using PWM to dim LED NEON STRIP

[MrDEB]

couldn't locate this thread but here is the issue
I breadboarded this circuit and it seems to work driving just an LED but I am driving a 12v led neon strip.
The circuit flashes the strip slowly to fast as per pot adjustment.
The 555 circuit works as planned with one LED (it is supposed to change the duty cycle
Thinking the two different supply voltages has something but what to do? Any suggestions?
NOTE THE 7555 IS POWERED BY THE pic
the MOSFET is logic level

 

[MrDEB]

LED NEON STRIP DESIGN
I have 3-Led strip projects
Design criteria
power supply = 12volts 5A max
LED neon strips = 12V
current draw = 0.8A/meter at 12volts (16.4 feet uncut length)
Desired current draw per Led strip section <= 1A ( .8A per 3.28feet) = .243A)
Each channel driving an Led strip per inch = .243A (1meter = 3.280feet / 12 = .273 inches
Keep sections of Led neon strips <= 3 feet = .8A per section
Using the posted pic of the Margarita sign as a sample

NEED: 9 outputs using Mosfets (< 2amps) for lettering so the lettering can be sequencely enabled
8 outputs for LED's, 20ma max per
1 Mosfet for the red straw
1 mosfet output for the glass
1 Mosfet for the lime
1 mosfet for the ice
7 mosfets for the green strips in glass





could use as suggested a ULN2003 (< = 500ma) but for future flexibility with the ability to dim sections of a design, the powerMosfets are desired.


Thought about using a MOSFET GATE DRIVER But for a betterFLEXABILITY and cost (< 25cents per channel), using the 555 timerdimmer circuit powered by 10 volts will increase the flexibility and ability to dim sections of a design and drive a standard Mosfet (thelogic level Mosfets have a higher Rds(on) than a standard Mosfet





Located a possible logic level with low Rds(on)https://www.taydaelectronics.com/datasheets/A-098.pdf


Considered using PWM on the micro but then I would need to reprogram or use the code that was suggested using ADC. Need to test if I can add more channels.

 

[For The Popcorn]

P = current * current * Rds
P = 4 * 9.5mohms

So what does power equal? Do you consider this to be a problem?

(1meter = 3.280feet / 12 = .273 inches

Does this make sense to you? 1 meter is about a quarter of a inch????

could use as suggested a ULN2003 (< = 500ma) but for future flexibility with the ability to dim sections of a design, the powerMosfets are desired.

How does the use of a ULN2003 prevent dimming? [via any rational means]

(thelogic level Mosfets have a higher Rds(on) than a standard Mosfet

This is only true for the logic level mosfets YOU have selected.

Thought about using a MOSFET GATE DRIVER

What does this accomplish? [Nothing for your applications]

But for a betterFLEXABILITY and cost (< 25cents per channel), using the 555 timerdimmer circuit powered by 10 volts will increase the flexibility and ability to dim sections of a design

Flexibility to SET a fixed brightness with a pot. Not flexibility to dynamically control brightness.

Considered using PWM on the micro but then I would need to reprogram or use the code that was suggested using ADC. Need to test if I can add more channels.

Please don't "FIX" the code written by somebody who actually understands what he's doing. Yes, it's expandable, but get it working with the three channels it's written for before you start making "improvements".

 

[tumbleweed]

Need to test if I can add more channels

Yes you can, but you've pretty much run out of pins for pots to adjust the levels.

If you don't mind one pot controlling the brightness of multiple outputs that can be done too.
Just give me a list of which pots control which outputs.

Try what's there before you muck around with it. It may need some tweaking since I tried to write it in a way that's easy to understand with no frills... for each channel take an ADC reading from the pot and use that directly as the PWM duty cycle.

 

[ChrisP58]

-

 

[MrDEB]

Well what to use for the 9,5mohms?
.0095, .095 or ??
I fail in math

 

[MrDEB]

I think I have this right but now what?
using this mosfet https://www.taydaelectronics.com/datasheets/A-098.pdf
Vgs or Vth =1-3v
P = current ^ current ^ Rds--4 ^ .0095 = .038
Pd = TJmax - Tambient / Rja -- 150 - 25 / 50 - 125 / 50 = 2.5 MAXIUM MOSFET POWER
Pd is the power dissipated? what does that tell me? this component will not need a heat sink? or what

 

[tumbleweed]

I'll admit this can get confusing... thetaJA, thetaJC, Pd, temperature, etc.

Here's a very simplified view:

Power Pd = current x current x Rds
Pd = 2 x 2 x 0.0095 = 0.038Watts (or 38mWatts)

To figure out a rough value for the temperature, look at Rtja (R theta junction to ambient).
Rtja is given in degreesC/W. What that basically tells you is how many degrees the part will rise for each watt of power.

So, to find the temperature rise take Rtja and multiply that by the power Pd:
Trise = Rtja x Pd, or in this case
Trise = Rtja x 0.038W

For a TO220 case, Rtja is typically around 60degC/W, so if you had one of those:
Trise = 60degC/W x 0.038W = 2.28 degC
If you're at normal room temperature (25degC) the part would be about 25 + 2.28 = 27.28 degC

BUT, that FDS6680A isn't in a TO220 case, it's in a little SO8 package.
If you look at the 'notes' section on pg2 they show Rtja values for three different PCB heatsink arrangements.
The last figure shows what the value for no heatsink, and that's 125degC/W.
A small plastic package gives a lot worse temperature... that makes sense if you think about it.

So, for the SO8 part w/no heatsink:
Trise = 125degC/W x 0.038W = 4.75 degC

The thing that saves your butt with that package is the low value for Rds.

For comparison, if you go back to the IRL520, assume Rds = 0.38 ohms and a TO220 case:
Pd = 2 x 2 x 0.38 = 1.52W

Trise = 60 degC/W x 1.52W = 91.2 degC
If you're at room temp of 25 degC, that part would be at 25 + 91.2 = 116.2 degC
Time to break out the marshmallows!

 

[MrDEB]

YES I viewed the diagram on page 2 thinking maybe put the package on the 1 sq inch but pins 1 - 3 can be connected together as well as pins 5 - 8 together, thus adding more heat sinking pad. pin 4 is the GATE,
curious why they didn't suggest this
Was hoping for a smaller package than the to220 pak. But at 77 cents per this might be the one to use. Need to convert the temp to degrees but is 4.75 degrees too much? IYO
according to the "addohms" work sheet, if P is bigger than Pd then a heat sink is required.

 

[MrDEB]

REVIEWING THE DATA AGAIN, MAX RATING FOR Rja is 50 and Rjc is 25
Tj = -55 to +150
We should be good hopefully.

 

[MrDEB]

Change of plans The SO8 is really small and needs lots of heat sinking.
Been mulling over the code Tumbleweed posted as I need to use an 18f43k22 to test then apply to an 18F2221 (all the 2221's I have are SMD and the 18f43k22 is soldered to a perf board (DIP socket)
"Meanwhile, back a the ranch, the lone ranger is racking his balls"
looked at GATE DRIVERS and discovered using a 555 timer to drive gate has been done.
Looking at using a bistable mode 555. Using a standard Mosfet IRF520 and applying 12volts to the gate and the 555 pin 3 drives the gate. Need to test for 2A and heat buildup.
A bench power supply sure would be handy but have none so contemplating building one using a LTC3780 DC ADJUSTABLE CONVERTER. lots TO DO

 

[For The Popcorn]

Get one of these. Connect a 12 volt or 19.5 volt power supply brick and you'll have something that works.

 

[tumbleweed]

The post in post 58 should work ok for the 18FxxK22.

...IRF520 and applying 12volts to the gate and the 555 pin 3 drives the gate. Need to test for 2A and heat buildup.

The IRF520 Rds is somewhere around 0.27 ohms, so that's Pd = 2 x 2 x 0.27 = 1.06W
With the TO-220 case that'll give about a 65 degC rise over room temp, so if you're in a 25 degC room it'll be roughly 25 + 65 = 90 degC sitting in free air. That's 194 deg F

That's warm, but the part should handle it. If you mount it to the PCB with some copper as a heatsink it should be ok as long as your room is < 65 degC.

 

[For The Popcorn]

I swear, I get dizzy watching MrDEB's "design process".

 

[ChrisP58]

I'll admit this can get confusing... thetaJA, thetaJC, Pd, temperature, etc.

Here's a very simplified view:

Power Pd = current x current x Rds
Pd = 2 x 2 x 0.0095 = 0.038Watts (or 38mWatts)

To figure out a rough value for the temperature, look at Rtja (R theta junction to ambient).
Rtja is given in degreesC/W. What that basically tells you is how many degrees the part will rise for each watt of power.

So, to find the temperature rise take Rtja and multiply that by the power Pd:
Trise = Rtja x Pd, or in this case
Trise = Rtja x 0.038W

For a TO220 case, Rtja is typically around 60degC/W, so if you had one of those:
Trise = 60degC/W x 0.038W = 2.28 degC
If you're at normal room temperature (25degC) the part would be about 25 + 2.28 = 27.28 degC

BUT, that FDS6680A isn't in a TO220 case, it's in a little SO8 package.
If you look at the 'notes' section on pg2 they show Rtja values for three different PCB heatsink arrangements.
The last figure shows what the value for no heatsink, and that's 125degC/W.
A small plastic package gives a lot worse temperature... that makes sense if you think about it.

So, for the SO8 part w/no heatsink:
Trise = 125degC/W x 0.038W = 4.75 degC

The thing that saves your butt with that package is the low value for Rds.

For comparison, if you go back to the IRL520, assume Rds = 0.38 ohms and a TO220 case:
Pd = 2 x 2 x 0.38 = 1.52W

Trise = 60 degC/W x 1.52W = 91.2 degC
If you're at room temp of 25 degC, that part would be at 25 + 91.2 = 116.2 degC
Time to break out the marshmallows!

I'd suggest this part: https://www.digikey.com/en/products/detail/diodes-incorporated/DMN2005UFG-7/4898854

The RdsOn at 4.5 Volts is only 5 milliOhm, so the power at 2 Amps is only 20 milliWatts. And with an RthJC of 4.2, the temperature rise is only a fraction of a degree.

 

[For The Popcorn]

Rather an impractical if not impossible package for somebody soldering with an iron don't you think? A 3mm × 3mm package with a power pad would be difficult for most people even with hot air or reflow.

 

[tumbleweed]

And with an RthJC of 4.2, the temperature rise is only a fraction of a degree.

That's useful if you're doing a heatsink design, but to me RthJA is a better indicator of what sort of temp rise you'll get using the part. With that, RthJA is 55 degC/W when following the datasheet assumptions.

Granted, the Rds is so low it's almost a moot point. You'd probably almost never know it's on.

 

[Pommie]

And MRDEB will use all of this information.

Mike.

 

[Pommie]

Maybe I should buy a lottery ticket this weekend. I'm feeling lucky.

Mike.

 

[MrDEB]

tHANKS FOR INFO IN POST#74
I have low melt paste and a heat gun so I Think I can manage to get it mounted hopefully.

 

[MrDEB]

yikes!
that package is 3.5mm square
not even going to attempt