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

Jump to Latest
#102
Sorry, I'm not used to HV caps and 10000 uF ones since I don't use them. 35mm caps barely fit 1U, which is my main enclosure size. I need to get used to a bigger ones that are suitable for HV, obviously.

Back to a topic though. Won't the 100uF upset the SMPS? Will it be able to start in this capacitance at all?
 
#103
You must have ultra low power equipment 😀 Even DACs and media players sometimes use a few 4700 or 10000 µF 16 ...35V when a good old linear PSU is used.

https://www.audiophonics.fr/en/hifi...-low-noise-2x-5v-1a-15v-1a-black-p-14032.html

The caps filter the rectified voltage and they are before the regulating part. I think you refer to caps connecting to the output of a regular SMPS.

BTW I regularly work with 2200 µF 450V, please check the size of these. I call them "cola bottles".
 
Last edited:
#105
My first DIY transformer. I hope it wont blow up.

IMG_20220809_150407.jpg
 
#110
I like the use of the SMPS for audio stuff. But i dislike the quality of the output of the chinees SMPS stuff.

Because every new build has other requirements it becomes more difficult and more cost(custom made transformers) to get the right transformers.

So i like to learn to build my own SMPS.

I do not want connect several devices to each other.
 
#111
b_force said:
Again, but for a preamp you also don't need an entire SMPS design.
Just a 12-15Vdc adapter with an high voltage boost converter will do just fine.

Personally I would go for an easier linear regulator as well, something with 1117 with a MOSFET pre-regulator for HV.
Click to expand...
Linear regulate a SMPS DC is not as easy as from Linear supply. Most IC regulators are optimized for 100/120Hz and have problems with 30 to 400Khz ripples. So this need an other kind of regulator.
 
#113
Koifarm said:
Linear regulate a SMPS DC is not as easy as from Linear supply. Most IC regulators are optimized for 100/120Hz and have problems with 30 to 400Khz ripples. So this need an other kind of regulator.
Click to expand...
I am aware of that 😉 Luckily ripples at those higher frequencies are also much easier to get rid of.
For decent performance, you need at least 100kHz (maybe 60kHz), anything lower will cause issues.
Ideally use some caps parallel on the output (of the transformer) to lower ESR, as well as putting at least some 100nF in parallel
At least 1206 or 1210, at least X7R or better (NP0), keep an eye on the max permissible voltage!
(there are high voltage available for certain packages)

It all depends if you really need perfect voltage stability.
Which is for audio never really an huge issue.

Otherwise just a simple capacitance multiplier will do incredibly well.
Unless you really care about 100-400mV voltage drop on 200mA full-load.
(which still seems a lot for a pre-amp, one can run a complete amp on that)
Since we are running only 200mA it's wise to use an isolated package for the MOSFET.
 
#116
Koifarm said:
This SMPS switches on 132Khz. About parallelling smalls caps for low esr is a good suggestion.

I will evaluate this design and if needed make improvements in the next version.

Next version will also have isolated heater supply.
Click to expand...
Also, with any board design with switching signals, a good solid ground plane is basically mandatory (because of EMI), as well as adequate slits to keep proper mains isolation.
I don't know about the current dimensions, but 4 layer boards are very cheap these days for everything that is below 100x100 mm
 
#117
Koifarm said:
Linear regulate a SMPS DC is not as easy as from Linear supply. Most IC regulators are optimized for 100/120Hz and have problems with 30 to 400Khz ripples. So this need an other kind of regulator.
Click to expand...
Linear reg speed depends heavily on the regulated currents! By the way...have you heard of capacitor multipliers at high voltage? They can be made fast enough and we can also apply them a basic zenner regulation.
The problem with using too much dry theory in smps design for audio is that it ignores our hearing capabilities...The real problem with switching circuits is the mains bleeding through the ground! Just listen to headphones on your smartphone while charging it from the mains through an ordinary cheap chinese smps charger! You won't hear any damn 50..100khz switching noise , but you will deffinitely hear mains hum harmonics:100...150 ..200hz!
https://www.eevblog.com/forum/blog/eevblog-1116-the-capacitance-multiplier/
The adp 150regs were used in a project i was involved to filter 400khz switching noise to reach a very complex 32 bits adc-dac and xmos board into electrostatic headphones amp and the only noise we heard was damn 100hz hum bleeding through both 230c ac:24v dc smps followed by a 400khz smps getting us the 3.3, +-15 and +500v lines we needed from 24 v .
Do you see those ADP regs very capable of reducing 400khz (precisely 470khz ) ripple?
Well ...we also used a very basic mosfet cap multiplier on the 500v line.
I remember telling my boss at the time how bad a cap multiplier should be after the 500v supply at 400khz and he looked back at me telling: do you know Batman? I'd like too meet him!
Apparently he designed bat detectors capable of analyzing and register individual sounds from each bat making bat sound signature libraries recorded and played back at lower speed in the digital domain helping bat protection organizations to identify each bat and keep a record , yet he never had the honour to meet Batman.Maybe you did! By the way...he's also one of the inventors of blue ray and sacd...I hope you can teach him a lessons on cap multipliers and smps psrr...I failed to pass him the fantastic knowledge of my romanian fellows modding cd players in their caves.
 

Attachments

  • Screenshot_20220811-114454_Word.jpg
    Screenshot_20220811-114454_Word.jpg
    219.7 KB · Views: 94
Last edited:
  • Like
Reactions: Koifarm
You must log in or register to reply here.
Top Bottom