meters (m.) mV= voltage drop across the shunt in milliVolts (mV.) Tc= temperature of the shunt wire in degrees Centigrade (”C.) N= the wire gauge size number (B&S American Standard). Note: use the following integers for these gauge sizes: for 0000 use -3, for 000 use -2, for 00 use -1, and for 0 use 0. In all other cases use the wire gauge number directly. This equation works for gauge numbers between 0000 and 40, even fractional gauges. I included this math data for techies with both DMMs and a computer (what a truly frightening combination). Inverter measurements In many of our systems all the power consumed passes through the inverter. Since the inverter plays such a crucial role, it is a very good place to make provisions for measurement. In some cases you may wish to use a dedicated meter. For example, I use an old surplus ($2) Weston analog ac voltmeter to constantly measure our inverter's voltage output. Inverter Input Current This is a place for a shunt. My favorite for this location is a Deltech model (50 milliVolts at 500 Amperes with surge to 1,000 Amperes). This shunt is massive and has very insertion loss (0.0001½), cost about $25 and has massive terminals with large bolts. If you don't need the accuracy of a precision shunt, then use the copper cable that connects the inverter to the battery as a shunt. Since this cable is in place anyway, measure off a shunt length anywhere along one of the inverter's cables, nick the insulation, solder the sense wires, and measure away! Inverter Output Voltage Inverters produce nonstandard ac waveform. Only certain types of instruments can accurately measure modified sine-wave inverter voltage or current output. These instruments will be specified to measure "True RMS voltage". For example the Fluke 87 will accurate measure an inverter's output. The Fluke 77 will not, and so won't most every DMM costing less than $200. We are interested in not only the RMS voltage of the inverter (should about 117 vac ±5%), but also the peak voltage of the inverter's waveform. Peak voltages should be within 15% of ±164 vac, and the plus and minus peaks should measure within two volts of each other. Oddly enough older analog meter will measure inverter output voltage with fair accuracy. For example, the old Weston iron-vane ac voltmeter I mentioned above works OK (±3%). These meters are basically too slow and stupid to notice the fine differences between sinusoidal ac (like from a utility) and modified sine-wave power (from an inverter). These older analog types are usually not phantom loads and will allow the inverter to go to "sleep" at night. Inverter Output Current The same rap about True RMS reading instruments applies to accurate measurement of the inverter's current output (ac amperage) also. A precision shunt or a copper wire shunt are used for this purpose with a True RMS DMM. Also analog ac meter movements are fairly accurate when used to measure inverter output current. Power Source Current We use a 50 milliVolt, 50 Ampere precision shunt inline with our PV array. This shunt allows us to measure the current output of our array with a DMM. This same shunt also functions with our Thomson & Howe recording Ampere-hour meter. This Ampere-hour meter has been on line for years now and totalizes the output of our array. We record its data daily in a notebook. This data has allowed us to very accurately measure long term PV performance in our neighborhood. See HP21, page 39 for a "Things that Work!" report on this recording Ampere-hour meter. Other power sources like hydros and wind turbines can also have current instrumentation. Once again, use a precision shunt or at least copper wire shunt. After all, the wire has got to be there anyway, and it's simple to use it as shunt. My experience with RE power sources shows me that operators are always curious about how much current is being produced. Are we done yet? No, when we get hooked on measurement, we're never done. From where I sit typing this in, I can see five digital meters and twelve dedicated analog meters. And I have three more meters waiting to be installed. I'm so badly hooked on data that I even write down the measurements in daily logs and them compile the data yearly on the computer. I obviously have an advanced case of data dementia. I am not suggesting that all home power producers get involved this deeply (although it fun and educational). Better use of our power comes from understanding and knowledge. Instrumentation gives us ability to measure the invisible electron's activity. Measurement is the key to our understanding and use of power. Access Author: Richard Perez, C/O Home Power, POB 130, Hornbrook, CA 96044 „ 916-475-3179. DMMs: John Fluke Mfg. Co., POB 9090, Everett, WA 98206 „ 800- 443-5853. Surplus Analog Meters: Fair Radio Sales. POB 1105, Lima, OH 45802 „ 419-223-2196. Also, C and H Sales Co., 2176 E. Colorado Blvd. Pasadena, CA 91107 „ 800-325-9465 or 213-681-4925. Shunts: Deltech, 13065-H Tom White Way, Norwalk, CA 90650 „ 213- 926-2304.