SMPS for 300B heater ?

[Jaap]

I bought two switched mode power supplies in a surplus shop. They can deliver max. 7A each.
Can I use these to heat a 300B tube (in a serious amp) ?

Has anybody tried this ?
?

[martinab2]

Try this thread

[Jaap]

I saw that, but I didnot want to go RF (too complicated for me) and I do not want to add another transformer in my amp for "regular" dc.
I bought these SMPS for a only a few euros so it would be handy if they could be used for the 300B heaters. Just curious if someone has tried this before and what were his experiences with it.

[martinab2]

I have never used smps for tube heaters although I have some experience of their use in transistor audio applications. The main issue is that you will be dealing with RF frequencies whether you want to or not! Most switch mode supplies operate at frequencies well above audio range and throw out quite a lot of noise.
The thread I linked to covers many of the issues involved
Probably nothing to lose by giving it a go but I'm not sure about the logic of introducing a source of rf noise to an audio amplifier when most designers are trying to do the opposite!
Also, have you tried using plain old ac heater supplies first? I've found them to work rather well in most applications especially with the use of a hum balancing pot.

[Jaap]

I have a hum balancing pot, there is always remaining a little hum.

I think I will try the smps some time and put it in a tin.

[astouffer]

What do you mean by "quite a lot of noise"? Most modern supplies will have noise in the millivolt range. We use switching supplies for opamps all the time at work without noise problems. I somehow doubt its going to have much affect on a 300B.

[analog_sa]

I somehow doubt its going to have much affect on a 300B.

Yes, it will. There will be no hum but without extensive filtering post smps the sound will be very unpleasant.

Comparing 300B to opamps is very funny though.

[martinab2]

What do you mean by "quite a lot of noise"?

Ok, so a rather vague description but the problem is usually the nature rather than amplitude. Most smps are very quiet at audio frequencies but produce rf noise over a wide bandwith (presumably harmonics of the switching frequency) - visible on a good scope. The other issue is radiation from components within the power supply and cabling.

Your point about opamp circuits is valid in that the noise produced by smps is insignificant in many applications. Most mass market audio equipment uses this sort of technology for reasons of cost, efficiency and reliability versus ultimate sound quality.

Anyone building a 300b amp is looking for the sort of sound quality that mass market stuff can't deliver (£1000+ spend on components alone) and at this level small design details have a dramatic effect on the final results.

[astouffer]

How is a few millivolts of RF noise going to affect a tube with a plate potential of several hundred volts and a grid with a 50 volt signal? The amount of filtering needed should be minimal due to the frequency. If theres no ill effects on opamps with 50db of gain, whats it going to do on a 300B?

[thoriated]

How is a few millivolts of RF noise going to affect a tube with a plate potential of several hundred volts and a grid with a 50 volt signal? The amount of filtering needed should be minimal due to the frequency. If theres no ill effects on opamps with 50db of gain, whats it going to do on a 300B?

How about potential magnetic coupling to the output transformer? Audio band optimized core materials usually do not behave nearly as well in the RF range. I've seen large negative and measurable effects from radiated RF hash on op amps such as noise floors raised by 10 db and internal RF slew limiting with attendant increases in audio band distortion, particularly with op amps with less than linear open loop behavior from much smaller amounts of radiated noise such as from dumpy slow old LSTTL, in a mixed signal environment, than the generous amounts of magnetic and electrostatic noise a beefy open frame switcher throws out. It may or may not have any effect on the 300B sonic qualities itself (maybe you can develop a test with a suitable control and let others at DIYAudio know more definitively), but this RF noise could easily be conducted through your interconnects and could also radiate out into your source componentry without some good shielding and filtering (e.g. ferrite beads, etc.), so those variables would need to be controlled also. Then you may also eventually be able to determine its effects on your source componentry. Which is not to say that you shouldn't use a switcher at all for a filament supply, but much care is probably called for that it is not influencing the sonic quality at some point in the audio chain.