Problems with 5v regulator

Post by **marshallh** » Sat Dec 10, 2005 11:06 am

I've built a 5v switching regulator based off a LM2672, like b52down. However, it's got problems.

I am using the SNES 10v power supply.
When no load is applied, the output is 5.01v, pretty good.

When the SNES is wired to it, the output drops to 4.28v and the SNES won't boot.
I've tried to keep the wiring as short as possible, any ideas why it won't keep a steady output?

Any ideas?

Kurt_ · Post by **Kurt_** » Sat Dec 10, 2005 3:16 pm

Newb talking, probably best to just ignore me, but...

Ever think that 4.3v isn't enough to run a SNES?

Post by **Sparkfist** » Sat Dec 10, 2005 3:49 pm

Many (myself included) think 4.8v is too little to run an SNES. I'd say if you want to power the SNES and make it simple, but an LDO that is rated to output 5v. You can put 6v in and so there will be less heat generated and you'll only have a single IC to replace the 7805.

Post by **gannon** » Sat Dec 10, 2005 5:10 pm

I'm not sure why the 5V circuit is dropping so much under load. I'll check out the datasheet for the chip later.

Post by **gannon** » Sat Dec 10, 2005 5:56 pm

Looks like you have everything right.
Values seem a bit off, but not sure if that'd be a big problem or not.
Cin = 22uF, 50V Tantalum
Cout = 47uF, 25V Tantalum
D1 = 3.3A, 50V Schottkey Rectifier
L1 = 68uH Sumida
Cb = 0.01uF, 50V Ceramic

Post by **Triton** » Sat Dec 10, 2005 6:52 pm

for the mini snes with bypassed 7805 it will run at a bare minimum of 4.3v

Kurt_ · Post by **Kurt_** » Sat Dec 10, 2005 7:16 pm

You just got PWN3D by a newb

Sometimes it's better to think simple before thinking crazy complicated. Simple is always better

Post by **marshallh** » Sat Dec 10, 2005 7:26 pm

Erm, for all you people who unfortunately think I don't know anything about a SNES, It can run from 4.7v on up to 5.2 or so.

To those people saying "Use a LDO linear regulator", well I could but what's the point?
Linear regulators are incredibly inefficient, while switching regulators have about 92% efficiency. See a difference?

That's why a switching regulator matters, and that's why I'm building one. Doesn't 30mins more play time sound pretty good?

Gannon: what datasheet did you get those values from? I'm using a National chip, with parts straight from the datasheet.
http://www.national.com/ds/LM/LM2672.pdf
That's the parts I used, the reference design is on the first page.

Post by **gannon** » Sat Dec 10, 2005 7:33 pm

I used the national one from http://www.ortodoxism.ro/datasheets/nat ... LM2672.pdf (first result from google)

Post by **Triton** » Sat Dec 10, 2005 8:02 pm

i know for a fact that the snes mini will run without problems as low as 4.3v it loses sound at 4.2 and crashes at anything lower than that. i dont know how high unregulated it can go but it worked fine at 5.6v

timmeh87 · Post by **timmeh87** » Sat Dec 10, 2005 10:25 pm

should be working. i doubt youre drawing more than 1.5 amps off it.. right? maybe the SNES power supply is too noisy. I've seen some pretty lousy power supplys, that appear to just be doing straight rectification of the ac.( ie. the voltage/time graph looks like a roller coaster.)

try hooking it up to a 7.2v RC pack?

your diode is like, forwards and everything right?

b52down · Post by **b52down** » Sun Dec 11, 2005 12:21 am

I noticed that your output cap is 63V 68 uF--is your input cap have a lower or
higher rating than 63V? If so, it might be that when you put a load on this setup, one of the two caps has too high of a voltage rating. If you look at the testing setup in the documentation for this regulator, the input cap voltage rating is 50V, while the output cap voltage rating is 25V.

If you refer to page 10 of the documentation for this regulator, it will go step-by-step on how to select the components.

I hope this helps.

nos_slived · Post by **nos_slived** » Sun Dec 11, 2005 2:08 am

I suggest trying one of TI's switching regulators. Free, and pretty efficient.

Post by **marshallh** » Sun Dec 11, 2005 9:51 am

timmeh87 wrote:should be working. i doubt youre drawing more than 1.5 amps off it.. right? maybe the SNES power supply is too noisy. I've seen some pretty lousy power supplys, that appear to just be doing straight rectification of the ac.( ie. the voltage/time graph looks like a roller coaster.)

try hooking it up to a 7.2v RC pack?

your diode is like, forwards and everything right?

Good point, the SNES power supply is unregulated and does fluctate quite a bit. I'll try a battery pack.

b52down wrote:I noticed that your output cap is 63V 68 uF--is your input cap have a lower or
higher rating than 63V? If so, it might be that when you put a load on this setup, one of the two caps has too high of a voltage rating. If you look at the testing setup in the documentation for this regulator, the input cap voltage rating is 50V, while the output cap voltage rating is 25V.

Good point! My input cap has a 40v rating! I suppose I should find one with less?

A wire slipped and 12v ran through the SNES board. Oh crap. Time to hit ebay!

timmeh87 · Post by **timmeh87** » Sun Dec 11, 2005 3:17 pm

as far as i know, the votlage rating on caps is irrelevant. it just the max voltage before the thing explodes. a 10uf cap rated for any voltage still only holds 10uf of current. so why not always just use ones rated for a million volts? because they are bigger and more expensive.