Question on calculating dissipation

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
shotgunefx
Posts: 20
Joined: Mon Mar 08, 2004 1:01 am
Contact:

Question on calculating dissipation

Post by shotgunefx »

I'm new to electronics and I hope this isn't too stupid a question but I'm just having a really hard time wrapping my head around some things.<p>Take the following circuit
Image
In each instance, how do I calculate how much power R1 needs to dissipate at it's maximum resistance, how much at it's lowest resistance, (say 1 ohm)?<p>Do I calculate with 5v or is the effective voltage 3v (5v - led 2v drop)?<p>From my limited understanding, the leds are going to limit the amount of current that can flow. <p>If this understanding is correct than Circuit A will be limited to 20ma and Circuit B will be limited to 40ma maximum.<p>So when the pot is turned to it's maximum resistance of 10k are the following calculations correct?<p>Circuit A
P = EI = 5 x 0.020 = 0.1W
Circuit B
P = EI = 5 x 0.40 = 0.02W<p>What about at it's lowest resistance?
P = I^2R
Circuit A
0.020 * 0.020 x 151 = 0.06W
Circuit B
0.040 * 0.40 x 76 = 0.1216W<p>I'd appreciate it if any could enlighten me.<p>Thanks,
-Lee
User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Question on calculating dissipation

Post by Chris Smith »

Hey, thats my handle, Err, I mean thats my email address? <p>Voltage in minus Voltage led
over
Amps led<p>5 volts minus 2 volt drop led = 3
over
20milli amps = Series Resistor value [150]<p>Parallel units follow other ohms laws of halving or doubling. <p>
for example<p>Also parallel Diodes are not recomended on a single resistor, as they tend to draw individual currents, based upon batch qualities of production.<p>[ March 25, 2004: Message edited by: Chris Smith ]</p>
cato
Posts: 366
Joined: Sun Sep 07, 2003 1:01 am
Contact:

Re: Question on calculating dissipation

Post by cato »

The LEDs WILL not limit the current.<p>V=IR<p>P=VI<p>When the pot is set at 10K the current flow will be very low and the drop across the leds will be almost 0V, so:<p>5V = I (10,075) or I= 5V/10,075 or 0.0004962 Amps or less than 1/2 a milliamp (0.4962mA)<p>When you have the pot turned the other way a lot of current will flow. Assuming your 2V led drop is correct, there will be 3V to drop across the resistor and the pot (76 ohms)<p>Again V=IR <p>3V=I(76)<p>I=39.47 ma<p>Power in the pot :<p>P=IV
V=IR
P=I(IR)
P=0.039A X 0.039A X 1 ohm (pot resistance)<p>P=0.00156W<p>The leds will not share the current evenly. You should use two 150 ohm resistors(one in each leg) instead of the 75 ohm resistor.
shotgunefx
Posts: 20
Joined: Mon Mar 08, 2004 1:01 am
Contact:

Re: Question on calculating dissipation

Post by shotgunefx »

Thanks for the help. I meant to say that R2 was limiting the current rather than LEDs.<p>This is actually part of something larger I made for my car to control RGB leds. I currently have 3 1/4w (R,G, and B) slider pots but they are hard to find in the right physical size with the proper maximum resistance. I've only been able to find 20K audio taper so the adjustment range is only a 1/4 inch. I was going to replace them with smaller rotary pots but they only have .1W dissipation and was trying to figure if they would burn.<p>I know it's better to attach the resistor to each LED but at least in some of the instances I don't physically have room.
User avatar
Externet
Posts: 1888
Joined: Fri Feb 01, 2002 1:01 am
Location: Mideast USA
Contact:

Re: Question on calculating dissipation

Post by Externet »

Hi Lee.<p>Circuit A
At maximum resistance:
3V ÷ 10150 þ = 0.3mA.... will hardly shine.
Power dissipation on R1 = 0.0003² x 10000 = ~1 mW
At minimum resistance:
3V ÷ 151 þ = ~20mA... OK
Power dissipation on R1 = 0.02² x 1 = 0.4 mW<p>Circuit B
At maximum resistance:
3V ÷ 10075 þ = 0.3mA... will hardly shine.
Power dissipation on R1 = 0.0003² x 10000 = ~1mW
At minimum resistance:
3V ÷ 76 þ = ~40mA... OK
Power dissipation on R1 = 0.04² x 1 = 1.6mW<p>To lit 2 LEDs at 20mA, better put them in series to a 50 ohm resistor fed by the 5 V. Again, a 10K potentiometer will reduce the current too much to lit.
If what you are after is to vary the brightness, the potentiometer should be of much less resistance, about 1K, and its exact value depends on the chosen LED efficiency.<p>Miguel
- Abolish the deciBel ! -
shotgunefx
Posts: 20
Joined: Mon Mar 08, 2004 1:01 am
Contact:

Re: Question on calculating dissipation

Post by shotgunefx »

Thanks Miguel.<p>So if there are 20 Leds than the following would be correct?<p>ower dissipation on R1 = 0.400² x 1 = 0.016 Watts? <p>So a .10 watt pot should be more than able?
You're certainly right about the pots being oversized. The new ones are 10K but offer enough adjustment.<p>[ March 25, 2004: Message edited by: Shotgun ]</p>
User avatar
Externet
Posts: 1888
Joined: Fri Feb 01, 2002 1:01 am
Location: Mideast USA
Contact:

Re: Question on calculating dissipation

Post by Externet »

¿Twenty LEDs? If you use them in parallel as per your circuit 2, chances are brightness will be very uneven; that is NOT the way to do it.<p>And if you want each to conduct 20mA in parallel, it will be 20mA x 20 = 0.4Amperes.<p>To feed that current from 5 Volts;
3V ÷ 0.4A = 7.5 þ as R2<p>With R1 at 1 þ minimum resistance:
3V ÷ 8.5 þ = ~0.35 A
Power dissipated at R1 = 0.35² x 1 = ~0.12 Watts<p>Miguel<p>¡ Damn Ohms symbol !
- Abolish the deciBel ! -
shotgunefx
Posts: 20
Joined: Mon Mar 08, 2004 1:01 am
Contact:

Re: Question on calculating dissipation

Post by shotgunefx »

Acutally the eveness is not as bad as you would think. I've wanted to get into electronics for some time and this seemed like a good starter project. It ended up being much more complicated than I thought.<p>What I actually have is a 5v 317t based regulator that feeds the three control pots. Each control pot goes to a box and feeds six 4-wire RJ45 jacks. (So I can have up to six "strands" of leds. The box has 3 20turn pots for each strand so I can adjust the current for the number of lights connected to it. <p>As I was new and didn't know what the hell I was doing, using the 20turn pots seemed the easy way out as I could adjust them easily (and add/remove lights).<p>The biggest qualm I have about using individual resistors is I don't have a lot of space. I couldn't quite think of a good housing scheme that would fit. Can you heat shrink over a resistor? I would think it would interfere with it's ability to dissipate the heat. <p>Here's a pic of what it looks like
Image<p>They are also in the guage cluster and the fog lights. I'm probably going to end up going to PIC based at some point and use PWM to control the dimming (and add some other features)<p>Thanks again for all the help.<p>[ March 25, 2004: Message edited by: Shotgun ]</p>
User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Question on calculating dissipation

Post by MrAl »

Hello Shotgun,<p>Adding to the other posts...<p>Normally when you start out learning electronic circuits you
start with linear circuits and work your way up to non linears.
This is a non linear circuit so it's a little more complicated
then a linear circuit, but it isnt too bad.<p>
The analysis of this circuit goes something like this...<p>The power in the resistor R1 is low when R1 is low and also
low when R1 is high, so first we need to calculate the max
power point of the resistor. In other words, when R1
dissipates the maximum power what is it's resistance?<p>To accomplish this we start by writing an equation for
the circuit in terms of current and voltage, then solve
it for current and plug that result into the power
equation for R1 which is simply P=I^2*R as you know.
Doing this leads to this result:<p>P=R1*(V-Vd)^2/(R1+R2)^2
where
P is the power in resistor R1
R1 is the variable resistor
V is the source voltage (5v in your circuit)
Vd is the voltage across an LED (2v approximated)
R2 is the fixed resistor<p>To find the max power we take the first derivative of
P with respect to R1 and set the result equal to zero
and solve for the solutions, or simply graph the
equation for P above and look for high point maximums.
Doing this and generalizing the results we find that
max power is dissipated in R1 when R1=R2.<p>Noting that last result, we also note that when R1=R2
we also have half the current flowing through R1,
so this allows us to generalize once again about the
current flowing through R1:
The current flowing through R1 will be half the max
total LED current. With the number of LED's equal to N
and each one drawing 20ma nominal, the current through
R1 max will be
iR1max=0.02*N
where
N is the number of LED's in parallel.<p>Also noting that R2's value depends on the number of
LED's and if we assume a 5v source voltage we find that:
R2=150/N
where
N is again the number of LED's in parallel.<p>Since we now know what R2 is we also know what R1 is when
the max power point occurs:
R1=R2=150/N<p>Since the total resistance will be Rt=R1+R2=2R2 when the max
power point occurs the current will be half or 0.010*N amps.<p>Knowing what the current will be and that the resistance will
be 150/N we can now calculate the power and solve for the
maximum number of LED's for a given pot rating:<p>P=(0.010*N)*(0.010*N)*(150/N)
P=0.0001*N*N*(150/N)
P=0.015*N<p>and now we can see how high N can go assuming we can utilize the
full power rating of the pot of 0.1 watt. Setting P=0.1 watt in
P=0.015*N we get:<p>0.1=0.015*N<p>and solving for N we get:<p>N=0.1/0.015
N=6.667<p>so the max number of LEDs should be about 6.<p>
Take care,
Al
LEDs vs Bulbs, LEDs are winning.
rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: Question on calculating dissipation

Post by rshayes »

I think that an important point is being missed here.<p>The potentiometer is being used as a rheostat. The power rating of the potentiometer only applies when the power is distributed over the full length of the resistive element. It must be derated when less than the full element is being used. For example, when the resistance is set to 10% of maximum, all of the power dissipated is being dissipated in 10% of the resistive element, and the power rating is about 10% of the normal rating.<p>There may also be a maximum current that can be carried by the moving contact, and this may be a limit in low value potentiometers.<p>It is better to think of the potentiometer as having a current rating. Using P=I^2*R for a 10K, .1 watt potentiometer gives a current rating of about 3.2 milliamps for the potentiometer. A 1K potentiometer could handle about 10 milliamps.<p>Non-linear potentiometers, such as an audio taper, have to be rated based on the highest resistance section of the potentiometer. This can easily reduce the allowable current by a factor of 5 or 10, depending on the taper.<p>In circuit A, when the potentiometer is set to zero, the current is about 20 milliamps. The potentiometer should be rated at at least 4 watts. In circuit B, the current can be twice as high, so the potentiometer should be rated at at least 16 watts. (P=I^2*R)
User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Question on calculating dissipation

Post by MrAl »

Hello again,<p>You might be on to something this time stephen,
as the power rating probably cant be the same
over the entire length of the element.
I would think the power rating would decrease
as a linear function of position because the
element is most likely similar to a plate.
The only problem is, it depends on how the
manufacturer rates their pots and it appears
that each individual model pot has different
ratings.
I couldnt find anything on rating a pot after
it's been cranked a few turns, but i did find
some other interesting things that will also
play a part in this application. That is,
the rating that specifies how many times the
pot can be cranked one end to the other before
it starts to wear out. I was surprised to find
that some pots are only rated for 20 cycles!
Yeah, that is amazing. After that, no matter
what the current/power rating is the pot begins
to wear out. Apparently some of the trim pots
are made to be adjusted one or two times with the
service life of the appliance that that's all.
The good ones are all wire-wound.<p>All i can suggest at this point is to look
up the specs on the exact pot model and go
from there, but certainly the one to get will
be the one with some 10,000 possible cycles of
use, not 20 or even 200, if the pot will be
continuously adjusted during normal use.<p>Take care,
Al
LEDs vs Bulbs, LEDs are winning.
shotgunefx
Posts: 20
Joined: Mon Mar 08, 2004 1:01 am
Contact:

Re: Question on calculating dissipation

Post by shotgunefx »

Thanks for all the info. (I ended up using the .1W pots before your replies).<p>So far with 100ma going through, they don't even get warm. There may well be some other aspect of my setup that accounts for this, but so far so good.
Post Reply

Who is online

Users browsing this forum: No registered users and 49 guests