Electronics help - TPS63001

[samski]

Hey, i am using one of these ICs to provide a 3.3V power supply from a li-ion battery. it works well but has an undervoltage lockout of 1.8v, much too low for a lithium ion battery.

here is the datasheet:
https://www.electro-tech-online.com/custompdfs/2009/02/tps63000.pdf

now it says on the pin descriptions that Vina is the control stage supply... this is also the pin that the undervoltage lockout is measured from. so could i create a potential divider for this pin so that when the li-ion voltage drops to ~2.8V, the chip locks out? what should be the minimum input current for this pin? i cant find a value given in the electrical characteristics -there is a circuit on page 11 which uses a 100R to the vina and other pins. using this and a Vin cut off of 2.9, i calculated 61ohms for the paired resistor. but this gives 18ma current leak, a little high if the battery is left for any time... (its a 1Ah)

sorry if this is in wrong forum

thanks for any help

sam

 

[duffy]

Vina is the supply voltage for the control circuit stage, I wouldn't mess with that. Yes, there's a circuit on page 11, but that's just a simple RC filter.

Do what you're trying to do on the Enable line. It switches between .4V (off) and 1.2V (on). The current is less than a microamp. If this is a one-off device, you can easily hit your lockout voltage using a 10k trimpot across Vina and gnd, wiper on En.

 

[samski]

thanks, i was looking at EN but was a bit unsure about the hysteresis etc and wondered if it would be a quick turn off or gradual as the input voltage dropped... like what happens between 1.2v and 0.4 as it drops and what if i turn the device off and on again.. eg if i have it trimmed for 2.8v in gives 0.4v at EN (ie 1/7th), then at 4.2 v it is still only 0.6v at EN

thank for clearing up page 11, i thought the parameter measurement circuit would give alright performance... the circuit is repeated on p 14 but without values too...

i also wasnt too sure what control stage meant (not done my electronics degree... yet )

thanks

sam

 

[duffy]

En won't have hysteresis ("snap action") but it is digital, so there's a trip point in that span. If this was for a production circuit, you would have to add a comparator and voltage reference.

Control stage, analog stage, digital stage - these will often be on a separate filter from the power electronics to keep the noise out. In a radio, you will see the sensitive stuff with an RFC - Radio Frequency Choke - to isolate it from noise.

 

[samski]

yh op-amp could be used, but this circuit has to be as small as possible.

ideal solution would be using lithium protection ICs but again size, and i cant be doing with hand soldering more surface mount stuff (plus deadline is getting closer)

anyway, ill stick in a pot and tune it. just have to hope the default for EN between 0.4 and 1.2 is ON...

also, when i first hooked the circuit up i had a little trouble with my external logic somehow interfering with the IC and it wouldn't properly turn on, so i had to delay turning on the external logic until the 63001 was on. (meant flicking 2 switches one after the other)... i created a delay using a mosfet with a capacitor and resistor on the gate but this made me wonder if the circuit on page 11 my do something like a softer startup... if i could elimit the mosfet, it would make the circuit smaller for sure...

thanks for the help mate,

sam

 

[duffy]

Yes, it will give you a soft start, because it takes a while for that capacitor to charge. If you use that 100Ω resistor they suggest, but make the capacitor a lot bigger - like 100µf - that circuit will give you about a 10mS delay.

 

[samski]

great, might be worth a try, not sure if it will be the right solution, its almost as if the external logic starts drawing current before the voltage has ramped up and the ic shuts down, as if it is detecting a short circuit... infact, even with the 2 switches i need couple of ohms in series with the load (the mosfet ended up being a replacement for that aswell)

perhaps it is better to leave it as it is (if it aint broke...)

 

[duffy]

Yeah, if that's what it's doing, you would want the opposite. Are you using a microcontroller? There's a "start-up" timer on some of them that's helpful with Vcc glitches. Also just adding a delay after reset might help.

 

[samski]

no, its a xbee and a gps unit...

perhaps the mosfet is the best way to provide the delay, and it provides the in series resistance (equiv of a couple of ohms)

 

[samski]

hey, just thought id post an update.

got it working with a sharp switch-off. a pot alone didnt work, i ended up with a pot with wiper attached to a diode which was in turn attached to EN. dont know why it works but it does!

then 10 minutes later i accidentally plugged it into a supply next to mine that some idiot had turned up to 20v: POOF! looks like another 2 hours hand soldering another one! quite a cool little hole where the smoke came out! atleast i got a working circuit

thanks for the help, sam