[aprssig] Battery & Charger for APRS Station

[Ray Wells]

Steve,

I wonder if you misunderstood the function of the solar charge 
controller? It didn't float the battery at 14.4 or 13.4, they were the 
switch off and switch on limits for charging. It was a cyclic function. 
perhaps somewhat similar to the charger John refers to except time isn't 
a function, just voltage.

I can understand the need in a solar powered installation to maximise 
charge for maximum night time and inclement weather periods supply 
capacity and I suspect that's the reason for the cyclic charging. 
Solarex documentation didn't have much to say about it.

The problem with a 13.8v float charge is that a) the battery is only 
charged to about 85% capacity and b) the electrolyte doesn't get stirred 
up - the specific gravity at the bottom of a cell can be quite different 
to that at the top (not a problem with SLA!). The gassing/boost charge 
sorts that out.

I can only imagine that Solarex (and Ctek) are looking for the best of 
both worlds - decent charge level without using excessive water, and a 
form of battery conditioning as well. Whether it's ideal or not is open 
to debate.

Regarding RF pollution from chargers. I had a homebrew 15A/13.8v supply 
of standard design - bridge, big cap, 723 regulator driving external 
pass transistors. This supply produced RFI up to the 6m band and it was 
coming from the bridge rectifier. A 0.1uF disc ceramic across the 
rectifier DC tabs fixed it.

Ray vk2tv


Steve Noskowicz wrote:

>--- John Habbinga <kc5zrq at gmail.com> wrote:
>
>  
>
>>I am installing a digipeater in a location that I want to have
>>reliable battery backup power. ...snip...
>>    
>>
>
>Hi John,
>
>   Looks like you already spent the money on your APRS digi backup, but here's
>another 2 cents anyway.
>
>  As far as I am familiar with lead acids, all the other remarks don't conflict
>with what I know, but I'll add what I know in terms of basics.
>
>   The correct term for lead acid batteries is "float" not trickle.  Float
>implies a voltage in lead acid and trickle implies a current for nickel
>chemistries.  The standard and correct float circuit is a linear voltage
>regulated power supply, NOT a simple transformer and bridge.  The bridge type
>is probably what Ctek is referring to in their ripple page....Shouldn't even be
>seriously considered for floating - just use for charging, then stop it.
>  From everything I've seen, the RESTED, open circuit voltage is acceptable for
>measuring lead-acid charge state.  VERY, VERY roughly 11.8 empty, 12.3 full.
>
>Some thoughts, that you may or may not already know.
>  While there are some differences in the other posts (Ray Wells' 14.4 boost –
>13.4 droop cycle), my experience is the following.  In the ideal, Floating is
>done with a current limited, temperature compensated, regulated, constant DC
>voltage of around 13.8.  I believe it is a positive tempCoef (hotter = higher)
>and I mean pure DC.This is what is recommended by battery manufacturers and is
>used in telecom and cellular systems. 
>  Therefore, a fixed DC supply is not quite correct unless things are at a
>constant temp and the supply output voltage is adjusted for that temp.  I do
>NOT know what the effect is on life if you do not do the temperature
>compensation correctly, but over charging an SLA is the killer no-no.  
>Overcharging a non-sealed type makes the gases and removes water, changing the
>spec-grav.
>   The supply is connected to the battery and equipment.  The equipment
>normally runs on the power supply/charger while the battery is simultaneously
>float charged continuously.  Lead acid chemistry draws what it needs to stay
>charged when fed the *correct* DC voltage.  Lithium happens to be similar.
>  The charger must also have a current limit which limits the current going to
>the battery.  Each battery has it's own safe maximum charging current.  Sealed
>versions can tolerate much less charging current in general.  When charging,
>the chemical reaction is being reversed.  When over charging, it is only
>decomposing the water into H and O and generating heat and pressure( in the
>sealed types).  Only extremely  small amounts of the gasses can be tolerated in
>sealed types, therefore, floating is much more critical.
>  I wonder about the "equalizing" mentioned by  Ray McKnight.  This is a
>popular concept in the high end Li-Ion radio controlled aircraft batteries. 
>They charge cells independently.   
>   It seems to me that in the lead-acid case, this will over charge the lower
>A-H capacity cells just a bit and use more water in order to full charge the
>higher capacity cells (assuming series cells).  This means high-cap cells will
>be cycled not quite as deep as the lower capacity cells.  Versus letting the
>higher cap cells only reach not quite full charge.  If 95% charge is worse than
>using water, then it must be good.  However, you can't expect to over-charge
>SLAs, right?
>
>   Depending on how the Ctek units turn on and how your battery voltage drops
>under load, the radio/transmitter will run on the battery for a while, then on
>the charger when it senses the voltage drop to its turn-on threshold.  In
>either case, the battery will be supplying the rig for some time until the
>charger senses a need to charge.  This looks like many mini cycles rather than
>the classical constant float – then discharge only when power fails. 
>Unfortunately, I can't comment on this mini-cycle mode vs. the classical pure
>float mode.  I recommend asking Ctek – they obviously think it is better, or at
>least want potential customers to think so. 
>
>  Also, The float voltage should be what the manufacturer recommends for *that
>specific* battery.  For longest life, It is advised not using any simple rule
>of thumb float voltage for all lead acids.  There are variations due to
>construction.
>I can't comment on the relative advantage of the Ctek method, but perhaps you
>can talk to them and get their information.  
>
>   There are typos on the ctek page regarding ripple.  I wouldn't call the
>transformer/bridge a "linear" charger; I'd call it cheap, consumer.   To me
>"Linear" means linear regulator.  
>   Is it 2 or 40 amps ripple at the 9.7 amp charge???  They obviously refer to
>the bridge type which is junk.  How about comparing to the TempComp regulated,
>current limited type?
>   I don't think the 120 HZ. ripple of a bridge causes much RFI as they say,
>but switch-mode can obviously be a killer if not tamed.  
>   The yellow line looks red to me.
>
>This either helps or confuses. Sorry if the latter.
>73, Steve, K9DCI
>
>
>
>  
>