2 Resistors of 0 Ohm as it is now. Normally 47 Ohm suffices but 100 is certainly OK. No ugly/unreliable constructions necessary, just replacing the 0 Ohm resistors that are on the right spot where they should be. Also we meet the dreaded muting-IC (SGM3710?) of questionable reliability again. I have dealt with defective ones a few times and the suspicion is that these die from ESD on the RCA cables so try to avoid touching cables when disconnected.
Attachments
Last edited:
Putting two zero resistors after each opa 1611 is not good practice , agree .
Wouldnt it be a good idea to take the left and right signal right from the + out leg of each of the opa:s 1611 , with a series resistor of 47 ohm and a series polycap 2.2 uF , cut the + lead of the rca connector and solder the capacitor directly to that + connector ?
bypassing the dreaded muting IC completely ?
Wouldnt it be a good idea to take the left and right signal right from the + out leg of each of the opa:s 1611 , with a series resistor of 47 ohm and a series polycap 2.2 uF , cut the + lead of the rca connector and solder the capacitor directly to that + connector ?
bypassing the dreaded muting IC completely ?
Last edited:
I think you don't want to omit good working muting (yet). Maybe the SGM3710 has undergone a revision that makes them more sturdy (that is what I wish to think how things work and they sometimes do work out that way 🙂). Maybe it is not SGM3710 but a newer version as there exist a few types by SGmicro. Apparently there is no need for output caps as they are lacking in the design. Including them seems unnecessary and the best coupling cap is no coupling cap.
On paper SGM3710 seems adequate but I'd rather not see them in devices 😀: https://www.visvie.com/static/upload/file/20220121/1642747164583685.pdf
On paper SGM3710 seems adequate but I'd rather not see them in devices 😀: https://www.visvie.com/static/upload/file/20220121/1642747164583685.pdf
Last edited:
Agree, but the 1611 will oscillate with zero ohm resistors if using high capacitance rca cable - so the resistor mod of 47 or 100 ohms is absolutely nessessary. Easiest done using the inside of the rca connector of the cable at the source end .
Users of power amps like hypex ncore and others without DC protection on the input should beware of the risk without output capacitor .
Users of power amps like hypex ncore and others without DC protection on the input should beware of the risk without output capacitor .
No not easiest and sloppy. Right is to replace R245 and the other one for 47 or 100 Ohm resistors and you are done.
The circuit and the SGM3710 are designed for direct coupling so DC offset should be very low. Knowing is measuring but it would not surprise me that it is about 1 mV. Of course good power amplifier designers and clever DIYers know that about the only useful/necessary place where coupling caps belong is as "DC fuses" so input caps in the amplifier. Just 2 caps are insurance against catastrophe. Having again 2 caps in the Ultra would make the dreaded "2 caps in series" scenario.
The easiest, most convenient and least possible faults mode is to have the sources direct coupled and the power amplifier protected with input caps.
The circuit and the SGM3710 are designed for direct coupling so DC offset should be very low. Knowing is measuring but it would not surprise me that it is about 1 mV. Of course good power amplifier designers and clever DIYers know that about the only useful/necessary place where coupling caps belong is as "DC fuses" so input caps in the amplifier. Just 2 caps are insurance against catastrophe. Having again 2 caps in the Ultra would make the dreaded "2 caps in series" scenario.
The easiest, most convenient and least possible faults mode is to have the sources direct coupled and the power amplifier protected with input caps.
Last edited:
Ferrite beads, can be good choices if they are in the supply lines. I think you would do a good job drawing the circuit with also the omitted parts. One can assume the guys knew what they were doing hence the 0 Ohm resistors and omitted parts but still. They also omitted the stop resistors.
Last edited:
Personally, I dont like ferrite beads in any analog stage, not even at the power stage, - it makes the sound strangely thinner, lacking details.
The measurements, however, often seems to improve using those.
Ferrite beads in digital parts is another story.
I guess a WiiM ultra with Teddy pardo LPS dont need ferrite beads in analog stage ?
The measurements, however, often seems to improve using those.
Ferrite beads in digital parts is another story.
I guess a WiiM ultra with Teddy pardo LPS dont need ferrite beads in analog stage ?
Last edited:
They are simply needed in RF rich environments. Like them a lot as they simply do what they should do when chosen right. The designers would not have implemented these if they were not necessary as they are relatively expensive.
Judging RF by ear and in fact all measuring by ear is deceiving (tube pilot territory) but let's say the influence of RF wandering through audio circuits is detrimental. More detrimental than having the beads. If you like experiment you can inject RF in an amplifier and then try to hear what the effect is. I know you will very often if not always hear an undesired effect. Without reading papers, necessary explanation, Henry Ott or sleepless nights you will then be convinced that RF is your audios enemy.
Judging RF by ear and in fact all measuring by ear is deceiving (tube pilot territory) but let's say the influence of RF wandering through audio circuits is detrimental. More detrimental than having the beads. If you like experiment you can inject RF in an amplifier and then try to hear what the effect is. I know you will very often if not always hear an undesired effect. Without reading papers, necessary explanation, Henry Ott or sleepless nights you will then be convinced that RF is your audios enemy.
Last edited:
SGM3711 (and 3710) have a closed resistance of 11ohms, so the measurements of the output z of the Wiim at 11ohms makes perfect sense. Haven't poured over the DS of the two parts so I haven't seen what differentiates between them
You can just make out '3711' on the image posted above
You can just make out '3711' on the image posted above
Last edited:
Desoldering those ferrite beads at the buffer stage ( first making sure they really are in series with the power supply to the 1611 ) and instead put 10 Ohm resistors at the same place will probably make the sound more ”analog” , or more earfriendly ( my five cents…)
Standard work for Marantz CD players from -90.
Standard work for Marantz CD players from -90.
Simply NO. That old slow logic stuff did not have very high frequencies running around. The beads very likely have way higher impedance in the relevant frequency band. It may have worked then but it will probably not work the same now as the environment is totally different with wireless and bluetooth, gigabit ethernet, a color display with PWM backlight (?) and a quad core CPU with RAM all running on various high frequencies. All in a small box with short distances to each other. One could defend the Eversolo having a larger casing 🙂
But go ahead with decisions based on emotions.
Last edited:
That is true for the SGM3710.
But for the SGM3711, which is used here, it is 11 ohm for both paths.
OK, did not know it is a SGM3711 (overread it, I need a larger screen on my phone) but then the measurements seem to be right.
I would then use 47 Ohm for the stop resistors.
I would then use 47 Ohm for the stop resistors.
Last edited:
Had it in the Hotel the last week. Test setup as always. External China r2r dac with Burson V7V sounds way better than the internal one, but passive volume control is a must as this China dac is noisy at low level signal. No clean up after LM317/117 might be the main cause.
Sennheiser HD540 Reference II for Evaluation.
Sennheiser HD540 Reference II for Evaluation.