[NLRS] Battery charge regulator

[Dr. Gerald N. Johnson]

Astron supplies do nothing gracefully. And they tend to not get along 
with batteries.

First problem is the basic design uses a 723 IC that very nearly 
predates the 709 op amp. While it has provisions for current limiting, 
Astron doesn't use them. It has no thermal current limiting like we are 
accustomed to from the LM78xx and LM340 families for decades now. Then 
Astron uses a gaggle of bipolar pass transistors in parallel but the 
emitter resistors that are supposed to force current sharing are nearly 
an order of magnitude too small so don't function. That way the 
transistor with the highest current gain supplies most of the load 
current, until it burns open (often the transistors have no heat sink 
compound to connect them to the heat sink), then the transistors that 
remain let the remaining highest gain one take the current until it 
burns open. The progression goes through the gaggle of pass transistors 
until there are no more and the driver transistor (good for only a few 
amps) joins the pass transistors. Those current sharing resistors can 
work very well if they are set to drop at least a half a volt at rated 
shared current. When they drop 1/10th that they don't affect current 
sharing so that sharing is set by current gain and the base emitter drop 
that tends to decline as the junction temperature rises.

Then they run so high an unregulated voltage that they need a crowbar to 
protect the equipment if a pass transistor shorts and the overall 
efficiency is 50% or less. It does take lots of regulator headroom when 
you drop an added half volt in the emitter resistors to force decent 
current sharing and you have to account for three or four Vbe drops in 
the drivers and regulator chip.

I have a couple circuits using FETs for regulators, one has more than a 
decade of use, the other has one circuit built and tested and it was 
shown in the 2010 CSVHF conference proceedings in the paper on PA 
protection as much the same circuit can shut down quickly to take power 
off a PA. Its on line at: 
http://www.geraldj.networkiowa.com/papers/CSVHF2010/protection.pdf

My shack supply for a long time is shown at: 
http://www.geraldj.networkiowa.com/papers/fetreg.jpg
It uses schottky rectifiers and FET regulators and regulates down to 20 
millivolts headroom. Some day I need to update it with synchronous FET 
regulators to raise the unregulated full load voltage about a half volt, 
then I'd make better use of the 12 volts RMS each side of center on that 
power transformer.

I do not own any Astron supplies, but I've fixed or advised on fixing a few.

Think about using MOSFETS for switches, 20 millivolts drop is almost as 
good as a relay and the drive power is much less than a relay. Trouble 
is with the most common N Fets that have the best characteristics, the 
gate voltage needs to be 5 ro 10 volts above the source voltage 
(typically the output voltage of a power supply switch) but the current 
is essentially zero, the gate looks like a capacitor. And if run in 3rd 
quadrant mode, the built in diode will take over if the gate drive fails.

73, Jerry, K0CQ



On 9/9/2011 11:09 AM, Doug Reed wrote:
>
>
> <SNIP>


> In the next vehicle we just hung the batteries across an Astron power
> supply and adjusted the supply for the specified battery float voltage,
> leaving the power supplies on all the time. This tended to boil the
> batteries, 13.6 volt float voltage may have been too high. When we did
> finally drain the batteries, the power supply smoked when it tried to
> charge the batteries at max current. (Astron supplies do not current
> limit gracefully......)
>
> <SNIP>73, Doug Reed, N0NAS.
>
>