APC UPS 1000XL - Pulse Charging for Lead-Acid Batteries since 2003

[Kiriakos GR]

I own my APC UPS 1000XL since 2003, this accepts external battery packs.
At the year 2004 the internal battery pack were degraded , and I did replace it with equiv-eland 2x 12V 17AH Panasonic, with life cycle estimate of 10 years.
At the year 2010 the Panasonic battery pack were degraded too, in just six years that the expectancy of 10 years specification.
I did pay high price for the best battery of Panasonic ( the most expensive than all others),  and since their dead I am now using a pair of used lead-acid batteries.

     
Today 22-10-2021 I thought to setup my own measuring monitor so to measure and log of UPS calibration internal process, and of the new charge cycle.
UPS calibration internal process = battery pack this gets depleted down to 20%,  and UPS now has calibration data of how long is the new run-time in minutes.
Correct method of 1000W UPS calibration, this is the load to be at 60% of what the UPS this is capable to support as load.
And therefore I did load testing at approximately 528W at 230 Volts. 

The final test results are extremely surprising, this specific  APC Smart UPS XL series, it does come with pulse charger ...
I was unaware of that detail in all these years.

Charger operation:
With the battery pack this be depleted, constant current mode (CC) this activates for the first 35 minutes, then all the rest charging process this is made due pulse charging.
I do not care for reverse engineering of how the APC pulse charger works in detail, but I did get good clues from my good in quality and fast clamp meter.
As example, Peak of the pulse could be up to 10A, but most part of active pulse duration, this is half of that.
Pulse time alive this is 10 seconds, then another one will appear.
If I would care for reverse engineering I would use my oscilloscope instead.   

The battery pack this is 2x 12V at 45AH Sealed Lead-Acid Maintenance Free.

The reason of this experiment
The new (used) battery that I found for my battery pack, this shown signs of weird behaviour about this to get fully charged, and to also about this to have the expected termination voltage after a full charge.
I did read an article about improper battery activation, that this might be the issue, because about everything else the battery seems as undamaged.
Proper battery activation  this should include a few charge / deep discharge cycles so the battery plates to become fully active, especially when the NEW battery comes from storage (manufacturing date up to be sold and get installed).

The battery which I got it might came from small boat that did not see much of action within 18 months time, and then it was replaced by a new, for 50 Euro retail price of goods, no one will spend time or money for tools, for long tests and or about to try to restore it functionality.

APC UPS 1000XL this is a great test bed for such battery restoration experiments, and my measuring setup this is the second strong point, which makes possible battery analysis, logging and conclusion generation in to such high detail.
 
Pulse Charging
The widest known benefit, this is less heat generation while charging.
The unknown downsize, this is that the battery needs several days to top-up.
The UPS feels happy within 2 hours of charging by reporting battery full, but the battery gets really full much more later, and this is what I am now exploring so to find out.
 
   
UPS discharge / Calibration process
At 24V the load at the battery pack (2x12) this was 22A.
The test duration was 39 minutes = the battery pack when down to 20% of available run time, and then charging this started automatically.
The UPS it is truly smart, and the engineers too which they designed the hardware and it software for the PC. 

While the UPS performs calibration at constant load, the PC software will report larger run-time when actual PC load in Watts,  this is lesser than the load which was there when the UPS did it calibration.
For example, my computer this is consuming 220W at idle, the estimated run time is 152 minutes.
At my run-time calibration I did load the consumption with another 300 Watt, due 2x150W spotlight system with separate power switches ( options 150W and or 300W).

 
UPS charge process timing and notes
Charge stats ---------------------12:40   
06 Min:  13.050 --- 13.580   (12:46)
10 Min:  13.114 --- 13.592   (12:50)
20 Min:  13.259 --- 13.628      (13:00)       UPS bat level 32%
37 Min:  13.452 --- 13.720       (13:16)     UPS bat level 45%   CC charge off
40 Min:  13.490 --- 13.740        (13:20)     UPS bat level 47%   (Charge 8.5A) pulse
45 Minutes                                   (13:25)      UPS bat level 50%
50 Min:  13.607 --- 13.820          (13:30)     UPS bat level 54 %   
63 Min:   13.789 --- 13.950          (13:43)     UPS bat level 64 %   (Charge 7.5A) pulse
70 Min:    13.795 --- 13.939          (13:50)     UPS bat level 68 %   (Charge 6.20A) pulse DC Fan stop. 
80 Min:    13.690 --- 13.835         (14:00)     UPS bat level 74 %   (Charge 3.80A) pulse
90 Min:    13.709 --- 13.850          (14:10)     UPS bat level 82 %   (Charge 3.20A) pulse
100 Min:   13.651 --- 13.793         (14:20)     UPS bat level 89 %   (Charge 2.37A) pulse
110 Min:   13.653 --- 13.795         (14:30)     UPS bat level 95 %   (Charge 2.03A) pulse
120 Min:   13.655 --- 13.796         (14:40)     UPS bat level 100 %   (Charge 1.75A) pulse
140 Min:                                        (15:00)                                            (Charge 1.42A) pulse
150 Min:   13.653 --- 13.801         (15:20)                                             (Charge 1.15A) pulse
 
Those are my first recorded stats, and I am still recording.
And in a later time, I will deliver any new conclusions from my latest adventure at the ITTSB Blog lab.