If you ordered them from a reputable supplier (approved and authorised) there should be no doubt.
The data sheet doesn't give a min Iq unfortunately.
Input bias current is very much lower for the 4562 and so you could try add a 1meg resistor between ground and the + input and then comparing the DC offset for your 4562 vs the 5532.
The data sheet doesn't give a min Iq unfortunately.
Input bias current is very much lower for the 4562 and so you could try add a 1meg resistor between ground and the + input and then comparing the DC offset for your 4562 vs the 5532.
DC offset is the DC voltage appearing at the output of the amplifier. Configure it like this and compare the opamps.
The 5532 has an input bias current of around 200na typically. That will develop 200 millivolts at the +input (with the 1 Meg resistor as shown) and that will get amplified by the chip which is configured for a gain of 1.
The 4562 has a typical input bias current of 10na which would develop just 10 millivolts.
The 5532 has an input bias current of around 200na typically. That will develop 200 millivolts at the +input (with the 1 Meg resistor as shown) and that will get amplified by the chip which is configured for a gain of 1.
The 4562 has a typical input bias current of 10na which would develop just 10 millivolts.
Attachments
Most applications of opamps are limited by circuit design and implementation, not by the innate limitations of the opamps themselves. Usually, NE5532 and LM4562 will measure comparably in terms of distortion. I found the NE5532 capable of <<100dB distortion, have never been able to find distortion figures in a data sheet. Probably manufacturers don't want to admit how good they really are. I had circuits where an NE5532 performed better than an LM4562 (straight from RS) in terms of distortion. And radio wave pick-up can be a serious problem for the LM4562, whereas the NE5532 is difficult to faze.
Therefore, the test Mooly suggested, using input offset voltage through a multiplier, is the best way to make a quick determination.
But, if they are fakes, consider yourself lucky and use them by all means for audio (except of course for servo applications or other applications requiring low Vos).
As per the diagram, attached 1 Meg resistor at +input, then measure the voltage across the 1 Meg resistor; NE5532 shows 1.3 millivolts, LM4562 shows 112 millivolts.
Maybe I done it wrongly?
Maybe I done it wrongly?
Maybe you have... or... perhaps they are both fake. Try measuring at the opamp output rather than the input.
"Measuring Opamp output", with reference to your diagram
- measure "DC offset" - Red Arrow marker against 0v?
- measure "DC offset" - Red Arrow marker against 0v?
Measure between the 0V line and the opamp output.
I drew that circuit based on how you said you had them wired in post #39 with output connected to - input. Its the classic buffer configuration.
I drew that circuit based on how you said you had them wired in post #39 with output connected to - input. Its the classic buffer configuration.
Thks Mooly,
Alright - “reboot” did measurements again：
Supply +- 15v (using voltage divider from 31.0v)
NE5532:
Quiescent Current ~ 8 ma. (this is as per Iq in datasheet)
DC offset measured at Output ~ 1.3 millivolts (1 Meg resistor on +input)
1st batch of LM4562:
Quiescent Current ~ 8 ma
DC offset measured at Output ~ 110 millivolts (1 Meg resistor on +input)
2nd batch of LM4562:
Quiescent Current ~ 4 ma
DC offset measured at Output ~ 1~5 millivolts (1 Meg resistor on +input)
Hmm?!?
Alright - “reboot” did measurements again：
Supply +- 15v (using voltage divider from 31.0v)
NE5532:
Quiescent Current ~ 8 ma. (this is as per Iq in datasheet)
DC offset measured at Output ~ 1.3 millivolts (1 Meg resistor on +input)
1st batch of LM4562:
Quiescent Current ~ 8 ma
DC offset measured at Output ~ 110 millivolts (1 Meg resistor on +input)
2nd batch of LM4562:
Quiescent Current ~ 4 ma
DC offset measured at Output ~ 1~5 millivolts (1 Meg resistor on +input)
Hmm?!?
I have just tried this using known genuine devices and obtained the following results. Both opamps in the package were configured identically with a 1meg resistor. Supply was -/+15 volts although in practice this makes little difference.
NE5532: Iq = 7.50ma. Voffset = -450 mv
LM4562: Iq = 14.2ma. Voffset = + 27mv
I would say there is little doubt that all your devices are suspect. Were they from Ebay ?
NE5532: Iq = 7.50ma. Voffset = -450 mv
LM4562: Iq = 14.2ma. Voffset = + 27mv
I would say there is little doubt that all your devices are suspect. Were they from Ebay ?
My sense is that both batches are unreal. My way of testing is a bit different from that of Mooly, in that I don't know how bias compensation schemes in opamps would influence measurements. Input bias is not a constant, it can influenced by factors such as temperature.
The way I test this kind of thing is by shorting the positive input and having the opamp provide high gain. This gain times Vos from the specsheet should then be ballpark what you measure as DC offset.
The way I test this kind of thing is by shorting the positive input and having the opamp provide high gain. This gain times Vos from the specsheet should then be ballpark what you measure as DC offset.
Similar these opamps configured identically with a 1meg resistor (just switching the opamp in socket)
NE5532 are pulled out of socket from a FX502Spro (build TPA3250) &
2 batches of LM4562 are bought from aliexpress.
Last edited:
Not the best source for high end opamps. Price too low, can't be real. Use RS, Farnell, Mouser or Digi-Key. Not that you are likely to notice any difference unless you measure pretty deep.
The numbers tell the story though. A genuine NE5532 can not possibly develop just a couple of millivolts or less of DC offset when configured like this.
Get some samples from the companies vacuphile mentioned and try the test again.
Another good test (not that we need it now though) would be to look at the HF response (squarewave test) with a scope and see how these devices shape up.
Get some samples from the companies vacuphile mentioned and try the test again.
Another good test (not that we need it now though) would be to look at the HF response (squarewave test) with a scope and see how these devices shape up.
That NE5532 is a curiosity though, something with relatively high quiescent but effectively FET input or good IBC. The run current doesn't seem low enough to make it a TL082/72 type.
A relabelled OPA2132? surely not. ( I suppose it could be relabelled surplus of some esoteric type that doesn't sell very well )
A relabelled OPA2132? surely not. ( I suppose it could be relabelled surplus of some esoteric type that doesn't sell very well )
I am wondering why your measured results were so much higher than the predicted numbers from the datasheet---????
Were they ?
For the 5532 an offset of -200mv to -800mv would have been within range. -450mv looked pretty much middle ground tbh. Iq looked pretty much correct to.
The 4562 looked reasonable too. Are you looking at Iq on that one ?
(last two images are for LM4562)
For the 5532 an offset of -200mv to -800mv would have been within range. -450mv looked pretty much middle ground tbh. Iq looked pretty much correct to.
The 4562 looked reasonable too. Are you looking at Iq on that one ?
(last two images are for LM4562)
Attachments
I guess I was looking at the "typical" without noting the "max" on bias current. The 4562's quiescent current is a bit high, though. Eh?
- Status
- Not open for further replies.