Jose Da Silva wrote:

>On August 30, 2005 01:25 pm, Thomas C. Sefranek wrote:
>  
>
>Seeing "now" that the solar panel is providing a respectable amount of 
>power of 3.2amps, there is no need to go this way. The 7805 is just 
>fine but the switching power supply appears over-kill.
>Based on the suggestion of using a switching power supply to increase 
>efficiency, this may be a slightly better compromise than the 7805 if 
>the power from the solar panel was miniscule, which it isn't.
>
>Okay, now that you pointed-out the disadvantages, here are some 
>advantages:
>
>Lower starting voltage than 7805. The 7805 has a 2v dropout.
>
Not according to Motorola specs...

@200ma., 60c, I get 1.5 volts min. 7805
Well that's not fair, your circuit can't do 200 ma. @ 6.5 volts in.
.4 volts/100K* gain of 100 (darlington 3904s) @ 200 ma. = huh? .4 ma.
I'll grant  that you can get a bit more,...  (lower current , higher 
gain.), no where even close to 200 ma.

So, say 30 ma. 25c, I get 1.5 volts min. 7808
.4 volts/100K* gain of 4,900 (darlington 3904s) @ 30ma. = huh? 19.6 ma.
I'll grant you that some 3904' s have more than 60 gain at  30 ma, but 
not enought to make a difference.

>Peak input voltage is 40Vdc on a 2N3904.
>
With a 6 milliamp maximum load ability.

Look at the 7805's 35 volt "limit", 5 volts less,  at  60c it has a load 
limit of 300 milliamps,
vs. 7 ma. for a 3904.

>Less wasted power than 7805 but still sufficient for a PIC and the PIC 
>shouldn't be doing anything active until it is needed (basically 
>everything is in off-mode until charger is overcharging).
>Your overhead would be (16v - 6.1v)/100k=0.1mA if you have nothing 
>hooked up on the 5v output, but the 7805 will use 5 to 8mA of current 
>while doing nothing (see quiescent current here).
>http://www.fairchildsemi.com/ds/KA/KA7805.pdf
>  
>
And source only 10 ma., one LED!
A 78L05 of the same package as a single 3904 has 3.5 ma burden, but it 
can source 100 ma.

>You still need to deal with noise on the inputs and outputs, so that 
>means you still should have a choke on the front, an input cap and an 
>output cap regardless of this circuit idea or a 7805. The choke isn't 
>to deal with voltage ripple from the solar panel, but to deal with 
>spark generated noises that you would get from an on board engine, 
>otherwise, voltage ripple isn't likely an issue here. you don't need 
>flat/perfect voltage regulation, so whether the voltage drifts 0.8mV 
>(for a 7805) or 20mV is not going to make much difference here.
>  
>
With a 7805 and it's 70 db of rejection, a choke is unnecessary, 
provided the bypass capacitors
are correct.

The difference between the 7805 is you have 2 transistors, 1 resistor

>and 1 zener, which is more complicated than a 1 7805 but simpler and 
>less noisy than a switching power supply. since it's a hobbyist 
>project, it may be more interesting to build than a plain 7805 type 
>circuit, but less complicated than building and testing a switching 
>power supply circuit.
>
>In terms of heatsinking, what type of circuit are we talking about?
>I'm guessing on the active side, a PIC for sure, maybe 1 or 2 NPN 
>transistors switching some hi-voltage relay. The relay(s) +ve should be 
>connected to the 12v and not the 5v supply and the PIC should control 
>the relay(s) using an NPN between ground/relay(s).
>With no sleep modes, we are probably talking about 2mA if the PIC runs 
>continually, or let's more than double it and say maybe 5mA.
>
>2mA * (16v - 5v) = 22mW
>5mA * (16v - 5v) = 55mW
>
>2N3904 at 60degrees celcius max dissipation=210mW
>
>It's a sailboat, so you don't get that much engine vibration, so clamp 
>an alligator clip to the head of the transistor if you think it's not 
>enough heatsinking.    ;-)
>  
>
On a sailboat I'd opt for the best, reliable, least parts circuit.

-- 
  *
  |  __O    Thomas C. Sefranek   WA1RHP@ARRL.net
  |_-\<,_   Amateur Radio Operator: WA1RHP  
  (*)/ (*)  Bicycle mobile on 145.41, 448.625 MHz

http://hamradio.cmcorp.com/inventory/Inventory.html
http://www.harvardrepeater.org


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist