Op-Amp Help Please

students and

sense

as the

LM358

either

the

input

(after a

volts

can

I'd

what

opamp

the

Depends on what's on the other side. A brief spurt of 15-25 watts can pack quite a punch. I have seen SOT23 violently explode after being hit with a lot less.

sheet

students and

to

sense

sense

as the

LM358

either

switch

to the

input

(after a

volts

can

supply. I'd

roughly what

opamp

the

Yep. Bye, bye lm317, or at least damaging it. The sense resistor and the sense opamp should survive. What else have you got?

sheet

students and

(transformer,

to

sense

sense

as the

LM358

either

switch

(+)

to the

input

(after a

volts

that can

supply. I'd

roughly what

opamp

so the

Abs max diff is 15V and it seems Jim's design is right at 15V max, plus and minus tolerances.

Most likely it'll survive, with the emphasis on "most likely". At least in my fields I can't design that way. It would be like scaling a cliff without ropes.

Thanks for the help climbing out of the hole I dug, gang. The TL-082 SMD device worked just fine with one small hiccup. Even with matching the resistors as close as I could using the "take the inventory box and the ohmmeter and find two out of the ten thousand in the box that are close" method, I still wind up with about a one to 1.5 volt output with zero current. No amount of loading the output (except, obviously, a short circuit) or fiddling with "trimmable resistors" will cause that offset voltage to go to zero, or at most a hundred millivolts or so.

Any thoughts about what is causing the output voltage to be above zero with zero difference at the inputs?

Jim

It's because the TL082 output stage uses common collector (i.e. emitter follower) configuration. It can't possible get closer to the supplies than a Vbe drop.

I posted another drop in from Linear Tech (I can't recall the p/n). Its output stage is common emitter, so it can get much closer to the supplies. It should work much better than the TL082 in this regard.

Bob

Hi Jim,

I haven't been following this thread closely... fretting over my own weirdly behaving circuit ;-)

Can you post schismatic? Is it single supply? That might be the issue, I don't think TL082 can output to rails.

...Jim Thompson

A resistive pull-down to ground on the TL082 output _might_ also fix it.

...Jim Thompson

That's my vote. If not, add a couple diodes or even an LED in series with the output...

James Arthur

That's true, and I could handle 500-600 millivolts. A volt and a half is a little large to accept.

Jim

-- "It is the mark of an educated mind to be able to entertain a thought without accepting it." --Aristotle

No amount of loading the output (except, obviously, a short

Schematic posted at abse. Bonehead simple 1.5-15 volt freshman power supply that is intended to power the student's projects while at school. Voltage to current converter is the U2 at the upper right hand corner.

Jim

Try 1K to ground from the output of U101A (I see no U2).

Lose the buffer.

...Jim Thompson

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

I was also thinking that a pull-to-minus load would solve the problem, but after looking more closely at the (simplified) schematic of the TL082 in the National data sheet, I'm not sure this will help too much because the top NPN device may be biased closely enough to being on that it will start conducting and thus keep the output from falling any more toward the minus supply.

In fact, per the initial post, the OP's load (the meter) is already hooked from the opamp output to the minus supply - although his latest schematic on ABSE doesn't show the meter hookup. He's also now showing a buffer stage, so if the meter is connected there then it won't help the previous amp's output.

Besides, adding a resistive pulldown is cheating. :-P

Bob

I took it all the way from 10K down to 470 with no better results. The buffer is there to introduce the students to something that does no "work" but provides isolation between the active circuit and the outside world. I can take it out, but I'd prefer to leave it in. What harm does it do? They come 2 to a package, and I suppose I could make a square wave oscillator with the buffer to flash the on-off lamps, but I'd prefer to leave it as a buffer.

Jim

Per your indication that a 470ohm pulldown doesn't help at all then you should really try something like the Linear Tech LT1366CS8. I would bet a week's wages that it fixes your problem.

Bob

offset the Vee pin with some - voltage generated from something simple like a 555 timer or an extra op-amp from another chip. I guess one could even use a small cheap DC-DC converter for this. or use a dual rail supply circuit.

if this is using it's own supply like a battery or xformer, you can slip a diode in line with the (-) rail supplying the Vee of the chip. The Vee then would become the common for external reference.

This will swing the output down ~ 0.6. 2 diodes of course will do double that. Just remember that you also lose out at the high end when you do this. This is a common trick for single supply circuits that employ there own supply and need a true 0.0 or less output for a reference.

Yep, that's right. A couple diodes or an LED in series with the output plus resistor to ground would fix that.

Or the LM358, with the sense resistor on the low side. I posted an ASCII up thread.

Or, to drive a meter, offset the outputs and use the spare op amp to drive the meter in bridge.

Cheers, James Arthur