SMPS for valve circuits

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

jacomart

Well-known member
Joined
Sep 13, 2017
Messages
140
Location
Tuscany, Italy
This one (perhaps with a little additional filtering) could be the solution to many problems: anode, heater, phantom ... who knows if it works! 🙄

Cheers
JM
 
You'd probably have to house the PSU in a separate enclosure, in addition to filtering?
 
I have a Recom R12-150b in my parts bin I’ve been meaning to try. 12v in, 200v out. You can stack 2 for higher voltage and trim the lower one for adjustable b+ according to the data sheet.
 
You'd probably have to house the PSU in a separate enclosure, in addition to filtering?
One of the comments say that the amp it is used to power is very noisy, so I guess some additional filtering and shielding is necessary. But, isn't it the case with a linear PSU?
The problem is many people think it just takes to put an smps in palce of a linear PSU, but actually, it takes some thinking.
 
Almost certainly not.

Regarding Bo Deadlys SMPS Checklist, issue #2 is a concern unless you just happen to be building a 200W power amplifier (with phantom power?) but my guess would be that #1 would be a show-stopper.

But you never with SMPS until you just get one and try it so don't let me stop you. But personally I wouldn't even bother. I would just "stack" individual modules that are the right power and use ones that are known to be good and low noise.
 
One of the comments say that the amp it is used to power is very noisy, so I guess some additional filtering and shielding is necessary. But, isn't it the case with a linear PSU?
The problem is many people think it just takes to put an smps in palce of a linear PSU, but actually, it takes some thinking.
Yes, the lack of "thinking" is often at the center of many problems. ;)
 
One of the comments say that the amp it is used to power is very noisy, so I guess some additional filtering and shielding is necessary. But, isn't it the case with a linear PSU?
The problem is many people think it just takes to put an smps in palce of a linear PSU, but actually, it takes some thinking.

Just by instinct, having a high frequency oscillator feeding a transformer, at those voltages, makes me think of quite a bit of EMI being 'broadcast' by the power supply.

But maybe it's not so bad, I haven't used any.
 
Just by instinct, having a high frequency oscillator feeding a transformer, at those voltages, makes me think of quite a bit of EMI being 'broadcast' by the power supply.

It all depends on the layout and chosen parts. An SMPS is effectively a tuned circuit. If the manufacturer cares to make it low EMI they can and over the years manufacturers (at least MeanWell) have managed to make some very quiet SMPS. But in my experience you do need to keep them some distance away from any high gain, high impedance bits more or less depending on the amount of current being thrown around.
 
I would have thought the biggest concern to initially jump is the AC mains input related safety. Who's to know what level of safety compliance is designed and built in to such a pcb module - until it comes with an appropriate standards compliance certificate then the user takes all that risk. In my country a retail outlet can't sell a product that is plugged in to the AC mains unless it first passes such a base level of safety compliance for the user (let alone the other main issue for EMC compliance).
 
My concern is what happens inside a tube at power-on when a SMPS feed of B+ arrives? I have seen high voltage output SMPS able to produce an extremely short duration very large voltage spike before settling. Would this give rise to a premature eroding at a close proximity bond wire in the tube or other elements? It would say an extra level of management as to when the voltages are applied to the expensive valve would be a good idea. Worse though is just how 'quiet' can any sort of SMPS ever be not only their terrible RF interference they generate but just how 'quiet' can the DC produced ever be filtered without making the SMPS more expensive that a linear.
 
My concern is what happens inside a tube at power-on when a SMPS feed of B+ arrives? I have seen high voltage output SMPS able to produce an extremely short duration very large voltage spike before settling.

There would have to be some filtering on the output. Currently I advocate using a simple capacitance multiplier. That will take RC time to reach voltage so any "spike" won't make it through.

Worse though is just how 'quiet' can any sort of SMPS ever be not only their terrible RF interference they generate but just how 'quiet' can the DC produced ever be filtered without making the SMPS more expensive that a linear.

SMPS have come a long way over the years. Some of the MeanWell LED supplies are known to be very quiet. But there are a bunch of rules about picking the right size, how to keep it out of hiccup mode, filtering the output and so on. If done correctly it's pretty easy to get completely noise free power.
 
There would have to be some filtering on the output. Currently I advocate using a simple capacitance multiplier. That will take RC time to reach voltage so any "spike" won't make it through.



SMPS have come a long way over the years. Some of the MeanWell LED supplies are known to be very quiet. But there are a bunch of rules about picking the right size, how to keep it out of hiccup mode, filtering the output and so on. If done correctly it's pretty easy to get completely noise free power.

There would have to be some filtering on the output. Currently I advocate using a simple capacitance multiplier. That will take RC time to reach voltage so any "spike" won't make it through.



SMPS have come a long way over the years. Some of the MeanWell LED supplies are known to be very quiet. But there are a bunch of rules about picking the right size, how to keep it out of hiccup mode, filtering the output and so on. If done correctly it's pretty easy to get completely noise free power.
Please offer us any make/model from a reputable manufacturer... (not an Aliexpress special). I would love to get hold of such a unit that can produce a B+ in the range of 100VDC to 550VDC unit and do some testing mainly with tube based audio amplification.
 
Please offer us any make/model from a reputable manufacturer... (not an Aliexpress special). I would love to get hold of such a unit that can produce a B+ in the range of 100VDC to 550VDC unit and do some testing mainly with tube based audio amplification.
Currently I am investigating the Meanwell ELG-150-C500A. It is actually an LED driver but until the current limit kicks in it will provide 300VDC at up to 500mA. Only ÂŁ41 from Farnell. Datasheet here: ELG-150-C500A

Cheers

Ian
 
Please offer us any make/model from a reputable manufacturer... (not an Aliexpress special). I would love to get hold of such a unit that can produce a B+ in the range of 100VDC to 550VDC unit and do some testing mainly with tube based audio amplification.

I'm actually not a tube guy but maybe watch the Ripple Filters thread. It seems Ian is exploring the various options right now.

However, one issue I can see straight away is that the higher voltage supplies tend to be a bit large. And that's a problem for smaller projects because one of Bo Deadlys SMPS Checklist items is that you have to load it enough to keep it out of hiccup mode. Hiccup mode is when the SMPS modulates the output at a low frequency to improve efficiency when low power is drawn. So if your project is relatively small, that's a problem. I'm not aware of a smaller SMPS that puts out high voltage.
 
Back
Top