Well, this sounds like a similiar issue I have....
and I agree with almost everything that has been posted so far.I will echo Ross' advice to get a calibrated lab grade hydrometer. If you have one of those $5 types that has reads only to the nearest 0.005 or worse yet has the colored floating balls, you really need a good hydrometer.The Francis Freas Glass Works manufacture an excellent line of high quality, calibrated, and easily readable temperature and hydrometer sets for about $20-25 delivered to your door. See the attached link.However you may have another sneaky issue happening here. I think that the Link 2000 is similar to the Link 20 with the exception of the inverter/charger control. I have a Link 20 that has tricked me into thinking my batteries were charged when in fact they were nearer 50% discharged.The Link 20 (and perhaps the Link 2000) considers a battery bank to be fully charged when the voltage is at 13.2V or higher and the charging current drops to less than 2% of the AH cap of the bank for a time period of 4 minutes. These are default values and can be changed.In my case I charge my batteries from a charger that is fed from my genset. The problem occurs if I have some deficit AH still to replace and the charger is supplying more than 2% of the AH rate. If I then shut down the genset (and charger) and start my auxillary to leave an anchorage, the alternator on the aux will be typically supplying 13.8V at perhaps something close to 2% or lower of the AH bank size. The Link 20 now thinks the batteries are fully charged and zeros out the remaining deficit AH. It then recalculates the charge efficiency factor for the next charging cycle. Now I have a battery bank that is really still down some number of AH but the instrumentation claims they are fully charged.This cycle can repeat quite often and each time the batteries sink lower in actual charge. Over a month or part of a season, you have batteries that are perpetually undercharged and sulfation begins. Batteries that are sulfated will not achieve the specific gravity readings that would indicate fully charged banks, and the batteries will "appear" to be fully charged based upon the terminal voltage and the charge current.The tipoff is that the batteries will seem to have much lower capacity. In your case of (3) 8D's, you may not notice the decrease in bank capacity.I think the only recourse in this case is to equalize until the specific gravity no longer increases. In my own situation I am going to look into the capabilities of electronic battery de-sulfinator devices.I hope some of this may be of help to you.