We don't know of any technical subject that can be learned ``on the fly'' without making a lot of avoidable mistakes along the way. Most certainly the things we know most about, electronic circuit design, computer programming, and battery systems, required some study and practice before we tackled a real project.
For some reason, people in general tend to underestimate the complexity of a lowly battery system. Over the years, we accepted the fact that most people purchasing Ample Power products will not be fully aware of how complex such systems are. And, we realize that most people want to be users, not technical experts in the secrets of battery operation.
While we can forgive the mistakes of the uninformed user, it's much more difficult to overlook people writing about batteries who are uneducated about their subject, and worse yet, pretend to be knowledgeable. And yes, we have to suffer through articles written in many publications about products from the magazine's major advertisers which are little more than advertising hype in a different format. But it's most disconcerting when a publication which accepts no advertising presents information that is of poor quality and doesn't even come close to reporting meaningful information.
The Regulator Report
Practical Sailor1 published a report on alternator regulators in their February 1, 1997 issue. The first issue of PS Reports introduced a contest regarding the regulator article, and offered a free Next Step Regulator to the first person who found what we called a spectacular blowout ...a major blunder. From the responses, we discovered that very few people know how to charge a battery properly, including many who are selling Ample Power products.
We can't rebut the regulator report on a point by point basis, because it totally misses the point of proper battery charging. So bear with us through the following material ...and carry this issue of PS Reports with you for as long as you own battery systems and review it as needed!
Introducing Figure 1
Figure 1, (shown at the end), is a graph showing voltage and current as a battery is being charged. The battery is an 8D of about 200 Amp-hours. This is the eighth time the battery has been charged ...seven times previously, the battery has been discharged to 10.5 Volts using about 9 Amps of current. For more information about breaking in new and old batteries in this manner, refer to the Ample Power Primer.
The data for the graph was obtained using an Ample Power monitor with a connection to a PC. Close to 29,000 samples were made of voltage, current, and battery temperature and logged to the disk drive during the charge process. Post processing took 256 evenly spaced data samples and imported them into a graphing program to make the graph shown in Figure 1. Why 256? The graphing program isn't able to use more than 256 samples.
The monitor used for voltage and current measurements was checked against a high-end Fluke digital meter prior to the logging session. The charger used for this test was the original prototype of the Ample Power Smart Charger ...parts screwed and glued to a piece of plywood ... not pretty but effective, as you'll see. The test took place during the winter, with the battery on a cold concrete floor and other equipment spread around a table in the warehouse.
Interpreting Figure 1
Making a double Y-scale graph in only black and white can lead to interpretation problems ...which is which? The voltage scale is on the left Y-axis, (vertical dimension). Voltage plotting starts at 12.25 and quickly reaches close to 13.50 Volts. The line that starts there and can be traced to the right where it eventually goes slightly above 14.50 Volts represent the series of voltage measurements taken during the charge process.
The Y-scale on the right represents battery current during the charge. The line showing Amps starts on the left-top of the graph at about 35 Amps, and stays fairly constant until 321 minutes have passed. Minutes of charge are shown on the X-axis, (horizontal dimension), at the bottom of the graph.
Figure 1 shows the complete charge cycle from beginning to end. A lot of information is packed into a small space. To make interpretation more meaningful, we've used measurements from the beginning and end of the charge process to make Figures 2 and 3.
Interpreting Figure 2
To make Figure 2, we extracted 256 evenly spaced data samples over the first 126 minutes and imported them into the graphing program. The very first voltage measurement was just a little over 12.30 Volts as battery current begins to flow. As shown, current quickly rises to 35 Amps, and stays there for the duration of the graph.
The most interesting part of Figure 2 is the voltage plot. Note how the voltage quickly climbs close to 13.50 and then begins to decline! We've fielded more than one phone call from users who think our Next Step Regulator has failed because they continue to run their engine and battery voltage keeps going down. As shown in the graph, it takes more than an hour before voltage begins to rise again. Just what's going on here?
Recall that the battery had been previously discharged to 10.5 Volts using about 9 Amps ...the battery was 100% discharged! The monitor used to record these measurements was programmed to dischage the battery to 10.5 Volts, disconnect the discharge resistor, let the battery rest for 8 hours, and then turn on the charger ...our crude version of a battery life cycle tester. As the battery rested, the sulfate that formed during the discharge began to harden into crystals. The internal resistance of the battery increased as this process developed.
Observe that the charger is current limited to about 35 Amps. In a charger, current limiting is done to prevent excessive currents through the transformer and solid-state control devices. An alternator is inherently current limited. That is, an alternator which is not connected to a battery can be shorted between positive and negative, and the current produced will be the maximum rating of the alternator. All charge sources are current limited, including solar panels and wind generators. This is an important point to remember because it explains why a charge source can't drive battery voltage to an arbitrarily high setpoint immediately.
Once charging commenced, battery resistance was high, and so the voltage across the batteries rose as necessary to force 35 Amps to flow. By using Ohm's law, we can even calculate the resistance of the battery. Divide 13.5 Volts by 35 Amps to derive 0.39 Ohms of effective battery resistance.
As the ionic activity gets in motion, battery resistance begins to decline. After 10 minutes, the decline is obvious. Since current is limited to a constant 35 Amps, as battery resistance decreases, the voltage across the battery must also decrease. Note how long battery voltage continues to decline ...for about an hour! We can certainly understand why a typical user would wonder if their regulator was working. We can be sure it is, however, because we can observe battery current as well as voltage.
At about 70 minutes into the charge, battery voltage begins a slow ascent. The middle part of the charge is pretty boring with a constant current, and not much happening with the voltage. The end of the charge is where things begin to happen!
Interpreting Figure 3
To make Figure 3, we extracted an evenly spaced 256 samples from the end of the roughly 29,000 samples in the log. As shown, these 256 points cover from 300 minutes to 420 minutes of the charge process.
At 300 minutes into the charge, voltage is rising rapidly. Current is still hanging in at about 35 Amps. At 320 minutes, something dramatic happens! The absorption voltage setpoint of the charger is reached at slightly over 14.5 Volts. Current begins to decline quite steeply. This is the beginning of the absorption cycle.
What has happened? The charger has reached a predetermined absorption voltage setpoint, and will not allow the output voltage to go any higher. The battery is reaching a full charge, and as it's internal resistance rises, current must decline with a constant voltage applied to the battery ...Ohm's law works!
Before continuing, let's answer the question of why the absorption voltage is slightly above 14.5 Volts. Not shown on the graphs is the battery temperature. The actual absorption setpoint was previously set for 14.4 Volts at a temperature of 77 F. Battery temperature at 320 minutes is in the mid-60s, (F). Note that as the charging continues, the voltage across the batteries declines. It does so because the battery temperature is rising, and the charger is temperature compensated. Once a battery nears a full charge, more and more of the current through it goes into heat instead of restoration of capacity. If voltage does not decline as temperature rises, the battery may go into thermal runaway ...an increasing common occurence as more people are using multi-step alternator regulators without temperature compensation. More on this subject can be found in the Ample Power Primer.
From prior discharge tests, we know that the 8D being charged is about 200 Amp-hours in capacity. In previous publications, including Living on 12 Volts with Ample Power, Wiring 12 Volts for Ample Power, Power News, and the Ample Power Primer we've suggested that the absorption cycle be terminated when the battery current falls below 5% for engine charging, and 2.5% for battery charger operation. Those points are illustrated in Figure 3.
Note that the 5% current point for 200 Ah of capacity, (10 Amps), is reached at 48 minutes after the start of the absorption cycle at 320 minutes. It takes 73 minutes to reach the 2.5% point. So how long should the absorption cycle be? Somewhere in the vicinity of 48 to 73 minutes in this actual test, depending on whether you're running an engine to charge, or are hooked up to shore power. Those familliar with the Ample Power Smart Alternator Regulator will recall that absorption time was 30-60 minutes depending on how long it took to reach the absorption setpoint. Forty-eight minutes is darn close to the center of that time!
Harvesting the Data Mine
Graphs are great to look at, but too often people just read the obvious plots and fail to glean information lurking behind the lines. Let's take a look at some of the things we can deduce from the graphs.
Notice in Figure 3 how current declines quickly during the first 48 minutes of the absorption period which starts at 320 minutes. The rate of change in current is definitely slowing down at 48 minutes, and the rate of change continues to slow as time marches on. Do we even need to know how big a battery is being charged to determine when it is full? NO! If we have been observant, and we know battery current, then the rate of change in battery current is sufficient to determine when to stop charging at the absorption voltage. With nothing but voltage and current, with a history of current over time, we can see that the absorption voltage should be terminated. Forget Amp-hours during charge! Volts and Amps is sufficient.
So how many Amp-hours went back into the battery. For 320 minutes, battery current hovered just above 35 Amps ...35.7 Amps was the highest reading in the log. Current was just below 35 Amps prior to our 320 minute marker. Using 35 Amps as average, for 320 minutes, about 187 Amp-hours went back into the battery. Not all of this was effective due to charge efficiency.
After 320 minutes, current is no longer constant. To avoid a lesson in exponential arithmetic, let's cheat a little bit and call the current decline a straight line between 35 Amps at 320 minutes and 10 Amps at 368 minutes. We can estimate the average current between those time markers at 1/2 the difference plus 10. In other words the difference between 35 and 10 is 25 Amps, 1/2 of that is 12.5, added to 10 yields 22.5 Amps. If you put a ruler on the current line at 320 and 368 minutes, you'll see that average current is shy of a straight line. If you want more accuracy, you can take shorter and shorter segments of straight lines and compute the average currents for each of the segments and add them up ...in the process, you'll have discovered the fundamental theorem of calculus!
For our purposes, lets just use 20 Amps as a ballpark average current between 320 and 368 minutes. During that time, then, an additional 16 Amp-hours is put into the battery, for a total of 203. As you can see, adding more Amp-hours as time goes on is becoming increasingly difficult. Running an engine for another half-hour would only add about another 3 Amp-hours ...why bother?
With approximately a 200 Amp-hour battery, the charge current of 35 Amps is 17-18% of capacity. In the past, we've suggested alternators that are rated 25-40% of battery capacity. Suppose you have 440 Ah of batteries and an alternator rated for 100-110 Amps. The alternator may only produce about 75-80 Amps at reasonable engine RPM ...or 17-18% of battery capacity. With that kind of ratio, you should expect to see the same kind of charge profile shown in Figures 1-3, assuming that the regulator and batteries were working a well as shown here.
Synopsis of Charge Process
Here's what we should be able to conclude about the charge process.
The Regulator Test Equipment
So what does the February 1, 1997 issue of Practical Sailor have to do with all of the foregoing? Not much actually, but let's review their report. Details are scant, and we have to read between the lines to glean much out of the report, so hang in there. The apparent intent of their article was to evaluate charging performance of three alternator regulators. From the article, we can see that they were interested in knowing how full the regulators managed to charge. How well they acheived their apparent goal will soon be clear.
Practical Sailor ``used four Trojan golf-cart batteries wired in series for 12 volts.'' We know, or will at least give them credit for meaning to say, wired in series-parallel, so those who responded with this line as a major blunder won't be winning the free Next Step Regulator. Practical Sailor goes on to state, ``these were new batteries with 220 amp hours of capacity ...''. We'll also give them credit for meaning 220 Ah each, for a total of 440 Ah at 12 Volts for the complete bank, and again, finding this error will not be sufficient to win the Next Step Regulator.
They also used a Powerline alternator rated at 100 Amps. We don't know if that's the hot or cold rating, or some number in between, but this glaring lack of detail isn't a major concern, although later we're going to wonder why the alternator apparently performed so poorly.
The test procedure also used a ``Cruising Equipment Amp-Hour Meter''. Now we know from above that we don't need to know anything about Amp-hours consumed, or even Ah capacity of a battery to decide when it is full. Perhaps some of you can see the smoking gun that the editor of Practical Sailor used to shoot himself in the foot.
The Balmar ARSII Regulator
The ARSII was the first regulator tested. You should, of course, review the full anecdotal report on the regulator. Some pointed out to us that Practical Sailor said, ``190 amps had been discharged from the batteries''. We know they meant Amp-hours, but we're not sure they know the difference. (No, winning the Next Step Regulator isn't as simple as pointing out this error.)
We're going to present the ARSII test in a tabular view of what we think it says. In the table, NG means not given. The way the information is anecdotally presented in the original report leaves much to be desired ...we're too accustomed to seeing data presented in tables, charts and graphs which allow a whole sense of things to be grasped quickly.
We're a little confused about the data at the 190 minute mark. It appears that 20 Amps was recorded about that time, then the regulator hit the 14.1 absorption setpoint and was 17 Amps. Forty-five minutes later, the regulator tripped to float. It appears that battery Amps was down to 4 Amps at the time, although we can't be sure that 4 Amps isn't the current after going to float ...more on this subject later.
The Alpha Regulator
Details of the tests on the Alpha Regulator from Cruising Equipment are even more sparse, although Practical Sailor does say, ``similiar to others in the bulk stage ....''
At somewhere around 3 hours the Alpha went into the absorption cycle and current was logged at 17.2 Amps. The test was terminated one hour later although the Alpha regulator has a factory setting of 2.5 hours for the absorption time. Apparently they thought that the regulator after one hour wasn't going anywhere so what was the point of continuing. As you know from the graphs presented earlier, it was a good decision to stop charging one hour into the absorption cycle, even if done for the wrong reasons.
The Aqualine Regulator
Practical Sailor found quite a few features of the Aqualine Reglator interesting. Anyone reading the data sheet for the Ample Power Smart Alternator Regulator, introduced in 1992, and the data sheet of the Aqualine regulator announced in 1993 and available at least a year later would be astonished at the similiarity of features. One might even draw the conclusion that Aqualine chose to knock-off the best example of proper battery charging that existed then.
Writing data sheets and designing, and improving regulator performance isn't quite the same ballgame, however. How did the Aqualine regulator fare in the Practical Sailor test? (Yes, we'll ignore the fact that Practical Sailor continues to say Amps when they should be saying Amp-hours.)
It appears that the Aqualine regulator started the timed absorption period at the 180 minute mark. From the anecdotal report, we think it must have gone to float at the 210 mark. According to Practical Sailor, the Aqualine voltage and amperages were similiar to the other regulators.
So Where's the Beef?
The things that don't add up in the Practical Sailor report are so numerous that it's hard to know where to start. Certainly, all the NG columns in the tables of information we've constructed from the article should make you pause and wonder about how thoroughly the tests were documented.
Practical Sailor initially made the decision to leave the regulators with factory settings instead of using manufacturer's instructions to set the regulators for the types of batteries being charged. Since some manufacturers choose to set their units for gel batteries, while others set them for liquid cells, it should only be obvious that it makes a difference, and that any test with unadjusted regulators is meaningless to begin with.
If you review the graphs presented earlier in Figure 1-3, you'll note that the current is almost exactly constant throughout the bulk charge cycle. Unexplained by Practical Sailor is the great reduction in battery current starting as early as 10 minutes into the charge cycle ...see the Table showing data for the ARSII. We'd expect the alternator to heat up, which would reduce its ability to generate current, but alternator heating can't explain the low battery currents recorded just prior to reaching the absorption setpoint. In fact, the tapering current can only be explained by the regulator circuits ...it appears all regulators have a similiar problem, which is not unusual for alternator regulators in general.
As stated by Practical Sailor, the batteries were new. Whether they were new-new, or had a few cycles isn't apparent. Unless those batteries have been broken in according to the guidelines presented in the Ample Power Primer, we would expect a low acceptance rate. With 440 Amp-hours of nominal capacity, the 63 Amps that the ARSII regulator maintained for about an hour is only 14% of AH capacity. Most certainly the batteries weren't being pushed very hard. Recall that in the test for the graphs presented above, charge current was 17-18% of capacity in Ah.
With the report being substantially anecdotal, it was somewhat difficult to determine precise times where absorption and float cycles commenced, and the data at those points is not as clear as we would hope. From the graphs presented earlier, however, we expect a sharp point where voltage ceases to rise and current immediately starts to decline. After another 48 minutes, we expect to be substantially charged. It certainly doesn't look like Practical Sailor came even close to that, but we'll have more to say about that later.
Gleaning the Report
As we've noted, Amp-hours need not be known for proper charging, and the use of an Amp-hour meter for prime instrumentation of a regulator charge test is a boner. Let's look at data from the ARSII test. Interpreting anything reported by Practical Sailor is risky, which is convenient for them, because they get to re-state things later as it would appear best for them. From the report, simple things like series connections versus series-parallel connections can be read through as sloppy grammar. Likewise, confusing Amps and Amp-hours is so often done, that we've almost given up trying to explain the difference.
So, when we say here that it appears the battery current was down to 4 Amps when the ARSII tripped to float, as opposed to after it got to float, we're relying on the precedence of words in the following statement which appears on page 17 of their report. ``We observed this output for the 45-minute setting, at which time charge current was reduced to 4 Amps, and the regulator went into float mode.'' Forty-five minutes looks curiously like 48 minutes. Did the ARSII regulator go to float at the right time?
So we have 440 Ah of battery capacity, and battery current has fallen to 4 Amps, which is less than 1% of capacity. Referring back to Figure 3, and the rate of change of current at that point in the charge cycle, it seems utterly incompetent not to recognize a full battery! Is an absorption cycle of 45 minutes sufficient? Sure ...if the battery is charged slowly enough, the absorption cycle can be a big fat zero, and with the tapering that is shown in the tests, charge rate is indeed slow.
So what about that Amp-Hour meter? It says the ARSII was still at minus 29 Amp-hours. Voltage and current indicate the battery is full, and the Amp-Hour Meter says it isn't. Volts and Amps don't lie, and only uninformed people who haven't read the Ample Power Primer and the manuals for the Energy Monitor/Controller would bother to regard Amp-hours as a consideration for determining a full charge.
Practical Sailor uses the Amp-Hour meter as if it were gospel, and even presents a chart showing the percent of full charge attained using the various regulators. Most people know that a wrench is no substitute for a hacksaw ...we wish more people understood appropriate uses for Amp-hour meters.
Yet More Gleaning
If the batteries were full when the ARSII tripped to float at the end of a 45-minute absorption cycle, what does that say about the accuracy of the Amp-Hour Meter? Assuming that the meter was accurate when it showed 190 Ah of discharge, and the batteries are full with only 161 Ah shown to be replenished, then what about the missing 29 Ah? Dividing 29 by 190 gives the answer ...the meter is inaccurately recording charge Amp-hours by 15.26%. But who really knows if it recorded the right amount of Amp-hours of discharge? It's safe to conclude that any of the Amp-hour measurements are highly suspect.
And Still Gleaning
While we suspect that the Amp-Hour Meter is lying, let's just assume that 190 Ah had been depleted prior to the ARSII test. To fully charge the battery requires that we replace 190 Ah plus a little for the inefficiency of the charge process. Just for ballpark reasoning, assume that we had to put back 210 Ah. The ARSII ran 4 hours and 15 minutes and apparently charged the batteries fully. This yields an average charge rate of 49.4 Amps. For a 440 Ah battery bank, that's an effective charge rate of 11.2%. Is this what people expect of a ``smart regulator''?
Using the data from Figure 1, the time to restore 210 Amp-hours is around 400 minutes for an effective charge rate of 31.5 Amps.
How does all this add up?
Practical Sailor starts out with an alternator they claim is rated at 100 Amps and it restores 210 Ah at an average rate of 49.4 Amps. That's less than 50% of the rating!
On the other hand, we use a charger limited to 35 Amps and can put back 210 at an average rate of 31.5 Amps, or 90% of the rating.
What's going on here? Where is the Practical Sailor analysis of their test results? The one graph they present shows Amp-hour measurements from a meter we have shown to be highly suspect, and totally inappropriate for the test in the first place. All the regulators tapered and achieved roughly the same average charge rate. Is this how they really work, or is there something wrong with the test procedure? The alternator? The batteries? Why would anyone want to purchase a regulator with such apparent lack of performance ...especially when they don't even compensate for battery temperature?
Maybe in our charge example there's some Ample Power magic at work!
The Final Synopsis
The problems we found with the Practical Sailor report on regulators are numerous, and not necessarily all listed below.
So Where was the Ample Power Regulator?
Over the years we read many articles about electricity in Practical Sailor and grated our teeth at what we consider to be inconsistent, sloppy, and incompetent work. We've written letters to them pointing out some of the problems we've found, and explaining other questions they posed, like how a fuse holder can cause a fuse to blow, (answer can be found on page 112 in the first edition of Wiring 12 Volts for Ample Power and page 126 in the revised edition). We think the readers might have been better informed had Practical Sailor printed some of our letters.
So where was our regulator? If we were a world class winemaker, would we ask the drunks at Pioneer Square to test our wines? After all, they drink plenty of wine and ought to now which one is best. In the past we've ``loaned'' Practical Sailor equipment to test that never seems to find its way home. We've sent them quite a few letters which are ignored. In the meantime, Practical Sailor regularly finds ways to sneak in editorial cookies for the competition. (For an example of this, see the final paragraph of their regulator report.)
Until Practical Sailor presents competent and unbiased reports on electrical system equipment, we'll take our chances without their evaluations, thank you.
And the Winner Is ...
We received some thoughtful responses to our contest that obviously took considerable time to produce. We were both amazed and dismayed that no one took up the issue of when a battery is full, and how that can be determined from simple Volts and Amps measurements. Under the circumstances, it would seem arbitrary to declare a winner, particularly after considering how many people have become victims of bad reporting and don't even know they are victims.
So the winner is the American Cancer Society. Ample Technology has made a $249.00 donation to the Society ... the list price of a Next Step Regulator. We thank all of those who called or wrote to talk about the regulator report, and especially welcomed the enthusiastic support we received for daring to challenge the ``gospel'' of Practical Sailor.
We expect the information presented in this issue of PS Reports will be useful, and everyone is welcome to copy it for their own personal use, and for the personal use of their friends. However, this issue, including all data presented in Figures 1-3 is the exclusive property of Ample Technology and may not be used commercially in any form without the express written permission of an officer of Ample Technology.
Ample Power products are manufactured by Ample Technology,
2442 NW Market St., #43, Seattle, WA 98107 - USA