Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 02 Jun 2015, 16:30

P1140801As-copy2.jpg
P1140801As-copy2.jpg (82.87 KiB) Viewed 117459 times
We would like to give our warmest welcome to Mr. Ivo Linnenberg from LINNENBERG-ELEKTRONIK to participate on the forum. It would be invaluable to having the designer himself to introduce his product concept, and other audio related matters.

Many thanks to Ivo !
There is always room for improvement. 日日新 苟日新

Linnenberg
Posts: 5
Joined: 17 May 2013, 17:19

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by Linnenberg » 02 Jun 2015, 20:49

LINNENBERG AUDIO is the brand name for audio products from LINNENBERG ELEKTRONIK. With the help of a handful of qualified employees, I manufacture fine purist high-end components. Since 1994 the company designs the most demanding digital and analogue circuits. We at LINNENBERG are in business for more than 20 years; our experience expands from chips to tubes and from rugged professional audio (PA-systems) to delicate high end audio.

It was always my intention to design and build components with heart and soul having a solid technical background. Manufactures claiming that sound quality is the only measure are wrong as well as those are wrong only looking for the numbers on their meters. A perfect balance is the key for a successful audio component. I think we did some great achievements here, but you, the customer will the final judge. May I encourage you to give these fine instruments a listen? Mr. Tsang will be the ultimate guide on your ride.

This forum is the perfect place to answer technical questions and give insides to any of our products. Please do not hesitate to ask.

Regards

Ivo Linnenberg CEO

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 03 Jun 2015, 14:29

Let me ask a few questions, it is good if you could provide us some insight ...


1. I see that you have been design various DAC products in the past. May I ask, from your experience, what elements are technically important when working in dac design ?

2. First off, I see that there is a strong hype on master clock accuracy in recent years. Obviously it is not simply the crystal oscillator, which takes the sole responsibility. If you please elaborate a bit, hopefully in a not too technical way, the other factors that influence the high accuracy.

3. I suppose a lot of DAC chips must have been compared before you decide on the ESS 9018 chip. But in general dare I say the 9018 DAC chip is not having some very good reputation sound wise, so may I have your reasoning in choosing the 9018 in your VIVACE ?

4. Last time we talked about the overload consideration in the analog stage when using the 9018 chip. You mentioned that it is this overload problem in analog stage, or at least it is one of the reasons, that gives the 9018 some bad reputation in terms of musicality. Could you give us some insight please ?
There is always room for improvement. 日日新 苟日新

Linnenberg
Posts: 5
Joined: 17 May 2013, 17:19

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by Linnenberg » 04 Jun 2015, 15:27

Ok, let me explain some design choices …

1)

The first commercially available product from LINNENBERG AUDIO in 1994 was the POWERDAC, unquestionable a DAC product. The DAC chip used was the PCM1702K, a 20bit ladder DAC in the highest selection grade. Over the years, I worked with nearly all chips the semiconductor industry could deliver. Leading parts from Analog Devices (AD1852, AD1853), Crystal Semiconductor (CS4391, CS4396, 43122), Texas Instruments (PCM1794A, DSD1794A, PCM5102A) were used, to name but a few. Although bit depth has increased from 20 to 32 bit, max. sample rate increased from 48kHz to 384kHz; although some devices featured a current output, while others had a voltage output, there is something all these chips have in common:

It is not the chip that makes the music, it is how it is used that determines if the end product sound right or just like crap.

2)

As described in the previous part, the processing speed of the digital circuits increased considerable. Remember, in 1994 the operating speed of an © Intel Pentium processor was 100MHz, the max. clock rate in audio circuits was a mere 12MHz. The digital revolution demanded tighter and tighter timing of digital signals. With today’s DAC chips, one needs a very sophisticated clock in order to take advantage of the capabilities of the new technologies. The old PCM1702 was much more forgiving (acoustically) than those new multibit sigma delta converters, even when industry claims other. To answer your question, it is completely useless to employ a very low jitter clock oscillator when not taking care about signal transmission. A few cm of board traces (or even worse a flying wire) ruins the signal quality of even the best atomic clock if not transmitted in a controlled environment (constant characteristic impedance to eliminate reflections). This holds true for each and every digital signal inside a DAC.

3)

As of today, under the DAC chips I like the most are the Texas Instruments series (PCM1794, PCM1792, DSD1794 and DSD1792) and the ES9018 from ESS Technology. The Texas Instruments (Burr Brown) chips are simply derivates from the same die. Please note, that the 32bit PCM1795 does not belong to that family. It is part of the lower quality PCM1796 / 98 series. Why TI has chosen an inferior series member for their 32bit version remains their secret. Generally it does not make sense.

In contrast, the ES9018 is a no compromise chip. It is programmable, with great user flexibility giving the user the chance to voice his product. The major drawback is a 4 times higher price tag.

4)

There are two different overload conditions encountered with the ES9018. The first one is similar with every other DAC chip on the market featuring a current output: A DAC is practically a digital piece of hardware and not a mixed signal component! The analog waveform is created by incredibly fast switching current sources which either are fully on or fully off. These charge packets generate a voltage (our desired waveform) when operated in a virtual ground connection. The limited speed of an OP I/V converter and the presence of some capacitance around smooth out the digital nature of the signal making it “analog”. Generally, an OP I/V is too slow for following such digital steps, creating nasty intermodulation distortion, giving a somewhat harsh sound. The HF spectrum of the ES9018 is somewhat more aggressive than the spectrum of for instance the PCM1794. For this reason, the VIVACE uses a discrete amplifier stage especially constructed for the application.

Another overload condition is created by letting the ES9018 sink or source to much DC current at its outputs. Many commercially available products as well as DIY approaches do not bother with this fact.


Ivo Linnenberg CEO

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 05 Jun 2015, 00:56

Today the CAS enjoys considerable popularity in domestic hifi equipment, by this it picks out the importance on the USB interface. No doubt that all the more enthusiastic products are giving strong attention on the USB input, for example the use of as-synchronize data transmission.

5. regarding this, please talk a little about the design particulars in your USB input of the VAVACE, and if you don't mind I stretch a bit, how do you see the add-on USB/SPDIF converters on this matter ;

6. let's come to the power supply section, what do you want to achieve when partnering your VIVACE with the UNISONO powersupply ?
There is always room for improvement. 日日新 苟日新

Linnenberg
Posts: 5
Joined: 17 May 2013, 17:19

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by Linnenberg » 05 Jun 2015, 23:18

As we are in the mood for some technical background, let is continue our “inside story” …

5)

Besides vinyl playback, PC audio has emerged as the source for audiophile stereo systems. Not long ago, CD playback had the dominant position, but that is diminishing more and more. For this reason, a DAC with a state of the art USB input is mandatory these days.

Our offering, the VIVACE has one of the most sophisticated USB inputs on the market, regardless of price. It features an asynchronous data transfer, an isolation barrier and a low noise clock management. Ok, asynchronous data transfer, meaning the transport (PC) is slaved to the DAC has become quite common in the past two to three years. So, that’s not unique anymore. But the VIVACE USB input has more to offer: a complete galvanic separation of the PC-ground from the DAC-ground leading to a huge reduction of transmitted HF noise originating from the PC circuitry.

Most digital equipment manufacturers leave out the galvanic isolation, the vast majority for cost reasons, and the rest claiming that galvanic isolation introduces a new jitter source. We at LINNENBERG are not willing to trade off one thing (noise reduction) against another (mild jitter generation). With our specially developed double slave USB circuit we can have it both: no additional jitter + 100% noise reduction.


6)
In order to explain our decision to introduce the UNISONO power supply, we have to go into greater detail here as well. Numerous compact sized high end audio components use a wall adapter these days. There are very good reasons for that:

- eliminating the noisy and mechanically vibrating transformer inside the gear
- mains voltage range from 100 – 260 VAC, making export easy
- low power consumption, green technology

When using a highest quality wall adapter, like the one we supply with our gear, the result is very satisfying. Nevertheless, we won’t keep quiet about the fact, that there are drawbacks using this technology:

- a wall adapter, even the very best has some residual HF noise
- the industry standard barrel connector, common to all low voltage DC adapter enables exclusively a single voltage due to the two contact (poles) limitation.
- limited current capability, moderate impedance

Some third party vendors as well as some original manufacturers figured out soon, that replacing the wall adapter with a conventional power supply increases the sound quality of the actual audio device, because two of the above listed drawbacks are eliminated.

When you go “plug and play”; and that’s what everybody offers, a serious problem remains: Your complete audio electronics is operated by a single voltage. 5V, 12V or 15V are ordinary voltages. You just have one of these and that’s it, while your gear needs a bunch of different voltages; some positive, some negative (referred to ground 0V). Now, one needs to know, that every audio component on the market using an outboard wall adapter or an open frame switch mode supply inside the gear MUST use additional DC-DC-converter in order to generate all those different voltages an audio circuit NEEDS. That’s fact, regardless of price. A well known Swiss company, once famous for their tiny tape recorders – as well as other manufacturers which we won’t name either – goes so far to generate tube voltages (anode voltage, heater voltage) out of a single DC voltage power input. Remember: a DC-DC-converter is a switch mode power supply, and you need lots of them!!! What ever you do, some residual HF noise will leak in to your audio signal.

To cut a long story short; UNISONO in conjunction with the DPP models dismissed all switch supplies inside the gear, eliminating a possible source of grain.

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 08 Jun 2015, 17:36

這裡略作簡短節錄 :

1)
根據過往經驗, 那些方面是設計解碼器之重點呢 ? :

我方第一個商業產品是1994年推出的POWERDAC, 它採用當時最頂級的20bit解碼IC, PCM1702K。這麼多年來, 所有在半導體市場上曾經出現過的品種, 差不多全都實際去試用過。 事實上它們大致可分為兩類, 分別是電流輸出類型, 或者電壓輸出類型。不管怎樣, 所有這些解碼IC (集成電路) 都有一個相同本質 :

它們本身並非一塊全功能工作的音頻元件, 它們唯一能夠實行的工作任務只是進行數據處理, 而不能單靠其本身去發出任何聲音信号。
任何一款解碼器, 由數據輸入, 然後轉換成聲音信号, 當中使用的任何解碼IC都只是負責其中一個環節, 整個流程之中必要經過週邊電路和其他元件共同配合來工作, 才能完成。 最終我們所聽到的聲音, 聲音信号, 是整台機器的複合成果。

簡而言之, 不管什麼解碼IC, 都不能單獨發出聲音。

如何為這些解碼IC選取符合目標要求的工作方式, 在電路上將其施展, 去發揮, 才是要點。

利用優良的綜合佈局, 讓電路上各環節扣應合宜地工作, 各部元件各施其職, 才是決定因素。


2)
近年大家都強調高精準度時鐘的重要性, 我雖然不是技術派, 也知道不是單純依靠一個高精度晶震這麼簡單。在針對極高準確度這方面, 您看有那些關鍵因素 :

確實是的, 現代解碼器之所以能夠達到非常高速的數據處理能力, 必須依靠極之先進和精密的時鐘基準來進行工作。 由此, 時基的傳送方式必須異常嚴謹, 否則, 空有一個原子能時鐘也是無補於事。 在傳送路途上, 短至只有幾公分長度的電綫或者印刷電路都會影响時基的精確度, 沒能例外。
There is always room for improvement. 日日新 苟日新

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 08 Jun 2015, 17:41

3)
相信您們廠方也比較過各式各樣的解碼IC, 然後才選用上ESS9018, 但恕我直言, ESS9018的聲音樂感一般都於人垢病, 那廠方究竟為何選用這塊IC呢 ?

我喜歡的解碼IC包括有Texas Instrument的PCM1794, PCM1792, DSD1794, DSD1792, 和這塊ESS9018。 這幾顆Texas Instrument (Burr Brown) 解碼IC基本上是來自同一款式的集成電路管芯 ... 反之, ESS9018就屬於一款毫無妥協的高端制品 ; 它的工作程式可以根據要求而去加以編輯, 從這方面來看, 在實際使用上有很大彈性。


4)
上次您曾經談及到, 在電路上不謹慎地使用 ESS9018時經常會出現過荷現象, 可否給我們深入講解一下 ?

首先, 正如上文所說, 任何解碼IC都只是一塊數據處理硬件, 並非全功能工作的音頻元件, 它們不能直接把數據轉換成我們所需的音頻信号。

ESS9018跟其他所有’電流輸出’類型的解碼IC一樣, 其輸出必要經過一個強而穩定的I/V轉換電踣。 當ESS9018以極高速度去進行脉冲式開關工作, 而一般的op amp I/V 轉換電路沒法有足夠良好之瞬變反應slew rate, 容易出現過荷並產生令人惡厭的互調失真。再者, 相比其他IC (例如PCM1794), ESS9018有更廣闊的頻响幅度, 對I/V轉換電路的工作穩定性容易造成侵擾。 由此, 我方的VIVACE採用了由分立元件定制之I/V電路, 來對應上述情況。

另一個發生過荷的原因, 是這樣 ...
ESS9018的各項工作參數必須得到謹慎協調, 在為這芯片進行電路設計時, 要選取合適的工作電流量, 避免芯片過荷輸出 ; 我須指出這一點, 適度協調的重要性, 通常很多電子工程人員都掉以輕心。
There is always room for improvement. 日日新 苟日新

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 09 Jun 2015, 01:09

5)
可否談談有關usb的設計意念 ; 若能更進一步的話, 您怎樣看外置式usb-S/PDIF轉換器 :

VIVACE的usb輸入包含了下列先進設計 :
asynchronous data transfer 非同步傳輸, 現今也差不多是必然做法了吧 !
此外還有 low noise clock management 即上文已經指出的, 謹慎去處理時基傳送 ;

另一樣同樣重要的, 也是VIVACE其中一項出類拔萃之處, 輸入以完全電氣隔離方式, 這令到電腦所產生的高頻雜噪不能輕易侵擾, 大大提高音質水平。


6)
您的外置供電UNISONO又是什麼原意呢 ?

首先, 任何音响器材都需要多組直流電壓去進行工作。
無論是外插式火牛, 或者是藏於機器內部的開關式電源(簡稱電子牛), 都只能提供一個既定直流電壓, 例如5V, 12V, 15V ...

當見到上面那些東東, 我們便應知道, 在機器內必定有其他DC-DC轉換器來進行變壓工作。 這些 DC-DC轉換器 (位於印刷電路板上不同位置又體積細小) 同樣也是開關式工作, 而且數目眾多 !! 由它們產生的雜噪干擾, 無法避免 ...
從VIVACE進而開發出VIVACE dpp, 同時加入UNISONO這個外置供電, 目的是為機器提供正負兩組獨立電壓去工作。此外, 電量供需穩定性獲得明顯更強, 對聲音有很大幫助。


另外補充一下 :
現今的確有很多體積細小的器材都使用外插式火牛, 原因有幾方面 : 杜絕機內火牛的電磁性和機械諧震干擾, 耗能環保, 與及可自動調節電壓方便產品出口。

我方產品雖然已經用上了一款高質素的外插式火牛, 但必須指出這方式仍然有其他缺點 : 就算經過多重隔濾, 始終還有小許殘餘雜噪 ; 另外這種普遍採用的直流接頭只可單一電壓適用 ; 及有限的電流量。

雖然市面上有各式各樣的plug & play後加電源配置, 可惜全部都受制於單一電壓傳輸, 實質改善有限。
There is always room for improvement. 日日新 苟日新

User avatar
skyblue
Site Admin
Posts: 187
Joined: 19 Mar 2013, 22:46

Re: Ivo Linnenberg from LINNENBERG-ELECKTORNIK, Germany

Post by skyblue » 12 Jun 2015, 15:35

7. Speaking of dpp, it stated that the VIVACE dpp has a revised I/V converter that results to several times lower distortion level. For the users, what improvement do we expect in sound ?

8. And one more question before we leaving the DAC topic, what measures do you use in order to reduce jitter at the coaxial S/FDIF input ?
There is always room for improvement. 日日新 苟日新

Post Reply