Четверг, 19.05.2022, 11:11
Приветствую Вас Гость | RSS
Главная | Регистрация | Вход
.......... ALEXMOD: projects
[ Новые сообщения · Участники · Правила форума · Поиск · RSS ]
  • Страница 2 из 2
  • «
  • 1
  • 2
Форум » Portable audio / решения для портативного аудио » DAPs / портативные источники » End Game или страдание аудиофила? QLS QA361
End Game или страдание аудиофила? QLS QA361
AlexДата: Среда, 27.04.2022 | Сообщение # 16
Автор сайта
Группа: Администраторы
Сообщений: 2985
Репутация: 120
Статус: Offline
2.5 Особенности аудиотракта. Выбор ОУ LPF

In a few days, I compared some SMD dual op amps on the QA361, including: MUSE8820, MUSE8920, AD8620AR, AD8620BR, OPA2134, LME49720, OPA2107, OPA1662, OPA1642, OPA1612, LME49860, AD823, LM4562, LME49722, AD80866, LT149866 , OPA2350, ADA4841-2, LMH6643, OPA2365, AD8616, in order to make each op amp play in the best state, some models also have modified peripherals to make them more compatible.

In the end, OPA2107, AD8620AR, AD8620BR and OPA1612 are all good op amps. Among them, AD8620BR is an upgrade of AD8620AR. The sound is slightly warmer and more delicate. It is completely different from OPA2107 and OPA1612 in style. It can be said that under QA361, OPA2107, AD8620, The three op amps of OPA1612 are not absolutely superior, but the styles of the three are completely different. Which one is used mainly depends on the sound selection under tuning and the overall matching.

The MUSE8820, which is very popular, has a completely unhealthy and unbalanced sound under our circuit. The vocals are very prominent, and the sound details are seriously lost. The entire sound field cannot be filled due to the lack of details, and the whole scene only exists in the middle part. (under AD8620 as reference).

For example, OPA2107 is an op amp with a very full sound and warmatmosphere. I finally chose it after comparing AD8620 and OPA1612 repeatedly. I think its sound is most suitable for QA361. Its price is very high (more than 100 yuan ($15) per piece), but there is a problem that the parameters are not high enough compared to these new models, so the parameters of the QA361 are actually the parameters of this chip (the parameters of this chip have become the bottleneck), the above DATASHEET The standard value of distortion is 0.001%, and there is no minimum and maximum value. I didn't care much about this before, because the distortion rate of some samples used in engineering machines is still relatively good.

Yesterday, I started testing batches of boards. Some of the numbers were lost and I was dumbfounded. A large part of the defective products were slightly distorted (not meeting our quality requirements), and the ratio was quite high. The engineer worked overtime all night to find the problem, and finally It turned out to be a problem with OPA2107.

This morning, we contacted the supplier for the first time. After a while, the supplier's FAE called back, saying that this distortion parameter is within the normal parameter range of this
chip's data sheet, and it is not a defective product. Dizziness  not within the scope of our quality requirements, what should we do? We can only urgently order another batch of chips and wait for them to be replaced when they come back. If they do not meet the standard, they will be replaced, and they will be replaced after testing. The newly ordered chips will have to wait for delivery, and we are under more time pressure. (Of course, after we have this
experience, we will test all the chips with a jig in the next batch, and then put them on the machine if they are qualified, but each batch will have to order a lot more of these chips, and the cost will be much higher. )

Note: In fact, if the OPA1612 with a much higher index was used at the beginning, there would not be these problems, the index of the whole machine would also be much higher,
and the price would be much cheaper. ......
But what is the use of high light index, QA361 with OPA1612 sound is not very good. (It is also worth spending more money to make detours for a good sound) In general, it is relatively smooth at present. The goal of starting shipments at the end of the month or the beginning of next month should be OK from the current situation. Everyone can wait for the goods with peace of mind, and don't rush us. I'm busy enough.

Haha,within 0.0008%, it is estimated that only the model can be changed (for example, OPA1612, which is too much cheaper, can have better indicators, but the chip on QA361 is not very good.), we use OPA2107 as our standard. The median value is 0.0008%, and a reasonable positive and negative deviation is given. If the deviation is too large, it is considered to be a defective product. The chip data sheet is 0.001% as the median value, and the maximum and minimum values ​​are not marked. (Actually, this batch has a high rate of deviations, and these have to be replaced).

Don't worry about replacing a SO-8 chip. Our technology and equipment can be replaced at least 20 times to ensure that the board is completely normal, not to mention only one replacement. (In the production of electronics, such replacements are all too normal.)

2.5.1 Выбор типа LPF

Recently,I have experimented with various LPF circuits, as well as adjusting the filter form, adjusting the crossover point, adjusting the values ​​of R and C to further improve the parameters & sound, and comparing the use of different OPAs...

Experiments found that LPF has an impact on the sound style. The influence is great, and some filtering forms will also slightly affect the quality and quantity of the low frequency. From the perspective of theoretical and test parameters, it is impossible to affect the low frequency, but it sounds like this.

Adjustingthe ratio of the capacitor can also adjust the sound between hard and soft, and can also bring out a better vertical air and line feeling. Some adjustments make the sound very punchy but with a lot of detail; some adjustments make the sound soft and audible but slightly less detailed; some component values ​​make the sound directly violent; some component values ​​make the sound subtle and subtle; some adjustments make the sound field larger; It will make the sound field small...

Some of the amplitude-frequency characteristics and phase characteristics are adjusted with the help of filter design and simulation software, but they actually sound and sound average; some phase characteristics obviously look slightly worse, But it sounds loose and airy. Sometimes only a small adjustment is made to the crossover point, and the sound style is completely different...
I have tried to dismantle for several days, but in the end I can't say which one is the best solution. It is based on the final sound of the whole machine and the direction of tuning.

Love the MUSIC, not equipment!
High-end portable now has a simple solution: aleXmod+RE272
AlexДата: Среда, 27.04.2022 | Сообщение # 17
Автор сайта
Группа: Администраторы
Сообщений: 2985
Репутация: 120
Статус: Offline
2.5 Особенности аудиотракта. Выбор выходного ОУ

The QA361 amp uses 6(!) OPA1622 high-performance operational amplifiers

The performance parameters of the OPA1622 are so excellent that they have appeared frequently in recent years, including some 10,000-dollar players. It is also used in the amp circuit. Most of the machines are Use 1 or 2 as headphone amplifier, QA361 uses 6 OPA1622 in the headphone amplifier circuit, the chips perform their respective duties to ensure lower distortion rate, better current output capability and better capacitance load driving сapability. The distortion degree (THD+N) of this chip is very good.

The distortion degree is only 0.000025% when the load is 600 ohms, and the distortion degree is only 0.000214% even when the load is as low as 16 ohms (few op amps with low resistance loads can there is so good).
Of course, its price is not cheap (only one chip is very cost-effective, but the QA361 uses 6 chips, and the price is very expensive compared to the op amps used by other amp circuits.): Yes The selected ultra-high current output is supplemented by our optional high-voltage mode. The RMS output voltage of the amp exceeds 10V, which basically takes most of the headphones on the market.
In terms of level and slew, we have actually measured that even the output level of 300 ohm HD650 to 2V is too loud for human ears, and we generally use this headset with a normal sound pressure of only 1V For more levels, we only need 4V level slew if we leave 200% margin for 2V, and our output RMS level exceeds 10V in high voltage mode, so audiophiles can choose a wider range of headphones , Impedance is not a problem that needs to be considered.
On the contrary, I prefer the sound style of most high-impedance headphones. Of course, if you like low-impedance headphones, the large current output capability of the QA361 amp and the large capacitive load driving force can deal with large Most low-impedance headphones are also very easy. But there are some headphones that are difficult to push for large desktop amps, such as  HD800, T1, AKG K701... This type of headphones are difficult to push, so don't expect QA361 can be pushed.)
Such low-distortion op amps perform their respective functions. The final measured distortion data of 6 pieces is the same as 1 piece, but the capacitive load driving capability and the optional super-current output capability are brought to the final headphone driving force
and are not certain. A single chip can achieve it, and users can also choose the working mode of a single op amp or multiple op amps in the menu according to the needs of their рeadphones.

About OPA1622

Theperformance parameters of OPA1622 are very good. In addition to a portable player with a selling price of nearly 10,000 yuan and the 4.4 balance port of PAW GOLD TOUCH, there are also many mobile phones called HiFi. The mobile phone is not tall enough, so there is a general statement about this chip.

Int he past few years of QA361 design, we have tried the QA360's separate amp circuit, AD8397, TPA6120, OPA+LME49600, OPA+BUF634, OPA1622, and a certain AD8017 called SuperAMP.....

AD8397 is in The QA361 is slightly thin, and the low frequency is the weak item. At the same time, the capacitive load carrying capacity of the AD8397 is only 270pF at a gain of 2, which is much weaker than the OPA1622. The amp circuit originally used by QA360MOD/LE is good for high impedance headphones under QA361, but when dealing with ultra-low impedance headphones less than 10 ohms, the driving force is obviously inferior to OPA1622.

TPA6120 does not support low impedance headphones on QA361. , Pushing low-impedance headphones is not as good as high-impedance headphones, and its capacitive load driving ability is very weak.

OPA+LME49600 sounds a little irritating when listening to some headphones on the QA361, the high frequency is not very controllable, and the sound is not enough to listen to. I have also tried to adjust the OPA, but it is still difficult to achieve a satisfactory level.

OPA+BUF634 is not as delicate as OPA1622 on QA361, and it is a little messy and not stable enough. I tried to replace several OPAs, but I gave up without getting very satisfactory results.

The AD8017 also has poor capacitive load drive capability and does not support high-voltage power supplies.

Note:The above comparison results are only valid under QA361, and the situation may be different under different circuits or different machines.

In the end, OPA1622 is the most suitable under QA361. It can drive high, medium and low resistance well, and the capacitive load driving ability is very strong, and the sound is balanced and delicate.

In order to further improve the driving ability of capacitive loads and realize the optional ultra-high current output capability, we use 6 OPA1622s, each of which divides the labor to achieve balanced positive and negative current expansion...
AD8017 is an op amp that many people may not have heard of.
This is a very powerful op amp chip known as "SuperAMP".

In fact, the test thrust is much stronger than the very hot op amp chip AD8397, and the high frequency -3DB cutoff point of AD8397 is 69Mhz, while the high frequency cutoff point of AD8017 is 160Mhz, the slew rate of AD8397 is 53V/us, and the AD8017's high frequency cutoff point is 53V/us. With a slew rate of 1600V/us, the AD8017 can have an output voltage of nearly 10 V pp even with an ultra-low impedance load of 10 ohms.
The short-circuit current of a single AD8017 can be as high as 1500mA.

These parameters can be described as "abnormal".
If you like DIY amps, I strongly recommend you try the AD8017.

Some people will ask: Why does QLS not use such a good amp chip?
This is a good question. We have been researching this chip for a long time. The reason why it is not used in the product is mainly as follows: Reasons:

1.We actually heard that although this chip has a strong thrust, the sound is not delicate enough, not natural and delicate enough.

2.Although AD8017 is very good with resistive load, the driving ability of capacitive load is still relatively weak, so if it is used as an amp, it will be relatively hard to choose headphones. With some headphones, users will be amazed at its thrust. It is called a large desktop computer, but some headphones may be very unpleasant and polarized.

3.The working voltage of AD8017 can only be up to 12V or plus or minus 6V, so it cannot be used in some occasions with high working voltage
Прикрепления: 6450423.jpg(26.7 Kb)

Love the MUSIC, not equipment!
High-end portable now has a simple solution: aleXmod+RE272
Форум » Portable audio / решения для портативного аудио » DAPs / портативные источники » End Game или страдание аудиофила? QLS QA361
  • Страница 2 из 2
  • «
  • 1
  • 2

Форма входа
Second-hand audio
xDuoo X20
FAQ по доработкам
Блошиный рынок

Copyright MyCorp © 2022 I Сделать бесплатный сайт с uCoz