op amp sub...?

[acorkos]

can i replace a njm2068 dual op amp with a 4558? the amplifier is a hartke a100 bass combo. thanks for your help!

[DrGonz78]

The basic input and the voltage ratings are perfect matches in many regards. However, the input bias currents are way too different in my opinion to allow the current to draw through the 4558 chip in the correct manner for this configuration. The op amp NJM2068 that you have in there now has a lower input resistance than the 4558 chips, typically speaking, and that will cause the chip to heat up more than the original 2068 op amp component. Typical input resistance for the JRC 4558 chip says 5 ohms and the NJM2068 rate typical resistance at 300 kilo-ohms. That is about 17 times less resistance (njm2068) than when you use this 4558 chip in your circuit. To me this means the amperage(nA) readings will then draw less current w/ the new chip. That energy then will build up in the 4558 chip causing it to heat up more.

This is just me trying to research your problem a little and I have a very small understanding of all this subject matter in many terms. However, I stand by my readings in regards to the two data sheets that I was reviewing. I did find other data sheets on similar 4558 chips such as the K4558AI chips to see their specifications, but I referred data sheets of JRC NJM2068 and JRC NJM4558. This may give different data for all of these chips so I tried to look at a make of chip (JRC) that I see most typical. I would think that if you put this chip in that the amp would still operate and work. All I am saying is that the 4558 chip will draw less current and might cause the chip to heat up more. Hope I am right about that and would love to know from someone here where I might be right or wrong on this advice... Thanks and good luck.

Edit: I also wanted to add two chips that I would consider swapping if I did not have a 2068 laying around... NE5532 fairchild looks like a winner and surprisingly enough, how I do not like NTE parts, the NTE891M chip looks like a good choice. The NTE chip is way more than you need but matches up perfectly with 2068 in so many ways. Just thought to add this as this question really got me thinking tonight.

[Enzo]

But of course, unless you happen to have that NTE chip laying around, you;d have to order it. And if we are going to order the part, why not just order the original part. We will then know it is right, AND it will cost a fraction of what NTE will charge.

I am not sure how a part drawing less current will get hotter. That 4558 input resistance was 5 MEG ohms. I don't think curremnt draw will be a large issue with either 5M or 300k.

I think of these two parts and I think the 2068 will be a lower noise op amp. But the differences may be minor in some applications. I see no reason a 4558 wouldn;t work there, at least enough to see if everything else was OK. I do stock 2068 in my shop, as well as 4558.


A 5532 should work as well.

It is just the one chip, so the actual power supply draw differences will be moot.

[DrGonz78]

Yeah I wasn't confident that current was even high enough to really be a factor, but was curious. Thanks for that followup it clears up my thinking even more. Like I always read on here this is not rocket science, so some parts are easily swappable. I did get into reading about dc bias in general and it got me reading some good information on the subject. My thinking was that as you increase resistance the draw of expected current can stress a component more. Agree>>> NTE is a rip! Just sometimes I am guilty of going to Fry's and getting a part or two. I try really hard not to go that direction but shipping charges/time waiting are factors for me.

[Enzo]

Ohm's Law.

For a given voltage, the higher the resistance, the lower the current.



Each of these op amps has its own characteristics, and a case can be made for subtle reasons to prefer one over another. Or even darn good reasons. I recall hearing from Marshall once about op amps in some Valvestate model. I had installed the "better" 4558 in place of the stock 1458 during a repair. The amp became unstable. Apparently the older 1458 had less inherent gain,. and the amp design was sort of on the edge of stability already. My "better" part had a little more gain available, and the amp squeaked.


But by far, in the majority of cases, within my experience anyway, just about ANY of the dual op amps will work in place of just about any of the others.

There are things like the 5532 - a low noise op amp - drawing about twice the current from the power rails as a 4558 might. In some mixer or amp, an extra couple of milliamps won;t matter, so subbing in a 5532 is OK. But if someone got the bright idea that if one low noise op amp was good, then the entire mixer full of them would be even better, well, they might not be so happy when the entire mixer full drew twice the current compared to stock.


But that example is about durrent draw from the power rails. In the signal path, the current demands on the I/O pins is negligible.

[Steve Conner]

Input bias current has nothing to do with the chip running hot. If you really want to know, read The Art Of Electronics. To summarise: it is a tiny DC current that leaks out of the op-amp's input pins, exactly like grid current in a tube. And again like a tube's grid current, it has to be provided with a DC path to ground for proper functioning of the op-amp.

It is almost negligible in FET input op-amps like the TL072, a bit bigger in general purpose bipolar op-amps like the 4558, and bigger still in low-noise bipolar amps like the 5532.

From an audio repair point of view, it means that you shouldn't substitute between these three categories, or the circuit could do weird things.