For determining variances "is/was" a $5 bobber would be fine. For precision in zeroing-in on exact numbers i.e. exactly 50% soc then exact means right on the number. 1.202 instead of 1.200 for example. To assume a five dollar hydrometer can do this is quite an assumption. I built a shelf with hole and below a ledge with v-notch to hold my hydrometer both steady and absolutely straight. With a strong light and magnifying glass incredibly tiny incremental variations in bobber height can be discerned. Freas built me a custom set. My favorite is an axpanded scale 1.255 - 1.265
It had to be N.I.S.T. traceable which means documented. Quite the instrument. It is housed inside a schedule 80 ABS pipe which is clamped to the wall. But the regular Freas unit is far and away good enough for guesstimating state of charge.
Always record temps if record keeping. I have an assortment of fresnel lens truangular Ashcroft lab thermometers. The 0c to 50c is the most used.
At work I had a National Controls Corp A to D data acquisition board connected to a 3100 dollar Compaq SX computer. 155mb hard drive, 25 MHz processor 2Mb RAM with 400 dollar additional 2 Mb RAM. An Indian programmer from Uddar Pradish set up the program then spent 2 weeks de-bugging it. Pure buggy whip. Setting the alarm for 3:20AM. Living on site while Quicksilver was being outfitted. Going to conventions as a rep. Rebuilding alternators at night. No wonder I have to stay busy down here in the land of cacti. Total withdrawal may prove fatal...
