Hi, I'm working on a circuit that outputs a pulse for 40 seconds and I had a few questions. I've attached a picture of the simulation.

What steps can I take to improve the power efficiency of the circuit? I'm trying to maximize the life of the batteries of the circuit.

Also, when actually building the circuit, what direction would I place the capacitors if they're electrolytic?

I don't want the circuit to blow up on me during testing, but the place that the capacitor fits in doesn't seem to really fit entirely positive or negative.

Thanks!

 

hi elecl,
Consider that the Base voltage of transistor Q2 can not be higher than say +0.7V, that should give you a clue.
Post your asc file and show the transistor types.
E

 

hi elecl,
Consider that the Base voltage of transistor Q2 can not be higher than say +0.7V, that should give you a clue.
Post your asc file and show the transistor types.
E

Hi Eric,

Oh that's true. Since the base voltage cannot go beyond +0.7V and the collector of Q1 can reach +5V, it's safe to assume that the capacitor's negative terminal should connect to that point right?

I actually haven't tested the circuit with specific npn transistors, but I think I'll just go with a typical npn like the 2N3904.

 

hi,
Your sim shows that Q2 Base voltage swings to approx -4v and Q1 collector is always positive.

Have you considered a MOSFET PG, use a much smaller cap and less charge current required.

E

 

hi,
Your sim shows that Q2 Base voltage swings to approx -4v and Q1 collector is always positive.

Have you considered a MOSFET PG, use a much smaller cap and less charge current required.

E

I was considering MOSFETs, but I was not sure how to replace Q2. Q1 can be changed by just swapping it with an NMOS. What is a MOSFET PG?

I tried using smaller capacitors, but it seems that if I switch it, the circuit starts discharging in between pulses. I'll attach a picture to show an example of what I am referring to. When I try smaller capacitors, I compensate by increasing the R3 resistance to reach the 40 second on requirement.

 

This is a very rough idea for the MOSFET P-ulse G-en

 

This is a very rough idea for the MOSFET P-ulse G-en

That seems to fit the specifications, but for some reason when I try to simulate it, I get much smaller of a pulse duration.

I'll probably just increase the resistance on R3. It seems to do the job.

Switching the transistor models from the default MOSFETs to actual MOSFETs made a huge difference. I really should stop using the generic LTspice model.

Are there any other changes I should consider if I want a power efficient circuit? Would having an even smaller capacitance improve the efficiency?

 

hi,
What is the source of the Input pulse and what is the 1K Voutput resistor driving into.?
E

 

hi,
What is the source of the Input pulse and what is the 1K Voutput resistor driving into.?
E

My intention was to connect the input with a battery and a button. When the button is pressed, the input pulse would be provided to the circuit.
The Vout is supposed to provide power to an MSP430.

 

hi,
You could use the push button/battery source to charge the timing capacitor, like this option.
E

 

Note that you can use 1/2 of a CD4093 package (2 gates) for the same circuit with four fewer resistors, but with a negative-going trigger edge and negative output pulse. The other two gates can be used as inverters to get back to positive in and positive out.

ak

 

hi,
You could use the push button/battery source to charge the timing capacitor, like this option.
E

That's great, it fixes the big capacitance issue.

Would the 2N7000 be a viable replacement for the 2N7002? The 2N7002 is only available in SMD form and I already have a set of 2N7000s.

Also, when I try simulating the circuit, the Vout takes a lot longer to fall for some reason.

 

Note that you can use 1/2 of a CD4093 package (2 gates) for the same circuit with four fewer resistors, but with a negative-going trigger edge and negative output pulse. The other two gates can be used as inverters to get back to positive in and positive out.

ak

That sounds interesting. Would this approach be a lot more power efficient?

I haven't done much work on logic gates.

 

That sounds interesting. Would this approach be a lot more power efficient?

Yes, by a factor of approximately 5000 to 1. The circuit in post #1 has a constant current draw of approx. 5 mA, plus the capacitor charging current during a timing cycle. With CMOS logic gates the timing current is the same (if using the same timing components, but the static current is around 1 uA.



ak

 

Yes, by a factor of approximately 5000 to 1. The circuit in post #1 has a constant current draw of approx. 5 mA, plus the capacitor charging current during a timing cycle. With CMOS logic gates the timing current is the same (if using the same timing components, but the static current is around 1 uA.

View attachment 134464

ak

Wow that's amazing. Even with the circuit improvements suggested by Eric, the circuit I've been working on has a much higher current draw.

I'll definitely look into this design some more.

 

hi elec,
You say that the output of the pulse gen will power a MSP430 and what other circuitry.?
What will the total load current be during the ON period.?
E

 

hi elec,
You say that the output of the pulse gen will power a MSP430 and what other circuitry.?
What will the total load current be during the ON period.?
E

I'm planning on having a RGB LED and a servo motor to act as a lock and unlock mechanism.

Thinking about it now, even the RGB LED will take 20mA for each color. I'm uncertain about how much the servo will require.

I might actually need circuits that use a decent amount of current huh?

Perhaps my ideal design will need to provide the exact amount of current required to activate the outputs of the uC.

 

I might actually need circuits that use a decent amount of current huh?

Morning elec,
The last circuit I posted is capable of supplying current.
Could you post your asc file of that circuit, I accidentally deleted my original copy!
I will do some current load tests.
E

 

Morning elec,
The last circuit I posted is capable of supplying current.
Could you post your asc file of that circuit, I accidentally deleted my original copy!
I will do some current load tests.
E

That's true. Yeah, here's my copy of the circuit.

 

hi,
Make R3 470R and R5 100R, check current thru R5.
E