Occasionally, some of you are tempted to experiment with different TDA1541(A) in the Audial DACs. While the TDA1541A pieces recently installed in the S5b are a very good series, it is still generally fine with me if you do that. Please, however, note that such experiments are normally not recommended for anyone lacking a basic understanding of how these things work, and, in addition, there are still several cautions you would have to keep in mind while doing this.
The first is about the procedure of the physical replacement. Practically all Audial DACs use a socket for this chip, so this is not hard to do; however, care is highly needed, so you don't bend the pins, and thus possibly render the chip, or maybe even the socket, unusable. Fortunately, there is a video, recently made by Matt Philpot, which demonstrates how this should be done. As a lever, you can also use a flat-head screwdriver of an appropriate size. Also, if you did not ground yourself properly, it is recommended to avoid touching the pins by hand, to prevent possible electrostatic discharge.
https://www.youtube.com/shorts/1bRsQyBZGE0
And the second is about the output DC offset. Normally, the output DC offset is factory nullified with the factory-installed chip; however, each TDA1541(A) sample has a somewhat different "zero" (mid-) point, resulting in a voltage offset that can vary between different samples up to approximately 200 mV. For the I/V stage and output buffers, this is not critical (although their performance may also benefit somewhat from trimming this mid-point as close as possible to zero), but such a relatively high offset may be a real problem for the units using the output transformers. The output transformers used in the S4/S5/S5b DACs have quite low winding resistance of 13 Ohms, and the DC currents higher than 2.5-3 mA running through their primary will deteriorate their performance and the performance of the output buffers, which will be forced to supply this current. DC currents above 10 mA may start producing audible distortion.
So, you can check this output DC offset by measuring it across the primary leads of the output transformers, and these are the black and red leads. The DAC input should be set to USB, and the source should be stopped or paused, or you can completely disconnect the USB cable from the DAC. But it is important to set the input to USB, because this is the simplest way to provide the proper clocking signals and frequencies to the TDA1541A, which also ensures the correct DC operating points for the TDA1541(A). Otherwise, its operation is not "under control", and its output acts erratically in DC terms, so the DC will become "undetermined" – the results measured under such conditions would be practically meaningless, and the DC set that way will most probably be wrong.
The Voltmeter should be set to mV range (or to auto-range), and you should be careful to ensure good contacts. The soldering flux residue can form the isolation layer here, so the probes with sharp tips are the best. Also, instead of trying to tap a proper point for both negative and positive probes at the PCB, you can use the ground terminal at the back plate to put the negative one. The error voltage (due to internal traces' resistance) will be negligible, almost always below 1 mV, which is acceptable. So, you will have one hand for the (positive) probe, and the other to tune the offset, by using the blue multi-turn trimmers beside.
It is good to check the DC offset right after powering up, but it is most important to set it finally only once the parts reach their final working temperatures, so probably 20-30 minutes after the DAC is powered up. The offset trimming circuit itself has changed a bit between different devices and series, but you can mostly expect thermal drift of about -30 mV. I.e. it is negative, so as the unit temperature increases, the output DC offset voltage will decrease. Ideally, the final voltage seen here should be about +5 mV. This is because, once the top plate is on, the temperature will rise a bit further, and produce additional drift of about -5 mV.
If you need more details, feel free to ask.
Occasionally, some of you are tempted to experiment with different TDA1541(A) in the Audial DACs. While the TDA1541A pieces recently installed in the S5b are a very good series, it is still generally fine with me if you do that. Please, however, note that such experiments are normally not recommended for anyone lacking a basic understanding of how these things work, and, in addition, there are still several cautions you would have to keep in mind while doing this.
The first is about the procedure of the physical replacement. Practically all Audial DACs use a socket for this chip, so this is not hard to do; however, care is highly needed, so you don't bend the pins, and thus possibly render the chip, or maybe even the socket, unusable. Fortunately, there is a video, recently made by Matt Philpot, which demonstrates how this should be done. As a lever, you can also use a flat-head screwdriver of an appropriate size. Also, if you did not ground yourself properly, it is recommended to avoid touching the pins by hand, to prevent possible electrostatic discharge.
And the second is about the output DC offset. Normally, the output DC offset is factory nullified with the factory-installed chip; however, each TDA1541(A) sample has a somewhat different "zero" (mid-) point, resulting in a voltage offset that can vary between different samples up to approximately 200 mV. For the I/V stage and output buffers, this is not critical (although their performance may also benefit somewhat from trimming this mid-point as close as possible to zero), but such a relatively high offset may be a real problem for the units using the output transformers. The output transformers used in the S4/S5/S5b DACs have quite low winding resistance of 13 Ohms, and the DC currents higher than 2.5-3 mA running through their primary will deteriorate their performance and the performance of the output buffers, which will be forced to supply this current. DC currents above 10 mA may start producing audible distortion.
So, you can check this output DC offset by measuring it across the primary leads of the output transformers, and these are the black and red leads. The DAC input should be set to USB, and the source should be stopped or paused, or you can completely disconnect the USB cable from the DAC. But it is important to set the input to USB, because this is the simplest way to provide the proper clocking signals and frequencies to the TDA1541A, which also ensures the correct DC operating points for the TDA1541(A). Otherwise, its operation is not "under control", and its output acts erratically in DC terms, so the DC will become "undetermined" – the results measured under such conditions would be practically meaningless, and the DC set that way will most probably be wrong.
The Voltmeter should be set to mV range (or to auto-range), and you should be careful to ensure good contacts. The soldering flux residue can form the isolation layer here, so the probes with sharp tips are the best. Also, instead of trying to tap a proper point for both negative and positive probes at the PCB, you can use the ground terminal at the back plate to put the negative one. The error voltage (due to internal traces' resistance) will be negligible, almost always below 1 mV, which is acceptable. So, you will have one hand for the (positive) probe, and the other to tune the offset, by using the blue multi-turn trimmers beside.
It is good to check the DC offset right after powering up, but it is most important to set it finally only once the parts reach their final working temperatures, so probably 20-30 minutes after the DAC is powered up. The offset trimming circuit itself has changed a bit between different devices and series, but you can mostly expect thermal drift of about -30 mV. I.e. it is negative, so as the unit temperature increases, the output DC offset voltage will decrease. Ideally, the final voltage seen here should be about +5 mV. This is because, once the top plate is on, the temperature will rise a bit further, and produce additional drift of about -5 mV.
