my experience so far with the TI dc-dc

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 8:17 pm

im sure many of you have opened this thread excitedly, hoping to solve your problems. alas, i have no answer, only more questions.

i started trying to sort my power today. i got a temporary (much smaller) 7.4v lithium battery (which has no protection circuit :O!), and i was trying to hook it all up with the dc-dc and everything. so here are my observations

system:wall / screen:wall
all is as it should be

system:battery / screen:wall
this means that the dc-dc is plugged in and operating. i see some faint diagonal lines that move quickly and continuously across the screen, but i figure i can fix it, and even if i cant, im willing to pay that price for efficiecny.

system:battery / screen:battery
EXTREME SCREEN DISTORTION. man. the whole thing goes to hell. right from the moment you turn it on.

so. allow me to discuss my theories.

the DC-DC, as we all should know, creates voltage ripple. when i have the system on battery and the screen on the wall adapter, i can see evidence of this on the screen - the diagonal lines. they are diagonal since the frequency of the noise is differnet from the vertical frequency of the screen. makes sense. since my screen's ground is tied to the system's ground in no less than FIVE places - 4 of them being the mounting brackets, the ripple on the system's ground shows through a little.

now, when i hook the screen up to the battery, im doing two things.

1- im completing a path for the noise to come through
2- im doubling the load on my batteries. (no led mod)

is it one of these things? both? neither?...

lets go back to the dc-dc converter for a minute. what it is doing is rapidly switching the power on and off. now, the voltage coming off my battery pack is directly related to the current draw. due to internal resistance, the more current you draw, the more the voltage drops. so the voltage on the unregualted side is probably taking a roller coaster ride. nothing that a cap shouldnt be able totheoretially fix though.

also, the regulator is designed for NINE VOLTS. people around here seem to keep forgetting that. NINE volts, and im giving it 7.2ish. and thats nine volts minimum, it will operate at anywhere to 15. so its not like we're giving it a little less than it needs... we're giving it a little less than the minimum amount it needs.. subtle difference. you think that if it would work at 7 volts, or even 8, that they would have included that in the range of voltages that are specified.

-- some things to think about anyways

so later this week, im going to bring it to school and set it up on a power supply. there, i can give it whatever voltage i want to, and see how it responds. also, the new digital scope came in today, and you know im just itching to give it a try. ill update you in a few days time with all i have learned.

for my own laziness, whats the max voltage that the screen will tolerate?

vb_master · Post by **vb_master** » Tue Mar 28, 2006 8:19 pm

Maybe this is why they are free - because they don't work!

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 8:20 pm

har-har

codeman · Post by **codeman** » Tue Mar 28, 2006 8:24 pm

ooooh...what kind of scope is it?

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 8:26 pm

i actually know nothing about it, other than its digital, and i think it has 4 channels. ill find out tomorrow

do you have any ideas codeman? you seem to know your stuff

Turbo Tax 1.0 · Post by **Turbo Tax 1.0** » Tue Mar 28, 2006 8:32 pm

what about rf sheilding of some sort?
would that do any good?

EDIT: wait a minute i just realized that iot probably wouldnt work

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 8:40 pm

thought of that, i just moved the converter far away

Post by **Skyone** » Tue Mar 28, 2006 8:41 pm

Sooo, what do you think is causing this? The voltage amount?

Krepticor · Post by **Krepticor** » Tue Mar 28, 2006 8:55 pm

didn't segasonicfan have the same problem with the pstwop?

Didn't he fix it with a 40 uH inductor? on the cathode end of the battery

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 8:58 pm

skyone: i pretty well posted everything i thought.... i dont know yet.

krepticor: i dont know. can you find the thread? it might help..

Krepticor · Post by **Krepticor** » Tue Mar 28, 2006 9:00 pm

there isn't a thread Its on his website

www.freewebs.com/segasonicfan/

codeman · Post by **codeman** » Tue Mar 28, 2006 9:06 pm

i can make up something intelligent sounding if you want

I've only really had a brief introduction to switching regulators, but I know that they have to be very finely tuned to keep them stable. Looking at the datasheet they show at least a 4V minimum difference between the input and output voltages, which seems a bit excessive. My only guess would be that they made tradeoffs between voltage drop, current capability and component size. If they increased the inductor size they could source more current, but the voltage across the inductor (input - output) would have to be greater to ensure that the inductor could recover fast enough at high loads. If the voltage across the inductor were only 2.4v or so, it can probably only source half it's rated current with an acceptable ripple...which might explain why it somewhat worked powering just the nintendo.

Anyway, I just based that off the fact that for an inductor V = L*di/dt, so a larger inductor would cause a slower current recovery at a particular voltage. A bigger voltage would would allow for a faster recovery, reducing the current and voltage ripple and increasing the current capability.

...long winded and cryptic, but does it make sense?

Post by **Skyone** » Tue Mar 28, 2006 9:06 pm

Well this makes us look stupid.

-Vertical line fix: using a 40uH inductor on the cathode input I was able to remove the vetical line interference caused by one battery operation! This is what enables the whole thing to run off one 12v batt with crystal clear video =)

Krepticor · Post by **Krepticor** » Tue Mar 28, 2006 9:08 pm

thank you Segasonicfan (even though he would have hated helping out a nintendo forum!)

timmeh87 · Post by **timmeh87** » Tue Mar 28, 2006 9:21 pm

well believe me, i saw what i saw and a 40uH inductor isnt going to do anything. i bet the wiring alone has that much inductance. I have a more serions problem on my hands. maybe once i fix that, and i still have a vertical line problem... maybe then ill need an inductor.

codeman: i happen to be responsible for all these equations on my upcoming physics exam, so i better understand it! (and i sorta do... i didnt go to many of those lectures). anyways. ill see what happens when i give it like 10 volts.

will the psone screen tolerate 10 volts? anyone?