Does anhone use those switching power supply for the heaters?

[pchw]

I was looking for some parts in MPJA and came across the power supply section. Something like this
http://mpja.com/productview.asp?product=16918+PS
cost less than 10 bucks and can easily provide the DC power supply for 4 6SN7's heaters (series and parallel wiring). These seem to be a handyway to offer VDC to the heaters of small tubes.
Did anyone use this kind of desktop power supply for such application?

Thanks,

 

[alexmoose]

I have heard before (however I have never tested myself) that using a swtich-mode power supply is too noisey to use for DC heaters. What I would do personally, would be to run that through a linear regulator, that would elimante the noise put in by the SWPS.

-Moose

 

[JoshK]

I think it might be interesting to get some cheap but substanitive surplus chokes and place them before and after the SMPS and shield the SMPS and see how it sounds. Try both CMC and series Choke.

 

[nhuwar]

I'm thinking about using Some higher end smps for a amp I'm building.
Although I dont know how they sound yet.

But I was always told it was easier to filter out higher frequencies then lower.

Nick

 

[pchw]

What about using one of these:
http://mpja.com/productview.asp?product=15654+PS
as following:
PS <=> 4700uF/25V <=> LM317 <=> o12.6VDC
|
0.01uF
|
ground.
Will that help reducing the noise?

 

[tjl]

Switching power supply for tube filament heating will be more convenience and also save space of linear power transformer,

At lower heater current voltage amplifier stage , I use LM317 linear regulator like PCHW mention about before, but LM317 need big heatsink for heat dissipation.

AT power amplifier stage, I prefer Switching power supply for the huge current of tube filaments.

As my experience: Switching power supply for tube filament will not inducing hum or noise, because its operating frequency is about 28KHz to 60KHz far away from audio frequency range.

12V 3.3A switching adaptor may be suit for 4 X 6SN7 tubes by series and parallel connection, but if add a 1000uf 12V capacitor on the output terminal will boost current and more stable for 12V voltage output.

 

[Brian Beck]

I've used in-line laptop power supplies that look similar to the supplies in the MPJA links, but have output voltages in the 18 to 20 volt range. These were various Dell models. Then I've used linear regulators (in the LM317 family) to drop the voltage down to 12.6 volts and to filter out noise. You need passive filtering to remove all the switching noise since the LM317 isn't fast enough to keep up with 50KHz noise. I suggest adding a ferrite clamp onto the lead going from the in-line supply to the chassis, if the unit doesn't already have one (many do), as well as chokes and caps inside the chassis before the LM317.

If not well filtered, the switching noise, even if it is ultrasonic, can intermodulate with digital artifacts from CD players to make in-band tones. I think it's best to include linear regulators and to filter well with passive parts.

The advantage of the in-line laptop switchers is that they can be separated physically from the sensitive audio circuits. They tend to be very efficient, lightweight and cool-running.

 

[pchw]

What kind of choke is adequate for the job?

Thanks,

 

[Eli Duttman]

What kind of choke is adequate for the job?

Thanks,

I would think a few turns around a ferrite ring used for noise suppression on computer ribbon cables will do the job.

Also, put a 100 nF. film cap. directly across the I/P pins of the linear regulator IC.

Finally, 2X 7812 fixed regulators will do nicely for 4X 6SN7 heaters. The 0.6 V. "error" is well within tolerance.

 

[JoshK]

I am building a PP 6c33c amp, which amounts to 6.3v @ 12-14a per channel of heater needed (if running both cathodes on each tube). Or one can wire for 12.6V @ 6-7a, still a lot of heater juice needed. I think I will look further into the SWPS option for at least powering the output stage heaters.

 

[kevinkr]

You'll definitely want to heat both sections of the 6C33 otherwise the tube behaves in a very non-linear manner for ac operation. Had this problem with my old otl design using this tube. (See my site for the sordid details.. )

Probably 12.6V operation will be easier to manage from a wiring gauge standpoint but high current 5V supplies are easier to come by. You might have to add an additional turn or two on the smps transformer to get proper operation at 6.3V - and usually you will also have to change a feedback resistor as well. Remote sense is a good choice in this application to eliminate the effects of voltage drop in the filament wiring.

There is a huge benefit to doing this however, and that is in terms of efficiency and weight. My filament transformers weighed about 18lbs apiece.

 

[richwalters]

I regulary design using 10W smps for DC down-line front ends.. When using split load concertina or sim with high cathode voltage, I always lift (isolate) heater to roughly half cathode volts. i.e if cath =+100V use +50V and decouple with small 10uF electrolytic to ground. This will avoid heater/cathode noise esp if combination med mu triode and high gain pentode is used.
Additionally, common mode chokes may have to be fitted to both in/out to avoid noise in chassis. This is a common problem of charge-injection where a tab of a switching mos may be clamped to chassis via insulating washer.

richj

 

[gofar99]

Greetings, In regard to Brian Beck's post. I have had similar experience using laptop supplies for tube preamps. I recently assembled a kit with various mods (on the diyaudioprojects.com site "Beauty") and found that it didn't sound quite right. On a scope (photo in the article) you could see the 70KHZ spikes that got through from the laptop smps. I used a pair of 100uh chokes and .1uf caps as filters. Other solutions would also work. On general principle I will now scope any smps to make sure high freq trash doesn't make it to the preamp, amp or whatever.

good listening,
gofar99

 

[JoshK]

Thanks for the tip Kevin! I haven't made up my mind yet on how to go about this. I did get my hands on some surplus iron, but as you said, they are huge and heavy. I'll probably use them to start with and consider a change to SWPS when it is running.