I'm new to electronics and I have probably simple, maybe even stupid question, which gurus here could easily answer.
Suppose I need to design circuit that consists of 2 parts of signal processing.
First part includes some simple logic using NAND gate, while second part uses opAmp (A1).
Second part would only be needed if first part returns positive output and let's assume this positive output is +5V DC.
Now is the question: can I use output of +5V DC as a power supply to A1 (Vs1+ as shown here http://en.wikipedia.org/wiki/Operational_amplifier), so my second stage would only be involved in processing while first stage outputs positive signal. Otherwise, op Amp wouldn't be powered, therefore it's output would be probably at 0 (??? also interesting maybe Vs1- as shown here http://en.wikipedia.org/wiki/Operational_amplifier)

Basically, the question is: how fast op Amp becomes operational -- is it momentarily or it requires sometime and if yes: how much and where is it specified.

Thank you.

 

As a general rule, you do not power off circuits that are not in use.

Powering off a circuit can cause a number of problems such as:
Damage to the input stage of the device from then input signal overdriving it.
Oscillations as the circuit powers up and down.
Loading of the input signal so that another circuit using that input does not get a good signal when another circuit is powered down.

Having said that, there are op-amps that have input pins designed to power down the amplifier. You can look for information on these parts to get some insight.

 

You can power down the circuit when not in use but, to avoid the problems that Richard mentioned you need to protect the inputs from any signals that might damage it when it is unpowered (such as a sufficiently high value series resistor) and allow time for the circuit to power up before the output signal would be stable. That time will depend somewhat on the time-constants of the op amp circuit.

Also you may have to use a switch (transistor) to control the power to the circuit since the logic output voltage may not have sufficient current capacity for that.

 

As is often the case in circuit design "the devil is in the details" weather your idea will work or not depends on many things, such as the actual circuit, input signal levels, the actual op amp used and other things. The only way to know for sure is to design the circuit and build it on a breadboard and test it.

 

Thanks everyone who replied. Very helpful.

Another question: I want to use my circuit to drive 120v AC 15AMP Load using relatively high frequency of switching power ON/OFF (much higher then native AC 60hz frequency). I need switching device that would work almost like TRIAC, but without limitation of only switching OFF at zero crossing. Basically I need to have something like TRIAC, which gate could be opened or closed at any moment, without wait for zero crossing. Was thinking about 2 opposite SCRs in parallel and/or Thyristors.

So, what would you recommend for this particular application?

Thank you

 

Back-to-back MOSFETS could do the job, but generating the floating gate drive is a bit of a pain.

 

Could you elaborate on your proposal? What exactly is "back-to-back" MOSFETS? Could they hold 120V AC @15AMP? My gate voltage is regular +5V. How much power would they dissipate? Maybe exact MOSFET type, so I can read?

 

Hello igor, I will leave aside the question of power control as other members are giving you good ideas.

The issue of control of the system whilst the supply rails are stabilising has come to the fore in recent years with pc systems.
The modern pc power supply must power up or down in an appropriate sequence and connected circuit modules be disabled until correct and stable rails are available.

If this is not done, the circuitry will be in indeterminate states during power on or off, leading to unwanted outputs at best and actual damage at worst.

So yes, your thoughts in this area a re good but need developing.

I see a continuing increase in the requirement for smart supplies and modules as the need for power saving grows.

 

studio,
Amazing, but you immediately recognized what I'm trying to do, even I did avoid an idea of my project.
But yes, I want to come up with some circuit that will redirect surplus power from solar generation, so house electric meter will not rotate backward if solar production > consumption, but instead all extra power would be utilized to power on some opportunity load.
So, if surplus production of the system is higher then required power on opportunity load, there is no problem, as after turning opportunity load ON, the system would still be in surplus production and everything is just fine.
The problem appears when surplus production of the system is less then required power on opportunity load. Then, after turning ON opportunity load, system would switch from surplus to consumption and new signal would be generated to turn opportunity load OFF. Cycle obviously repeats itself.
Dimming opportunity load in this case is of no use, as not used start of AC wave would be registered on the meter as surplus current, and thus waisted.

I understand what you are saying power on/off instability and damage, but do we have any other solution for such a case?

Thank you.

 

I would not recommend your proposed solution for reasons already outlined.

The usual solution is to use a gated op amp and hold the inhibit/enable terminal down (or high) until needed.
This is usually synchronised to the rest of the circuit by a system clock - this is probably the easiest way to achieve this most times, which is why clocks are so common.

 

Back-to-back MOSFETS could do the job, but generating the floating gate drive is a bit of a pain.

Here is one way to switch AC that does not need the awkward back to back MOSFETs. The diodes waste some power, however. The diodes have to be rated for the AC line voltage and the load current.

 

http://forum.allaboutcircuits.com/attachments/ac-mosfet-png.30839/

This is a clever circuit- much lower loss and fewer parts