- #5874 [johnnykmtang], 26-08-11 18:28
- #5873 [niceday009], 26-08-10 23:15I see, why use PSE 845 when only set 760V plate voltage? In this case the PSE 845 only can provide 20W
To use 1100V and IT can get single 845 to produce 35W
Can't understand what is point of wasting 32.5W for another 845 filament power dissipation
最後修改時間: 2026-08-10 23:33:28 - #5872 [johnnykmtang], 26-08-10 10:24The guy run the PSE 845 with much lower B+ and idle current. He just use the LL1623 120mA which is commonly available.
A bit of waste using so many tubes with small output power.
Johnny -
- #5871 [niceday009], 26-08-10 02:35The Lundahl 1623 is too small for 50W SE output. For 845 PSE I would use 10kg PSE OPT

- #5870 [johnnykmtang], 26-08-09 23:34Another balance PP drive 845 PSE by Filenet from Germany .... So cool. However, I prefer taking out GNFB from the circuit so that it keeps the flavor of the 845 single-end spirit.
https://www.diyaudio.com/community/threads/845-pse-with-balanced-gnfb.443090/
- #5869 [johnnykmtang], 26-08-09 17:38Hi Jimmy,
I don't think the voltage doubler circuit for the B+ high voltage is specially made for "較聲". I believe the manufacturers used this because they want to reduce the cost and size of the power transformer. - #5868 [JimmyDIY], 26-08-09 01:28WilliamsonAudio兄,謝謝你!我記得Marantz 9也是用倍壓的,只是在單端機不多見,用這種倍壓方式而不直接用較高輸出電壓的券火牛是有有隱藏的較聲技巧在內?
最後修改時間: 2026-08-09 01:39:23 - #5867 [WilliamsonAudio], 26-08-08 20:06550V 交流,正常整流后,高壓不足夠給211約1千多伏。
所以他用了"倍壓"電路。把電壓提升一倍。
那支整流膽,並非針對"負偏壓",是信壓電路的一部分。
最後修改時間: 2026-08-08 20:09:05 - #5866 [JimmyDIY], 26-08-08 10:54WilliamsonAudio兄,謝謝你!530V下邊那份整流是提供負偏壓的嗎?它和佐久間駿提出用燈絲choke原理是否相同?
最後修改時間: 2026-08-08 10:59:10 - #5865 [WilliamsonAudio], 26-08-08 09:00也是DC, 只是。電源部份没有畫出来。
可以參考下圖的原理,也是Thomas Mayer設計
最後修改時間: 2026-08-08 09:03:40 - #5864 [JimmyDIY], 26-08-08 01:20我在外國討論區看到一個燈絲choke的例子是這樣的,只串一隻燈絲choke,沒有做直流,這樣會有交流hum聲嗎?

- #5863 [johnnykmtang], 26-08-07 04:511605 V4
Octal base with Filament CT.
- #5862 [johnnykmtang], 26-08-07 04:47Hi Jimmy,
If you really don't like having semiconductors in your new 1605 SE power amp, I believe using fullwave diode rectifier with CLC filter for the filament would be the best choice.
By the way, there is another version of the 1605 with Octal 8 pins. Also, it has the center tap of the filament. Then, you can connect the cathode circuit directly to the center tap instead having two resistors in series.
Johnny - #5861 [JimmyDIY], 26-08-07 02:42想問問大家,如果你們做一部直熱單端後級會用直流還是交流燈絲?如用直流會用原子粒穩壓還是普通CLC濾波?兩者聲音有什麼分別?從前我看過佐久間駿提出用燈絲choke,不知有什麼優點?因為我打算做一部1605後級用直流燈絲,但不想機內有任何原子粒.
- #5860 [bambino], 26-08-01 18:32yes, self-bias will automatically adjust its operating point to try to maintain similar plate current. so, just looking at the plate current, it changes very little over time, giving you an impression the tube is still in good health. when you see it has dropped significantly, the tube actually is in fairly bad condition.
when you say 300B ages rather quickly, you have to be more specific. for example, when plate & bias voltage are kept the same (at say 350V/-74V), how much plate current drops and how long it has been used before this happens. to be honest, i am not too worried about it as EH, GL & TJ are relatively cheap 300Bs. replacement cost is not that high.
最後修改時間: 2026-08-01 18:44:20 - #5859 [johnnykmtang], 26-08-01 18:18300B's are great but I found they got aged rather quickly. I have few pairs of EH, Russian Gold Lion, and TJ Full music mesh one are getting rather weak.
I used them quite conservatively with approx 350V plate-cathode and 60mA idle current.
Using self-bias may not be that obvious.
Johnny - #5858 [bambino], 26-07-24 10:11with the gain of 6J5 & 6V6, it looks like Komuro circuit can be used as integrated amp. although it's a 3-stage design, that's all you need. on the contrary, for a 2-stage power amp, you will normally need a preamp to get enough gain. it's also 3 stages on total. the upside is there's separate PSU in preamp for better noise isolation. on the flip side, signal path is longer, requiring an additional interconnect and also a power cable.
最後修改時間: 2026-07-24 10:11:49 - #5857 [johnnykmtang], 26-07-23 22:56Here's the simulation of the Komuro driver circuit

- #5856 [johnnykmtang], 26-07-23 22:35Using triode connected 6V6 with choke load is very good indeed. The simulation shows it can provide +/-70V drive with THD of 0.37%. The 6V6 input signal is approx +/-7V.
The distortion from the 6J5 is rather minimal at 0.1% with +/-7V output.
3 stages circuit design is a little more complex but it is much easier to achieve good result.
Johnny
- #5855 [bambino], 26-07-23 12:42don't forget this Komuro 300B amplifier. 3-stage DC design. it's relatively simple and easy to build. if you have some tube knowledge, you can modify it to use the input/driver tubes of your choice.

最後修改時間: 2026-07-23 12:42:12 - 下一頁 (1 of 294)
- 返回 ...

Running 760V B+ makes sense too. The 845 stage would produce super clean low distortion output.
Johnny
最後修改時間: 2026-08-11 18:30:22