Papa! I want to have Zen V5.

Status
Not open for further replies.
zenlite zen lite zen-lite zenv5 zen v5 zen-v5

Babowana said:

After all wire arrangement, tested the sound,
got wonderful sound,
and wanted to finish this project,
completing the housing of the PSU body.


congrats babo!
🙂


in my lineup mind
😎 Zen V5 RULES 😎
... when it comes to ultimate designs I have encountered in this forum
... too bad I did not come up with ZenV5 circuit before Nelson Pass
... it would have been something to accomplish in life


Same thing with Zen Lite 🙂
... why did noone else come up with that trick!



lineup
- thinks some nelson's stuff is not over the edge interesting
- but some of it Really Hit's The :bullseye: bullseye
 
Zen Mod said:
. . . in that case wires can be shorter . . .


Actually, there is a detachable lifeline formed by five wires
between the heat sink and the wooden box - this was my
real concern
.

Five wires are:

1 input wire to the gain mosfet gate
2 output wire from the gain mosfet drain
3 preregulated wire to the regulator mosfet drain
4 reference voltage wire to the regulator mosfet gate
5 regulated voltage wire from the regulator source

And, these five are tightly braided together forming the lifeline
(about 20 cm length).

So, I was worrying about unwanted high frequency oscillations
which could be possible due to the following causes:

1 capacitive and magnetic coupling from the output to the input
2 load lead inductance feedback from the power line to the circuit

My worrying was just a worrying, with luck of no such oscillation 🙂


By the way, the five wires should have been six including one
more wire, which should have been a protective earth line from
the heat sink to the protective earth point. But, my ill plan forgot
it. Therefore, I am using a separated wire having snakehead clip
at each end. One head is biting the heat sink while the other
biting the protective earth point.

Any comment or advice? Tks
 

Attachments

  • lifeline.jpg
    lifeline.jpg
    29.7 KB · Views: 392
Babowana said:



Actually, there is a detachable lifeline formed by five wires
between the heat sink and the wooden box - this was my
real concern
.

Five wires are:

1 input wire to the gain mosfet gate
2 output wire from the gain mosfet drain
3 preregulated wire to the regulator mosfet drain
4 reference voltage wire to the regulator mosfet gate
5 regulated voltage wire from the regulator source

And, these five are tightly braided together forming the lifeline
(about 20 cm length).

So, I was worrying about unwanted high frequency oscillations
which could be possible due to the following causes:

1 capacitive and magnetic coupling from the output to the input
2 load lead inductance feedback from the power line to the circuit

My worrying was just a worrying, with luck of no such oscillation 🙂


By the way, the five wires should have been six including one
more wire, which should have been a protective earth line from
the heat sink to the protective earth point. But, my ill plan forgot
it. Therefore, I am using a separated wire having snakehead clip
at each end. One head is biting the heat sink while the other
biting the protective earth point.

Any comment or advice? Tks


you know that I'm always full of-with advices


just make whole mess as I draw , make holes ditto between heatsinks and box,put wires already soldered to mosfets on sinks through holes and solder inside.
htsnks can be grounded on screw heads inside of box.

biggest work must be dislocating of in-out connectors from present back to present top plate


anyway- your decision is final :devilr: :Popworm:
 
Confirmation of input impedance


I have assumed that R1 is considered as the input impedance
if I assume the node just behind R1 as a virtual ground.
I have R1 of 500 (1K//1K).

I measured the input impedance of about 500.

I think my assumption was reasonable. Right?
 

Attachments

  • input impedance.jpg
    input impedance.jpg
    24.7 KB · Views: 316
seems all right

usuall way of measuring this is to use pot instead 1k ,and when you have same voltages across pot and input of stage , you can measure exact res of pot and then you know input resistance.

anyway,both ways are equally good

buda is in everything 😉
 
Surprise, surprise


Any one built or building Zen V5?
Please spend some $ to bypass the big electrolytic caps, C8 and C9 with 1 - 10 uF film unless you are using high quality of C8 and C9. It will very much improve the frequency response around top-end.

I am using Solen 4.7 uF, and get natural details of cymbals' delicate top-end sound.

If you don't agree with the result, I have intention to compensate your expense invested for the film caps 🙂

Sorry, Nelson, for copying your drawing without getting any permission
 

Attachments

  • bypass cap.jpg
    bypass cap.jpg
    48.3 KB · Views: 574
Walking-distance away from my place, we have two cinemas: Cinema A and B.
I often go to Cinema A whenever they play English-speaking movies, and enjoy both the big screen and the movie sound. This Cinema A seems to be using horn speakers. I hear the typical coloration of the horn speakers and somewhat hardness.

Last night, our apartment was under the blackout – no electricity. It was dark and cold so that I walked down the stair ways and went out. And, I found myself standing at the ticket widow of Cinema B playing The Guardian. But, the price was 14% more expensive compared with Cinema A for the same movie. I wondered reason why. Probably, the screen bigger, or the seat more comfortable, or the movie sound system better. Wow, actually the sound was much better. Keeping energy, the sound was smoother and warmer. The helicopter propellers were roaring realistic.

For me, the improved sound was worth the 14% higher price (ticket of about USD 10.5).

Audio-diyers’ ears are sharp! 😎
 
Babowana said:
Walking-distance away from my place, we have two cinemas: Cinema A and B.
I often go to Cinema A whenever they play English-speaking movies, and enjoy both the big screen and the movie sound. This Cinema A seems to be using horn speakers. I hear the typical coloration of the horn speakers and somewhat hardness.

Last night, our apartment was under the blackout – no electricity. It was dark and cold so that I walked down the stair ways and went out. And, I found myself standing at the ticket widow of Cinema B playing The Guardian. But, the price was 14% more expensive compared with Cinema A for the same movie. I wondered reason why. Probably, the screen bigger, or the seat more comfortable, or the movie sound system better. Wow, actually the sound was much better. Keeping energy, the sound was smoother and warmer. The helicopter propellers were roaring realistic.

For me, the improved sound was worth the 14% higher price (ticket of about USD 10.5).

Audio-diyers’ ears are sharp! 😎


:headbash:

those must be bad horns in cinema A

ya almost tell
horns=bad

:whazzat:




:yinyang:
 
Bad, bad, bad . . .
Something seems to be going stange with me.
What is wrong with me?

Last time I measured the damping factor of 5.
But, when I measured it again, I got 8.
Why measured again?
Because last time I measured it at too high frequency.
So, I wanted re-measurement at 50Hz.

On 50Hz signal
DF = 8 (8 ohm load)
Output impedance = 1 ohm

The original article says, DF of 4 . . .

Where I am wrong?
These days I am having too little sleep . . .
 

Attachments

  • damping factor - revised.jpg
    damping factor - revised.jpg
    39.7 KB · Views: 347
Math FFT frequency component (spectrum) measurement


Okay, I took this measurement.
Not to get the distortion figures, but just to have a look.
Why . . . ? Of course, for fun.

The fundamental frequency was 1 kHz of 20V(peaktopeak)
at the output to the 8 ohm load (about 6W). From this test,
my ZV5 shows both 2nd and 3rd harmonics. I guess that
if the power is varying, the portions of the 2nd and 3rd
harmonics would be also various.

If you find out any mistake in my measurement set-up,
I would much appreciate your comment.

🙂
 

Attachments

  • math fft measurement.jpg
    math fft measurement.jpg
    51.6 KB · Views: 308
Status
Not open for further replies.