TL081 weird glitch

[sheisthedarkness]

Hello, I bought a bunch of TL081CP from eBay and I got a different revision from the ones I already have. Making a simple negative gain circuit with any of the 15 I bought causes these glitches in both the input and output signal as pictured in the attachment. The exact same circuit with only the opamp swapped out for one of the older ones I have does not exhibit this problem, signal is clear as it should be. When the opamps are powered off the signal also gets  through perfectly.  The older revision has the code "86AEP2M" while the new ones have "18MDSCH". At first I thought they were low quality chinese parts but I found it weird that they'd all exhibit the same imperfection.
Does anybody have a clue as to what might cause this or what could be done to correct the problem ? Thank you in advance.

[magic]

weird glitch

eBay

Another one

https://www.eevblog.com/forum/beginners/opamp-input-offsets-working-in-the-opposite-direction-to-what-i-expect/

And yeah, welcome to the forum, I guess. We have those threads every other month now
Maybe eBay should pay Dave to put up a banner ad informing that opamps from auction sites are fake

[wraper]

You got not TL081 but something completely different. There is little chance getting genuine op amps on ebay, even piss-poor LM358. Especially when price is on lower side.

[sheisthedarkness]

You got not TL081 but something completely different. There is little chance getting genuine op amps on ebay, even piss-poor LM358. Especially when price is on lower side.

Oh, well that's unfortunate. I didn't think something like that would be worth faking. How bad is it that it can't reliably give me a gain of 2 at 1kHz ? Or am I using whatever it is wrong by using it like I would a TL081 ?

And yeah, welcome to the forum, I guess. We have those threads every other month now

Thanks, I searched "TL081" but couldn't find anything relevant so I made the thread. The link you sent is surprisingly similar lol.

[magic]

It apparently is worth faking
Somebody just buys those LM358/LM324 dies and has them packaged as single, dual or quad and prints whatever markings that are in demand given week on top.

Gain of 2 at 1kHz shouldn't be a problem for LM358. But do you have a problem with gain or is your only complaint the distortion near zero crossing? The distortion is crossover distortion, it's kinda expected in this opamp regardless of gain.

Not sure what to think about the input side glitches. Perhaps it's capacitive coupling from the output (what's the source impedance driving the amplifier) or perhaps it is somehow caused by input stage overload when the output stage is switching polarity and completely unresponsive. Try some legit LM358 from a reputable manufacturer if you have one.

[sheisthedarkness]

Gain of 2 at 1kHz shouldn't be a problem for LM358. But do you have a problem with gain or is your only complaint the distortion near zero crossing? The distortion is crossover distortion, it's kinda expected in this opamp regardless of gain.

Ah, you're correct, gain is as expected, it's really just crossover distortion. It's most likely the LM358 then ? Thanks, that explains it perfectly. I guess that rules out eBay for any future component purchases 

Try some legit LM358 from a reputable manufacturer if you have one.

I don't, but I will remember to if I get my hands on one

[wraper]

Somebody just buys those LM358/LM324 dies and has them packaged as single, dual or quad and prints whatever markings that are in demand given week on top.

They are not LM358/LM324, too expensive. They are some sort of cheapest Chinese opamps. And LM358/LM324 are counterfeited too. https://www.eevblog.com/forum/reviews/chinese-fake-op-amps/
https://zeptobars.com/en/read/Ti-NE5532-real-vs-fake-opamp

[magic]

They are the cheapest Chinese LM358/LM324

Classic chips like 709/741/071/358/74xx have been cloned all over the former commie bloc for half a century. You probably know.

[wraper]

They are the cheapest Chinese LM358/LM324

They are cheap but not LM358/LM324, at least most of them.

[iMo]

Try to wire a few kohm resistor from opamp's output against Vcc. It could be the glitch disappears..

[magic]

And what's cheaper than LM358?

Probably not even recycled chips. The prices of those tend to start around 30 cents on AliExpress, the LM358 fakes go for 10 cents

[wraper]

And what's cheaper than LM358?

You can make 5 times cheaper than LM358 if you want. Look at fake NE5532 die shot, it's certainly not LM358. And it's simpler than LM358.

[sheisthedarkness]

Try to wire a few kohm resistor from opamp's output against Vcc. It could be the glitch disappears..

Just tried, no luck, same thing. How could this help exactly ?

[magic]

It's RC4558 topology, another cheap classic opamp and more suited for audio. But I admit that in the case of fake 5532 I'm not entirely sure if it's an off the shelf die because it's just so tiny. I don't know if it meets the power dissipation specs of a real 4558 and for how long Though on the second thought, some discrete transistors are even smaller, so who knows, perhaps it is a legit 4558 rather than a custom job.

But you are right, going for custom tiny underpowered and oversimplified designs will probably be the final step required to break the magic 10 cent barrier

Just tried, no luck, same thing. How could this help exactly ?

By biasing the output in class A. But the current conducted by the resistor at all time needs to be higher than the output current of the entire circuit. What kind of load is this opamp driving and what are the supply voltages?

[sheisthedarkness]

What kind of load is this opamp driving and what are the supply voltages?

Input is 1kHz about 2V pk-pk, supply voltages are + and -12V. I don't really need to get it working, I'm going to buy some real ones from digikey, I was just curious about your post.

[magic]

It wasn't my post. But a few kΩ resistor to either rail should eliminate this distortion if there is nothing more connected to the output.

Okay, gain 2 so you have some gain resistors. Try unity gain or higher value gain resistors or lower value resistor to the rail, up to the point of cooking that resistor

[sheisthedarkness]

Try unity gain or higher value gain resistors or lower value resistor to the rail, up to the point of cooking that resistor

Hey, that worked ! 1k resistor from out to vcc eliminated the distortion, although I still don't really get why. Thanks !

[mikerj]

Try unity gain or higher value gain resistors or lower value resistor to the rail, up to the point of cooking that resistor

Hey, that worked ! 1k resistor from out to vcc eliminated the distortion, although I still don't really get why. Thanks !

It's because you are forcing the output stage to work in class A operation, i.e. only one of the output transistors is ever conducting.  This means you can only drive higher impedance loads however, once you load the output enough that the 1k resistor can't provide sufficient current the crossover distortion will be back.

[T3sl4co1l]

Oh, well that's unfortunate. I didn't think something like that would be worth faking.

Rule 34 of electronics sourcing: if it exists, there's counterfeit of it!

If it came from eBay, Ali or other unofficial source, assume it is counterfeit.  Or salvaged.  Doesn't matter what.

And agree, looks like LM358 crossover distortion.  Could be something even worse than that (a Chinese copy?), but not much worse without simply not working at all.

Tim

[RoGeorge]

See if this remedy works

[iMo]

Try to wire a few kohm resistor from opamp's output against Vcc. It could be the glitch disappears..

Just tried, no luck, same thing. How could this help exactly ?

The designers of that opamp forgot to layout the resistor on the die. Next time you should order that opamp together with option #1 - "1k resistor for crossover distortion elimination".. 

[Kleinstein]

Oh, well that's unfortunate. I didn't think something like that would be worth faking.

Rule 34 of electronics sourcing: if it exists, there's counterfeit of it!

If it came from eBay, Ali or other unofficial source, assume it is counterfeit.  Or salvaged.  Doesn't matter what.

And agree, looks like LM358 crossover distortion.  Could be something even worse than that (a Chinese copy?), but not much worse without simply not working at all.

Tim

In an other thread about fake OPs they did die photos, showing something similar to LM358 - probably a Chinese version. The performance was more like slightly better than expected for a real 358. Given that the OPs are not that expensive - they could as well sold the chips as what they are. There definitely would be a market for a cheap improved 358.

Someone around here even got fake 0 Ohms resistors.   So don't expect anything to cheap to get fakes.


The cross over distortion is kind of a compromise taken for lower supply current.

[Audioguru again]

Why bother adding an output resistor to eliminate crossover distortion when an LM324 or same opamps in an LM358 have trouble producing high audio frequencies due to the slow slew rate?
They are also noisy (hisssss).

[sheisthedarkness]

Why bother adding an output resistor to eliminate crossover distortion when an LM324 or same opamps in an LM358 have trouble producing high audio frequencies due to the slow slew rate?
They are also noisy (hisssss).

We're not certain what it is but I tested it with the 1k resistor from 0 to 20kHz, I did not see the crossover distortion appear again at any of those frequencies. I can't comment on noise as I haven't done any real testing but just by looking at it, compares fairly well to the real TL081 I have, which is why I was so dumbfounded 

[wnorcott]

I don't get it because a TL081 only costs about 50 cents at Mouser electronics in quantity 1 and  you know you are getting the real thing, from TI.  Or if you wanty to go the discount route and don't mind to wait and wait, get them from Futurlec for 30 cents each.  What is one minute of your time worth, that got wasted debugging bad chips?  It's worth a lot more than 50c.  I would not warm up my scope for fifty cents.