0
\$\begingroup\$

I have a project which runs on a 12v battery and it needs to draw as little power as possible when being "inactive". I have an ATtiny85 running on a lithium coin battery consuming on average about 5 µA. It is connected to a J-K flip flop IC (cd4027bc). The voltage from the ATtiny i/o pin is 3v which is not enough to "toggle" the flip flop which runs on a separate voltage source(see schematic below). So using two transistor voltage boosters I can toggle the flip flop from the arduino. The problem is that this circuit consumes a lot of power.

The current solution using a transistor voltage booster with two 2SC1740 transistors seems flawed, and I am not sure how to fix it since I am a rookie when it comes to electronics.

Below is the schematics for my circuit:

schematic

simulate this circuit – Schematic created using CircuitLab

(pins 9-13 are also connected to ground on FF1)

The flip flop runs on a 12v source separate from the microcontroller.

The goal is to control a raspberry pi, using as little current as possible when the raspberry pi is turned off. So, in other words. I want a switch controlling the 12V battery which will be connected to a 5V regulator that will in turn be connected to a raspberry pi. This switch will be controlled by the ATtiny85 running on a 3v lithium coin battery which is separate from the 12V battery.

\$\endgroup\$
7
  • 2
    \$\begingroup\$ Have you considered using a MOSFET that you can switch from the MCU? \$\endgroup\$
    – Ignacio Vazquez-Abrams
    Commented Aug 24, 2016 at 17:33
  • \$\begingroup\$ One thing that immediately comes to mind is to switch the transistors from individual I/O pins so they spend most of the time OFF and up the resistances to 100k+ \$\endgroup\$
    – JIm Dearden
    Commented Aug 24, 2016 at 17:34
  • \$\begingroup\$ There are a whole slew of high-side power switches from companies like TI which AFAIK are designed to be switchable by 3.3V or lower V devices. Likely one of those would be drivable instead of the flip-flop and MOSFET. \$\endgroup\$
    – gbulmer
    Commented Aug 24, 2016 at 17:47
  • \$\begingroup\$ @IgnacioVazquez-Abrams gbulmer are there any suitable for my application you think? \$\endgroup\$
    – Linus
    Commented Aug 24, 2016 at 18:50
  • \$\begingroup\$ I haven't done a proper part search; there are a lot of categories and parts for power management at ti.com, but this TPS1H100 is the sort of thing I was thinking of. There are so many options that it will take a while to work through enough of them to understand the right families. \$\endgroup\$
    – gbulmer
    Commented Aug 24, 2016 at 21:28

2 Answers 2

Reset to default
2
\$\begingroup\$

You are looking for a fix for a flawed design rather than a design that meets the requirements. Pop up two levels and explain what you really want to accomplish, leaving out how you think it should be accomplished.

One of Q2 or Q3 will always be on, so there is over 1 mA just for the collector pullup alone. Then whenever Q2 is on, there will be even more base current. All this is just wasted current that ultimately only drives a high impedance digital input.

You say the voltage from the microcontroller isn't enough to drive the clock of the flip-flop. This makes no sense. If the micro and the FF are powered separately from different voltages, then you should say so. You currently don't show the FF powered at all, so you really can't expect it to work properly.

It seems the ultimate goal is to control a low side switch from a microcontroller, using very little current in steady state. The obvious answer is to use a FET suitable for the purpose. If the highest voltage in your circuit is 12 V, then you have plenty of options, like the IRLML2502 for example. It goes down to 80 mΩ with only 2.5 V gate drive.

\$\endgroup\$
6
  • \$\begingroup\$ Sorry for not explaining that the flip flop and microcontroller was powered differently, I was sort of in a rush. I'll add the required information. The goal of controlling a switch from a microcontroller with as little current as possible is correct. \$\endgroup\$
    – Linus
    Commented Aug 24, 2016 at 18:21
  • \$\begingroup\$ I updated my question. From what I have read about IRLML2502 I can't find any information stating it is "toggleable". Do I still need the BC4027BC part if I use the IRLML2502? \$\endgroup\$
    – Linus
    Commented Aug 24, 2016 at 18:46
  • \$\begingroup\$ @Linus: The IRLML2502 is just a FET. However, you're driving it from a microcontroller. You can implement any toggle logic in firmware you like. You now say you want to control the 5 V power to a RPi. For that you don't want a low side switch as you are asking about. That changes the question too drastically. Again, step back two or three levels and explain what you want to accomplish. By asking about a supposed but flawed solution, and then leaving out info, you wasted everyone's time and didn't get the answer to your real problem anyway. \$\endgroup\$
    – Olin Lathrop
    Commented Aug 24, 2016 at 18:56
  • 1
    \$\begingroup\$ Sorry for not asking about my real problem. I relied too much on my very limited experience and knowledge and in a way it was helpful that I asked this badly written question because I learnt something very useful from it. Sorry that you feel I wasted your time. I hope by stating I learnt something from this makes you feel better. I'll ask a new question which I will try to include all details and be more verbose and not rely on myself. Thanks again. \$\endgroup\$
    – Linus
    Commented Aug 24, 2016 at 19:03
  • \$\begingroup\$ @Linus: Too verbose isn't a good thing either. The first thing to do is to think about your problem to decide what it really is. Once you've done that, you should be able to express it clearly in a paragraph or two. Explain clearly and succinctly what you want to accomplish. Leave the how to accomplish it to us. If you knew how to do it, you wouldn't be here asking. It's easier to answer a question if you don't first have to explain who some hairbrain scheme won't work. If you want to understand why some approach won't work, ask that separately. \$\endgroup\$
    – Olin Lathrop
    Commented Aug 25, 2016 at 15:34
1
\$\begingroup\$

if you choose 74HCTxx part with 5V and a "logic level" FET , it will be 3V compatible on inputs. Then a low idle current 5V LDO is needed.

If you did the toggle function in software, then you don't need a JK FF or transistors,, just drive the FET directly with a part with 1-1.5 Vgs

\$\endgroup\$
2
  • \$\begingroup\$ I'm doing the toggle function in hardware because I want to save as much power as possible. I want the ATtiny85 to be idle most of the time. \$\endgroup\$
    – Linus
    Commented Aug 24, 2016 at 18:52
  • \$\begingroup\$ So then you can use the 5V supply for a HCT74 D FF toggle out. Why a coin cell? if always connect to 12 V load or You could charge a 3V cell with a pullup R from 5V to offset the drain current. \$\endgroup\$
    – D.A.S.
    Commented Aug 24, 2016 at 19:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.