warning, this another geeky post, so if you are not interested in boat electrons, skip this one!
For those of you who have been following along, and yea, we need to post more often to e blog, you may remember the power issues we had coming down the coast from Seattle. Basically rough a combination of some corrosion on the battery terminals and a voltage regulator that went wonky (and we are not sure if one caused the other or not), we ended up frying the old house bank on Piko. Albeit this was an expense I figured we wouldn’t we wouldn’t have until later, Tyne bank was 4 years old and I knew it was going to have to be replaced at some point. Not going into details, the regulator started outing out too much voltage and fried Tyne batteries. They were replaced by Deka 8c12 batteries in the same size in San Francisco before we headed south.
Piko uses(d) J185 sized floor sweeper batteries when I bought her and that us what we replaced the failed batteries with. These are an industrial style 12v battery similar in design to a ‘golf cart’ style battery, but in a 12v configuration. They are 190 Ah at a 20hr rate, so with two of them in parallel, we get just under 400 ah total, or about 200 of that usable using the 50% rule.
Feeding juice into the banks is an ample power 120a alternator now being driven by a balmar mc-614 regulator. We also have about 300 watts of solar, a Kyocera kd135 and a smaller 40 watt panel in parallel being fed into a Morningstar sun saver MPPT controller on each side of the boat. This combination will output about 12-15a into the batteries down here in the sun. That is just below what the ‘specs’ say, but is about right when you temperature compensate for the conditions down here. We usually get this output from about 10am to 4:30 pm every day over e winter.
The last part of the system, which is what I believe, led to the early demise of our new batteries is the canters link 20 monitor. This is a little box that should tell us all e stats of the batteries, it had a current shut so we can monitor voltage, amps, and amp hours used on the batteries. I has been on the boat for many years and has worked great in our pre-cruising lifestyle.
Sadly, we found out the amp hr gauge on this unit doesn’t work well at all if you have any sort of renewable energy on the boat. We had been using the gauge to determine when we had to run the engine to supplement the solar on the boat. Basically whenever we got to around 40-50% of the battery capacity used, we would run the engine and get the batteries back up to 80-85% where the solar could trickle them back to 100 over the afternoon. This gauge has some logic built into that resets the Ah meter back to zero after it senses a charge. This routine works great for people who use a shore charger and it works pretty well in this configuration. The solar not he boat, apparently, tricks the unit into thinking a complete charge had occurred when it had not. Basically whenever the current and voltage flowing into the batteries stayed at a certain level for a certain amount of time, which on a shore power is a good measurement that the batteries have been recharged, resets the counter to zero. What I didn’t realize is that this same condition happens almost every day when charging using solar when the sun goes down. This meant that we had slightly, but repeatedly at times been undercharging the batteries.
We had noticed that out banks voltage, which is another measure of its charge, though harder to use since the only time it is accurate is after letting the bank sit for some time with no energy going in or out, was too low. Sadly on a boat like ours, the batteries are NEVER at rest for any period of time, so you have to use your best judgment and see what devices are turned on at the time and make an educated guess as to what e voltage means. This is expressly why I bought the battery monitor, but I was lulled into a false sense of accuracy of it ah calculations. What this all means is we would check e batteries in the evening and see -4 ah which to us, meant we had used 4 Ah that evening after the sun when down. A morning, we might have been 100Ah down, though the meter reset when the solar had only put in 80Ah that day. So in reality, we were 24Ah down, not 4. Over time, this built up and hurt the batteries.
Batteries don’t like being disbarred too deeply. Car batteries, which are designed to crank e engine hard in cold weather, might only last a dozen or so z50% discharges. The deep cycle batteries we use on boats should last longer. I don’t have the specifics for the 8c12’s we have now, but they should last somewhere around 4-550 cycles to 50%. If only discharged them to 80% capacity, they might last over a thousand, and if brought them down to 25% only 100. Since these batteries are about 3 months old, they have about 100 cycles on them. Most were less than 50%, but there were a few days before thanksgiving that their voltages got down to 11.7 or so, which is pretty low. Now our voltages drop down to 12.1 after about 50Ah have been drawn, which is showing that our bank is only running at around 25% of its theoretical capacity.
Our solar controllers have a nice logging function on them, so not only can we see how much each string of panels is putting into the batteries over time, we can also see out voltages both highland low over time as well. This is where i saw the really low voltages the few days before thanksgiving. Sadly I was completely unaware that we we’re killing out batteries at the time since I was going off the link meters ah meter and though all was well.
We tried a single night in a marina a few days back in Ixtapa to try and run an equalization of the batteries. This is basically a controlled over charging of the batteries which could bring some life back into them. We did a full charge and let them stabilize, then used out outback charger to do the equalization with specs sent to us via Deka. This process makes the acid in the batteries boil, so you have to watch them very carefully when doing this.
Before doing this, I I’d check common things such as disconnecting the parallel batteries to make sure they individual voltages we the same, and checked GE specific gravity of each cell of each battery. The SG reading, albeit low even after compensating for temp, were all pretty much the same for all 12 cells, which mostly ruled out a short in one battery bringing the bank down. The connectors and batteries themselves were also clean and have been properly monitored and watered since install.
Sadly, the single equalization didn’t really seem to do anything and our batteries were not happy last night. We are going to spend some more money and try this a few more times in a marina before giving up hope in the batteries. We also might try using the PV on the boat to equalized each battery separately over the next few days since all the marinas where we are are wuitr expensive. Sadly, even with the mild abuse we gave them, they still should not be acting this bad in my mind, but getting warranty service on anything down in Mexico is basically next to impossible. Don’t get me started with the pain we have had with our new Asus laptop we bought before the trip!
If we do have to replace the batteries again, i think I am going to reconfigure the bank and use 6v batteries this time. The batteries we have now are not a common size and will be difficult to find down here, and next to next to impossible as we head further south. His would require us to rebuild the battery box to accommodate 4 smaller batteries compared to the 2 large ones we have now, but we should get slightly more capacity as well. If I can find t-145 sized batteries, we could get about 500Ah and theoretically a few more cycles out of the batteries before we run into this problem again. This is good for us because it give us more headroom before we get close to the 50% level and if we do, should get more cycles at that given level.
The reefer on the boat is the biggest consumer of energy. Down in zihuatanejo, where the water is over 80 degrees F in the winter and the air us even warmer, we are consuming about 80 Ah just keeping some meat frozen and the beer cold. I rebuilt the box itself with high efficiency insulation before we left, with that, these numbers seem in the ballpark. I can’ even image how much we would be drawing with the factory insulation! If we are sailing, we draw another 15-20 for the instruments and chart plotter, a few few Ah for lights and fans, maybe 10-15 for charging the laptops and iPods etc. We only run the water maker about once a week and it draws about 20-40Ah to fill up the tank depending on how low it was. That basically means that our 390 Ah of battery should be fine for our normal uses, but doesn’t give us a ton of headroom. Getting closer to 500 ah, which is a pretty large bank for a boat out size, would be really nice and would stop a few arguments between us if we both want to use our laptops at the same time if we are anchored in an area with free wifi.
Or we decide to give up on the freezer for the south pacific and only use it as a fridge. That would most likely drop or reefer budget by 25-35%. We are also thinking of getting an arch built for the boat to add another 1-2 large panels. As much as I hate e aesthetic of having a ton of crap on the back of the boat, adding another 135-200 watts of solar to the boat would also help a ton, and would give us something to make a little more shade in the cockpit than the boom tent that we currently use.
Here are some general numbers that talk about voltages at rest and how they relate to capacity and specific gravity:
Figure 1 - Battery State of Charge Charge Voltage Voltage Specific Level (12v) (6v) Gravity ------ ------- ------- -------- 100% 12.7 6.3 1.265 75% 12.4 6.2 1.225 50% 12.2 6.1 1.190 25% 12.0 6.0 1.155 0% 11.9 6.0 1.120
TOWARDS THE PACIFIC 
Ah, the fun never ends, eh!
We have had a Link20 for about 15 years now and it has never been reliable wrt AH. Extended charging rarely gets the counter to zero. It locks up frequently enough that I put a simple switch in the power wire to facilitate resetting the unit. The factory took a look at it and said “Its fine”. Have any of your fellow cruisers recommended a different monitor?
We have heard good things about the victron unit, but the chances of us finding something else at this point are pretty slim. We can live with it know, as changing bunch of h function values in the link seems to make it closer to reality, but in the end I think I just have to learn NOT to trust it and only use it as another data point.
I am so sorry to hear you are going through this again.
Your write up was *extremely* helpful to me. We have the Link 20 and we noticed it behaving erratically this last summer (it would go from -100 to zero in a short time) and started watching the voltage. We aren’t sure but we seemed to have the opposite problem – it would tell us we had used more amps than the voltage suggested.
So how are the TPPL’s treating you so far? I doubt I could even find them, but am curious none the less. If you google around, you can find many people who have had problems with the link monitors, and one site had some recommendations on how to make it batter, which seems to have helped. Basically you tell it no only reset if a much higher bar has been reached, which the solar chargers should not trip as easy…
Hey Lauren,
The short answer is that so far the batteries are working perfectly…but that I’m not sure how to use them the way that I had hoped (fast charge to 100%). The long answer is here:
http://thegiddyupplan.blogspot.com/2010/12/sears-diehard-battery-update.html
I’m going to look into the Link settings. I have a vague memory of changing the voltages when I reset the link on install because our batteries take higher charging voltages than others and that could be why we are having the opposite problem (it telling us we’ve used more than we think we have) and why it doesn’t reset as easily (which is good).
Also, Bob from Eolian said he replaced his Link 20s with Bogart Engineering Tri-Metric 2020 and had much better luck with them. Not that you want to be doing that right now…
Livia
Sorry Livia –
You must have me mixed up with someone else. We have had a Link 2000 on Eolian ever since I installed the inverter. And it works great!
bob
So sorry Bob – It was an RLW who commented on our blog and I connected it with you without thinking!
Hey Lauren(s)
Sounds like it is a super cruise so far, notwithstanding some challenges.
RE: Voltage readings: does a voltage check with a meter right on the battery posts match the reading you get with your monitor?
I agree, regular recordings of voltage, so you can monitor charge status that way is typically simpler, and more useful in determining charge regimen than the frustration which commonly comes from amp hour metrics.
The equalization phase you are doing makes sense, and hopefully will work. It should.
Glad to hear the refer insulation work is paying off, after all isn’t cold beer a necessity?
Keep the fun meter in the green zone!
Cheers,
Ken
Here are 2 things to try.
1. Your Morningstar MPPT controller is actually a modified PWM controller. It is a good controller but it conflicts with your generator. If your panels are outputting voltage your generator control module, which is PWM, thinks your battery is almost fully charged. You need to add a switch,
when using yor generator, on the incoming solar line to your solar controller to disable power to your batteries. Your generator will then fully charge them and not be fooled by the solar voltage.
2. You can reduce your refrigerator power consumption by 30 to 40% by placing a switch in the feed line from the auto defrost timer and the defroster heating coils. Do not disable the timer. You will run the risk of disabling the timer in defrost mode. The refrigerator will not work and you will use a ton of power.
Let me know if this is any help.
hey laurens, sorry to read of your electric challenges. hope all else is well…..
what is manufacture date of the batts you got in alameda? seems that equalization of good batts subjected to the conditions you describe would bring them back. i wonder if something more is going on…..
as far as e-meters go, like a gps, or a weather forecast: they’re all merely suggestions of what may be……
where you said you were 50ah down (suggested by emeter?) and at 12.1v, was that a fully rested voltage? otherwise it’s voltage under load and not reliable as to state of charge, but you already know this.
have you gotten out the clamp meter and measured amps going in/out say, at the height of solar charge/no loads? or at times freezer is working with no charge occurring? prob a good idea to measure at the bank. might give clues as to if your math/assumptions are good. i’d be taking as many and methodical direct observations as possible to try and paint the most complete pic possible. not that you haven’t been!
wonder if you’ve tried the timed-known-load-exercise and compared resultant at-rest-voltage to expected, assuming that you are confident bank is fully charged (specific gravity takes the guess work out).
mmmm……just the kind of thing that i would want to cipher on! if you know what i mean……i’d not be able to keep myself from trying!
best fortune as you continue the electron wars!
I am note sure of the manufacture date, I don’t see anything easily decipherable.
As far as amp in/out, the value from the shunt do correspond with the clamp meter.
We did another full charge and PV equalization today, and my temp compensated specific gravity reading show the batteries are at about 90% of full health. Sadly, I didn’t write down the SG readings last week, but they were much lower, so things are looking a little better. I am going to try a more traditional load test tomorrow and see if that data point correlates with the chemistry.
Either way, I think our bank is just a tad small for our use down here. My guess is in energy conservation mode, we draw about 100-150Ah a day, which is pretty darn close to 50% of our 390 on the high end. The solar does help a bit since it averages over 100 in while sunny. Another 100-200 Ah would be really nice, since then if we are hogs that day, we would only be drawing down to around 60%, or less than that if we are efficient. That would mean we would need another solar panel though to keep up too… Ahh, the choices, the money it costs and the increase in lifestyle… 🙂
The Link 10, 20, 2000, 2000R battery monitors will reset the Amp-hour value accumulated since the last reset based upon the following:
The value of kilo-Watt-hours (true energy removed) removed since the last reset has been returned to the battery
The charge current falls below the stated Amp-hour capacity multiplied by the stated percent given in the “charged parameters” stored. For example, for a stated 100 Amp-hour battery and a set value in charged parameters of 2% the charge current must fall below 2 Amps for the set period of time at or above the set charged parameter stated voltage.
The number of Amp-hours that “unwind” as charge current exists depends upon the stated CEF or charge efficiency factor, which is calculated with each discharge/charge/reset cycle.
For those cruisers who have a repeatable cycle of day-to-day use of energy I recommend setting the CEF to a set value, such as 95% in order to not have a variable false CEF which can happen in PV applications (and for those who use AGM or gel-cell batgteries) for several reasons. I also recommend that you at least temporarily observe the parameters recorded in memory of kW-hr usage along with Amp-hour usage in order to help zero in on establishing a confidence in the meter’s ability to tell you when the battery is truly full.
Assuming that the MPPT tracking PWM is not causing some interferrence with the meter’s ability to track Amp-hours accurately I believe that the biggest problem is with a variable CEF which can make the meter “lie” to you.
Because most shore power charger regulators do not deliver more than14.4 V the default charged parameter of voltage has been set to a value that is lower than ideal yet sufficiently high so as to allow an Amp-hour reset.
For those who have the ability to deliver a charge voltage above 14.4 V it would be better to set the charged parameter voltage higher than the default value. Recovery of lost Amp-hours to a cyclical use in deep-discharge will be vastly improved with acceptance voltages in excess of 14.4 Volts as long as the Amp-hour law is not excessively exceeded. That means you can charge at any voltage as long as the charge current does not radically exceeed the number of A-hrs “missing”. You can tell this with the monitor.