[time-nuts] HP 105B Battery the saga continues

[Perry Sandeen]

Hi,
The battery problem becomes more convoluted every day. Due to family medical issues I have not been able to do any testing/repairing.

The 105B charger is set up for nicads using some selected un-alterable charging rate so to use an internal lead acid setup requires a different ps.

My unit prefixed 2128A has a pc board non-adjustable ps that requires a different extender board to trouble shoot with the power on than the other boards require. Mine partially works as I have oven power but a current limited other voltages so I have to either buy the special extender board or rebuild the ps. Since I'm never going to use NiCads rebuilding the ps is a very viable option.

My other unit has the older style ps with a heat-sinked pass transistor mounted to the chassis. At present it in in the equipment rack and cooking nicely.
A number of battery back-up proposals have been shared, many are case specific.
In my situation here in So CA are rare occasional fractions of a second drop outs. I purchase a refurbished APC Pro 1500 for $100 delivered from ebay.
IF I get aroundtoit, the least expensive option for me is a 12 car battery and a buck-boost converter from china for $8 and use a trickle charger [Harbor Freight] to keep it topped off. I'd use a Shottky barrier diode for isolation from the ac ps and not use a relay to switch over. 
 Where the nicad pack was located one can put in 33,000 uF of Nichicon 105C caps for $20 for a buffer hold over. The oscillator draw is listed at 18W.  If I have a power outage for several days my 105B will be the least of my problems.
I've cobbled together a 105B manual for the 2128A prefix only using the errata sheet and some ski's from a list member who I apologize to for forgetting his name for attribution. It is 75 Mbytes in MS word and I have to re-install my PDF creator so I can get it to a size that will fit my attachment limit.  Although not complete it is a pretty good start.
The wiring method is the weirdest I've seen any time in my electronics career. The underside of all the pc board edge connectors are covered with a non-removable aluminium plate, so one has to probe through the top. It is the hardest piece of HP equipment that I've had to work on in 25 years of working on HP equipment.
The divider boards can easily replaced by TTL logic and a 10 MHz output jack can be added. Great oscillator but horrid implementation.
Regards,
Perrier