#1042

Episode 1042
I explain where the offset is generated and what to do to minimize it.
Be a Patron: / imsaiguy

[@rfburns5601]

I sorta knew this at a surface level, now I know it better. Thanks.

[@SS-mj2mq]

You can also use a microfarad cap on the offset null as well and yes you can use a trim pot and once you get it set you can put a resistor or two resistors in place of it to replace the trim pot. Some trim pots create noise. And using a microfarad cap or even a polyester film cap can help alleviate that noise.🤗 I'm self taught so I learn from experience LOL. A lot of times it is to ensure that the offset null is removed from the output so you set the trim pot at zero. But of course it all depends on how you plan to use it with the long tail pair. And also of course you can use a trim pot and a resistor together. And of course the trim pot will adjust itself after time so you will have to go back in side the amplifier and adjust the trim pot again.

[@byronwatkins2565]

At 8:00, input bias CURRENT generates offset voltage if the I_b R_i are different. Care must be exercised, however, because the bias currents are also not typically equal... see input offset current, I_OS. The nA of the OP07 are seldom important. But input voltage is due to different V_be for the differential pair. This ratio drifts with time and temperature limiting the efficacy of the NULL option... see input offset drift. You can always add DC voltage to one of the inputs even if the chip has no nulling terminals. Horowitz and Hill "Art of Electronics" has a section showing how input transistor IC layout is optimized in precision devices. We do what we can to minimize how much device imperfections affect our designs.

[@jj74qformerlyjailbreak3]

I found “INTRODUCTION TO ELECTRONIC AMPLIFIERS” by Rodney B. Fabre. -1971 I love this book already. God Bless.

[@jj74qformerlyjailbreak3]

Great talk Coach 👍

[@andrewkieran8942]

Thanks for the bonus IC de-lidding.

[@krish2nasa]

Excellent explanation, thank you very much.

[@MrBanzoid]

Great explanation, thanks.

[@Enigma758]

Basic question: How does a 0 volt input have any effect since the op amp's input transistors would need the base to be at roughly at least .6v (BE diode drop) to start conducting.

[@IMSAIGuy]

the emitter goes to a negative voltage (e.g. -12v), so zero volts on the base is already conducting.

[@Enigma758]

@@IMSAIGuy Ah, the VBE drop is why we can't fully swing to each rail. I suppose in the single supply case, 0v is generally a problem unless the part is specifically designed to extend to ground.

[@IMSAIGuy]

@@Enigma758 you got it

[@davidmonk7773]

Hello. Thanks for this video, I have an issue with and active speakers tweeter output having -23v Dc at the terminal, could this be from the opamp being unbalanced after a mosfet change in that circuit?

[@CABohol]

Nice!

[@bobdoritique7347]

Merci!

[@KG4JYS]

What's going on in the bottom right corner of the chip? Is that some kind of damage or just some dust or something obscuring the view?

[@IMSAIGuy]

just a little damage to the part when I took the top off and it sheared the bond wires

[@SS-mj2mq]

I can totally understand that and I agree with you great video. 🤗💯💖 Yes you have to use extra parts so you might want to go ahead and by the genuine one 😂 either way you're going to spend the money of approximately $6 or more

[@IMSAIGuy]

The fake OP-07 parts did not have functional offset adjust, so I can't even add parts to make them better. they are faker than fake.

[@SS-mj2mq]

@@IMSAIGuy wow not good. And yes I prefer precision op amps overall so I'm willing to pay good money for them.

[@marka1986]

@@IMSAIGuy there is a way to use a universal network to null any op amp by using resistors and a pot to inject a variable small voltage into the inverting input. Look up universal nulling techniques. Thought TI or National has an app note. Have some references in a couple books and some notes.

[@braveheart5280]

The spec sheet for the NE5532 shows diodes across the two inputs, the spec sheet states the inputs cannot exceed -vcc or +vcc , would'nt this much voltage burn out the diodes across the inputs? I am confused.

[@IMSAIGuy]

look at the polarity of the diodes. they will not conduct current unless > +Vcc or < -Vcc

[@braveheart5280]

@@IMSAIGuy Are we both looking at the transistors used as diodes across the Inv and non Inv inputs? I measure 800 mV across the two inputs in both directions with the meter set to diode check function.

[@IMSAIGuy]

sorry I was thinking of some other part. yes there is a diode across the inputs. so the inputs can not be more than .6v apart, but they both can be at +V and they can both be at -V

[@jackofalltrades3727]

What value resistors did you use

[@IMSAIGuy]

probably 10k

[@jackofalltrades3727]

both of them 10k-- I used a 100000k and a 100ohm and get 15.43 volts from ground to the output and divided by 1000 but you were reading in microvoltts for the op07

[@IMSAIGuy]

@@jackofalltrades3727 the op07 is laser trimmed to give good results, plus a much more complicated design

[@jjoeygold]

What about Vdd and temp variations once the pot is nulled

[@IMSAIGuy]

A little change, but the resistive ratio will stay

[@s_amoku]

Im not 100% sure but I think that it will be deminished.

[@jjoeygold]

What about heat and see?

[@peatmoss4415]

Oh pee zero seven