PSU for LM22673

[lowell]

Considering building this to power some filaments. Universal Vacuum Tube Filament Regulator

There's no info on how to supply the 4.5-42 volts to the input. It's an SMPS so I'm assuming I don't need a transformer...and hope that I don't cause that'll defeat the whole purpose. The TI data sheet doesn't even go into detail on this... other than mentioning the use of a Schottky diode.
http://www.ti.com/lit/ds/symlink/lm22673.pdf

This is all it says:


9 Power Supply Recommendations
The LM22673 device is designed to operate from an input voltage supply range between 4.5 V and 42 V. This input supply should be well regulated and able to withstand maximum input current and maintain a stable voltage. The resistance of the input supply rail should be low enough that an input current transient does not cause a high enough drop at the LM22673 supply voltage that can cause a false UVLO fault triggering and system reset. If the input supply is located more than a few inches from the LM22673, additional bulk capacitance may be required in addition to the ceramic bypass capacitors. The amount of bulk capacitance is not critical, but a 47 μF or 100 μF electrolytic capacitor is a typical choice.

So ideally I can get within range using a rectifier and a cap or two, or three, maybe some zeners or a regulator directly off the 120v mains.

[Enzo]

Simple as it says, you need to feed it 4-40vDC. If the system you are installing it into doesn't have that, then you will need to make a supply that does provide it. You can't feed it AC.

What would this do that a simple DC supply won't do? I see regulation as not necessary, the heaters have a steady current draw after the first few seconds. SO a simple 6vDC supply will get your heaters going. And that could be derived from the same source you would power this thing from.

This unit has its input and output grounds tied together, so you definitely would not want to tie it direct to the mains.

[lowell]

I want to build a tube amp using only a SMPS. Just for fun. I had thought this project was just that...but only for the filaments. Is it not? I think it must not be considering I'll need a regular linear supply for it. Ideally I can use a flyback SMPS buck type of dealy instead. I've been reading up on the supplies but definitely have more studying to do. I get the basic concept of the flyback air gap transformer, FET switch, feedback and PWM. But it's in the details....so I need to research a bit more.

[nickb]

I want to build a tube amp using only a SMPS. Just for fun. I had thought this project was just that...but only for the filaments. Is it not? I think it must not be considering I'll need a regular linear supply for it. Ideally I can use a flyback SMPS buck type of dealy instead. I've been reading up on the supplies but definitely have more studying to do. I get the basic concept of the flyback air gap transformer, FET switch, feedback and PWM. But it's in the details....so I need to research a bit more.

If you are serious about an entire tube SMPS, you might want to choose an easier path. You don't need the complexity of PWM, after all your current linear PSU manages just fine without and indeed, many solid state amps with a SMPS don't either. Plus not having that regulation means you retain the sonic properties that go along with the voltage droop. You drive your custom wound switcher transformer at a higher frequency and simply because the frequency is higher, the turns are less and the transformer can be made smaller.

I have had this on the back burner for a while: A half bridge MOSFET design using a ETD44 core driven by an IR2153 with secondaries for 6.3V, 60V and 400V-ish). I have all the parts and I even wound the transformer but that is as far as it's got as I'm been busy with other stuff. One of the problems was getting the transformer right due to the low number of primary turns. It has just two turns. It should good for well over 400W.

I think there are plenty of pitfalls. Large circulating currents and unwanted inductances spring to mind and are just waiting to vaporise parts and turn in all into a smoldering pile of junk

I really, really don't want to discourage you or to be wet blanket, but I do think you do need read and learn quite a bit more, as based on the discussion above there are some pretty big gaps in your understanding. OTOH, there will be huge satisfaction in doing all that learnin' stuff if you can pull it off and your enthusiasm will be rewarded.

It would be ideal to have tried and tested design to copy as a starting point. We can't be the first to think of this and I bet that quite a few people who read this forum have already done it and can teach us a thing or two

[Enzo]

Look at an existing SMPS design of roughly similar specs, and note the complexity. Every part in there has a purpose, they don't just throw in extras. Your design will probably need to look similar.