That diagram is a bit confusing. The WM8741 DAC chip features differential outputs for each channel. An NE5532 dual op-amp is being utilized for I/V conversion of each channel's complementary output phases. The TLO72 dual op-amp is being utilized to combine each channel's differential signal phases in to left and right single-ended signals. These then are sent out through RCA connectors. It's likely that those op-amp circuits also incorporate low-pass signal reconstruction filters. Where the diagram maybe gets confusing is that the TLO72 should depict two output signal arrows, one for each channel. Instead of appearing to incorrectly depict only a single channel.
Last edited:
Based on my own experience I reckon they got the opamps in the wrong order - better use TL072 first because its JFET and able to present a much lighter loading on the DAC without incurring noise issues. The 5532 has much better drive capability than TL072 so is much better suited to facing the outside world.
Yes, the DAC outputs 2x RCA,
the diagram's end is not showing it.
(or maybe it does, just in Japanese.. because thre's some Japanese text there in the box)
Interesting.
Makes me consider giving up trying this DAC..
BTW from what I noticed here in DIYaudio,
I think that NE5532P and TL072 are not exactly considered Hi-Fi op-amps..
This reminds me a question I am curious about:
Is my tendency to prefer USB Powered DACs and avoid Wall Powered DACs, prevents me from discovering even better sounding DACs than what I experienced so far?
Or if to phrase it differently:
If you take several DACs, based on several different DAC chips,
and hear them once when they are USB Powered (assume a good power circuit for that, stabilizer, DC-DC for Dual Rail, and what ever can be done to make the power good),
and then hear them again, using power from an AC Adapter, giving 12v DC for example.
Wwill you hear a difference?
(I am only talking about using the Line Out, for speakers,
and not about Headphones Out, which requires more current than Line Out)
Last edited:
If the power supply is a wild, outrageous one then it won't matter if its USB powered or mains powered.
Yes I think you are limiting yourself in wanting always to have USB power as that's only going to go up to 2.5W (not counting USB3 here). Such a low power limit is a constraint most designers won't want to impose on themselves.
Yes I think you are limiting yourself in wanting always to have USB power as that's only going to go up to 2.5W (not counting USB3 here). Such a low power limit is a constraint most designers won't want to impose on themselves.
I see.
Maybe I should open up to Wall Powered too..
BTW USB 3 indeed should provide more power (900mA according to the specification).
But I haven't seen a USB Powered DAC that requires that it must be powered from USB3.
(probably because makers of DACs don't want to limit the potential number of customers they have, via specifying their DAC will not be good on USB2).
BTW #2:
Many motherboards from recent years, are built to provide 1A even on USB2 ports.
This can be checked on the motherboard's specification.
This improvement was mainly done in order to enable people charge their phones faster.
Maybe my 3rd round of DACs experimenting should try some wall powered DACs then
Maybe I should open up to Wall Powered too..
BTW USB 3 indeed should provide more power (900mA according to the specification).
But I haven't seen a USB Powered DAC that requires that it must be powered from USB3.
(probably because makers of DACs don't want to limit the potential number of customers they have, via specifying their DAC will not be good on USB2).
BTW #2:
Many motherboards from recent years, are built to provide 1A even on USB2 ports.
This can be checked on the motherboard's specification.
This improvement was mainly done in order to enable people charge their phones faster.
Maybe my 3rd round of DACs experimenting should try some wall powered DACs then
Yeah, and there is zero need for a FET input there. Better off with 3 5532s.
Well, actually I’d pass on a purportedly high end design done after 2010 if it uses those op-amps.
I see.
Maybe I should open up to Wall Powered too..
BTW USB 3 indeed should provide more power (900mA according to the specification).
But I haven't seen a USB Powered DAC that requires that it must be powered from USB3.
(probably because makers of DACs don't want to limit the potential number of customers they have, via specifying their DAC will not be good on USB2).
BTW #2:
Many motherboards from recent years, are built to provide 1A even on USB2 ports.
This can be checked on the motherboard's specification.
This improvement was mainly done in order to enable people charge their phones faster.
Maybe my 3rd round of DACs experimenting should try some wall powered DACs then
Just because most USB hosts allow 1A doesn’t mean it’s a good idea to make a device that relies on that. For actual USB 2.0 compliance your device needs to draw no more than 100 mA before enumeration and 500 mA after requesting it from the host.
It’s a pain to do it right, so lots of devices cheat. I’ve never seen a host unwilling to deliver 500 mA, but I’ve seen several that will trip a polyfuse or load switch if you exceed 500.
Last edited:
Thank you Chris.
OK will skip this one.
Too bad, wanted to have a listen to a WM8741 DAC.
BTW, what op-amps do you prefer, for DACs from recent years?
LMH6643MA
LME49990
ADA4898-1?
Or any other?
Well, I speak in terms of measured results because I haven’t seen a convincing theory otherwise... but here’s a quick take:
It highly depends on the circuit position. 5532/4 isn’t bad btw, but I’d spend the extra couple dollars on LME49710.
OPA1611, AD797 and others are great for where you want a bipolar input and have a low source impedance. LME49990 very similar but discontinued. ADA4898 is in this group too but I haven’t seen any real audio band distortion measurements of it.
For FET input you have OPA827 and 828 now. OPA1656 if you don’t mind a little extra 1/f and want better output drive. ADA4625 if voltage noise is your main concern. OPA1642 for cost savings vs 827.
I don’t think you need anything crazy fast for even I/V conversion. In the “you better know what you’re doing” category, I find ADA4896 and ADA4899 to be interesting. ADA4807 also for low quiescent current vs performance.
For heavier loads, OPA1622 is great if you don’t want to use a composite amplifier or other buffer.
At the output of a voltage out DAC like the Wolfson? I’d probably start with OPA1611/2.
None of what I mentioned is particularly cheap - there are lots of other good options that are less expensive and only slightly lower performance, but I don’t really believe in trying to shave every penny in DIY. If you make a living wage you’ve probably spent the difference or more based on the amount of time spent thinking about it.
Last edited:
Are you sure about the I/V?
https://ctrlv.cz/shots/2019/05/22/Wqah.png
The first opamp NE5532 is working only as a filter (perhaps as LPF of a sallen-key topology).
The second opamp TL072 is doing job as an unbalanced output.
This sallen-key filter (or any active) will affect the sound and you will not get the true output of WM8741 (I prefer a passive filter).
Once again, it is not the way how to test the WM8741
You will test another DAC with another implementation of the WM8741
Are you sure about the I/V?
https://ctrlv.cz/shots/2019/05/22/Wqah.png
The first opamp NE5532 is working only as a filter (perhaps as LPF of a sallen-key topology).
The second opamp TL072 is doing job as an unbalanced output.
This sallen-key filter (or any active) will affect the sound and you will not get the true output of WM8741 (I prefer a passive filter).
Once again, it is not the way how to test the WM8741
You will test another DAC with another implementation of the WM8741
Miro, yes, you are correct. I see from it's data sheet that the WM8741 has 2VRMS voltage outputs. I wrongly assumed it to have current outputs. I just now also notice that the NE5532 circuit blocks are labeled LPF, for low-pass filter. I would guess that the entire three op-amp per channel analog section exhibits little or no voltage gain, but performs filtering plus differential to single-ended conversion.
Hmm..
If this is a common problem, why isn't that filter built-in the DAC chip?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.