PSU Question - (Newbie)

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
I buy most of my components from Jameco, Newark, Mouser, or Tayda Electronics. They only buy components from manufacturers or authorized resellers. Jameco and Newark have been including letters attesting authenticity in their shipments.
You've convinced me. Thanks!
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
Hi, folks!

I tried calculating what would be necessary to power the fan and an LED from my (finally stable) 15V branch (I increased it to 15V for the LED to work).

I cam up with the attached circuit. It worked fine when I tried it on the current-controlled buck/boost device. When I did it on the L7815C branch, my resistor smoked.

I had figured the following:

desired amps: .02A
desired voltage drop: 1V (15V-12V for fan) - 2V for LED
R = 1V / .02A = 50ohm
P = (.02A)^2 * 50 = .02W

That seemed like safe margin for 1/4W resistor.

What am I not getting?

Thanks!

Carlo

Led Circuit 1.jpg
 
Last edited by a moderator:

LesJones

Joined Jan 8, 2017
4,174
First it is a BAD idea feeding a 12 volt fan with 15 volts.
For your LED you have to drop 13 volts across the resistor . (15 - 2) You want 0.02 amps through the LED. So the value of the resistor will be 13/0.02 = 650 ohms (The nearest standard value is 680 ohms which will be near enough. Next calculate the power dissipation in the resistor. This will be 13 volts times 0.02 amps which equals 0.26 watts.
With the 50 ohm resistor that you used it's dissipation would be 13 * 13 /50 = 169/50 = 3.38 watts.
(The current would have been 13/50 = 0.26 amps which would require quite a high power LED Your current value of 0.02 amps would be a suitable value for a normal 5mm LED.) I don't understand your reasoning for your calculations.

Les
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
First it is a BAD idea feeding a 12 volt fan with 15 volts.
For your LED you have to drop 13 volts across the resistor . (15 - 2) You want 0.02 amps through the LED. So the value of the resistor will be 13/0.02 = 650 ohms (The nearest standard value is 680 ohms which will be near enough. Next calculate the power dissipation in the resistor. This will be 13 volts times 0.02 amps which equals 0.26 watts.
With the 50 ohm resistor that you used it's dissipation would be 13 * 13 /50 = 169/50 = 3.38 watts.
(The current would have been 13/50 = 0.26 amps which would require quite a high power LED Your current value of 0.02 amps would be a suitable value for a normal 5mm LED.) I don't understand your reasoning for your calculations.

Les
Les,

Thank you for helping me understand! My "reasoning" is a misunderstanding of the basics. I will study your response and try to figure a circuit that actually works (and safely). Back to 12V now that I understand 15V-12V for the fan doesn't make sense.

Carlo
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
Joy! I used 650ohm 1W resistor (1/4W seemed too close to the edge) and all works! Even better, I understood your info and why I was thinking wrong about it!

I think it's ready to be boxed up!
 

LesJones

Joined Jan 8, 2017
4,174
If you want to drop the 15 volts down to 12 volts the simplest way is to connect 5 forward biased silicon diodes i series with the fan. That would drop the voltage by between 3 and 3.5 volts. Any of the the 1N400x range of diodes would be suitable. (Th "x" is any number from 1 to 7.)

Les.
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
Why are you using a 15V regulator for a 12V fan?
If you want to drop the 15 volts down to 12 volts the simplest way is to connect 5 forward biased silicon diodes i series with the fan
It was a misunderstanding on my part. I thought I had to account for a 12V drop by the fan and a 2V drop by the LED. I realize now that didn't make sense. So I made the switch to the 15V regulator but went back to the 12V regulator once I understood. Everything seems to be working great now.

Thanks!!

Carlo
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
Based on your input, Dennis, I am replacing all of my resistors and capacitors and most other components through the vendors you mentioned. It also got me thinking about the buck/boost controller I was using from Aliexpress. I replaced it with this:

https://www.amazon.com/gp/product/B01LWXAC5E/ref=ppx_yo_dt_b_asin_title_o03_s01?ie=UTF8&psc=1

Interested in any feedback. I saw it mentioned as a reasonable item for hobby purposes on some electronics forums.

Thanks!

Carlo
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
IMG_20210305_082217120.jpgIMG_20210305_082224887.jpgIMG_20210305_082217120.jpgBoard soldered up and tested! Now to make the enclosure. Thanks for all your patience and help!

Carlo (Most Improved Novice)
 

dl324

Joined Mar 30, 2015
16,839
Board soldered up and tested!
Congrats! It's always rewarding to finish a project.

If you're buying those prototype boards from AliExpress; a word of caution. I've found that they have a defect (shorted pads) rate of at least 10%. After buying several lots of 10, I decided to go for broke and just get 100. The lot I received had shorts on about 50% of the boards. I got a partial refund, but it was still a pain to have to sort through all of the boards to mark the shorts and then send pictures to AE.
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
Congrats! It's always rewarding to finish a project.

If you're buying those prototype boards from AliExpress; a word of caution. I've found that they have a defect (shorted pads) rate of at least 10%. After buying several lots of 10, I decided to go for broke and just get 100. The lot I received had shorts on about 50% of the boards. I got a partial refund, but it was still a pain to have to sort through all of the boards to mark the shorts and then send pictures to AE.
Thanks, as always, Dennis. You've helped me decide to order from the "safe" suppliers and I've replaced most of my parts.

Let me try to reason how to tell a shorted board: the copper pads touch? If so, I inspected this one after reading your post and it looks ok. Let me know if there's something else to look for.

Have a great weekend!

Carlo
 

dl324

Joined Mar 30, 2015
16,839
You've helped me decide to order from the "safe" suppliers and I've replaced most of my parts.
There are some parts on Ali Express that are reasonable to use. You just need to make sure that whatever you use is appropriate. Components might not have the correct tolerance or temperature coefficient, but if it doesn't affect your circuit, they can be used.

I bought some lots of 100 AO3400 and AO3401 MOSFETs because I didn't want to pay 10-20X more for parts from more reputable sources (different part numbers). I bought 100 of each and did some random testing of parameters (mainly threshold voltage and 100-200mA drain current) that were important to me before buying more. I thought about buying full reels of them since both parts are end-of-life, but I happened to see a picture of a place in China where people go to buy counterfeit parts and I saw reels of SMT components. That made me decide not to buy a reel.

The counterfeiters will sometimes include a small number of good parts in a "batch" and I didn't relish the idea of having to do 100% testing on thousands of parts.
Let me try to reason how to tell a shorted board: the copper pads touch?
More than half of the boards in a lot of 100 had shorts. Some only one or two, but my expectation was 100% of the boards free of defects that prevented them from being used as intended.
clipimage.jpg

When I use these boards, I usually use them with the pads on the component side. That way I don't need to worry about shorts caused by routing bare wires:
clipimage.jpg
 

Thread Starter

CCinBR200

Joined Oct 4, 2020
80
There are some parts on Ali Express that are reasonable to use. You just need to make sure that whatever you use is appropriate. Components might not have the correct tolerance or temperature coefficient, but if it doesn't affect your circuit, they can be used.

I bought some lots of 100 AO3400 and AO3401 MOSFETs because I didn't want to pay 10-20X more for parts from more reputable sources (different part numbers). I bought 100 of each and did some random testing of parameters (mainly threshold voltage and 100-200mA drain current) that were important to me before buying more. I thought about buying full reels of them since both parts are end-of-life, but I happened to see a picture of a place in China where people go to buy counterfeit parts and I saw reels of SMT components. That made me decide not to buy a reel.

The counterfeiters will sometimes include a small number of good parts in a "batch" and I didn't relish the idea of having to do 100% testing on thousands of parts.
More than half of the boards in a lot of 100 had shorts. Some only one or two, but my expectation was 100% of the boards free of defects that prevented them from being used as intended.
View attachment 232244

When I use these boards, I usually use them with the pads on the component side. That way I don't need to worry about shorts caused by routing bare wires:
View attachment 232246
Thanks for sharing your experience and the tip about putting the components on the pad side!
 

dl324

Joined Mar 30, 2015
16,839
Thanks for sharing your experience and the tip about putting the components on the pad side!
Another thing I do is do a board layout so I can optimize component layout and avoid wire crossings (which lets me use bare wire). This is the layout for that board:
clipimage.jpg
I designed it for 4 discrete flip flops. I extended it to 6 flip flops on the board I wired so I could use the whole board.
 
Top