Festival of Lanterns元宵節

We had earthquake this morning Richter-scale 6.4 in Kaohsiung, near Tainan.
You can do nothing on such natural disaster except to seal the mouths of these furious  tigers.
The question is : Who will do it?
I either hooked the tiger or stitched her mouth.
Hope it will help.
Any suggested option?

The lantern show, as part of festival of lanterns; started yesterday evening at Taipei city.

We had the chance to  tour in the crowded Ren-I road, along the section no. 4 which stretchs for more than 1000 meters  long.

Here are some pictures I took on the first day. The show will be ended within 2 weeks.

The strung islands dividing the road into 4 tracks are fully decorated with lights.

The lights are not neon lights but the high tech LED threaded ones.

 Canada,  maple

 Greece, olive

 Holland, tulip, windmill

 Japan, cherry, castle, Fuji   mountain

 South Korea, rose of Sharon

 USA, rose

 Malaysia

 Palau 帛琉 , but not Bangladash as corrected by my grandson, at age  of 8

 

Italy

 

 What means by POWER IN ME as it is shown at the end of  road leading to Taipei city halls ?

POWER is well known to me in electricity, but to whom the ME is referring is difficult  to my understanding.

Perhaps this is a kind of puzzles waiting for cracking, but  in a new fashion.

In traditional festival of lanterns, another activity is the so called 燈謎 lantern puzzles, literally translated, or called射虎hunting tigers. The puzzles are usually written on a piece of paper and hung below the candle-lighted lanterns.  One who cracked the puzzle will receive a small prize as fun.

The lantern show is holding in the open ground which occupies a complete block between two main roads,  with a huge building Dr. Sun Yet-Sen  Memorial Hall in the field.

Standing at both sides of the entrance are this 4-meters high lantern of India shown above. 

In front of  the hall stands another not so elegant one offered by city Mayor.

 Macau are taking the advantage to promote their sight-seeing or gambling. It is a model of two stories high.

 Two drumming tigers are working hard in collecting money and in-feeding treasures as the composite character posted in between shown.

 This flying tiger looks furious when the mouth is open and teeth exposed.

According to tradition, the tiger of  lantern shall keep mouth closed otherwise some kind of disaster is inevitable. Although the expert warned in advance some days ago after finding this mistake done by modern artist,  it is never taken into account, but considered as superstition.

 I like this smiling tiger with orchid flowers.

All such lanterns are constructed on a plate form with the figures higher than one meter up to several meters.

All of them are sponsored by some private enterprises or  schools.

My grandson is  Spongebob fan海綿寶寶迷.

I have no idea what these figures actually are playing, but it looks that bob is in anger.

A woman vendor was selling the balloons with different  figures.

My grandson had two balloons, in figures of bob and star-fish with big belly.

We enjoyed this festival .

5687併聯多管機

 

Multi 5687 SE amplifier           by Stein  Feb. 18, 2010

0.緣起：

手上有Colin的5687單端機，聲音很好，兩年前diy過」五球十燈機」，併聯4屏5687及7119表現也不錯，只是給我推出去斬了，只留下10幾支管子。

庫存還有一對拆下來One Electron 美國的output transformer高音也很好。

虎年春節連續休假9天，待在台北，天天下雨，冷到10度。吃飽閒閒，動點腦筋，設計一台多管5687單端機吧。

Colin 5687單端機單屏1 watt出力，我diy湊8屏，出力也許會有8w，那才叫做 唬唬生風，5687＜偶又發機＞了。

1. 高頻跟米樂電容Miller capacitance:

 

我要是有電子魔眼，站在輸入端子，一頭看進5678柵極，就會有一個所謂Miller電容，這是柵極、屏極之間的電容Cg-p，因為放大A倍數的作用，膨脹為A*C-g-p，再加上本身的Cg-p，等於Miller capacitance (A+1)*Cg-p, 還看到Cg-k(C-in), Cg-p(c-out),  配線潛佈電容C-stray。這些電容可以看做併聯接地。

另一頭看輸入訊源，有音量Volume 可調電阻Pr(一般100k)，還有前級輸出的內阻（個位數的電阻可以忽視），當作R-out。

如上圖，形成一個低頻過濾器Low pass filter。

隨著頻率F提高，訊號流失下地的越多，當R-out等同於電容抗時，這個頻率叫 F-cut-off -3db，算式

F-3db=1/2-pie*R*C

5687單屏的電容依上圖,方便起見，算做76+14＝90 pF

R-out最壞、最大的情況是當Pr轉到中點，兩頭50k併聯，得到 25k。

如上圖紅色算計得到 F-3db=70kHz。

要問70kHz夠不夠好？

按照Mr. Morgan Jones計算F-3db= 131kHz時，20kHz衰減 -0.1db

日本 黑川達夫 桑diy 26 台amplifier，每台有測試報告。我查了單端機，沒有feedback的性能，F-3db一般在70kHz上下，少數到達100kHz以上，但這是整體含輸出變壓器的。

20k Hz的一個8度音程是40kHz, 兩個8度是80kHz, 3個8度是160kHz

所以說 70kHz只能說是勉強及格而已，最好有150kHz以上。

這也就是Colin一向主張Pr要小，用50KB 併聯30k，約相當於 R-out=50/4並30k=9k的道理。這樣 F-3db=196kHz。

如果8屏併聯，要達到 F-3db=150kHz,  R-out需要低於 1/6.28*150k*8*90pF=1.5kr

 

2.    輸出功率：

Amplificaton factor不大受plate milliamp的影響, 受Eb工作屏壓影響稍大。最受屏流影響的是transconductance 互導 及 plate resistance屏阻，特別是屏流低於15ma時候，屏阻急速上升，要避開。

 

按照T.S.的3個 class A1工作例，如上表列屏壓120, 180, 250 V。

最大的250 V*12 ma = 3000 m W還沒有用到上限,

到頂是250 V*15 ma = 3750 mW.

輸出變壓器接1.6k/8r反映到每屏負載是 1.6k*8=12.8k, 在上圖裡劃上綠色線（平行320v/ 25m a=12.8k。也可以把plate milliampers乘8，劃上1.6kr負載線）。

其實功率計算我看只能當參考而已。

因為喇叭的負載是電抗，變壓器也是。頻率不同，負載值也未必是1.6k。

大負載一般是低頻，功率因素低的大電流，B+供電內阻大、接續不上的話，也發不出大功率來。

不過以純電阻計算也有參考之用。拿最簡化的方式：

Ec最大擺幅0-24 V下，Ep/Ip分別是 70V/30ma—390V/4ma

P=(390-70)*(30-4)/8=1040 mw 每屏 ，8屏有8W。

另一種算法：(12/1.41) Vrms *14(V.G of 5687)*0.07(V.G. of output trans)= 8.34 Vrms@ OPT

8屏有8.34*8.;34/8=8.7 W

功率大，需要屏壓高、偏壓深。

如果以T.S.第二例 180 V, -7 V, 23 m a計算

（7/1.41）*14*0.07=4.87 Vrms,   4.87*4.87/8= 3W  8屏只有3W

 

5687換用7119也可以，放大倍數高24/18=1.3倍, 整體提高1.33＊1.33＝1.78倍＝+5 db、會更嗆。也許要調一下Pr2.，降低偏壓值，補升屏流。

 

4. 驅動級：

依前計算，驅動級的輸出內阻需要小於1.5kr.

最方便是一律用5687，工作在90V, Ec= -2V .Ip= 20 m a, 放大13倍，輸出阻抗 1.5k.

 

 

輸入靈敏度全8W輸出時需要0.68 Vrms（12.5V/1.41/13），因此Ec= -2 V＝ -1.4 Vrms已經有足夠超載的空間了。整體放大13倍＝22 db，勉強夠用。

直接交連的話，務需採用雙電源，否則發熱量太大。

用LED當陰極RC，之前做過，效果不錯，接線方便。

200V B+採用Zenner穩壓，驅動級工作點固定好，則功率級的偏壓可以調整Pr2到需要的電流。

全機9管5687燈絲串接110 VAC/9 = 12.2 V還在12.6 V誤差之內。最好是整流後，調到12.6*9=113 VDC。 這樣省下燈絲加熱變壓器。

LED改用正統的RC，可以測每屏的電流。R=10r, 1ma=10mv,  要是R再大，會增大內阻，有助於高互導管5687的安定。好壞難說。用RC則配置太繁雜了。

 

如果驅動級用417A取代5687，整體放大倍數29 db最恰當。

之前用 417A直交2A3總覺得倍數稍嫌不夠。

8*5687+ 2* 417A燈絲串連照樣可以，只要417A並個82R, 吃掉多的0.15A。

 

5.實作：

實作的話，雙電源較費事。需要隔離變壓器110/220V兩隻。一隻B+電流300ma不小。還要4個大電流的choke, 一個小電流choke。整體機重近乎 20公斤

這樣的架構，高電壓放大倍數當功率管併聯，比傳統的低放大倍數單功率管，不知道聲音如何？值得撩下去嗎？

6. 8*5842:

規劃時，起頭是2*6V6驅動直交8*5842-417A的。

像圖一樣。想做到2W＋2W。






因為我有現成的一台"發久久"單端機用8支JJ的ECC99 上頭還有兩個8腳插座給6V6。
原來是兩支整流管，因為電源變壓器5V燒了，我改成diode 整流，改名」發不久」機。
容易改機，不必打孔做無聊事。像照片一樣。
就怕5842震盪、起童。
網路上有人說，震盪會爆裂玻璃管，是真的嗎？
我想頂多是，不能工作而已吧

MMMC唱放終結機

從畫圖到試機花費兩個月，算是慢工出細活。



燈絲加熱一定是整直流。買現成的voltaage regulator 12VDC可以精確調到12.6V。
LED指示燈也取自這裡。
B+來自Hammond power transformer 先橋diodes整流，經過C-R-C-L濾波，由6BM8三極五極管 做 誤差放大、回輸串流電壓調整，擁有三大功能：

1. 穩壓，實測交流市電變化8％時，B+變化在1％。
2. 降低B+供電的內阻。
3. B+有緩慢啟動的效果，約有60秒以上的緩衝時間。

12A7跟417A第一二放大管的B+供電233V再經過 C-R-C濾波。
6BM8燈絲電壓掛在240V。
擔心的是Hammond Choke額定 400VDC已經超載，開機時還有超過500V的短暫高壓。壽命如何不知道。








架構沿用 沒完沒了機，變壓器輸出，不用 新生機的電容輸出。這是聽從 Dewen金耳朵的意見，不用電容輸出。木耳可以確定的是用輸出變壓器的哼聲比電容的小。

RCA輸入端子只有一對。MM-MC由指撥開關切換，變換負載值為 470R──51k，
靈敏度變換為 0.3 mV跟 1.2 mV。MMMC位置有LED 紅、綠燈指示。

12AT7做SRPP放大，輸出取自陰極，降低輸出阻抗到 1.6k.

RIAA補償線路、數值取自網路 KAB passive RIAA network computer
R1由上次、沒完沒了機、的33k降低為22k，可以配合現成的C1, C2數值，不必修正補償小電容容。
22K應該還有下降的空間。
417A放大管取代上次的12AU7，放大倍數高一倍，內阻1.5k，接線空間加大，配線容易很多。






輸出插孔，輸入RCA插座放在前頭。
兩側12AT7, 中間417A.




電源開關紅燈在側面。外接地線端子也在側面。三面平整。
大電容後頭是6BM8.電源變壓器遠離在後頭。





唱放新生機MMMC電容輸出還留著。




主要另件。除了voltage regulator, Hammond Interstage transformer, chokes，小濾波電容 拆自舊機之外，是新品。
輸出變壓器準備兩套， Hammond interstage 90k/10k N=3
跟Colon的interstage 20k/5k, N=2
準備做AB比較。








機殼是Hammond鐵盒子，150＊335＊50 mm.  這是老早就準備好的，之前用元山牌機殼做了三個機種，是因為容易加工打孔，遷就的結果是嚴重的麥克風效應。






這made in China的陶瓷管座不錯，很緊，不會有made in USA的New Old Stock有 牙週病。







Colin的 輸出變壓器排除出局，不是一山不容二虎，實在是空間太小，容不下。空出來的地盤，給了6BM8.




輸入切換開關銲上需要的電阻，縮短訊源路徑。
咖啡色線接LED指示燈











放大倍數太夠了。接71A-Philips RH 532只要轉一個鐘點就夠了。接71A-耳機只要轉兩個鐘頭也夠了。
12AT7試換過12Au7,放大倍數雖然還可以，但聲音比較乾瘦，放棄了。

45 degree LP stereo recording-B

 

Feb. 07 2010 under construction.

 

看看LP立體MM唱頭拾音的原理

看看LP立體MC唱頭拾音的原理






LP立體MC唱頭拾音的另一例：




特例：Denon MC唱頭 DL-103




看看源頭的LP立體刻片刀

http://www.sickoftalk.com/whyvinyl_two.html

Recording Discs
(continued from page one)
Stereo Cutting Head

   
    The stereo cutting head consists of a stylus mechanically connected to two drive coils and two feedback coils, which are mounted in a permanent magnetic field, and stylus heating coil wrapped around the tip of the stylus (Fig. 12-10). When the signal is applied to the drive coils, the alternating current flowing through them creates a changing magnetic field, which alternately attracts and repels the permanent magnet.

Since the position of the permanent magnet is fixed, the coil moves in proportion to the strength of the field created and the attached stylus moves with it. The drive coils are so wound and mounted so that energizing either one alone causes the stylus to move in a plane 45 degrees to the left or right of vertical, depending on which coil is energized. Feeding both coils an in-phase signal causes the stylus to move in the lateral plane, while feeding the coils out-of-phase signals causes stylus motion in the vertical plane.
    The feedback coils are attached to the shank of the stylus and are, therefore, moved whenever the stylus moves. The motion of these coils in the permanent magnetic field creates a current flow in them, which is an accurate representation of the stylus motion. By mixing the outputs of these coils out of phase with the input signal in a technique called 「negative feedback」, several advantages are achieved.

45 degree LP stereo recording-A

 

Myav討論串有說：

quote:

---

原發表者是 hckao2


何大英明, 是 LATERAL = L+R, VERTICAL = L-R
所以低頻, 人聲... 中央聲道者之震幅於 lateral 方向均倍增之(L+R).
反之於 vertical 方向均相抵之(L-R).

以下恕刪


.

---

Stein 質疑說：

那麼以下這位日本Yosh桑說的要如何解釋？：
Specific to stereophonic groove of 45degree modulations, the amount of lateral displacement is always same as that of vertical displacement irrespective of amplitude and frequency differences between channels.

http://www7a.biglobe.ne.jp/~yosh/recspecs.htm

Sample of stereophonic groove: please note that top width of stereophonic groove is changing momentarily and that the minimum top width is depending on the displacement (amplitude) and the nominal top width of plain groove. Specific to stereophonic groove of 45degree modulations, the amount of lateral displacement is always same as that of vertical displacement irrespective of amplitude and frequency differences between channels. Hence cartridge to trace such groove should have same compliance laterally as well as vertically - though lateral compliance is usually higher than vertical compliance in most stereophonic cartridges.

這下不小心踩到hckao2的痛腳，換來回應說：

stein 兄,

報歉, 我雖然很想回答這個問題.
不過也不希望論壇倫為吵架場所.
所以要問問題請注意態度.
我沒義務回答不禮貌者的問題.
不是第一次態度不有友善者.... 道不同.

 

接著hckao2又說：

Very easy,
Yosh san made a "mistake".

lateral = L+R, vertical = L-R 當然不同.

if L+R = L-R,
then R = 0.
>>> not a stereo modulation

 

我也頂回去了：

我就直接、友善的說：
too easy to make a mistake for hckao2!
Yosh san did not make a "mistake".
L, V是波動變化的，像是交流電一樣，不能簡化成直流電一樣，直接相加、相減。


請再仔細看一下原圖裡

黃lateral displacement
藍vertical displacement
棕right wall 45 degrees
綠left wall 45 degrees

棕、綠，相加 得黃
綠、棕，相減 得藍
雖然畫得不很精確，但很清楚，用意到了。

就是 lateral 跟vertical displacement 變化雖然不同步，但幅度一樣。

 

何大也質疑說：

quote:

---

原發表者是 hojuiyang111


the amount of lateral displacement is always same as that of vertical displacement irrespective of amplitude and frequency differences between channels

這句話對嗎?

如果 lateral = L+R, vertical = L-R .

如高兄所推演, L+R = L-R, then R = 0.

還是我們有何誤解嗎?

---

hckao回答：

這個日本san說 :
"管它的啥頻率, 振幅! 水平與垂直的移動量都要相同"
他如果沒錯, 那要表達什麼呢?
當然是"錯"的!!

不過他要說些什麼不重要.
lateral = L+R, vertical = L-R 是定義沒啥好辨的.
如果這個都不知道, 不用功察資料.
那雞同鴨講就不必再深入了!!
您說是嗎?

Stein只好查資料了。

http://www.sickoftalk.com/whyvinyl_two.html

 

Recording Discs
(continued from page one)
Stereo Cutting Head

   
    The stereo cutting head consists of a stylus mechanically connected to two drive coils and two feedback coils, which are mounted in a permanent magnetic field, and stylus heating coil wrapped around the tip of the stylus (Fig. 12-10). When the signal is applied to the drive coils, the alternating current flowing through them creates a changing magnetic field, which alternately attracts and repels the permanent magnet.

Since the position of the permanent magnet is fixed, the coil moves in proportion to the strength of the field created and the attached stylus moves with it. The drive coils are so wound and mounted so that energizing either one alone causes the stylus to move in a plane 45 degrees to the left or right of vertical, depending on which coil is energized. Feeding both coils an in-phase signal causes the stylus to move in the lateral plane, while feeding the coils out-of-phase signals causes stylus motion in the vertical plane.
    The feedback coils are attached to the shank of the stylus and are, therefore, moved whenever the stylus moves. The motion of these coils in the permanent magnetic field creates a current flow in them, which is an accurate representation of the stylus motion. By mixing the outputs of these coils out of phase with the input signal in a technique called 「negative feedback」, several advantages are achieved.