Looking for 2nd opinion: +12v & +5v only, no need for AT/ATX or neg rails

**Gibson99** · 08-22-2004, 12:44 PM

Originally Posted by rando

Rob's supply has a turn-on wire that can be fed from the carnetix so that they come on/off together.

duh, that makes sense. sometimes i overlook the most obvious things and make things more complicated than they need to be.

i guess i could simply tap the turn-on in the 5v section from the car's ACC lead inside the cnx, and the 5v out on the white wire on the cnx. no need to tap into anything or steal current from anywhere. the simpler the better!

Rob: what size (watts) is the 15kilohm resistor in your p/s? just something small like 1/4W just to drop the voltage down to whatever your PTxxxxs are looking for on that pin? also, if i'm reading your schematic right, i'll br bridging several pins on the PT6625 - 4-6, 7-10, and 11-13. pins 1 and 3 are the only ones with a single wire connected to them. right?

if you can't tell, i'm not that great at reading schematics, and my knowledge of electronic theory is greatly lacking. your PCB doesn't make any sense to me at all without components on it. but that's just because i'm not that familiar with circuits - more just components, such as installing radios, dvd players, alarms etc into cars, and basic home electrical wiring.

**rando** · 08-22-2004, 02:17 PM

A 1/4W resistor will be more than enough. The 15k & 5V zener path regulates the STBY voltage to a safe 5V level. Assuming you provide 16V max to that pin, your 15k resistor would drop 10V. P = V*V/R = 733microW.

Those pins are internally bridged on the PT6625. They correspond to input, ground, output. You should hook them all up to improve maximum current handling. Pin 1 is an output sense/feedback signal. Pin 3 is the standby (on/off) pin.

**Gibson99** · 08-22-2004, 02:46 PM

i just read the datasheet on the 6625 here: http://www-s.ti.com/sc/ds/pt6625.pdf

i'm confused. on page 5 of that pdf, it says that if you have an open circuit (which i interpret as no connection) on pin3, then the regulator will supply a regulated output whenever there's a valid Vin. i thought an open circuit was what you had when you turned off the ignition.

but then there's a little table below that paragraph...

Code:

Table 1 Inhibit Control Thresholds
Parameter     Min     Max
Enable(VIH)    1V     5V
Disable(VIL)    –0.1V     0.3V

to me, that table means that if you put voltage on pin3 it turns on the output, and no voltage (open circuit?) turns it off.

obviously since putting 5v(max) into pin3 turns it on, and no voltage to pin3 turns it off according to rob's schematic, then my definition of "open circuit" is different from TI's.

maybe they mean open circuit as a direct short to ground. since my ignition switch does not ground anything when it's off (and i've never seen one that does) should i add a DPST relay before the 15k resistor to force it to ground out when i turn off the key?

edit: ok, my green is showing... i was only looking at the upper half of rob's schematic and didn't see the 4v7 diode or 4k7 resistor between ground and pin3. durrrr someone please smack me. :P

and after you smack me, would you mind telling me what the heck a 4v7 and 4k7 are? i know zip about diodes except that they restrict current from flowing one way, and my guess at the 4k7 is a 4007 ohm resistor, but that seems an oddball impedance to me.

**Rob Withey** · 08-22-2004, 04:35 PM

Originally Posted by Gibson99

would you mind telling me what the heck a 4v7 and 4k7 are? i know zip about diodes except that they restrict current from flowing one way, and my guess at the 4k7 is a 4007 ohm resistor, but that seems an oddball impedance to me.

4v7 is a 4.7v zener diode.
4k7 is a 4700 ohm resistor.

Low wattage is fine throughout for the enable pin.

**Gibson99** · 08-22-2004, 04:44 PM

thanks to both rob and rando. i love learning new things.

now to figure out WHICH 6625 i need to request a sample of.

**rando** · 08-22-2004, 10:26 PM

Pin 3 is internally pulled to 5v. That's why the regulator stays on if you don't hook anything to pin 3. Rob's circuit uses the 4k7 resistor to change this behavior by pulling pin 3 low and thus the regulator stays off until you apply a positive voltage to the enable input. The 15k resistor and zener clamp the voltage going into pin 3 to 4.7V max allowing you to safely apply a higher voltage (i.e. battery voltage) to turn the regulator.

EDIT: I just realized that Rob was using a 4.7V zener, not a 5v.

**Gibson99** · 08-23-2004, 02:15 AM

thanks for the clarification rando.

i just got done drawing this:

does it look good to you guys? did i keep the proper bits from rob's psu and not leave anything out? will it work right the way i have the output ground going straight to chassis ground on the carnetix? i figure it should, esp since rob's grounds do the same thing in his schematic. ought to completely eliminate any chance of ground loop noise, don't you think?

and a little nit-pick... do i actually need that relay before the LED on my green switched line? it's there to protect the LED, but if the carnetix is there, even if it's on the coil side of a relay (or something) inside the cnx, it should be enough load to protect the LED from burning up. also, should i leave the led in series like that, or should i have it ground out before the carnetix, basically in parallel?

**rando** · 08-23-2004, 03:15 AM

Why do you need the LED? Also, I don't understand that part of your drawing. It shows a 1K resistor and two switches but you mentioned something about a relay up there? I don't see it.

Also, you should toss a diode across your fake-ups relay coil. This will help prevent kick-back onto your ACC line.

Finally, you should put a fuse in front of the 5V regulator and probably another at the output. IIRC, the PT6625 is not overload or short circuit protected. If you overdrive it or short the outputs there's going to be smoke and maybe fire. A fuse may not prevent damage to your regulator but it should keep your car from turning into a torch.

**Gibson99** · 08-23-2004, 08:05 AM

Originally Posted by rando

Why do you need the LED? Also, I don't understand that part of your drawing. It shows a 1K resistor and two switches but you mentioned something about a relay up there? I don't see it.

Also, you should toss a diode across your fake-ups relay coil. This will help prevent kick-back onto your ACC line.

Finally, you should put a fuse in front of the 5V regulator and probably another at the output. IIRC, the PT6625 is not overload or short circuit protected. If you overdrive it or short the outputs there's going to be smoke and maybe fire. A fuse may not prevent damage to your regulator but it should keep your car from turning into a torch.

the LED is there to remind me that i have manually turned on the carputer without the key in the ignition. 1k resistor is just to protect the led. only relay i have in that drawing is the one for the fake ups. i think you misread what i said about the load to protect the led - that whatever i'm triggering inside the cnx (i guessed that it was a relay) would probably be enough load to protect the led so that i didn't really need that 1k resistor.

anything to prevent my car from turning into a symbol of the olympics is a good thing. especially when fuses are so cheap! i'll update the schematic tonight. right now i'm off to work!

**Gibson99** · 08-23-2004, 09:36 PM

heh, i just re-read my original post there... now i understand your confusion rando

Originally Posted by Gibson99

do i actually need that relay before the LED

i was thinking resistor, but typed relay. it was late, i should have been in bed for a couple hours already when i posted that. it's just that when i get started doing something like this, i like to finish it. it's a bad habit and usually ends up with errors (like the relay/resistor mixup)...

i do have a fuse between the +B and the regulators (5v and cnx). are you suggesting a second fuse at the +B input of the 5v regulator, and then a third on its output?