|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
17th May 2023, 11:29 am | #21 |
Octode
Join Date: Mar 2016
Location: London, UK.
Posts: 1,051
|
Re: NE5532 Distortion - What Happened?
Thank you, all!
If I manage to get into the workshop today I will try and perform some of the suggested tests with the offending OA. Note that I replaced it in the Bax-EQ with another 5532 and the problem was gone (I've been using them on this PCB for several years). The 5532 then gave the same harmonic pattern in the test jig, wired as a bog-standard inverting OA (photo in post 1). It hadn't occurred to me that the harmonic pattern could be a sign of oscillation. If that is the case (my FFT machine goes to 250KHz), then it would suggest a fault occurred that damaged part of its architecture relating to stability. I also use the NE5534 and agree that the external compensation can be handy in some applications. I shall try: i) Phase comparison between I/O waveforms ii) Square response at HF iii) How is harmonic affected into 600 and 300R? iv) FFT up to 250KHz v) Different gains NB - sorry for the poor screenshot. I saved time by snapping it on phone and not transferring a file via USB stick. Odd harmonic is -68dBV, with peak amplitude at +5dBV (so -73dB in effect at this gain with 100mV input and 1K series R / 10K feedback R). It's what the test jig was set up for at the time and I performed a quick, crude test. |
17th May 2023, 12:50 pm | #22 |
Octode
Join Date: Jul 2018
Location: Worthing, West Sussex, UK.
Posts: 1,005
|
Re: NE5532 Distortion - What Happened?
OP Amps don't like capacitive loads, so put a resistor of 100-560 ohm on the output,
That should have read, they don't like inductive loads either, basically long leads cause problems without the resistor.
__________________
"Behind every crowd, there's a silver Moonshine" |
17th May 2023, 12:52 pm | #23 | |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,785
|
Re: NE5532 Distortion - What Happened?
Quote:
That's a good test |
|
22nd May 2023, 11:36 am | #24 |
Octode
Join Date: Mar 2016
Location: London, UK.
Posts: 1,051
|
Re: NE5532 Distortion - What Happened?
Managed to sneak a few minutes at the workshop on Saturday. FFT plot now shows a 1K tone, but aligned to 0dBV – with the harmonics showing to 250KHz (the limit of my AA machine). With the 1K / 10K gain, oscillator needs ~0.05mV to show 0dBV. I didn't have time to perform all of the suggestions, alas.
I compared phase by superimposing I/P sine and O/P sine, but could see no lag. The square shown came from my Lindos LA100, which has about the squarest square of any of my generators – this is @ 31KHz, the limit of the Lindos. It doesn’t look any different to the square taken from a perfect IC in an adjacent socket of the ZIF socket used in the tester. Rather annoyingly, I pulled out a load of NE5532s from PCBs that had failed due to some distortion or other (where I’d come to the conclusion that the issue was passive, track or soldering-related) and they all tested 100% (I was hoping that I'd missed the same 5532 issue and by pulling them out I'd miraculously be able to utilise the PCBs...). Some were 5532P and some were 5532AP, the AP supposedly being quieter. As wired in the jig with 1K / 10K inverting, I could not see any difference at all in SNR… There are relays that flip between the positions on the ZIF sockets, and turning the rotary switch to flip between ICs rendered no movement in the noise floor… I wonder just how one is supposed to see this lower noise floor of the AP? Maybe it’s only visible when wired as voltage follower? Maybe it isn’t visible, and TI charge twice as much for the AP based on a theoretical, but impossible to measure difference? |
22nd May 2023, 6:24 pm | #25 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,903
|
Re: NE5532 Distortion - What Happened?
Attempts to measure the noise of reasonable low-noise devices can be thwarted by the noisiness of signal sources and the floors of subsequent measuring instruments. It's normal to have to work at it quite a bit to be sure whose noise you're seeing.
It's worth giving house-room to a conventional RF spectrum analyser. You don't just need to see audio sorts of frequency regions, you sometimes need to be able to see all the frequencies the devices you use could hoot at. For general use this suggests at least a 100MHz analyser and >1 gig if you do HF/VHF RF work. They're good machines for those cases then things just don't seem to add up. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd May 2023, 8:40 pm | #26 |
Octode
Join Date: Jul 2018
Location: Worthing, West Sussex, UK.
Posts: 1,005
|
Re: NE5532 Distortion - What Happened?
TI datasheets and advice are almost endless, perhaps some of them indicate how they made the measurements?
__________________
"Behind every crowd, there's a silver Moonshine" |
22nd May 2023, 9:59 pm | #27 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,785
|
Re: NE5532 Distortion - What Happened?
I'd be interested in the results of the tests suggested in posts 17 & 18.
|