Hi guys I'm new here, I just wanted to ask your opinion about what opamp IC is best for designing a preamp? I 've been using 4558, I don't know if there are any better than this.
There are plenty better than the 4558 which is pretty much outclassed on all fronts by other devices.
If it is a normal preamp then look at the LM4560, NE5532 and the OPA2134.
We'll move this to Analogue Line Level I think 🙂
If it is a normal preamp then look at the LM4560, NE5532 and the OPA2134.
We'll move this to Analogue Line Level I think 🙂
It depends on your requirement.
4558 shall be avoided , 1v/us is slower than a snail plus noise is high.
NE5532 will beat up any amplifier in bandwidth and speed for the price. If you ask me , i will prefer a OPA2134. LM4562 sounds too sterile but noise is very low and specs are really good.
4558 shall be avoided , 1v/us is slower than a snail plus noise is high.
NE5532 will beat up any amplifier in bandwidth and speed for the price. If you ask me , i will prefer a OPA2134. LM4562 sounds too sterile but noise is very low and specs are really good.
Ya you're right, I've been noticing that, 4558 is really noisy, thanks for your responses, I'll try them. It would be cost effective than trying all the IC's my self 🙂
You can work out required Global Bandwidth Product op amp needs from your pre amp design.
GBP= pre amp gain * audio bandwidth.
Then add a bit for good measure.
GBP= pre amp gain * audio bandwidth.
Then add a bit for good measure.
That's only true when you don't care at all about distortion. Otherwise you want to have a large loop gain at all frequencies of interest, so GBP >> pre amp gain*highest frequency of interest.
Noisy?
Maybe perhaps you've used some "off brand" versions, but Texas Instruments TL072CP's that I've used are dead silent, and I've used them in various preamp circuits.
No noise, no hiss... silence.
The TL072 can sound superb and I think they are still one of the best devices going, particularly when used correctly. I do tend to limit my suggesting them on here though because so many have grown up with new devices and take comments they read on the web about the TL07x series at face value without actually using them.
The Tl07* aren't that bad. Except that they can only drive highish impedances properly, which can lead to more noise.
The opa134 serie is kind of moot today, unless you really need pdip package. The opa1641/2/4 do everything better and cheaper.
If jfet inputs aren't needed, the lm4562 is a great value for the money.
The opa134 serie is kind of moot today, unless you really need pdip package. The opa1641/2/4 do everything better and cheaper.
If jfet inputs aren't needed, the lm4562 is a great value for the money.
They aren't bad but why to use them when you got better ic at cheaper price like ne5532 njm2068.
Unless the application requires a FET input type.
Unless the application requires a FET input type.
It all depends on what you use them for. 18 nV/sqrt(Hz) results in a noise floor of just over -114 dBV(A), so for line level signals, that is dead silent. For a microphone amplifier for a 200 ohm dynamic microphone it's horrible: the noise figure is over 20 dB relative to 200 ohm, so you lose 20 dB of signal to noise ratio.
Anyway, this thread is in the line level section, so I tend to agree with you.
I picked up a none jfet high bandwidth op amp for a scope front end.
I was horrified when my pcb came back that it wasn't working right.
The output from the op amp was sat on a DC offset.
The 1 meg resistor on the input in AC mode was in conjunction with the input current causing a voltage shift.
I had to bodge it with a resistor to +ve rail to correct the offset.
Another problem with op amps is the bandwidth in the spec.
It only applies to a small signal.
So for large signals you need a much better slew rate.
At gain 1 to 5, 4558 & TL072 are fine. Like in graphic equalizers or line drivers. Main advantage of these, they don't oscillate without a local power supply capacitor and a feedback resistor bypass capacitor. Saves $.50 per section.
At 50 * gain 4558 & TL072 hiss. Like ribbon mike inputs or MM phone inputs. I use MC33078 at those gains, NJR2068 is equivalent but has to come from mouser that charges me 2 to 3 times the freight bill of digikey or newark. Then mouser air freghts the package which is a waste of jet fuel. LM4562 was $4 ea when I was buying, is down in price now. NJR4562 is a different inferior part. OPA anythng is pricey.
If you like surface mount, Peavey is using NJR4565 for both high gain input & line driving in PV8. It can competently do both.
At 50 * gain 4558 & TL072 hiss. Like ribbon mike inputs or MM phone inputs. I use MC33078 at those gains, NJR2068 is equivalent but has to come from mouser that charges me 2 to 3 times the freight bill of digikey or newark. Then mouser air freghts the package which is a waste of jet fuel. LM4562 was $4 ea when I was buying, is down in price now. NJR4562 is a different inferior part. OPA anythng is pricey.
If you like surface mount, Peavey is using NJR4565 for both high gain input & line driving in PV8. It can competently do both.
Last edited:
D. Self published tons of data on the tl072. From what he published, I wouldn't load them with less than 3K3 or, even better, with less than 10K.
There two ways to look at this : you can quite easily build a line preamp around the tl072 that is "good enough" in most applications. Or you can build a preamp that matches the specs of the best sources you could put in front of it but that will require other opamps.