tl;dr in last sentence ____When I saw a simple NE5532 based headphone amp for $20, I got curious... Off the bat I'm going to say this thing, while cheap, is surprisingly built: solidly, in an aluminum shell (PCB well secured and the box is strong with 8 structural screws) - not only that, it sounds just excellent with, at least (from what I've gathered thus far) high impedance headphones. I've only connected my HD650s 'til now (they might be good with lower impedances provided they're IEMs or extremely efficient).
I've only powered it so far with 9V (regulated with low low ESR 3302.1uf cap), and it clips at 5.6V peak to peak with it. If p-p is always 3.4V less than voltages at and above 9, and the 5532 is like most other op amps in that its max continuous current (not provided) is ~50-60% of its shorted rating (38mA), a 19V input voltage is perfect for maximizing the 5532s potential driving 300 ohm headphones. My HD650s are 320 minimum, so for their current draw to be maximizing the 5532s potential, 18.04V is needed at its input. Printed on the back chassis is "9-19V DC, tip pos." - perfect! For this reason, I think the AU-10X is best suited for average sensitivity (HD650 are middle of the road), high impedance headphones.
At the back left of the board (looking at the picture) is the standard-size DC input jack: 9-19V, tip pos.
The big red is RCA input/output, and on either side of it are corresponding 1/8" stereo jacks which are connected in parallel with the RCA jacks (not switchable/isolated from one another)
For power, it just so happened that I had a wall-wart style, transformer-based, 9VDC out, 120VAC power supply, outputting to a standard DC jack.
But between the adaptor and the jack, long ago I stuck a project board with a 7809 (rated 0.5A) + 2x 1uf metalized polypropylene capacitors + 0.1uf ceramic.
And connected to the jack with the wires, I also put a really low ESR, 16V 3300uf capacitor (one of those barrels with conductors on opposing sides).
Here you can see its rating of 3300uf/16V
And here you can see the janky mid-wire filter caps designed to eat any hf junk that wasn't taken care of by the wallwart's transformer and possible subsequent filtering before its output. I did this on purpose to use the inductance of the 8 feet of 22awg as a filter. No, I'm joking there, I did it after the full length of the wart's included wire to maximize length and save material (I didn't have any two conductor cables handy so for the next 10 feet I had to fasten two individual cables every 8 inches (you can see one clamp in the picture of the 3300uf cap lol)
This thing, I originally made for my RME Babyface Pro, but I've since switched to powering it through USB with Topping's HS02 USB isolator (it's a great device, isolates data and power! Power is up to about an amp on the output and about 20% more on the input(s) - I'm really happy with it!)
The wall wart thing - it wasn't ideal for the Babyface, and it's not ideal for the AU-10X, but I'm thinking of making it a bit better for the AU-10X by removing the 7809. Remember above about the 38mA short circuit rating of the 5532, from which I extrapolated a 20mA continuous rating estimate?
20mA and 320 ohms is 6.4 volts, 128mW
(6.4*2)/0.70707 = 18.10V input
----------------------------
So I stopped writing and did a test: amp to scope, how many volts does clipping start at?
2.01V RMS: When I give 9.07VDC I get 5.63 p-p from the 5532.
This means, for any given desired p-p output voltage, you need 3.4V more at the input to make it possible.
The 5532 datasheet is very sparse compared to other audio datasheets. Maybe this is because it's so old - anyway, the datasheet shows its recommended maximum voltage is +15V and -15V, and absolute maximum voltage is +22V / -22V. Another op amp, OPA1612 has a +18 and -18V maximum, and like many other op-amps, gets a second maximum rating: +36V and 0V. The 5532 didn't give a + only voltage, so it could be that the 5532 can be given +30V and 0, or it could be that 22V is the absolute maximum, so 19V is the actual maximum the chip should get (remember, 9-19V is printed on the chassis). It's probably more likely 22V is the actual maximum (it would have been nice if this was included if it's the case..)
I also took the dang thing apart!
..it does seem to be a little different than the original picture. I don't mind. Maybe...
In the picture above, you can see the actual layout of the board I have, and that there are four different values of capacitors - two 25V, and one 16V...
Are the 16V caps all on the signal side (or does it look like they would be?)
What is going on with this board? It's just.. not what I expected.
That's the daughterboard immediately south of cap 1... A63A is LM321 in the SOT23-5 package.
I don't know what it's for - it's probably not for the LED (see very first image, resistor is clear to see in original version)
Could this be for managing DC somehow?
1.) Are the 16V caps rated high enough for where they are?
2.) What's the 321 for?
3.) Which caps should I swap for the best sound?
4.) Are any other op-amps able to be directly swapped with the 5532 without modifying the values of any surrounding components??
let me know if other pictures would be beneficial - I have a few more
So those are my questions. This thing sounds pretty good, I think. An analogue source might sound especially nice. Definitely recommend for $20 or for use in the car - you could power it up with no conversion using the 14.8V from the cigarette lighter! lol.
tl;dr as promist: What's this board doing? Are the 16V caps going to blow up from the 19V? What's with the regulator? Which are the caps that, if replaced, might improve sound quality? The 5532... are there any other op-amps which could be dropped in its place and work just as well without changing any of the rest of the circuit? eeextra: if you open these pics in new tab, they're pretty massive. I have other angles, took abour 40 pics so if anyone wants/needs another angle
I've only powered it so far with 9V (regulated with low low ESR 3302.1uf cap), and it clips at 5.6V peak to peak with it. If p-p is always 3.4V less than voltages at and above 9, and the 5532 is like most other op amps in that its max continuous current (not provided) is ~50-60% of its shorted rating (38mA), a 19V input voltage is perfect for maximizing the 5532s potential driving 300 ohm headphones. My HD650s are 320 minimum, so for their current draw to be maximizing the 5532s potential, 18.04V is needed at its input. Printed on the back chassis is "9-19V DC, tip pos." - perfect! For this reason, I think the AU-10X is best suited for average sensitivity (HD650 are middle of the road), high impedance headphones.
At the back left of the board (looking at the picture) is the standard-size DC input jack: 9-19V, tip pos.
The big red is RCA input/output, and on either side of it are corresponding 1/8" stereo jacks which are connected in parallel with the RCA jacks (not switchable/isolated from one another)
For power, it just so happened that I had a wall-wart style, transformer-based, 9VDC out, 120VAC power supply, outputting to a standard DC jack.
But between the adaptor and the jack, long ago I stuck a project board with a 7809 (rated 0.5A) + 2x 1uf metalized polypropylene capacitors + 0.1uf ceramic.
And connected to the jack with the wires, I also put a really low ESR, 16V 3300uf capacitor (one of those barrels with conductors on opposing sides).
Here you can see its rating of 3300uf/16V
And here you can see the janky mid-wire filter caps designed to eat any hf junk that wasn't taken care of by the wallwart's transformer and possible subsequent filtering before its output. I did this on purpose to use the inductance of the 8 feet of 22awg as a filter. No, I'm joking there, I did it after the full length of the wart's included wire to maximize length and save material (I didn't have any two conductor cables handy so for the next 10 feet I had to fasten two individual cables every 8 inches (you can see one clamp in the picture of the 3300uf cap lol)
This thing, I originally made for my RME Babyface Pro, but I've since switched to powering it through USB with Topping's HS02 USB isolator (it's a great device, isolates data and power! Power is up to about an amp on the output and about 20% more on the input(s) - I'm really happy with it!)
The wall wart thing - it wasn't ideal for the Babyface, and it's not ideal for the AU-10X, but I'm thinking of making it a bit better for the AU-10X by removing the 7809. Remember above about the 38mA short circuit rating of the 5532, from which I extrapolated a 20mA continuous rating estimate?
20mA and 320 ohms is 6.4 volts, 128mW
(6.4*2)/0.70707 = 18.10V input
----------------------------
So I stopped writing and did a test: amp to scope, how many volts does clipping start at?
2.01V RMS: When I give 9.07VDC I get 5.63 p-p from the 5532.
This means, for any given desired p-p output voltage, you need 3.4V more at the input to make it possible.
The 5532 datasheet is very sparse compared to other audio datasheets. Maybe this is because it's so old - anyway, the datasheet shows its recommended maximum voltage is +15V and -15V, and absolute maximum voltage is +22V / -22V. Another op amp, OPA1612 has a +18 and -18V maximum, and like many other op-amps, gets a second maximum rating: +36V and 0V. The 5532 didn't give a + only voltage, so it could be that the 5532 can be given +30V and 0, or it could be that 22V is the absolute maximum, so 19V is the actual maximum the chip should get (remember, 9-19V is printed on the chassis). It's probably more likely 22V is the actual maximum (it would have been nice if this was included if it's the case..)
I also took the dang thing apart!
..it does seem to be a little different than the original picture. I don't mind. Maybe...
In the picture above, you can see the actual layout of the board I have, and that there are four different values of capacitors - two 25V, and one 16V...
Are the 16V caps all on the signal side (or does it look like they would be?)
What is going on with this board? It's just.. not what I expected.
That's the daughterboard immediately south of cap 1... A63A is LM321 in the SOT23-5 package.
I don't know what it's for - it's probably not for the LED (see very first image, resistor is clear to see in original version)
Could this be for managing DC somehow?
1.) Are the 16V caps rated high enough for where they are?
2.) What's the 321 for?
3.) Which caps should I swap for the best sound?
4.) Are any other op-amps able to be directly swapped with the 5532 without modifying the values of any surrounding components??
let me know if other pictures would be beneficial - I have a few more
So those are my questions. This thing sounds pretty good, I think. An analogue source might sound especially nice. Definitely recommend for $20 or for use in the car - you could power it up with no conversion using the 14.8V from the cigarette lighter! lol.
tl;dr as promist: What's this board doing? Are the 16V caps going to blow up from the 19V? What's with the regulator? Which are the caps that, if replaced, might improve sound quality? The 5532... are there any other op-amps which could be dropped in its place and work just as well without changing any of the rest of the circuit? eeextra: if you open these pics in new tab, they're pretty massive. I have other angles, took abour 40 pics so if anyone wants/needs another angle
Last edited:
