Who can/will help me with highvoltage SMPS design for tube preamp

The only real silent switcher is one followed by a linear regulator. Not absolute truth but nearly so.

What is exactly gained by using SMPS? It can not be low cost as a simple low power R-core transformer for a preamp is not that expensive.
True! I've seen ultra low noise switchers (10uV noise/10v output 10% variation over 10 days) used at LHC but they were low power (10...20 watts) anyway.Once you ask for more power and higher output voltages you ask automatically for higher noise.
 
A rare occurence where classic technology easily outperforms the modern replacement (in consumer grade electronics) except for some today important features: power efficiency + certifications and lowest cost in all aspects.

I would not loose sleep over the lifetime of the device and the difference in Watt(hours) of energy loss over that time when I get higher reliability and lower ripple voltage/EMI in return.
 
Last edited:
What is exactly gained by using SMPS? It can not be low cost as a simple low power R-core transformer for a preamp is not that expensive.
The absence of 50/100hz resp 60/100hz products. No 50/60hz transfer true housing to microphonic tubes.
My linear solutions with regulator were always more expensive as smps solution.

https://www.diyaudio.com/community/threads/phonodude-koifarm-edition.293719/

https://www.diyaudio.com/community/threads/mini-me-tube-phono-preamp-pcb.362704/

https://www.diyaudio.com/community/threads/hybrid-phono-preamp.362868/

https://www.diyaudio.com/community/...th-printed-circuit-board-design.206415/page-4

SMPS is smaller when using just one. See also my picture of a 2 x 400w class d with tube input buffer.

https://www.diyaudio.com/community/threads/universal-pcb-for-purifi-1et400a-amps.379211/page-3

But for now i like to build a SMPS.

IMG_20220619_230351.jpg
 
Last edited:
The absence of 50/100Hz resp 60/120Hz products. No 50/60Hz transfer trough the housing to microphonic tubes.
My linear solutions with regulators were always more expensive as SMPS solutions.

But for now I like to build a SMPS.
First the choice for an expensive buffer because of the choice for tubes. Then executing it cheapest way. That makes the tubes more a gimmick isn't it?

The choice for DIYing an SMPS will be expensive again so against that philosophy. A lot of work and no certification at all. If SMPS is the choice then ready made is the way to go as all certification/safety/EMI work has been done by the manufacturer. You will need equipment for measuring if the device is not an EMI transmitter for instance. A noble idea on itself but hardly practical/executable in various aspects but hats off if you persevere.
 
Last edited:
Actually I have been working on this for a while. Although a slightly different approach, more like a DC-DC boost converter, switching around 300-400kHz.
Input voltage around 15V as well as a heater PSU on board (12V and 6.3V). So a standard wall-plug adapter can be used.

The aim is to deliver around 100mA max.

Unfortunately because of chip shortages this has been on hold for a bit, but definitely planning to continue.
 
It seems that you don't need any advice...you know it all!
He needs advice on an SMPS circuit to build, not people trying to say it's a waste of time or why not use linear etc. 😛

I want to design something similar as I said. I'm thinking why re-invent the wheel? Use SG3525 and some 600V MOSFETS to switch. Just need an appropriate coil and that's where my design skills fall apart. I could design the transformer for 60Hz, but no idea about pulse transformers LOL
 
  • Like
Reactions: ruffrecords
My audient id 22 has asilent switcher with slew rate limiter lt 3439, a rather old ic:
https://khronscave.blogspot.com/2021/02/67-audient-id22-part-2-repair.html?m=1
You can check out Analog devices suggestions:
https://www.analog.com/en/product-category/ultralow-noise-regulators.html
There's one AN from Jim Wallace regarding its versions of low noise Royers there. You better follow the original baxandall placement of the damping coil though! Jim's version is a little bit noisier.
In 2018 I even prototyped some switchers for a hybrid phono preamp here, but never built them:
https://www.diyaudio.com/community/threads/1v2-phono-riaa-preamplifier.315824/page-2#post-5597656
 
  • Like
Reactions: Koifarm
It is in 2 of the languages I speak and it means exactly the same. Small mistakes happen when you're on an international forum. Minor issue, most will understand perfectly as it says "Regulated output with rimple <1mV" with regards to a power supply. It starts to happen when you start learning a third or a fourth language as usual here.

The dutch word is "rimpel" or, more correct, "rimpelspanning". As you can see we say "el" as it is spoken like this too. So "rimple" is a mix of English and Dutch. We would write "ripple" like "rippel".
 
Last edited:
  • Like
Reactions: dreamth
From a bit of a google. It’s more similar to wrinkle. Similar, but not the same.

anyway it wasn’t intended as criticism, just a possible reason for less results. That being said, low noise switching supplies aren’t commonly published. Those like that used in hi-end designs for audio and medical/technical/aerospace more recently are well guarded I imagine.

interested to see where this goes.
 
Last edited: