This one with 1K secondary termination

And apart from harmonic measurement, listening to no termination interstage always feel more bold and concrete. That is why I always recommend no pure R termination, if no obvious overshoot in 10kHz square wave, just leave it not terminated, otherwise use snubber network



最後修改時間: 2024-11-02 01:05:43

correction！！！

power amp input impedance of 100K should be reflected to primary load as 100K x 16 =1.6M ohm

with 1K secondary termination, primary reflexes as 16K //1600K=15.8K

1K termination come to 0.31 THD (at 2V RMS in secondary, x4 in primary)
no secondary termination come to 0.11 THD(at 2V RMS secondary, x4 in primary)

0.31÷0.11=2.8倍





最後修改時間: 2024-11-02 00:55:44

This is the one with 1K secondary termination

＞＞ having a low impedance resistor as load would cause a bit higher in distortion 


As low as 1K would not make just a bit more distortion, it would be a lot more distortion via the simulation of VTADIY


The terminated R will be reflected as load through transformer voltage ratio.
using 1K in secondary is reflected as 16K primary load through a 4:1 tranny. The line amp will finally connect to power amp assuming power amp input impedance as 100K, so 16K // 100K is 13.8K

Now define the output voltage at secondary as 2V RMS, but the primary need to swing 4x of 2V then it will be 8V RMS or 11.2V peak or 22.4V peak to peak, VTADIY take peak voltage as measurement

With simulation of VTADIY the one with 1K load secondary producing 2V RMS is THD 0.36 where 0.01 being the odd harmonic distortion

The one without 1K secondary termination is 0.13 THD and no show of odd harmonic

0.36 is 270% of 0.13 where also need to take account of odd one



最後修改時間: 2024-11-02 00:03:20

The schematic of the shunt regulator shown can be adjust down to 1Vdc with good voltage stability. The noise is very low which is good for use in low level signal circuit like the line amp.

I have tried the TL431 programmable shunt regulator IC but it generates too much noise.

Johnny

Johnny, the cathode shunt regulator schematic shown below seems to have got rid of zener, any improment?lower noise level and voltage stability ?

The very nice thing about shunt regulator is that it can adjust the bias voltage for lowest distortion.

Also, it doesn't have the phase distortion caused by the capacitor depending on the frequency especially at the low frequency end.

I have tried using the Shunt Regulator in line amp, and power amps. I really like it.

Johnny

＞＞是否用cathode CCS？

Barker 師兄，

For voltage amplification stage in cathode portion, cathode CCS is only applicable for differential VAS, not the single ended common cathode VAS. You don't need cathode CCS in single ended common cathode VAS.

In your 5687 preamp, for single ended common cathode VAS, it's cathode need to see low impedance cathode R for AC bypass, the shunt regulator keep 死 the cathode to ground voltage, any variation of AC then bypass, the cathode R for AC component can be as little as 1 ohm

While CCS in fact practically provide virtually large R. Exactly opposite, It is another way around

最後修改時間: 2024-11-01 19:24:14

Hi Steven,

I use the simple transistor "shunt regulator" to provide the cathode voltage for bias of the tube.

Johnny 兄，小弟用SLX-01/20，另外想知道您是否用cathode CCS？

I upgraded my parallel 6DJ8 - Monolith amorphous SLA-02 20mA line ouptut transformer configured 2:1, with 5842Q tube. B+ 120Vdc, idle current of 10mA.

I notice 5842Q (417A) tube is more linear than 6DJ8 and with even higher gain.

Real cool.

Johnny

最後修改時間: 2024-11-01 10:41:10

Hello Steven Baker,

Good to hear from you. I experimented number of interstage, line output and input transformers. I realized that having a low impedance resistor as load would cause a bit higher in distortion and less dynamic. The sound image opens up a lot without the resistor. However usually there's a high frequency response at around 20~30KHz which can be damped with use of an RC network.

What model of the Monolith Line output transformer you are using? The 20mA one?

Johnny

#14
Johnny 兄，請問您模擬後結果？1K用與唔用，性能分别如何？另外小弟只以用半支5687，工作設定與附圖一樣!

1K termination in transformer secondary may cause the loadline to be
significantly tilt, 5687 not beneficial to have constant current load

If there is no overshoot ringing in 10kHz square wave, I would say. to leave the secondary not terminated

In case obvious overshoot is detected, put in place the snubber network , 330p in series of R, it reduces the burden of secondary load (1K is heavy)and make loadline horizontal, Lundahl always recommend sunbber network

最後修改時間: 2024-10-31 23:12:09

Hi Mr. Steven Baker,

I tried to guess what the schematic could be for your new 5687 line amp. I worked out the following schematic. Most likely it is different from yours.

Johnny

＞＞次級並1K電阻

隻Monolith 牛次級如果唔並1k电阻係得唔得，抑或佢指定要电阻termination(有D牛overshoot )，唔並loadline係打横失真更低

如果唔一定要並电阻，5687係在A1工作佢只係需要望到足夠初次电感，幾K:幾K其實係冇意義，唔並次級电阻等於隻5687在constant current 工作而load 係極高阻抗綫性更高

最後修改時間: 2024-10-31 16:18:35

刚刚睇到呢個post，想問吓制作概念

1. 請問綠色訊號線係乜型號？旁邊对喇叭乜型號？

2. 用5687 接4:1前級牛只得3至4倍gain, 點解唔用你之前隻EC8010 可以出正常gain

3. 300B 或300B 变種胆，用兩級common cathode 6SN7其實已經夠哂gain做合併機，可以用passive volume 令訊號純度更高，請問加多部前級声音是否有其他变化而透明度及純度也有变化

最後修改時間: 2024-10-31 15:53:36

正....! 高手~

網友部EML-300Bxls後級，亦都同時升级，換上同廠Amorph芯輸出牛!

.