Modification

The main idea to modify receivers had been created due to three important suggestions:

1. We know that the modification of CDP gives us a great improvement in sound quality. The main elements for modification are the power supply, clock, OPAs (most important, especially in I/V converter) and then coupling capacitors.

Let's take a look at the block - diagram of a modern CD player (CDP). We can discuss every detail and block here, though our goal is - to define what are the "bottlenecks" of CDPs, what can be upgraded and why. The most critical elements are shown in red. They are (in the order of importance) - the power supply (incorrect design causes non-filtered digital noises and interferences), the clock (caused so called "clock jitter"), DAC (digital-to-analog converter), I/V (current-to-voltage converter after DAC, one of the most critical parts), post filter (the filter after digital-to-analog conversion) and, finally, an output stage. Both channels are equal, so we will take a look at the channel one. What can be modified (tweaked) in CDP? Almost everything, we can replace all elements to the best ones, but will this way be the most effective? Unfortunately not, as modification of one element can improve the resulting sound quality to 20%, and modification of another element - only to 0.5%. So after some experiments tweakers had found the most critical parts and blocks and offer their modification for CDPs. They are - clock, mods for I/V and post filter (rarely for output stage), power supply (better filtering of the supply voltage plus filtration of digital noises in the power supply and ground bus for DACs). It is extremely rare when any tweaker offer to modify the DACs, even when they are not the best ones, because such operation requires (in the most cases), changes in the circuit design and cannot be made that easily. We have some exceptions, though, but they will be discussed in the future modifications. Such modification can improve the sound quality enormously, and as better are the DACs, used in a particular device as better will be the final results. No less important then DACs are OPAs in I/V stage after DACs, and in many cases they are "the bottleneck" for the sound quality, and not the DACs. Tthe key factor are "the slew rate distortions" and related to the slew rate "settling time", as stated by many tweakers. Is it true? One can make the necessary calculations, based on provided by Queen University of Canada information (PDF file). Another good source can be found here (PDF file). Results of those calculations, for 20kHz sine wave and 2V output RMS signal it is enough to have only 0.355V/mks slew rate in OPA to avoid slew rate distortions. It looks like we are on the wrong way, thinking about the slew rate distortions, we have plenty of slew rate reserve here in modern DAC stages, which have OPAs with 6V/mks like JRC2068, or even 13V/mks, as JRC5534, used in 49TXi. But WHY the practical results are different, then the calculations, based on the theory?

It has already been proved (by the majority of DIY specialists and tweakers) that the OPA in I/V stage with the low settling time and high slew rate (the parameters in OPA specification are showing how fast is this specific OPA) provides quite audible better sounding results, (for example commonly used for such purposes AD825) and it is better to have it with the lowest possible noises and harmonic distortions. What is also quite important - to have in the used op amp the high open loop gain. In addition we have to take in mind, that OPA has to be fast enough to handle not only audio signal amplification, but the amplification for spectrum of quantization errors, as shown here. For 24/192 signals the bandwidth is 96kHz .... In this case we will have less signal distortions after the post filter.

If the incoming signal is below the open loop bandwidth limit, there will be no TIM, since the amp has a bandwidth large enough to pass the signal without having to wait for the feedback. For I/V stage we have no high frequencies filtering of the DAC output signal. We cannot get rid of the negative feedback in this particular DACs board, but we can lower it's influence, using faster OPAs with high open loop gain. We can't make the measurements of TIM at the moment, but we have another interesting suggestion, why the modified board produce definitely better sound quality then the same board without a modification.

Following are some links to confirm the above information:

2. Based on the ideas of modification for CDP, some tweakers started to offer the modification of DVD players. It is more complicated, as we usually have impulse power supply in DVD players (except of the expensive top models of some producers), DSP (digital signal processor) for Bass Management, Time Alignment and decoding of different formats like DD, DTS, etc. Tweakers again offer the modification of OPAs - or operational amplifiers - (usually used in I/V and post filter), coupling capacitors, power supply (exchange of capacitors and rarely - diodes, and usually - analog power supply for DACs and output stages, including I/V and post filter). The problem is - to lower the cost of such modification, they offer it through the front channels, and it is absolutely not enough for the modern 24/96 6 channels signals.

Folowing are some links to confirm the above information:

3. An AV receiver (or processor) has the same DACs, the same I/V stages and the same post filters for all supported channels (though sometimes they are already implemented inside the DAC or combined DAC/ADC chips, and in this case the modification is impossible or not so effective). Why shouldn't we be able to improve the sound quality in the same way as in CDP? We also have the great opportunity to receive high-resolution signals almost without jitter by i-link, which makes necessity to modify only D/A conversion stages, exactly as in CDP. The modification of DSP or software in DSP (like for example implementation of Dolby Pro-Logic IIx algorithms) requires much more investments and is not easy even for producers of such DSP, and therefore will not be discussed here. But we can have much better sound quality (SQ) with the existing DSP, especially if such DSP has High-resolution processing. So the first modified receiver was 49TXi.