There is about the beats beween them are at 1kHz and harmonics at 2, 3, 4...
But when you are listening MUSIC, the beats beween notes are SUPPOSED to be there, so it is OK to be played. I wasn't reffering to those but the "extra" ones above 20kHz range. Listen to a sine shaped 14kHz signal and then one that is square.
Anyway, I am sure that LM4562 will be fine even with a 30kHz signal in the I/V or filter stages of a DAC.
But when you are listening MUSIC, the beats beween notes are SUPPOSED to be there, so it is OK to be played. I wasn't reffering to those but the "extra" ones above 20kHz range. Listen to a sine shaped 14kHz signal and then one that is square.
Anyway, I am sure that LM4562 will be fine even with a 30kHz signal in the I/V or filter stages of a DAC.
Here is my recent post on the wiki for e-mu mod :
"
I would be very interested to see photos of the pcb in the article. It will ease understanding and help choosing this platform for mods.
I am thinking of removing all regulators and replace them by good ones, that is no switching regulators. It should be even possible to put batteries for some portions (li po seems to be the buzz).
Can we see the clock ? Is it possible to uprgade it with good ones ?
"
Anybody can post good photos of the pcb, or part of it please ? We need to see the different regulators and the clock.
Thanks for your work and attention.
"
I would be very interested to see photos of the pcb in the article. It will ease understanding and help choosing this platform for mods.
I am thinking of removing all regulators and replace them by good ones, that is no switching regulators. It should be even possible to put batteries for some portions (li po seems to be the buzz).
Can we see the clock ? Is it possible to uprgade it with good ones ?
"
Anybody can post good photos of the pcb, or part of it please ? We need to see the different regulators and the clock.
Thanks for your work and attention.
Huge thanks to Audiodesign for suggesting those simple mods.
The improvement is almost unbelievable. I used an LL1674 connected as 1+1:2 and later on added an external regulated 5v. Both had a substantial, entirely positive audible effect. The only additional and relatively easy step i can think of is raising the 5v analog supply pin of the dac chip off the board and supplying clean regulated voltage directly to it with possibly some nice local decoupling. The dac is now comparable to my 1794 dac with resitor/transformer IV stage. Probably less natural sound with colder mids but surprising detail, separation and dynamics. Mid bass, as commonly the case with transformers is perhaps less incisive than it was with the opamps but acoustical bass instruments are really great. I just cannot believe something built so poorly can sound so well. Sadly, the board is really not mod-friendly and replacing the horror dc-dc convertors seems next to impossible.
The improvement is almost unbelievable. I used an LL1674 connected as 1+1:2 and later on added an external regulated 5v. Both had a substantial, entirely positive audible effect. The only additional and relatively easy step i can think of is raising the 5v analog supply pin of the dac chip off the board and supplying clean regulated voltage directly to it with possibly some nice local decoupling. The dac is now comparable to my 1794 dac with resitor/transformer IV stage. Probably less natural sound with colder mids but surprising detail, separation and dynamics. Mid bass, as commonly the case with transformers is perhaps less incisive than it was with the opamps but acoustical bass instruments are really great. I just cannot believe something built so poorly can sound so well. Sadly, the board is really not mod-friendly and replacing the horror dc-dc convertors seems next to impossible.
I found a nice photo of the complete card.
Here it is : Modding a E-MU 0404 USB | Jacken's Blog
We can see some clocks, and it should be good to try more precise clocks, even with lipo power for good i2s before the dac.
Anyone tried to remove switching regulators to put thier own good regulated power ?
Here it is : Modding a E-MU 0404 USB | Jacken's Blog
We can see some clocks, and it should be good to try more precise clocks, even with lipo power for good i2s before the dac.
Anyone tried to remove switching regulators to put thier own good regulated power ?
good question. duno. I never used it before. I use EMU 0404 USB for measurement through analog output only.... pity. good thing. I'd like to repair it.
Last edited:
Emu 0404 USB modifications
So at last I found a *second* cheap (used) Emu 0404 USB and I can start something I was waiting for: I wanted to modify the first unit I've had for a while, and that I use mostly as a USB DAC out of a "music server", but I thought that to really check the effects of different things, I needed a stock unit to use for reference.
So I have started looking at what is what, and I am updating the E-MU 0404 USB mod wiki with my findings.
Just "diagrams" and description of the unit, for the moment.
I am posting in this thread because it seems the most recent specifically dedicated to the subject. (Sorry Artmaster, I will not really go into the spdif part... hope you solved your problem.)
It's probably an old subject, but who knows, maybe somebody is still interested, and anyway that keeps me busy while waiting for a certain long-pending group-buy for another DAC...
Enjoy.
-
So at last I found a *second* cheap (used) Emu 0404 USB and I can start something I was waiting for: I wanted to modify the first unit I've had for a while, and that I use mostly as a USB DAC out of a "music server", but I thought that to really check the effects of different things, I needed a stock unit to use for reference.
So I have started looking at what is what, and I am updating the E-MU 0404 USB mod wiki with my findings.
Just "diagrams" and description of the unit, for the moment.
I am posting in this thread because it seems the most recent specifically dedicated to the subject. (Sorry Artmaster, I will not really go into the spdif part... hope you solved your problem.)
It's probably an old subject, but who knows, maybe somebody is still interested, and anyway that keeps me busy while waiting for a certain long-pending group-buy for another DAC...
Enjoy.
-
Wow!
I started testing.
The setup:
MacBook Air playing flac on Songbird.
USB to the Emu 0404, line outputs (not balanced) into my headphone amplifier (JLH "circuit 2") into AKG K240 Monitor.
I couldn't really put together an A-B switch, so in fact I have to disconnect and reconnect every time. I also don't have all my equipment here, so the volume level matching between the "DUT" and "reference" is not really done, and this can fool comparisons, I know.
The modification I tested is one that's extremely easy to do:
Replace the supply to the op-amps of the output stage.
You only need to remove two resistors, and solder three wires into easy places (look here).
Well, you also need a bipolar supply...
I just connected two 9V batteries.
My conclusion is that the result is well worth the mod.
I was careful not to be fooled by a volume difference (but +/-9V intead of +/-6V shouldn't change the gain of the op-amps, right?) and I must say that the "clean" supply is distinctly audibly better:
the instruments can be better distinguished, there's "space" and width, and more dynamics. Not "night and day" (the stock unit is already decent, I think) but certainly an improvement.
I used Mahler's 6th by the SF Symphony conducted by Tilson Thomas, in 24bit 96KHz. I have heard this symphony live in concert halls too (that's still a whole other planet...)
I think this is a nice start.
-
I started testing.
The setup:
MacBook Air playing flac on Songbird.
USB to the Emu 0404, line outputs (not balanced) into my headphone amplifier (JLH "circuit 2") into AKG K240 Monitor.
I couldn't really put together an A-B switch, so in fact I have to disconnect and reconnect every time. I also don't have all my equipment here, so the volume level matching between the "DUT" and "reference" is not really done, and this can fool comparisons, I know.
The modification I tested is one that's extremely easy to do:
Replace the supply to the op-amps of the output stage.
You only need to remove two resistors, and solder three wires into easy places (look here).
Well, you also need a bipolar supply...
I just connected two 9V batteries.
My conclusion is that the result is well worth the mod.
I was careful not to be fooled by a volume difference (but +/-9V intead of +/-6V shouldn't change the gain of the op-amps, right?) and I must say that the "clean" supply is distinctly audibly better:
the instruments can be better distinguished, there's "space" and width, and more dynamics. Not "night and day" (the stock unit is already decent, I think) but certainly an improvement.
I used Mahler's 6th by the SF Symphony conducted by Tilson Thomas, in 24bit 96KHz. I have heard this symphony live in concert halls too (that's still a whole other planet...)
I think this is a nice start.
-
Tried another easy one:
Pick the output before the volume potentiometer, skipping output buffer and muting.
This time I put together a sort-of A-B switch to be more comfortable.
Result: Cannot hear a difference.
I had put back the stock (noisy and unprecise) supply to the opams, though, to test one thing at the time.
So this could mean one of the following:
- The difference, if any, is masked by the bad opamp supply
- The difference, if any, is less than my system's and my ear's discriminating power
- The "signal path" is something more complicated than we think, and the fact that the downstream chain was still connected, makes its "effect" stay (I just connected before the pot, without interrupting the rest of the connection).
In fact, I was almost expecting that without the output buffer it would sound worse, since the way through my A-B switch makes double length of interconnects...
So for now I would conclude that removing the pot or muting transistors is probably not the first modification to do...
Next: replace coupling caps at the outputs of the DAC.
_
Pick the output before the volume potentiometer, skipping output buffer and muting.
This time I put together a sort-of A-B switch to be more comfortable.
Result: Cannot hear a difference.
I had put back the stock (noisy and unprecise) supply to the opams, though, to test one thing at the time.
So this could mean one of the following:
- The difference, if any, is masked by the bad opamp supply
- The difference, if any, is less than my system's and my ear's discriminating power
- The "signal path" is something more complicated than we think, and the fact that the downstream chain was still connected, makes its "effect" stay (I just connected before the pot, without interrupting the rest of the connection).
In fact, I was almost expecting that without the output buffer it would sound worse, since the way through my A-B switch makes double length of interconnects...
So for now I would conclude that removing the pot or muting transistors is probably not the first modification to do...
Next: replace coupling caps at the outputs of the DAC.
_
Replaced the 47uF coupling caps (C80, C81, C82, C83) with 56uF Sanyo OsCons. These were the "best" I have around.
Result?
I think I can hear an improvement in symphonic, the sound is "wider", the strings nicer, more precise and "silky-creamy". Not huge difference, though.
With baroque the difference is less pronounced. Maybe attacks are more precise and decays a tad longer...
With rock I cannot really hear a difference.
(Comparing to my other unit, kept 100% original. So here the only difference are the coupling caps, all previous changes had been undone).
I guess that now I'll build two dual supplies for the opamps, put that in both units and proceed from there...
-
Result?
I think I can hear an improvement in symphonic, the sound is "wider", the strings nicer, more precise and "silky-creamy". Not huge difference, though.
With baroque the difference is less pronounced. Maybe attacks are more precise and decays a tad longer...
With rock I cannot really hear a difference.
(Comparing to my other unit, kept 100% original. So here the only difference are the coupling caps, all previous changes had been undone).
I guess that now I'll build two dual supplies for the opamps, put that in both units and proceed from there...
-
Hi kannan,
sorry, I didn't trace much on the input part, and I know even less about ADCs than I know about DACs...
There are several opamps near the input connectors, and they are probably used as input buffers of some sort. I would expect that they also benefit from a cleaner supply (remember: both the positive and negative supplies to the opamps are obtained from dc-dc converters that to build 6V from the 5V of the supply, use high frequency switching of some sort; the negative supply seems to be especially "noisy").
Trace the circuit starting from pins 4 and 8 of the opamps, and see if there is an easy place where the supply rails are "picked" (like the two 0ohm resistors R149 and R157 of the DAC output stage). In that lucky case, you can cut this connection and use it for a better supply; if there are local reservoir caps, that's an easier soldering point.
Looking at the board, candidates could be R134, R380, C61, C63.
But I don't know how the phantom supply (48V) gets in the game...
Make sure it's off, unless you know how to handle that "detail"
I am using this soundcard also for the input, and digitizing vynil at high resolution to see what it does is also in my thousand-plans list.
I have been recording instruments, but for "musicians" high fidelity seems to be less of an issue... (it's not like I am really a musician anyway...)
For the moment I'll concentrate on the output part, but who knows, maybe later on.
Cheers,
pilli
-
sorry, I didn't trace much on the input part, and I know even less about ADCs than I know about DACs...
There are several opamps near the input connectors, and they are probably used as input buffers of some sort. I would expect that they also benefit from a cleaner supply (remember: both the positive and negative supplies to the opamps are obtained from dc-dc converters that to build 6V from the 5V of the supply, use high frequency switching of some sort; the negative supply seems to be especially "noisy").
Trace the circuit starting from pins 4 and 8 of the opamps, and see if there is an easy place where the supply rails are "picked" (like the two 0ohm resistors R149 and R157 of the DAC output stage). In that lucky case, you can cut this connection and use it for a better supply; if there are local reservoir caps, that's an easier soldering point.
Looking at the board, candidates could be R134, R380, C61, C63.
But I don't know how the phantom supply (48V) gets in the game...
Make sure it's off, unless you know how to handle that "detail"
I am using this soundcard also for the input, and digitizing vynil at high resolution to see what it does is also in my thousand-plans list.
I have been recording instruments, but for "musicians" high fidelity seems to be less of an issue... (it's not like I am really a musician anyway...)
For the moment I'll concentrate on the output part, but who knows, maybe later on.
Cheers,
pilli
-
Last edited:
Done that, and not just those but all 'lyics of the PCB. Nothing fancy, I just used panny FC that at least are long life and well regarded soundwise.
I don't know what to do with the volume pot. After near 3 years it has started to fail, sometimes it makes loud crack and pops on the speaker when you turn it up/down. Meanwhile I don't use it but instead a preamplifier (buffer) after the emu. Which in turn I wouldn't need, because it adds no gain, just volume controll, it's a unity gain buffer.
Haven't tried yet a better psu rather than the cell phone charger though...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.