ZiShan T1 Hi-Fi Player Thread

Is this the best DAP under $100?

  • Yes

    Votes: 17 23.3%
  • No

    Votes: 20 27.4%
  • Way above the price range!

    Votes: 6 8.2%
  • Could be

    Votes: 30 41.1%

  • Total voters
    73
Sep 16, 2019 at 11:18 AM Post #436 of 652
You were right you can play with uf values even the resistors to shape the filters but they are necessary and they too can be fine tuned and can be charted; You are missing this golden opportunity to get the balance right!

The signal coming from the analog board is at a higher frequency then it’s down sampling that what through you off about the lpf and it’s measurements. Because there’s is much noise shaping going through the circuitry ONLY the audible signal is supposed to be coming in the output!

The Digital filters are based on pulse outside the audio signal which effect bass response etc. it’s actually in the Datasheet.

Yes there are other application for the chip but this is how the DSD Pro DAC was made.

DSD= Direct Stream Digital aka SACD

Hence, “Zishan DSD”
This is factually wrong, on every point.
But as I said before, I see no point to explain how things actually work as it proves to only be a waste of my time and energy.
How’s points 1-6 going?
Following those will allow you to clear me of unfounded accusations.
 
Sep 16, 2019 at 11:47 AM Post #437 of 652
“The process of creating a DSD signal is conceptually similar to taking a one-bit delta-sigma analog-to-digital (A/D) converter and removing the decimator, which converts the 1-bit bitstream into multi-bit PCM. Instead, the 1-bit signal is recorded directly and, in theory, only requires a lowpass filter to reconstruct the original analog waveform. In reality it is a little more complex, and the analogy is incomplete in that 1-bit sigma-delta converters are these days rather unusual, one reason being that a one-bit signal cannot be dithered properly: most modern sigma-delta converters are multi-bit.“

The Akm4497eq goes much higher then DSD64 which is mentioned here:

“A DSD recorder uses delta-sigma modulation. DSD is 1-bit with a 2.8224 MHz sampling rate. The output from a DSD recorder is a bitstream. The long-term average of this signal is proportional to the original signal. DSD uses noise shaping techniques to push quantization noise up to inaudible ultrasonic frequencies. In principle, the retention of the bitstream in DSD lets the SACD player use a basic (one-bit) DAC design with a low-order analog filter. The SACD format can deliver a dynamic range of 120 dB from 20 Hz to 20 kHz and an extended frequency response up to 100 kHz—though most current players list an upper limit of 80–90 kHz.“

The Zishan Pro lpf goes up to 186Khz because the firmware plays DSD256!

https://en.m.wikipedia.org/wiki/Direct_Stream_Digital

7B8797E8-BB90-4AAC-95BA-EE213D2F09E5.jpeg
 
Last edited:
Sep 16, 2019 at 1:50 PM Post #438 of 652
Not really, you need SMD.
But to be honest, good starting point would be using maybe not as precise, but cheaper oscillators that fit well. Also they are tricky to solder, do you have hot air station or someone who has one?
Maybe ask @HeyManslowdown97 what oscillators he used, if he can provide links?


@fabien32 you can all three frequencies here!

https://www.diyinhk.com/shop/audio-kits/35-ndk-nz2520sd-20ppm-ultra-low-phase-noise-oscillator.html

(Wasn’t aware of the third oscillator)

45, 49 and 24 for the CPU?? Not sure about the 24 but I can confirm the 45, 49 do work and they are fantastic! You can also replace the windbond flash like I did on my DSD Pro it improves the cache I believe. It did with my DSD Pro and made the UI much snapper when I replaced it.

Also changing the ceramic caps with less % off and low esr will improve the noise levels in the clocks. In fact replacing all the ceramic caps with higher voltage will also make the player better then in stock but I don’t think it’s truly necessary.

Having ultra low phase noise allows for better balanced in stereo since the T1 doesn’t provide the offset option.

In the schematics of the ak4497eq the lpf is using 2x Opa1612 which isn’t a Class D op amp.

They are traditional Class AB.

Class D isn't popular due to EMI issues (can't do BTL output!). Not even the Opa275 is a Class D op amp in the Zishan DSD Pro lpf.
 
Last edited:
Sep 16, 2019 at 6:05 PM Post #441 of 652
can you give for those a link? if so then the best one money can buy

I used Arrow.com. However, it’s changed a great deal and maynot be cost effective for you since they got rid of free shipping internationally and in the Us you can get free overnight shipping over 50 dollars.

I don’t want to give you the wrong stuff since I did these mods on the DSD Pro and not the T1.

The “Windbound” forgot the exact name but it’s written on the chip with you open your T1 just copy the number and do a search in arrow. If you change the flash you will have to reboot the firmware. It’s not hard but it’s risky and I can’t recommend it to someone that isn’t somewhat experienced with a hot air soldering on the DAC.

The DSD Pro The low esr around the clock are sized 0805 0.1 ufs but even 0605 will do. I'm not sure what cap values are on on T1 so you'll have to measure their values but honestly the T1 sound great as is its not necessary to kill yourself with lots of mods.

Start by changing the op amps on dip8. And work gradually with mods. One I got my first DSD Pro, i was like you and wanted to change everything and spent lots of time & $ in parts, etc at the end of the day it’s really only about sounds 15%-20 better. You wanna get the op amp that sound great to you and meet your own preferred audio signature.

Also I’m not the greatest technician when it comes to actually doing the soldering and hot air soldering I’m only good enough to “get the job done”.

Disclaimer: Modding really isn’t for everyone and you can damage your T1 (and more importantly hurt yourself). Pay lots of attention to this disclaimer I was given the same advice before I did any modding on my DSD Pro and it is truly sound advice because I did so really stupid mistakes that could’ve ended up hurting me in the end even though luckily nothing ever happened it’s is somewhat dangerous to do this type of work on a PCB board. Please take this to heart. I was somewhat cocky when I first started b/c of beginners lucky and I almost made really dumb and careless mistakes. Take EVERY precautionary steps to prevent injury to yourself and others!

I don’t wanna give false claim that everything will work out or that my mods will work for you as well. There is a lot of risk involved! I try to stay within my own skill set and don’t see the benefits of doing crazy overhauls to the Dac that I’ve seen on these threads and I have found better results and more worthwhile with risk-free to fix the actual source music files so they play nicely on all my music players.
 
Last edited:
Sep 16, 2019 at 8:55 PM Post #442 of 652
In a conventional transistor amplifier, the output stage contains transistors that supply the instantaneous continuous output current. The many possible implementations for audio systems include Classes A, AB, and B. Compared with Class D designs, the output-stage power dissipation is large in even the most efficient linear output stages. This difference gives Class D significant advantages in many applications because the lower power dissipation produces less heat, saves circuit board space and cost, and extends battery life in portable systems.


It’s digital (Bjt) because DSD stands for Direct Stream Digital (PWM) and transmits the digital into the analog for both Circuitry designs for the DSD pro and T1 we must agree on this that the source music is digital (format) come from the source files. How it is read and then turned into analog is what a DAC does!

The Class D Op amp has the switches frequencies baked into microprocessors circuitry and doesn’t require the digital Bjt. It’s not a Class D design as misinterpreted of the two separate PWM Modulator in the Delta Sigma.

This is important because Ivan is confused about the circuitry design.

DSD = Direct Stream Digital

The signal is digital (bjt) that’s being transmitted to the output (analog) aka Digital to Analog (DAC)



644F0D77-1C3A-47CE-9EB4-0E1226A1BA56.jpeg
BAABDDAB-FAF8-4F0A-B977-7F39BA0B1A76.jpeg
 
Last edited:
Sep 17, 2019 at 3:27 AM Post #443 of 652
@fabien32

The Zishans are DACs. Which means they convert digital-to-analog signal. A lot of the mods that you see fail to see the Delta Sigma Modulation is basically converting a digital signal into an analog one.

These particular players have the Ak449x chips which specifically the decodes DSD signals into an analog signal. It plays DSD natively.

The only really improvement you can make is eh digital signal to the analog signal. The output stage as “IvanTT” in this specific application convert the digital into an analog signal.

Filters, output stage and lpf circuitry and even op amps have a wide variety of other application. The “specs” of the op amps in the modern world for exceeds the usage in the audio. So asking someone do describe the function of all the circuitry designs it is unnecessary litmus test. We are not electricians or here for electricity.

Basically what’s in the digital files (information) is decoded. Think of them as your DNA what written on them remains the same constantly. Unless there are major issues with the sound you can only improve the DAC design to certain point. Better power supply, clocks, battery life, etc)

The reason many Zishan uses are experiencing very hot T1 is because there files are harder to read and requires more cpu power to convert the signal.

There’s are lot more “hands on” circuitry when it’s an analog player like vinyl that does with the physical elements for instance that effects the sound but in this case it just reads binary code ones and zeros into an analog signal.

That’s why people when they make improvement say that the digital sounds cold, clinical and analytical.

DSD files in my opinion are the exception and sound the most analog of all of the formats but nobody plays them on these Zishan threads! I’ve also heard the PCM played on a R2R DAC sounds organic more natural.

I believe that many folks on these threads who talk endlessly about the circuitry can’t see the trees from the forest and that’s where the misinformation stems from.

I have mentioned on several occasions that if Ivan is interested in the circuitry, he lives in New Zealand which has access to the US markets to buy all the offical Evaluation boards to make things like headphone amps, dacs etc. All the world renowned companies who make the components are extremely generous with their knowledge of instructional videos of each electrical topic and even have forums for these very discussions. Even the arrow.com where many Zishan users have bought their parts have these forums.

I’m not saying they should be censored in fact talk about it all you want I just think they feel a sense of chip on their shoulder. In the digital age software is king! (Why do you think the most profitable companies in the world are in the Silicon Valley?) And I have yet to see in these discussion how to improve the firmware or open source it with more applications for audio. These are of course hacks and the Zishans are not development boards but they do exist and so do all the components found on the Zishan in the US markets can be programmed to do what Zishan. If these folks are who they say they are then why not build one from scratch!
 
Last edited:
Sep 17, 2019 at 3:44 AM Post #444 of 652
The only really improvement you can make is eh digital signal to the analog signal. The output stage as “IvanTT” in this specific application convert the digital into an analog signal.
Apologies, but this is factually incorrect.

The output of DACs that handle digital to analogue conversion (as the name implies) is analogue, so all signals starting from DACs chips analogue output pins is analogue, including LPF, subtractor opamp and discrete transistor output stage, including DC-DC converter that provides power to analogue circuits and has nothing to do to actual digital to analogue conversion. There’s no Class D circuits in any of Zishan our there either, it is a simple fact (obviously, not to you) that proves impossible to prove as it is pretty much axiomatic.

I appreciate that it’s an honest mistake on your behalf and you believe that your understanding is true and correct, but it is not, unfortunately. It also makes you believe that you are personally corrected or criticised, but actually the incorrect and flawed understanding of circuit and how it works is being (unsuccessfully so far) argued with.
 
Last edited:
Sep 17, 2019 at 4:06 AM Post #445 of 652
Now, speaking about modding...

Zishans are fantastic for DIY, as they are easy to open, offer quick gain when stock components are replaced/upgraded (in the world of HiFi we have a rule of diminishing returns, where each % of subjective sound quality improvement is more and more difficult and expensive), but even inexperienced person can make those improvements easily enough.

Zishans also proved to be fantastic development boards/test beds where various ideas/circuits can be easily tried and tested and even worst case scenario (total loss) is not that financially catastrophic.

That’s why on these forums we had a very nice and supportive and friendly community forming over the years and there were few most inspirational tweaks developed and implemented, seeking this elusive goosebumps inducing sound.

Unfortunately, the spirit of this community is incompatible with loud aggressive confrontational vibe, so this thread is unlikely to support creation of such a community around T1 player and its mods at all levels, from basic to intermediate to advanced. I lament lost opportunity, to be honest.
 
Sep 17, 2019 at 4:25 AM Post #446 of 652
Apologies, but this is factually incorrect.

The output of DACs that handle digital to analogue conversion (as the name implies) is analogue, so all signals starting from DACs chips analogue output pins is analogue, including LPF, subtractor opamp and discrete transistor output stage, including DC-DC converter that provides power to analogue circuits and has nothing to do to actual digital to analogue conversion. There’s no Class D circuits in any of Zishan our there either, it is a simple fact (obviously, not to you) that proves impossible to prove as it is pretty much axiomatic.

I appreciate that it’s an honest mistake on your behalf and you believe that your understanding is true and correct, but it is not, unfortunately. It also makes you believe that you are personally corrected or criticised, but actually the incorrect and flawed understanding of circuit and how it works is being (unsuccessfully so far) argued with.

Sadly you continue to misinterpret what I said and have your own agenda.

There is support for DSD through I2S! Ie:

DSD uses the microcontroller STM32F7 + CPLD + dual audio crystal to achieve I2S

The digital signal is being transmitted through the power supply. The conversion is being done by the CPU. It’s not analog it’s steaming digital and decoded at the output stages. It’s not playing PCM, it must be converted you said this yourself!

There is no Class D I said this fifteen million times and yet you keep repeating it.
 
Last edited:
Sep 17, 2019 at 4:36 AM Post #447 of 652
Jesus Christ, the thread is out of control. This is serious. The thread is already losing 2 (or more) kind persons. Just because of 2 persons, the thread can get locked. I really hope that this thread can help people about the Zishan T1. Why won't you two just start a PM and argue there? I've tried asking you two to stop this kind of argument politely, but guess what? It's no use I think. 1 more heated argument and that's it. I'll take things seriously. This thread is not just for me or 2 people arguing. But, as I already said many times earlier, this thread is for a lot of people.

People are reluctant to read when a thread that they need is full of pointless things. Should I copy and paste all of the HF rules here and through PM?

I want people to share their opinion and ideas here not hatred and misinformation. Be mature. Think about others. Respect opinions and the people. I will still tolerate you two to be peace to each other. If it's impossible, then start PM. Fight there. And, if it's not enough, find somewhere to pick a fight.

Come on guys, this is head-fi
I actually enjoy the debate, it helps me understand things, you all are very informative and friendly (to me) my only problem is that it became hostile.
i live i a country that is in constant war, when i surf the net i try to get a break from that. so really guys, lets be friendly
 
Sep 17, 2019 at 7:19 AM Post #449 of 652
2. Find out the name of Allwinner CPU from the same photo (A). Google its darasheet (B) and find out what clock frequency it usually uses (C), compare to photos of PCBs that use same CPU (D), compare to T1 (E). Post your findings here clearing me of your misplaced suspicions.
A. Allwinner F1C100s

B. Datasheet, 1st page in google results

C. Page 16 of Datasheet, Clock: 24Mhz Crystal input

D. Google image search, first result:
LicheePi-Nano-Board-Large.jpg


E. Clearly same CPU, same crystal in the same location, conclusion: 24Mhz crystal clocks T1's CPU.
 
Last edited:
Sep 17, 2019 at 7:41 AM Post #450 of 652
DSD uses the microcontroller STM32F7 + CPLD + dual audio crystal to achieve I2S
Much better!
I2S bus is used for both PCM and DSD data stream and clock in some implementations, specifically AKM.
The digital signal is being transmitted through the power supply.
Not there yet, digital data is sent over I2S bus, definitely not over power supply!
Specifically for AK4497 pins 3-5 and 61 (page 5 of datasheet).
The conversion is being done by the CPU.
Digital to Analogue conversion is performed by Digital to Analogue convertor, DAC.
Digital data (PCM or DSD) + clock are sent to DAC via I2S bus and is converted to analogue signal (AK4497: pins 31-34 and 47-50, AOUT (for Analogue Output), L and R, Positive and Negative).
It’s not analog it’s steaming digital and decoded at the output stages.
Yup, it is digital at DAC's (Digital to Analogue Convertor) input and converted to analogue signal by DAC (Digital to Analogue Convertor) and then stays in analogue domain all the way through LPF, Subtractor op-amp and discrete transistor headphone output stage. For the avoidance of the doubt, from analogue output pins of DAC and all the way to headphone output signal is analogue.
It’s not playing PCM, it must be converted you said this yourself!
Absolutely, what I said is PCM stream (say 16 bit 44,1kHz) is converted to DSD data inside of Delta-Sigma DAC and is converted to analogue domain just like DSD data stream would, which does not need conversion (unless DAC's volume is enabled which it is in Zishans except for 4495).
This process involves Interpolation, Delta-Sigma modulation and digital filtering (these are the filters that as user switchable via software and they relax the requirements for analogue filtering compared to Nyquist reconstruction filters. All done inside DAC's chip and result in analogue domain leaves DAC via analogue output pins as above.

Power supply is not controlled by CPU (except for supplying power when Zishan is booted and turning it off) and data stream not only does not enter power supply, it is highly undesirable as this will introduce noise in analogue path: LPF, subtractor opamp and discrete headphone output stage.
 
Last edited:

Users who are viewing this thread

Back
Top