I have asked in other places about the noise of MOSFETs compared to BJTs and FETs but never got a definite answer. Are there any measurements of MOSFET noise in the audio region and how do they compare?
I have some stuffed Paradise boards without PSU. AS Audio build and calibrated them.
I build a very simple PSU, double mono, MosFet Z-diode multiplier and RC filtering.
This stage is super quiet and sounds just fine. I may publish pictures and circuit diagram.
I build a very simple PSU, double mono, MosFet Z-diode multiplier and RC filtering.
This stage is super quiet and sounds just fine. I may publish pictures and circuit diagram.
This stage is super quiet and sounds just fine. I may publish pictures and circuit diagram.
Hi Joachim,
Please do.
Also, have you tried using other transistors in the input of the Paradise? Like the low Rbb ROHM and such?
Cheers,
Mogens
Hi Mogens ! Yes, i tried other transistors. The good old japanese ( ROhm, Toshiba, Hitachi ) are more or less unobtanium. There are some interesting Zetex. I think we discussed this.
On the other hand there are 8 transistors parallel in the Paradise so noise is really close to the physical limit. Alo running each pair at ca. 2mA makes sure that the 1/F noise is not shifted up. More current brings even less voltage noise but there is a point where the 1/f noise shifts upwards. The optimum is ca. 10mA and that is aproximate the total current in the input stage. BC327-40 and BC337-40 also have high HFE. That reduces the input offset. On the other hand the Paradise has an input offset trim. I think it is good as it is.
On the other hand there are 8 transistors parallel in the Paradise so noise is really close to the physical limit. Alo running each pair at ca. 2mA makes sure that the 1/F noise is not shifted up. More current brings even less voltage noise but there is a point where the 1/f noise shifts upwards. The optimum is ca. 10mA and that is aproximate the total current in the input stage. BC327-40 and BC337-40 also have high HFE. That reduces the input offset. On the other hand the Paradise has an input offset trim. I think it is good as it is.
The good old japanese ( ROhm, Toshiba, Hitachi ) are more or less unobtanium.
Hi Joachim,
Yes, I know. I have only very few of the Toshiba Type. The Rohm Type I have a little more of. However, unfortunately most of the Rohm type I have are from different hfe group.
Thank you for the schematic for the alternative power supply. It actually is great as I have been thinking about building a Paradise RIAA as a present for a friend of mine. Could make the build a little easier.
Cheers,
Mogens
Many people are asking me if i still have Paradise parts. Going though my vaults i found more then i thought.
Here come some pictures and i explain later.
Here come some pictures and i explain later.
A pair of Calvin buffers, ready to play. 4 Unstuffed Paradise boards R31b. 4 Paradise R3 without shunt.
Pre regulator stuffed, transformers and resistors missing. Stuffed Paradise R3 stuffed, semiconductors missing. FdW shunt regs, unstuffed boards.
Attachments
Jaochim, can I take a blank set of Paradise main boards and the Calvin's off you, I'm not bothered which rev the main boards are.
Simon
Simon
Frans and i are quietly working on a new phonostage. The goal is to make something really simple but high performance. The reason is that there are some ultra low noise transistors availlable that do not require paralleling. Stay tuned.