Схемку можно глянуть, откуда передирали? Сомнения гложут народные массы.
Что касается идеи измерить и измерять "класс Д" вообще и трипатх в частности.
В упор не понимаю, зачем? Все графики ЕСТЬ в даташитах. В случае Трипатха - графики IMD, THD vs Freq, АЧХ значения SNR, первого полюса и т.д. кошмарны. Просто волосы на голове шевелятся от первобытного ужаса. Они THD+Freq вообще только в одном даташите приводят, по-моему. И на него смотреть аж больно.
Что же там у остальных детищ?
График IMD он приводит в нелинейном масштабе спектра, чтобы "ежик" на маленькой картинке почти сливался.
Не верите что все так плохо, что ли?
Смысл измерений какой?
Или хотите доказать, что звук и измерения связаны как-то иначе, обнаружить зависимость, предложить собственную методику? Так вроде много кто такой позиции придерживается. И не безосновательно.
Чивера, например, достаточно прочитать. Уже все украдено. До нас.
Если Вам нравится, вообще вопросов нет. Нафиг этот велосипед?
Про трипатх еще в 2002 году было сказано:
Tripath employs a rather strict definition of class D (namely a fixed-frequency PWM based one) in order to award themselves a new class name, class T. Nonetheless, for all intents and purposes it may be classified as class D without omission.
They were the first to deliver a serious chip to the class D scene, a 10W device targeted at multimedia. Another product is a driver module to build higher power amplifiers with. Power stages found in Tripath's products are half- and full-bridge setups.
Operating principle
The control system is a 3rd order switched-cap loop controller (the first order forcibly converted to continuous-time to prevent aliasing of switching errors and noise) with hysteresis modulation to create a time-quantised PWM (of variable repetition rate) signal. The time-quantisation implies that the PWM will have quantisation noise on it. For a third order loop (now a noise-shaper) to sufficiently control this noise inside the audio band a high sampling rate (switching frequency) is required. Indeed, the (idle) switching rate is frighteningly high at 1.4MHz. The result is that even Tripath are reluctant to claim much better than 80% efficiency.
Distortion figures are good, but are usually spoiled by the coils of the output filter which is outside the control loop. The figures inspire Tripath to make claims toward "audiophile" sound quality, but the reality is that soundwise it is hardly better than anyone’s "first".
Reportedly, Tripath has not always been able to satisfactorily solve customers' EMI problems.
Good reasons to use these chips
The idle spectrum shows "noisy" harmonics, making audible intermodulation with AM carriers less obnoxious. Most other designs produce a clear squeal whereas here a more noise-like interference is produced.
Not good reasons to use these chips
• "Digital Amplifiers". If you're into these, tripath are 100% analogue.
• "Audiophile sound quality". Not more or less than most and even simpler implementations.
Good reasons to not use these chips
• Second lowest efficiency around (Sharp wins this one)
• Sometimes problematic to obtain EMC.
Rating
• Idea: 4/10
• Execution: 5/10
• Sound: 5/10
• Performance: 6/10
В 2003 я его послушал и понял, что плохие измерения могут сочетаться не только с отличным звуком, но и с ужасным.
))
В то же самое время, про сигнальные образцы UcD было сказано:
This circuit is now replacing SODA. It was originally somewhat immodestly called "Ultimate Class D", now "Universal" (just as immodest imho).
Operating principle
The modulation scheme is some phase-controlled self-oscillating design with feedback taken only after the output filter. No details are given but the info says something like "second order loop control with zero order response" - beats me what this should mean apart from a smoke curtain. Anyhow, the results are very decent THD figures that remain constant with frequency and an unusually low output impedance (lower than most linear amplifiers), resulting in a marvellously flat frequency response independent of load. The circuit is a half bridge with two N-MOS this time and fully discrete drive circuitry (including the comparator), resulting in phenomenally low cost.
Comments
Sound quality is simply the best I’ve ever come across (including megabuck class A amps etc), by not a small margin. EMC performance is comparable to that of the Philips chips. As this is a discrete circuit even more than SODA, the same cautionary notes apply.
Good reasons to use this circuit
• Best sounding amplifier ever
• One of the lowest cost solutions around
• Good measured performance
• Excellent EMC (when properly applied)
Not good reasons to use this circuit
"Zero feedback": The company tries its best to convince you that the amp’s behaviour is identical to a zero-feedback amplifier. It has feedback, OK?
Good reasons not to use this circuit
No precooked modules are available and trying to "build your own" based on the technical documentation is unlikely to succeed easily.
Rating
• Idea: 8/10
• Execution: 9/10
• Sound: 9/10
• Performance: 8/10
Удивляюсь резону "ковыряться" с горетрипатхом.
Или я чего не догоняю?
