Odyssey AGM and solar experience?

[PNjunction]

NOCO 3500 ac charger results

As promised, I tested the 50% DOD 7ah powersonic agm with the NOCO 3500 in 85F ambient temps.

Note that the charger has a high and low output. The manual clearly states that for my 7ah size, I should be using the low (800ma output) rate. However, I tried the high rate (3.5a max), under SAFE conditions.

It started out at a steady 3.5A (0.5C) for a few minutes, and then ramped down to a steady 2.9A (about 0.4C) until 40 minutes later, the battery reached 13.2 volts, and dropped the current to 880 milliamps which remained steady until reaching 14.7 volts, and then dropping to a trickle of 80 milliamps in the space of a little over 2 hours. Sidenote: NOCO calls this transition from 13.2 to 14.7v absorb, although I would have preferred it to actually absorb longer at 14.7 volts.

At no time was the battery more than 3 degrees hotter than ambient temp. A discharge test shows it maintaining capacity as it should. I guess I'm not afraid of going to 0.5C SAFELY, which might come in real handy when my solar insolation is only like 2 hours and the battery is no lower than 50% DOD to start with. The battery won't be fully charged, but maybe 90 - 95% charged will get me through to the next day.

[PNjunction]

As usual - great info for me to chew on. I also like Sunking's advice, although instead of the battery, I should put my head in the fridge.

One thing I may be overlooking too is that I am starting from no lower than a 50% DOD when I pump 0.5C into the general purpose agms. As I understand it, internal resistance is lowest when fully charged, and highest resistance when fully discharged. Maybe that is the reason I seem to be getting away with it as I'm not at the highest resistance but somewhere in the middle when I start charging - but this is pure speculation on my part.

Anyway, I'll do more testing, but will give you guys a break and thanks for input - it is definitely appreciated.

Maybe the best move is to stop piddling around with these general purpose agm's, and just enjoy life with the Odyssey's as they'll take anything I can throw at them.

[Sunking]

These two possibilities are one very good reason that you need to check the individual battery or cell voltages during Eq.

Just one little problem with that. You cannot do that with any SLA battery and is why they should never be EQ. Well if you have 2 volt SLA batteries you can monitor voltage.

[inetdog]

During the bulk phase, I'm well below the absorb voltage setpoint - in practice, I'm in the "float" voltage arena during the bulk process, where larger amounts of current beyond the common 0.3C shouldn't be a huge issue. By the time the battery reaches the fixed absorb setpoint where the CC is now in control, I'm already well below 0.3C. So two controllers are in charge here - in bulk the battery runs the show until it reaches absorb, then the CC takes over signaled by reaching 14.4v (or whatever you have set it to). The CC is now overcharging the battery on purpose, but nowhere near the current levels seen in bulk.

??

I'm not trying to prove a point, just trying to um, absorb it all in my head if you will.

Well, you can try absorbing this, if I state it correctly:

1. When you are in Absorb, you are not yet overcharging the battery. Your CC thinks that it has found the point at which constant current might lead eventually to overcharge and so it switches to constant voltage mode. When the current drops to a certain level at the correct Absorb voltage for the batteries, you are on the threshold of overcharge, and going to the Float voltage will allow the CC to just replace internal losses in the battery and maybe charge it slightly.
An controlled overcharge (AKA Equalization) will happen when you increase to voltage to the Equalize voltage, which is higher than that for Bulk, Absorb, or Float.
This voltage will produce a substantial current through the batteries even after they have fully charged and will expend all of its energy in electrolyzing water and heating the battery.
2. The complication, which makes it more difficult to understand in detail, is that you are usually working with a series string of cells, either within one battery or in the form of a string of batteries. Since the SOC and capacity of different cells may eventually vary, even if they have the same life history, applying the Equalize voltage to each of the cells simultaneously is not going to happen. They will all get the same current, but may be at different voltages. As a result, if one cell is still, effectively, in Absorb or Bulk in terms of accepting charge, the other cells in the string will be getting an even higher voltage applied than just Veq/N where N is the number of cells.
Or in the other direction, one of the cells may not have been discharged as far as the others and will be fully charged while the others are still at the Bulk or Absorb level. In this case, the voltage across that one cell has to increase to make up for the lower voltages at all of the others. It is now getting a super overcharge.
These two possibilities are one very good reason that you need to check the individual battery or cell voltages and watch temperatures on all batteries during Eq.

[Sunking]

Put the battery in the fridge and charge them.

[PNjunction]

Right on about the 14.4 absorb and 13.6 float. Although I think my CC's temperature compensation is playing into that at nearly 100F. (I really need to go with remote temp sense to do it right rather than just ambient).

Maybe I've spent too many hours in the sun testing yesterday - my brain feels like lead-sulfate now. I think I see what that guy is pointing out and why a single word of "overcharge" aka absorb is significant to me here:

During the bulk phase, I'm well below the absorb voltage setpoint - in practice, I'm in the "float" voltage arena during the bulk process, where larger amounts of current beyond the common 0.3C shouldn't be a huge issue. By the time the battery reaches the fixed absorb setpoint where the CC is now in control, I'm already well below 0.3C. So two controllers are in charge here - in bulk the battery runs the show until it reaches absorb, then the CC takes over signaled by reaching 14.4v (or whatever you have set it to). The CC is now overcharging the battery on purpose, but nowhere near the current levels seen in bulk.

Then again, maybe that's in a perfect world? Poor quality manufacturing may be the limiting factor to 0.3C ? (ie local hot-spots from manufacturing abnormalities acting as a thermal runaway trigger, and not so much from a pure engineering standpoint) ??

I'm not trying to prove a point, just trying to um, absorb it all in my head if you will.

[Sunking]

You mean 2.45v/cell, or 14.7 volts right?

No Sir 2.38 to 2.40 vpc. Set your charger to Absorb 14.4 and Float @ 13.6

[PNjunction]

GREAT point which I didn't catch!

Btw, the 7ah test battery is holding at 13.21 volts SOC after a 9 hour rest. Doesn't say much until I load test it, but it is a data point.

You mean 2.45v/cell, or 14.7 volts right? Unfortunately, my charge controllers only go as high as 14.4v. Aside from the safety issue not being fully tested, at this point I'll just be walking-down my capacity it looks like.

However, I will rerun the test later this week with a 3.5a NOCO charger that does go to 14.7! Again this will be done with safety first in mind. Maybe I'll just do the NOCO on it monthly if the safety test with it passes.

Edit: YIKES - you just made me realize I'm not reaching gassing voltage with solar on my expensive Odysseys either! Guess I shouldn't quit my day-job huh?

[Sunking]

The battery didn't immediately complain and seemed not to notice that it had been over 0.3C for about 45 minutes. No hissing, gassing, whining or tantrum thrown!

You never reached the gassing voltage.

[PNjunction]

Ok, guess I'm drifting away from Odyssey's, but did a test today on a general purpose Powersonic 7ah agm with currents larger than the common 0.3C to check for heating and/or gassing. Objective was to take it to 0.5C, which was met, and test was a success - although this is just one time, and long-term results are yet to be seen.

Test subject: Powersonic PS-1270 (12v, 7ah) agm. Discharged to 50% DOD as per the "watts up" meter and verified again after 12 hour rest. Known-good unit - not pulled from a shed 10 years ago powering a weedwacker and rejuvenated with a 50ah wheeled charger. I treated this new battery nicely for about 50 cycles.

Test Gear: Fluke 87V and Extech 470 multimeters with IR and type-K thermocoupler.
(Yep Russ, the IR consistently reads about 5 degres higher than the thermocouple. Also why didn't I listen to Sunking and get a real shunt earlier?? )

Panel: 60 watt , 12v nominal

Charge Controller: Morningstar Sunsaver 6 PWM, unjumpered for 14.4v. Ambient temp compensation only. 12-foot round trip #18 gauge from CC to panel. 3-feet round trip #18 from CC to battery.

Environment - 85 to 105 degrees fahrenheit from start to finish of test. All gear outside and in the shade, except for panel. Panels ramped up to 160F pretty quickly. No clouds, and panel ground mounted, but angled towards sun a few times. Body temp at wrist 95F. Felt like being on the surface of Venus.

Results: (note that the battery temperature never rose more than 5 degrees above ambient - even at the 0.5C rate, so I'll leave ambient temp out of the table.

Time | Charge current | Terminal voltage | Battery temp |

1045am | 3.45A | 13.46V | 85F
1100am | 3.38A | 13.72V | 85F
1115am | 2.50A | 14.00V | 90F
1130am | 1.68A | 14.12V | 91F (this has just crossed below the recommended 0.3C rate)
1145am | 1.05A | 14.20V | 93F
1200pm | 0.78A | 14.22V | 95F
1215pm | 0.64A | 14.23V | 95F
1230pm | 0.49A | 14.18V | 97F
1245pm | 0.40A | 14.18V | 99F
1300pm | 0.35A | 14.13V | 101F
1315pm | 0.30A | 14.08V | 105F
1330pm | 0.29A | 14.09V | 105F

and so forth climbing it's way down until I ran out of good insolation.

The temp probe and the IR of the Extech was fun to use. I searched all over to see if there were any hot spots developing (front, back, sides, rear, bottom). There were up to maybe 2 degree differences in temp depending on where I measured, the hottest being near the terminals, and generally the top. 2-3 degrees difference maybe all around with IR.

The battery didn't immediately complain and seemed not to notice that it had been over 0.3C for about 45 minutes. No hissing, gassing, whining or tantrum thrown!

Thing is, this is a one-test shot, so again long-term, I don't know. But now that I'm armed with some temp probes, I plan on testing further at the 0.5C rate with my larger powersonics.

[Sunking]

Comments from the experts?

You have to reach the Gassing Voltage of about 2.39 volts per cell before any gassing occurs. This is why voltage regulation is mission critical when charging VRLA battery types. This is why monitoring the temperature of VRLA batteries is important to monitor to prevent thermal runaway. In the event a cell fails shorted (common failure mode), the voltage drops and fools the charger into thinking the battery needs more charge current and this causes the remaining individual cell voltages to rise above the gassing point and where heat begins to build up causing the voltage to drop even lower which causes even more current and heat. The effect is regenerative feedback and called Thermal Runaway. If left unchecked can cause a violent explosion and fire.

Thermal runaway is not as serious of an issue with flooded lead acid batteries because of the large thermal mass encountered in FLA batteries. Nor can a FLA explode from over pressure.

[inetdog]

One thing that I also take issue with is the statement that no oxygen or hydrogen are being produced until the cell voltage and SOC reach some hard and fast limit. This is not necessarily the case, since IMHO irregularities in the electrodes could still result in some localized electrolysis even though the cell is not yet gassing violently. That should be handled by the recombination catalyst, but it still will not result in a zero pressure on the vent.

Comments from the experts?

[russ]

Comments within the text in bold

V I picked up an Extech EX470 multimeter that has IR and a type-k thermocouple. I bought a cheap Extech MiniTec 26 - it is a couple of degrees C off. It is repeatable so no problem for me but I suggest you check your against a known temperature. Type K is perfectly suitable.

To let you know where I'm going with this - The Odyssey's are doing fine so far with at least .4C or more - no sweat. But now I'm going to experiment with the typical Powersonic ups-style agm's, and see if I can't take them up to .4C or maybe even .5C at the most without blowing vents or going into thermal runaway. Got the temp meters and the facemasks ready. Powersonic says NO to anything above .3C at any phase of the charge cycle, but I've got to find out first hand I guess. SAFELY that is.

[PNjunction]

Thanks Russ - good to know about the relative IR inaccuracy. I lost the temp probes for my Fluke 87V, so I picked up an Extech EX470 multimeter from RS on short notice that has IR and a type-k thermocouple. I won't be taking it into my work lab - but it beats a shirt-pocket model that's for sure.

To let you know where I'm going with this - The Odyssey's are doing fine so far with at least .4C or more - no sweat. But now I'm going to experiment with the typical Powersonic ups-style agm's, and see if I can't take them up to .4C or maybe even .5C at the most without blowing vents or going into thermal runaway. Got the temp meters and the facemasks ready. Powersonic says NO to anything above .3C at any phase of the charge cycle, but I've got to find out first hand I guess. SAFELY that is.

[russ]

Whoa! Thanks inetdog - I overlooked that.

In fact, I'm off to get some temp probes and perhaps an IR meter right now! You just saved my butt, and most likely my house.

The IR meter is not terribly accurate - it can be go - no go or is it getting hotter but to get a precise correct temp isn't easy.