Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: zamboniman60 on December 01, 2006, 05:43:14 PM

Title: Lab power supply
Post by: zamboniman60 on December 01, 2006, 05:43:14 PM
I'm working on a simple transistor-regulated (series-pass) power supply. I don't expect to pull a lotta current off this sucker, maybe 500mA, but I'm using a 4A 12V transformer, and four 1F supercapacitors in series to make the base raw power supply. Does anyone know how much ripple I can expect?
Title: Re: Lab power supply
Post by: Admin on December 01, 2006, 10:37:50 PM
the basic equation to calculate how long a capacitor can hold 50% of its charge (after about 50% it basically stops working):

time (sec) that cap can power something = C*V*.5 / I

so . . .

1/(1/1F + 1/1F + 1/1F + 1/1F)*12V*.5/.5 = 3 seconds, if I did that right . . .

I wouldnt expect any ripple at all . . . would your 500mA draw do anything at really high frequencies?
Title: Re: Lab power supply
Post by: Militoy on December 05, 2006, 10:41:48 PM
If you are asking about the ripple on your raw supply to the front end of the regulator, you should get around 10% ripple (more or less), from a full-wave bridge with a capacitor equal to at least the "critical" capacitance. With a 60 Hz input, that's around 1000uF per amp of load. Any good linear regulator circuit should easily cancel out that much ripple. You can actually end up with less than 0.1% p-p ripple on the output, with a 723-based linear regulator, and proper compensation. IMHO, 250,000uF of capacitance (4 1F caps in series) won't buy you much more than a single 1000uF electrolytic cap will at the current you're running - but it will buy you an enormous inrush current at turn-on - and maybe nuisance tripping of your input fuse or circuit breaker. As a couple of quick notes; you can expect around 15 VDC raw rectified output from a 12 VRMS transformer running through a FWB rectifier using a capacitor filter - but bear in mind that transformer manufacturers design for full-load output. This means that you should plan for higher raw voltage at light loads, and rate your components accordingly. With commercial transformers, regulation can vary widely. At only 40 degrees C temperature rise, this can mean up to 50% regulation (higher no-load voltage), if the manufacturer has gotten aggressive with his design margins. Best to test the transformer on the bench before making any assumptions. You should allow around 3 VDC compliance for your regulator (15 VDC raw voltage for a 12 VDC regulated output). Hope this helps.
Title: Re: Lab power supply
Post by: Admin on December 16, 2006, 01:47:22 PM
Ive been looking into supercapacitors some more . . . basically, they arent as neat as I thought they were . . .

Imagine a 1 ohm resistor in series with the cap. What happens if during a sudden peak, you try to draw 2A from that cap? 2A*1ohm = an internal voltage drop of 2V i.e. useless. Also how much heat? 2A*2A*1ohm = 4 watts of peak heat generated inside the cap = boom.  :o

(exploding capacitors make a lot of smoke and shoot stuff in the air, and is why I always wear safety goggles the first time I turn on a new circuit)

Check the resistance of those caps in the datasheet . . . will probably be around 30 ohms . . .
Title: Re: Lab power supply
Post by: Militoy on December 16, 2006, 09:28:49 PM
Ive been looking into supercapacitors some more . . . basically, they arent as neat as I thought they were . . .

You're right - there's no real magic to them - though they can be useful under the conditions they're designed for. Under quick discharge conditions, the ESR (equivalent series resistance) of a lytic cap isn’t usually such a big issue in and of itself, as the time period that the current is passed is very short. Rather than concern with dissipated power (watts) and resulting temperature rise in the cap, the biggest concern under this condition is the energy dissipated in joules (watt-seconds) through the internal lead contact points. If the energy dumped is too high over too short a time for the internal lead connections of the particular cap, one or more of the leads will be damaged or may fail, internally. A cap with an ESR of 30 Ω, that is charged to 10VDC will limit itself to a discharge current of 0.33A under a short-circuit condition. Most “super caps” are designed with relatively high ESR, to limit their discharge current to what the leads will handle - There are plenty out there that are not, though.  ESR and ESL (equivalent series inductance) can have a very detrimental effect under high AC ripple conditions, especially at high frequencies. Super caps are usually used with relatively ripple-free DC, for that reason. They make dandy rechargeable batteries at low voltages – but the user has to watch his charge/discharge cycles carefully, and keep the ripple down.