Does This Circuit Destroy an Op Amp

[eev_carl]

Hi,

I pulled this schematic from "Op Amp Circuits: Simulations and Experiments" by Sid Antoch.  I replicated the results with an LT Spice simulation but when I breadboarded this with a 14588, I noticed that my current protection kicked in after about a minute.  Two chips seemed to respond ok initially in terms of current and voltage.  However, they both started pulling in 150mA which is my power supply cap.  I watched this happen to two ICS.

Do I need a current limiting resistor on the V+ for this to work?  I verified each of the three voltages (-10,10,5) from the power supply with a DMM.

Thanks,
Carl

[T3sl4co1l]

Did you forget bypass caps?

Tim

[eev_carl]

I didn't have any on.  Should I put a pair of ceramic 0.1uF ones on V+ / GND and V- / GnD?

[T3sl4co1l]

Yes, or 10uF electrolytic, or both is even better.

You should probably pay attention, to when things are getting... hot, and smelly and burned.   It sounds like the chip was, I assume oscillating (without an oscilloscope, it's not easy to prove), causing excess current draw, causing it to heat up.  Eventually it got past maximum temp rating and failed shorted.

Tim

[capt bullshot]

Did you turn on the 5V last?
If you apply a low impedance voltage source to the Op Amp inputs without the supply voltage you might damage the IC. A series resistor to the + input isn't necessary in theory, but helpful in practice (1k ... 10k should do the job here).

[Zero999]

Did you turn on the 5V last?
If you apply a low impedance voltage source to the Op Amp inputs without the supply voltage you might damage the IC. A series resistor to the + input isn't necessary in theory, but helpful in practice (1k ... 10k should do the job here).

Yes, the 5V supply should be switched on last, otherwise it might damage the op-amp. If that's not possible, connect the input pin to both supply rails, via diodes and connect the signal source via a 10k resistor to limit the current.

Note that the diodes might already be included in many modern op-amp ICs, but it doesn't do any harm to add some. Irrespective of whether the diodes are build-in to the op-amp IC or not, the current limiting resistor is definitely necessary, if the signal can be applied before the power.

[rstofer]

Thinking in terms of an ideal op amp, we know that the inverting and non-inverting input currents are zero.
Therefore, there  is no reason not to have series resistors on these inputs. The value isn't terribly important but I would gravitate toward 100k Ohms.  Try that with the LTspice and see how the output maintains the required value.  If I thought 100k might be too high (electrically noisy), I could back down to 10k.

Use

Code: [Select]

.step param RL 100 1000 100
.op

You put the .step line in the schematic as text and the .op as the simulation command.  I put the .step command just above the .op line.  I don't know that it matters...

This will provide a graph of the output voltage as a function of the load resistor while the resistor iterates from 100 to 1000 Ohms in steps of 100 Ohms.

Relabel the VALUE of the load resistor as {RL} (including the {}) instead of some numeric value.  The RL inside {} needs to match the RL in the .step line.  You can, of course change the name (in both places) to anything you want.

It turns out that just yesterday I was doing this very thing on a simulation for a +-10V reference output from a +-15V system.  The main difference is that I had a physical +10V reference device, two op amps and two pass transistors.  I played with various values of input resistors and transistor drive resistors to get the values as high as possible and still maintain the proper output from 500 mA down to 5 mA.

[eev_carl]

You should probably pay attention, to when things are getting... hot, and smelly and burned.

I have the current limiter on to <150mA and was keeping an eye on the power supply.

[eev_carl]

Did you turn on the 5V last?

I think that may have been a difference.  For one of the ICs I used the "all on" button for the three voltages.  I didn't pay attention to the order in the subsequent trial.

[eev_carl]

I wired up another 14558 and powered the 5V V+ source last.  It's been working for the past 10 minutes without a spike in current on the power supply.  Thanks for the tips regarding the power up sequence.  I'm glad I breadboarded it.

[GigaJoe]

opamp consumption approx 3-10ma, not 150. This is a voltage follower. An idea, that opamp doesn't consume current from a source (on input), therefore don't change an input voltage ( or let say an input voltage potential), and do output the same as input but capable with much great load.

[eev_carl]

opamp consumption approx 3-10ma, not 150.

The LT Spice simulation and my (revised) breadboarding has this at 5 mA.  The 150 mA figure I'm throwing around is the current protection setting on my power supply, indicating that it's drawing far too much.

[nrxnrx]

Are you powering the 14558 with +-10V?  https://www.njr.com/semicon/PDF/NJM14558_E.pdf says absolute max is +-7.5 .

[eev_carl]

Are you powering the 14558 with +-10V?

Yep

(Quickly puts in a 4558)

[rstofer]

opamp consumption approx 3-10ma, not 150.

The LT Spice simulation and my (revised) breadboarding has this at 5 mA.  The 150 mA figure I'm throwing around is the current protection setting on my power supply, indicating that it's drawing far too much.

The problem has no doubt been solved but I would suggest that if you know the circuit only consumes 5 mA that you set your current protection a lot lower.  Maybe 20 mA.

I like to keep it tight.  If there's an "oopsie", I want it caught before things get warm.

Sometimes, when there are large capacitors involved, the voltage will ramp up slowly with the low value of current limit.  This may, or may not, be a problem.  Regardless, the limit will still prevent the magic smoke from leaking out.

[T3sl4co1l]

Also, note that power supplies are DC only.  What they may do for short durations, has no bearing on the settings of the dials.  Most supplies have a big fat capacitor at the output, so that the short-circuit current peaks extremely high for a brief moment (10s of microseconds?) before settling to its steady-state value (which may still take milliseconds).  Peak currents like 100A!

This is why you cannot test LEDs with a bench supply set at, say, 10V and 20mA.

Well, you can, but you need to remember to use the power supply as a current source.  Whereas a voltage source delivers zero power into an open circuit, a current source delivers zero power into a short circuit.  So for testing LEDs, you must always short the leads together (so the circuit is carrying, say, 0V and 20mA), connect to the LED, then un-short.  Then short again when disconnecting, move to the next LED to test, and so on.

Loose cables have an inductance in the ballpark of 1 uH per meter of length, not at all a negligible amount -- that 1uH has a reactance of, say, 50 ohms at 8MHz, and 8MHz is right around the transition frequency of most jellybean op-amps, where supply sensitivity and risk of oscillation are greatest.

An electrolytic capacitor, placed at the load, has a dampening effect, which helps to counter this.  That's why it's a good idea, or at least some smaller capacitors, but those could actually make things worse, too (1uH resonates with 0.1uF at 0.5MHz, and the impedance can be in the same risky range (>10s of ohms) if the Q factor is modest, which it probably will be using film or ceramic capacitors here).

So that's the technical explanation of why I worry about oscillation, it's quite plausible.

(Now that we know it probably was a power sequencing problem -- in effect, testing the op-amp's input diodes like an unfortunate LED! -- it probably wasn't oscillation after all.  Good to know. )

The other thing you can do, to prevent surge currents, is simply wire up everything with resistors!  For powering op-amps, a series 10 ohms is more than enough, say.  A kohm is even fine for setting the op-amp input voltage, and limits the 5V supply to a short circuit current of merely 5mA.

Tim