Hi, my dad gave me a SONAR DC POWER SUPPLY model: PS-2515 (its pretty old), and I used that as a power supply for my PIC16F628 programmer, but there is some problem (no I didnt nlow any chips) I am smart and did not connect the IC's until I was certain that the voltage regulator/1N4001 diode combo worked properly. Anyway the voltage regulator seems to be only supplying ~3V. I need somewhere close to 5V if I want to power the IC's.

I was wondering my power supply seems to turn of the voltage LED (now the voltage knob is inaccessible) and turns on a current LED when it detects a load on its output lines, Should I set the appropriate voltage first (input voltage for 7805 voltage regulator seems to be ~7V ) then connect the wire to the circuit?

Also I noticed that the voltage regulator puts out quite a bit of heat, not enough to burn ones hand but it gets warmer than it is at an idle state, is this is a warning sign?

Thank you all for your help

Markings on voltage regulator:

L7805CV
88219
Posted on 2003-04-28 20:26:54 by x86asm
The 7805 regulator has to sink current to maintain voltage if there is no load. So it will still run warm. If it's holding at 3v then, yes, just adjust it to 5 and you should be OK assuming it can supply your circuit with enough juice.
Posted on 2003-04-28 20:38:11 by drhowarddrfine
Here is a piece of electronic knowledge that should be understood well and placed safely in you heart and sole:

GOOGLE: L7805CV ;)

Heres what i found : L7805CV

Pay some attention to the curve Figure 6 on page 15, as well as its specs.

There is also some simple circuit arrangements their too, good place to start learning stuff from (feel free to ask question if one stikes your fancy ;) )

PS: Old cpu power supplies make EXCELLENT test bench supplies for beginners. You dont even need to open it up; comes with an on/off switch; high current with good voltage regulation; a cooling fan; standard power cord (to keep your self from being electricuted); and rail voltages for Op-ams (+12 V / - 12V) as well as TTL (+5V) and ground!

:alright:
NaN
Posted on 2003-04-28 21:29:58 by NaN
If your PS-2515 power supply works like most do, the voltage light is on for 'normal' operation, and the current light on for constant current mode. You usually set the constant current for slightly higher then you expect the circuit to draw, that way you have some limiting in case of a problem.

(I say 'norlam' nad 'usually' because you just may want to control the current and not care what the voltage is, but thats somewhat rare to do.)

If the supply has both voltage and current meters, you can set the current limit by shorting out the output, and adjust the current for what you need. A PIC and a few LEDs would only need in the 10's or 20's of milliamps, so set it for say 50 to 100 mA. That will reduce blinding arcs when you get your first short <grin>.

The 7805 is a fixed 5 volt regulator. There are some tricks you can play to get a higher output voltage, but if you grounded the ground terminal, you get 5 volts.

If there is no load, it should be cool. Its quiescent current is near zero, so with no load current, no power to dissipate.

If a 7805 gives out only 3 volts, something is wrong. First check the input, give it 10 volts to start here, thats guarnteed to let it run with 5V out. If still at 3 volts, recheck the wiring. Check it a fifth time, I'm pretty sure its wrong. I personally have transistor dyslexia (the complete inability to properly wire a three terminal device in less then 6 tries) so I'm pretty well aquainted with how parts work when you wire them helter-skelter.

If still bad, there is a chance the 7805 is oscilating. Yeah, they can still do this. Add some 0.1 uF caps from input to ground, and output to ground. Add a third 10uF electrolytic or tantalum at the input is you have more then 6-12 inches of wire between the 7805 and the PS-2515.

If all that fails, go to Radio Shack and get a new 7805. Its dead Jim.

ONe nice thing about these parts is once properly wired, they are near indestructable. If you draw too much current and they overheat, they will cut back the output voltage to keep themselves safe.

About the only way to blow one in circuit is to reverse bias them. That can happen on power-off if the input voltage goes to zero before the output caps discharge. If this worries you, add another 1N-xxxx diode from output to input so its normally reverse biased.

Ya gotta have good power before you can continue (as you know).
Posted on 2003-04-28 23:06:02 by Ernie
Can you provide the actually details on the PS-2515; google doesn't know what you're talking about.

Can you post a schematic of your supply?

How exactly does the LED know that there is current flowing. I hope that it is not in series.

Capacitors are a must. If you have a well regulated supply, and a small circuit, you might be able to get away with no caps (excluding those in the Power Supply). You should have a 100uF on the point of entry, and .01uF, .1uF, & 1uF at every chip. The smaller values are needed as they react faster than larger value capacitors. If you have a really large circuit, you may add some 10uF and 100uF every so often.

A word of caution with using an old PC supply. Some of the switching supplies used in PCs are not designed for no load conditions. They will do their best to generate and infinite voltage; but some insulator will break down and you'll let the smoke out of something important. This problem may not exist in any of the supplies today, but it is very common in older (pre 2000) supplies.
Posted on 2003-04-29 01:00:35 by eet_1024
I fixed the problem, Thanks guys I think my wiring was wrong so I rewired the whole power part of the circuit.

I have a 100microfarad, 16V electrolytic capacitor I think that will do the job that I want it to do. Should I connect the cap between the L7805's common ground and output or common ground and input?

I fixed the problem it was because I didnt turn up the voltage properly, I put it up to around 5~7VDC and then the voltage regulator gave out a nice round even 5.01V which is good enough to power the digital logic in this circuit.

Which PC PS are you talking about NaN a ATX or AT or could I use either?

hahaha nice joke Ernie BTW, I like blinding arcs =P

I believe I was forcing too much current through the 7805 I'm lucky it didnt blow, but in a sense I have about another 4 of these right behind me so no bigge for me :) .


I'll take a look at the datasheet of the 7805 and look for circuit arrangements thanks guys :)

I only wired the 7805, the 1N4001 diode and tested to make sure output voltage is ~5V so I dont blow any chips :)
Posted on 2003-04-29 19:28:03 by x86asm
Time for a quick lesson in capasitors (Ernie feel free to add any experience tips ;) )

Capasitors are energy storage devices. Think of them as a big barrel coulombs (like a barrel of water).

I have not gotten into the frequency side of electronics yet in my tutorials, but here is an early start. Its a well known property of reactive devices (capasitors and inductors) that their overall impedance (or resistance to current flow) is found by:

X(c) = Reactance of Capasitor = [ 2*pi*freq*CapSize ]^(-1)
X(l) = Reactance of Inductor = [ 2*pi*freq*InductorSize ]^(+1)

This means all things being constant. the higher the frequency, the smaller the resulting reactance will be for a capasitor. Reactance is a component of the overal impedance (resistance to frequency based current flow).

So this in short means, higher frequencies will pass through with less hassle than lower frequencies for the same capasitor!! To limit this result you can make your capasitor small such that only really high frequencies will take advantage of this property. This is why you will find a small capasitor on the output of things like regulators to ground. They are there to allow only high frequency ripples and noise to shunt to ground (removing them from your circuitry)! The lower frequencies and the overall output are not shunted so easily. The result is a cleaner output signal. (The higher frequency ripples took the easy way out ;) )

Now then why the big capasitors. Well they are there to take a less techical spin on your circuit. They are BIG tanks of coulombs. They are their to provide energy (or current flow) when needed spontaneously from a chip. A normal circuit should provide the current anyways, but sometimes heavy switching in a chip can drain its internal store of coulombs to ground faster than the supplying power circuit expected. Having a LARGE "coulomb tower" (1uF cap) near by will easily dump its payload into the chip (and surrounding circuitry) before any "odd" chip behaviour happens.

This is why you often see people insist on putting caps across the supply/ground pins near every chip. Remember that chips are really little "cities", where your circuit is more like barren country roads interconnecting these "cities". To have a local "water tower" for each city helps them operate more smoothly than hoping water will come in time from "across country" ;)

Hope this helps.
:NaN:


PS: x86asm, dont forget that PIN 18 is your most important connection in your entire circuit ;)
Posted on 2003-04-29 20:32:59 by NaN
OK NaN :-)

Not all capacitors are created equal. Browse any catalog and you will see tons of styles, packages, ect.

For most low voltage work, you'll only need 3 or 4 types (meaning thats how many I get away with).

First is the ceramic capacitor. Available as leaded or SMD mount. Very close to an ideal capacitor, very low series resistance, leakage, and inductance (especially in the SMD package). Values from 1pF to .1 uF, vooltage rating to 200 volts, though you will find the higher values have a lower voltage rating.

Ceramic caps come in two main variants: XR7 and NPO. I forgfet what they stand for, but XR7 are for general purpose use, while the NPO's have a very stable capacitance over temperature, thus making them good in timing sircuits and the like. XR7 will be cheaper. Ceramics are also great for bypassing the power supply at each high current IC.

Next is the good old electrlytic. Values from about 1uF to lots (seen 100,000uF I can recal), voltage ratings from 6V on up to a few hundred. Think of these caps as DC caps, they are terrible at even modest higher frequencies due to series inductance. Best use for things like power supply smoothing caps.

Tantalum caps bridge the gap between electrolytic and ceramic caps. They have big capacitance values, and pretty decent high frequency performance too. They are also low leakage so they do good in timing circuits (sorry, don't know offhand how they do over temperature; if I need a temperature stable cap, I always pick an NPO).



For the circuit at hand (and thanks for putting the schematic up), I'd say put your 100uF at the input to the 7805 after the diode. I'd personally use 10uF there, but 100uF should be fine. Also put a .01uF ceramic (XR7 is fine) parallel to the electrolytic (for high frequency filtering), and a 2nd ceramic on the 7805's output.
Posted on 2003-04-30 00:12:35 by Ernie

OK NaN :-)

Not all capacitors are created equal. Browse any catalog and you will see tons of styles, packages, ect.

For most low voltage work, you'll only need 3 or 4 types (meaning thats how many I get away with).

First is the ceramic capacitor. Available as leaded or SMD mount. Very close to an ideal capacitor, very low series resistance, leakage, and inductance (especially in the SMD package). Values from 1pF to .1 uF, vooltage rating to 200 volts, though you will find the higher values have a lower voltage rating.

Ceramic caps come in two main variants: XR7 and NPO. I forgfet what they stand for, but XR7 are for general purpose use, while the NPO's have a very stable capacitance over temperature, thus making them good in timing sircuits and the like. XR7 will be cheaper. Ceramics are also great for bypassing the power supply at each high current IC.

Next is the good old electrlytic. Values from about 1uF to lots (seen 100,000uF I can recal), voltage ratings from 6V on up to a few hundred. Think of these caps as DC caps, they are terrible at even modest higher frequencies due to series inductance. Best use for things like power supply smoothing caps.

Tantalum caps bridge the gap between electrolytic and ceramic caps. They have big capacitance values, and pretty decent high frequency performance too. They are also low leakage so they do good in timing circuits (sorry, don't know offhand how they do over temperature; if I need a temperature stable cap, I always pick an NPO).



For the circuit at hand (and thanks for putting the schematic up), I'd say put your 100uF at the input to the 7805 after the diode. I'd personally use 10uF there, but 100uF should be fine. Also put a .01uF ceramic (XR7 is fine) parallel to the electrolytic (for high frequency filtering), and a 2nd ceramic on the 7805's output.


Thanks everyone I will take your suggestions, interesting tut NaN :D
Posted on 2003-04-30 07:35:21 by x86asm