1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

ZiShan T1 Hi-Fi Player Thread

Discussion in 'Portable Source Gear' started by IcedFrosty, Jul 27, 2019.

Is this the best DAP under $100?

  1. Yes

    11 vote(s)
  2. No

    10 vote(s)
  3. Way above the price range!

    4 vote(s)
  4. Could be

    18 vote(s)
16 17 18 19 20 21 22 23 24 25
27 28 29 30 31 32 33 34 35 36
  1. HeyManslowdown97
    I do too but the basic research and applied science have such wide propose for innovation. It’s really more important for lifesaving stuff, or whether or not there's life on a mars- Literally!
  2. HeyManslowdown97
    I understand the difference but by ripping the lpf you also disabled the digital filters. The msop operates at high frequency that's why the DSD is using 1.2 Hz switch and goes through those filters. It operates on Pulse Density and PWM is what makes the filters special.

    I put these on DSD dac and ran better in the DC to DC boosters!



    I don’t see why you can't put the AD8397 in either in either stages in the T1 since the op275 is now in the SE and The lm4562 are in the lpf.
    Last edited: Sep 15, 2019
  3. Merlin-PT
    I've read they were the same silicon, but wanted to confirm with someone that had both.

    What's the other model for the 1688?

    I'm curious to see how they will implement
    ak4499, it has 4 channels and I'm not sure if 4 dip8 opamps+their circuits will fit the board, maybe soic8 can fit.
  4. Ivan TT
    Please take no offence or see it as a personal attack, but this understanding is wrong.
    Digital filters are an integral part of delta-sigma convertors and are built into AK4497 DAC and enabled (and switched) via firmware.
    Please refer to figure 1 of AK4497 datasheet, SCF in the block chart.
    Please take no offence, but what you refer as MSOP is a DC-DC convertor and it is a part of power supply and it has nothing to do with DSD, DAC or filters (except for providing the former with +/-9V power supply).
    AK4497 is x256 oversampling DAC (page 1 of datasheet), and operates at just over 11MGz sampling rate internally. The rule of thumb related to oversampling DACs is 6dB improvement of DR for each quadrupling of sampling rate, mainly due to delta-sigma modulation noise spread over wider frequency range. This noise is then attenuated by digital filters (please refer to datasheet as to how much attenuation they provide, which is significant).
    PWM is a actually a problematic part of DC-DC convertors, as this noise is coupled into the power supply rails, hence inductors and 3x caps DSD has (and T1 lacks). Also, one of the simpler ways to reduce this noise coupling is to add 1206 0.1uF capacitors on top (ideally - below as this reduces inductance) of 12uF caps on top and bottom of all opamps. Please take no offence, I'm not criticising your personally but making comments on what I believe incorrect understanding of the principles of operation (that result in personal attacks directed at me, unfortunately).
    Yes, you don't. But it's an easy fix, don't worry about it.

    But my question was:

    In my mod 1622 is in SE output stage, not in LPF. Do you understand the difference?

    Can you give me direct answer, as this will allow me to help you to develop understanding of why removal of LPF is not as atrocious as you believe (we already cleared the digital filters part, awesome!) and how different applications (in DSD and T1: LPF, Subtractor, potentially output stage) call for different Op Amps, making some of them more suitable and some - less for each of these stages. But let's start with LPF compared to output stage, what qualities are beneficial (and what not that important) for either stage?
    Last edited: Sep 15, 2019
  5. Ivan TT
    I think it's called "rebadging", not rebranding as I previously stated.
    Why not to use only 2 channels?
    Merlin-PT likes this.
  6. Merlin-PT
    When I read it had 4 channels, I thought that each 2 channels could be used in some kind of mono mode to give better 2 channel specs. Like when you use 2 dual channel ak4497 dacs in mono mode.
    Last edited: Sep 15, 2019
  7. HeyManslowdown97
    Please take no offence, but what you refer to as removing the lpf entirely is a fact (as well the HF). Let’s not deviate and have you reframe the question and introduce other arguments. The datasheet clearly shows lpf it is schematics can you tell why you did this? Please refer to page 38 of the datasheet as it clearly shows the lpf

    If when you removed the lpf what is now the lpf the opa1622 replacing the op1612 on the evaluation? If so what used as lpf of your mod. Let’s start there since it your mod that’s goes against the circuitry design? Then we can define what the difference is are.

    Please take no offense if this is considered to be in an Orthodox modification since there’s no mention of scarlet book or anything you mentioned in the data sheet unless there is? If so I’ll learn something new. If not, I don’t want to deviate from the original modification in question with a direct answer which isn’t factual and not speculative or an opinion.

    It’s unorthodox because it’s not in the datasheet and nobody I know has done this besides you. Isn’t this problematic to go against what is the schematics of both the AKM4497 EQ and the DSD pro? Again, please don’t take offense since I haven’t seen anyone do this ever nor is the this modification common anywhere. 33D63486-4A03-4538-B285-8F3EF5FB059E.jpeg
    Last edited: Sep 15, 2019
  8. Ivan TT
    Yes, but unlike what you believed and expressed above I did not remove digital filters, or PWM-based DC-DC convertors don't have ANYTHING to do with the operation of DAC or filters (except for providing power for opamps). I note that you did not acknowledge these errors in your knowledge, but in response reinvigorated attack at me, why would you expect ANYTHING different in response?

    You are very negative and critical of approaches like removal of LPF, but how objective and factual this criticism is if you believe that this disabled internal DAC filters? Or that PWM in DC-DC convertor is somehow good? How much would you trust your own opinion?

    I explained the rationale behind LPF removel several times, but I'm not wasting any more effort and time on explaining it to someone who will only try to find (or invent or imagine) more and more flaws based on some sort of personal agenda, and I'm only interested in improving quality of knowledge and understanding (mine and others), not personal agendas.

    Ah, by the way... Datasheet clearly states that the circuits are EXAMPLES, it's not a bible that everyone has to abide to, using datasheet circuits is just a way to simplify development, so datasheet is not orthodox, it's just an example...

    And do me a favour...
    Open OPA1622 datasheet, page 1 left bottom corner where it says "OPA1622in a High-Fidelity Headphone Driver Application" and tell me where LPF is in the circuit? (hint: it is there)
    Last edited: Sep 15, 2019
    Alex3221 likes this.
  9. HeyManslowdown97
    We are not talking about the LPF of the Opa1622 which is designed for headphone amps.

    Do you know the differences between a headphone amp and a DAC?

    Does the AKM4497 have an Headphone Amp design or is it a DAC design? Please show where it says it’s a Headphone amp application. (Hint: it’s usually found in the top of the datasheet)
    Last edited: Sep 15, 2019
  10. Ivan TT
    OK, last attempt - I don't think you are willing to understand as you are to busy trying to prove me wrong as per personal agenda (and entertaining personal agendas is well outside of the scope of my interests) But let's if I'm wrong, I will be massively pleased if I am wrong here.

    Please navigate to Figure 3 in AK4497 datasheet and tell what attenuation built-in digital filter will provide @ around 20kHz?
    (insert answer here)
    Then navigate to Figure 75 and tell what is the order and cut-off frequency of LPF and what attenuation internal digital filter ALREADY provides at this frequency?
    (insert answer here)
    Then navigate to Figure 52 in OPA1622 and tell if it has LPF and what is the order and the cut-off frequency of this filter?
    (insert answer here)
    After you done that, go to AK4497 evaluation board datasheet and tell what is the order and the cut-off frequency of the subtractor circuit?
    (insert answer here)
    Then, finally research what happens to the quantisation noise in delta-sigma convertors relative to the sampling frequency, what the sampling frequency of AK4497 is for 44.1kHz PCM data input and specific effect integrator, decimator and digital filter have on this noise (it the simple terms, what is the output noise of delta-sigma convertor output, before any external filters?
    (insert answer here)
    Finally, what is the noise floor of AK4497 DAC for 2.6 VRMS output?
    (insert answer here)
    This will allow you to backtrack my rationale, provided you have sufficient knowledge and understanding to interpret the result (and provided I'm wrong about the first statement above).
    Last edited: Sep 15, 2019
  11. Merlin-PT
    AK4499 block diagram shows it has 2 dual channel dacs inside:

    The datasheet says that each one of this dual channel dacs can be configured to work in mono mode (pag. 73).
    Also pag.1 specs are: "Dynamic Range, S/N: 140 dB (Mono), 137 dB (Stereo), 134 dB (4ch)"

    If I understood it correctly, using the 2 dual channel internal dacs (each one in mono) for stereo output, gives a better dynamic range "137 dB (Stereo)" and uses all the silicon for stereo output.
    Using the 4 channels separated only gives "134 dB (4ch)" and would only use 1 internal dac for stereo output, the other one would be unused.
    Last edited: Sep 15, 2019
    Ivan TT likes this.
  12. HeyManslowdown97
    The Zishans aren’t Developmental Boards. Like I said If you want to make your own original designs based on the Datasheet because you have extensive knowledge in doing so there are threads and forums made just for that.

    As far the AK4497 vs AK4499 it usually takes time for seasoned engineers to figure out how best to use the chips. The AK4497 has been out for a while so lots has been worked out to make them sound better. The T1 is no doubt an upgrade from the DSD Pro made three years ago. I like to learn why but I’m really more interesting in what sounds better not all the science stuff unless it clarifies a concept for me better if I do wish to one day make my own designs.

    I’m sure some of my information post are misinformed and if some professional engineer and an experienced diy whose built their own dacs, speakers, etc are laughing to themselves but they don’t feel the need to correct me and probably know it’s part of the learning process. You seem to act as if you at this level and yet haven’t even attempted to build your own dac or headphone amp and yet if anyone tells you that the zishans aren’t developing boards you become dismissive and scold them for ever questioning you or your experience in the first place.

    If you feel that you are above this thread and only take one or two members who also feel experienced you should all buy some Ak4497/ak4499 chips and start building your own dacs and headphone amps. The chips are sold openly so there’s no need to hack the Zishan with original designs if you claim like you do that to have this vast knowledge and have worked out your ideas to make them. It’s not like you are buying a PlayStation and there are proprietary purposes as to why you can’t build your own to hack the Zishans.

    I will also buy the T1 ak4499 version but I’m not qualified to discuss it in-depth on this thread and unless I am creating a dac from the scratch about what channels do what and whatnot there are forums and threads just for that.

    The Zishans are consumer-based entry level portable music players that performs well above its worth for those who love it which also happens to be mod-friendly what its not is a development board.
    Last edited: Sep 15, 2019
    fabien32 likes this.
  13. fabien32
    Hallelujah :)
    on another subject, do you know when (and where) we can buy a AK4499 ?
  14. Ivan TT
    OK, I gave you a chance to understand the rationale behind LPF circuit removal you were so critical about, but you are not seeking to understand, you have a personal agenda of criticism and character assassination, that settled.

    To my Zishan that I own I do whatever I please, so it is a development board to me, why not?

    PS: my 1st DSD does have an LPF filter similar to Figure 76 of datasheet, the only difference is that mine has 1st order filter and cut-off frequency of around 150kHz.
    Last edited: Sep 15, 2019
    Alex3221 likes this.
  15. Ivan TT
    Yes, that's my understanding too, but given the limited space full 4 channel +/- out circuit is unlikely to fit on the daughter card.

    One possibility I expect is an arrangement similar to the internal structure of ES9038PRO, which has 8 channels but only 2 outs, so outputs could be paralleled BEFORE LPF and I/V circuit, maybe requiring balancing resistors. Or only 2 channels will be used, after all "only" 134dB is a fantastic figure and 3dB improvement over AK4497... Or 4 SOIC opamps, these could fit I guess...

    PS: nope, datasheet says " IOUTL1P/R1P/L2P/R2P current and IOUTL1N/R1N/L2N/R2N current cannot be summed. The differential outputs are summed externally after I-V conversion", so scratch #1...
16 17 18 19 20 21 22 23 24 25
27 28 29 30 31 32 33 34 35 36

Share This Page