Peukert actually works

[booster]

I had been messing with the settings on the Trimetric battery monitor to get it more accurate on the recharge cycle (you have to set a charge efficiency). Last trip it went to 98% full on the recharge before reset, so that was pretty good. When we go home, I decided to let the batteries run down to 50% before the recharge to see if it still was accurate when charged from there. It would also give me a chance to check the specific gravity to see what the actual % left in the batteries was, compared to the Trimmetric. I let the van sit with the 12v power on, running the frig full time and some lights and stuff part time as I did things in the van. Loads were showing as low as .3 amps when just the detectors were running to about 3.1 amps when the frig was running, to 5+ amps with the fluorescent lights on. It took 8 days to get to 50% left on the 375AH of batteries. Actually it got to 47% by the time I checked the specific gravity.

Specific gravity checked right at 1200, which per the Trojan chart would put the batteries at about 62% charged, quite a difference from the Trimetric. The Trimetric showed that 195AH had been used, which matched pretty well with its 47%. I did a backwards calculation to see how many AH I would need to have 62% left after taking out 195AH, and got 513AH. This is about 35% more AH than the rated 375AH. The rating is at 20 hours, so we were way, way, slower than that, and even beyond a 100 hour rating, which I haven't found for my batteries yet. I was very surprised at how much you gain at the low discharge rates. That is good for us, as all our loads are pretty small. We rarely, if ever, use the inverter, and the TV only pulls about 2.5 amps even with a DVD playing. The rest is led lights and device charging. Only the air compressor pulls over 10 amps. It also points out that we will need to do another of these tests in the real world of camping to see how where we land then for capacity. I would hate to have the Trimetric say we were killing the batteries below 20% if we really had 40% left, especially if we didn't want to leave where we were right away.

Time to plug it in and see where it comes out % wise on the recharge cycle.

The same specific gravity test allowed me to see if using unmatched batteries was as bad as some say, or not bad as others say. We have two 260AH 6 volt GC batteries and one 115AH 12v deep cycle battery, all Trojans, all test good, with the 12v 3 years older than the 6v ones. All the cells, in all 3 batteries, checked right at 1200, with barely any variation, so at least at the slow drain, they are equally sharing the load. The batteries had not been watered since spring, and the 12v looked to be lower than the 6v, so that was a bit worrisome, as there is talk of the charge going to the lower internal resistance battery and over charging it. To be sure, I measured how much water each of the batteries took to fill, using the hydometer to put in the water from a measuring cup. As it turned out, all 3 batteries took nearly the same water, which, theoretically they should as they are similar in size (1380 watts for the 12v and 1560 watts for the 6v). It is deceptive because the 6v have a lot more reserve electrolyte than the 12v does, so they need watering less often. I am much less worried about the mismatch than before, as there doesn't seem to be any big deal going on.

[markopolo]

Will you be able to calibrate or adjust the Trimetric once you have the info from the recharge cycle?

[booster]

Originally Posted by markopolo

Will you be able to calibrate or adjust the Trimetric once you have the info from the recharge cycle?

I will be able to adjust the charge up side if it doesn't come up to 100% on the meter when done charging by changing the charge efficiency. The only way I could adjust the discharge side would be once I figure out what our real world discharge rate gives us for Ah available. Then I could tell the Trimetric that the battery bank was bigger than where it is rated (20 hour rating) to compensate for the extra Ah we get from discharging slower than 20 hr rate.

[booster]

Just did the recharge on the coach batteries to see what the Trimetric says. They were down to about 45% on the Trimetric, with the specific gravity saying they were at closer to 60%, due to the previously discussed low drain, Peukert stuff. They are showing that they are charged (flashing light), but the % full says 83%. It will reset to 100% as soon as it starts to draw down, but it only does that if it has reached full charge first.

I am really starting to question the usefulness of the Trimetric as a remaining capacity gauge. On the discharge side, the reading is heavily influenced by the rate at which the power is used. In this test,at low drawdown rates, it was 12% state of charge off at 50% on the Trimetric, which is closer to 25% of the reading. That is somewhere around a day or two of use for us, and could be more time for folks who don't have a compressor frig. Going back up in charge, it was off a similar amount. If you didn't get to a full charge, which would be very typical if you are using solar (we do), the Trimetric would continue to read lower and lower with every cycle (as it would not reset like it does after a full charge), so you might think you had hit 20% state of charge, and danger area, when you really were at 50% or more.

I could adjust the battery bank size setting to compensate for the discharge error, but it would again be based on a discharge rate. Unfortunately, out discharge rate will change based on weather (frig and fans), how much we are away from the van, etc. Add the solar during the day, so the discharge rate probably will go to zero pretty often, and there is still a lot variability.

Similarly, the recharge side works off the battery bank size, so it would adjust the same amount as discharge. But it charges at the same rate all the time on the charger, but variable on solar, so the Trimetric would think it had a bigger battery bank to charge from a given % and it would read even lower on recharge because it would think it needed more amps to full. The compensation would be to increase the charge efficiency %, but it would have to go over 100%, and I will have to check to see if the Trimetric can even do that.

All and all, there are just too many variables in all this to be confident in the % capacity left reading on the Trimetric.

The method used by Campskunk, and others, of knowing how much capacity is left by putting a known amp draw on the batteries and checking the voltage is looking better all the time. There is a bit of upfront work to determine what the voltages are at the fixed draw, versus the state of charge (by specific gravity), but then you are done if you can read the amps from the batteries and voltage.

The method of using a fixed load to test is tempting, but the amps used will vary with the battery voltage and influence the reading. I think it would work if we pick a load, like the inverter with the surround sound amp on, and then use the lights to tweak to the actual amps we want. The Trimetric will be very good for doing this, as it reads amps and volts, right at the batteries, on the display. I will probably do a chart at 20% state of charge increments with what the set amps voltage will read at that state of charge.

It is kind of a bummer that the Trimetric doesn't give very accurate information, as it is so easy to use. There must be lots of folks out there that are relying on it for keeping track of their batteries that are getting fooled. The good thing is that it appears the errors are toward the "safe" side of having more battery capacity left, than having less than thought, so at least your batteries are safe.

[markopolo]

There are big differences in the state of charge charts between battery types and also battery manufacturers. Can you select the battery type in the Trimetric setup?

Look at these two state of charge charts: http://www.classbforum.com/phpBB2/vi...tart=11#p16496

12.06 volts is 50% SOC on one chart and 12.25 volts is 50% SOC on the other chart. (12.25v on the first chart would be around 65% charged)

[booster]

Originally Posted by markopolo

There are big differences in the state of charge charts between battery types and also battery manufacturers. Can you select the battery type in the Trimetric setup?

Look at these two state of charge charts: http://www.classbforum.com/phpBB2/vi...tart=11#p16496

12.06 volts is 50% SOC on one chart and 12.25 volts is 50% SOC on the other chart. (12.25v on the first chart would be around 65% charged)

I have seen all the different state of charge vs voltage charts also, and they make little sense to me. I can see a difference between types, wet, AGM, gel, etc, but not within groups like we see. The variations is one of the things that makes the Trimetric look nice, as it doesn't care what type of battery it is, or what they say the voltage vs state of charge is. It calculates the state of charge based on amperage used/put in versus the battery capacity. The only voltage they use, to my knowledge, is one you set for the full charge voltage. You set voltage (that it has to be above) and amperage (that it has to be below) and a time to hold if you want. All three have to be met before the fully charged light comes on.

As we know, it takes very little load on the batteries to mess up the voltage reading, and thus the state of charge calculation. Plus there should be a no use wait time before checking, making using voltage a difficult way of evaluating state of charge. That is where the Campskunk, and others, method of checking under a known load comes in. This method would be very specific to the batteries in the van, their condition (so the method would vary as they wore out), and probably other conditions, so you would have to do testing on your specific application for it to work. The good thing is that the testing would not be all that difficult. All you would need to do was watch what you voltage read, with your fixed load, and check the specific gravity at the different voltages as the batteries drained. Then make a graph and you would be done.

As long as the van is just sitting in the garage, under good consistent conditions, I think I will do the drawdown, specific gravity, testing, and voltage under load test and chart. As long as I use existing loads in the van no wiring or other work or devices are needed other than checking the specific gravity at various times and making a chart. I think it will be very interesting to see how the chart matches the trimetric reading in actual use, on the road, when we throw in the variable loads, solar, etc.

I have a feeling that the Trimetric will turn out to be very valuable for monitoring loads, charge rates, etc, but much less useful in keeping track of state of charge, but we will see how it all turns out,

[markopolo]

Have you gotten into the "charge efficiency factor" settings?

Choosing values of "charge efficiency factor": In the “L1” mode this value is fixed at 94%. If you wish to adjust it to
a different value you will need to switch to “L3” mode. The actual "charge efficiency" factor for lead acid batteries
while they are not at the top of charge is generally higher than 94%, so this will usually give a conservative value for
"amp-hours from full"-- that is the meter will generally slightly under-estimate the “% Full” value for the batteries
which will give a useful practical result for knowing state of charge.

Trimetric note about Trojan batteries re: "charge efficiency factor" setting

We have measured this to be 95-98% with Trojan wet cell batteries

.

It looks like you should up the "charge efficiency factor" setting because you have Trojan batteries.

additional info:

4. As the battery discharges, the TriMetric “amp hours from full” gradually go negative at a rate depending on the
number of amps, and go positive when charging. For example if the batteries are discharging with “amps” =
minus 10.0, the "amp-hours from full" display gradually decreases by exactly 10 for every hour that goes by.
Similarly, when amps are positive 10 amps (charging) the "amp-hours from full" display gradually increases by
almost 10 amps for every hour. The "almost" has to do with "charge efficiency factor" which can be adjusted by
program P10. If the efficiency factor is set to 100%, the amp-hours will go up by exactly 10 per hour , however if
the efficiency factor is 94%, the "amp-hours from full" display goes up by only 94% of 10, or 9.4 amp-hours for
every hour. The purpose of this is to account for the fact that you don't get as many amp-hours out of the battery as
you put in, so that the TriMetric "amp-hours from full" display will give a reading of amp-hours which reasonably
closely estimates how many amp hours have been removed.

It is not clear to me if adjusting the "charge efficiency factor" will affect the reported state of charge.

[booster]

Originally Posted by markopolo

Have you gotten into the "charge efficiency factor" settings?

We upped the charge efficiency to 98% a while ago.

Trimetric note about Trojan batteries re: "charge efficiency factor" setting

[quote:10usan7m]We have measured this to be 95-98% with Trojan wet cell batteries

.



It looks like you should up the "charge efficiency factor" setting because you have Trojan batteries.

Yep

additional info:

4. As the battery discharges, the TriMetric “amp hours from full” gradually go negative at a rate depending on the
number of amps, and go positive when charging. For example if the batteries are discharging with “amps” =
minus 10.0, the "amp-hours from full" display gradually decreases by exactly 10 for every hour that goes by.
Similarly, when amps are positive 10 amps (charging) the "amp-hours from full" display gradually increases by
almost 10 amps for every hour. The "almost" has to do with "charge efficiency factor" which can be adjusted by
program P10. If the efficiency factor is set to 100%, the amp-hours will go up by exactly 10 per hour , however if
the efficiency factor is 94%, the "amp-hours from full" display goes up by only 94% of 10, or 9.4 amp-hours for
every hour. The purpose of this is to account for the fact that you don't get as many amp-hours out of the battery as
you put in, so that the TriMetric "amp-hours from full" display will give a reading of amp-hours which reasonably
closely estimates how many amp hours have been removed.

It is not clear to me if adjusting the "charge efficiency factor" will affect the reported state of charge.[/quote:10usan7m]

That is how it works, as I understand it. If you are starting at the correct state of charge reading on the Trimetric, and it is a big "if" based on what I have seen of the problems in the discharge side of the readings, your state of charge readings will be off by the amount the charge efficiency is incorrect for the batteries you are using. I had originally thought that was the issue with our setup, as it never came back to 100% on the charge cycle, even when I changed the efficiency to 98%. As it turns out, the efficiency is probably not the issue, and it is related to the discharge side that is measuring state of charge by comparing actual amp hours against the 20 hour rating of the batteries. If you are off 30% on your starting point state of charge, the 2-4% change in charge efficiency is much less of an issue. On multiple recharges that don't get to full (where the Trimetric resets), you get 30% more error in the discharge side, on each cycle. If your solar was taking you to 80% every day, so it never reset, you would just keep dropping state of charge on the Trimetric, while the batteries were actually much higher.

Trojan charts show about an 11% increase in Ah capacity between the 20 hour and 100 hour rating, so that would put our 375AH bank at 416AH. Normally, we could go well over 4 days without hitting 20% on the batteries, so we are at least 20% more capacity than the 100 hour rating, so I think I will change the battery bank size to 450AH in the Trimetric and monitor it as I do the voltage/fixed load chart. Since I will be checking the specific gravity anyway, it should be interesting information to have. It "should" come out much closer than the last test, but will be dependent on the discharge rate I use in the test.

At this point, there is so much error in the discharge side, that messing with the charge efficiency probably isn't worth the effort.

So much for a gauge to make things easier and more accurate

[booster]

Still testing and trying to see how the whole thing ties together. I did a full rundown from 100% SOC to 25% randomly picking points to check specific gravity to get actual SOC. I used a 5 amp load, per the Trimetric, at each point and read the voltage the Trimetric. I also noted the voltage at our base load, which is about .6 amps. Also noted what the Trimetric thought was the SOC, with the battery bank size set at 450AH for our 375AH bank. The results were very linear, with the .6 amp load .1 volt higher than the 5.0 volt load. Pretty surprised at that. Trimetric bounced around, compared to actual SOC, but was always within 7%. At the end it said 31% vs 25% actual. I think the jumping was due to the varying discharge rates I put on over the test. It used 309AH for the test, which matched the Trimetric 31% of a 450AH bank.

I was really interested in seeing what the Trimetric would do on recharge, and how many A it would take to go back up to full. Charge efficiency was set at 98%, and it came back to 96% SOC, which was about where it was before the full reset to 100% at the test start. So it shows you do put back in more AH than the battery bank is rated at, if you use that many by discharging slowly.

It looks like the SOC chart would be simple and linear, and would probably be accurate enough at our .6 amp base load (detectors and solar controller), which surprised me.

I think I will now do a bunch of discharge, charge, cycles at random rates and depths, and recharge levels, to see if they match the test chart.

So far it looks like that we could easily use the chart to tell SOC very quickly at our base load, or at 5.0 amps, and that the Trimetric SOC reading is good for rough readings, or a prod to check with the chart. One good thing is that all the information, except the chart,is read right off the Trimetric and only takes a few seconds to do.