SWITCHING POWER SUPPLY

[SMPL]

I have Iron with 50vac taps not being used. I would like to use them for a switching supply. I have read about to many heater filament tap issues and do not want to go that route. Should I purchase a smaller transformer for my 5v omrons or is there a good schematic I can use to utilize the 50vac taps not in use and step down the voltage after rectification. I have seen Brown notes supply and the #124 supply with a regulator. Any thoughts?

Thx!

[Colossal]

I have Iron with 50vac taps not being used. I would like to use them for a switching supply. I have read about to many heater filament tap issues and do not want to go that route. Should I purchase a smaller transformer for my 5v omrons or is there a good schematic I can use to utilize the 50vac taps not in use and step down the voltage after rectification. I have seen Brown notes supply and the #124 supply with a regulator. Any thoughts?

Thx!

SMPL,

You could use your 50VAC supply but you might consider dropping the voltage down to something more reasonable like 12VAC before rectifying it. If you plan to regulate 5VDC for your relays, you wouldn't want a regulator like say a 7805 to have to dissipate much power if you didn't have to. It's better to get the voltages closer to say 12VDC before regulating. A small, standalone power transformer might be a better fit. In a pinch, you could just use brute force by rectifying your 50VAC tap to 70VDC with a bridge, then filter and drop it down to 12VDC across a fat resistor, then regulate to 5VDC with a 7805. A 7805 doesn't want to see any more than 35VDC on the input.

[DonMoose]

Switching supplies are magnetically - and electrically - noisy unless you are exquisitely careful with return paths and switch node grounding. Then you design for a target load to give you maximum efficiency (and lowest noise) at that load. At lower loads, the efficiency falls off and the noise may increase.

Besides which, that 50VAC winding is probably good for 100mA max - that's 5W if your switcher is 100% efficient (switchers are like transformers - constant power (VAin = VAout * efficiency ). Probably enough, but what you'd pay in parts and board and futzing around, you'd do at least as well with an aux transformer + recto + linear reg.

tl;dr - I know.

[C Moore]

If you don't mind me asking.......What is the advantage of S-M Power Supplies.? To look at them, they seem to be kind of complex and with a lot of parts.
Thanks

[XgamerGt03]

If you don't mind me asking.......What is the advantage of S-M Power Supplies.? To look at them, they seem to be kind of complex and with a lot of parts.
Thanks

Extremely high efficiencies. You can get up to 96% efficiency with a good quality SMPS. It also allows you to control the output for voltage or current. Its great for battery chargers and other items where you know what voltage you want and need to maintain a constant current, or have a maximum limit.

However, they are extremely noisy if you do not know what you are doing. They have to have very well thought out layouts or you will not only induce noise in other items, but you will have an unstable power supply. You also would have to have a pretty large input filter to keep the switching from back feeding into your transformer.

You may be able to find one of those "switcher in a chip" ICs that will regulate your 70+ volts after rectification to 5 volts. You would be looking for what is called a DC/DC converter chip. Those tend to be a bit less noisy as they don't need transformers and other items.

[Super_Reverb]

If you don't mind me asking.......What is the advantage of S-M Power Supplies.? To look at them, they seem to be kind of complex and with a lot of parts.
Thanks

Switchers use pulse width modulation to control output voltage based on load and input voltage. These power supplies turn a power device full on or full off (to minimize power) and then adjust duty cycle (on time) in order to control output voltage over load conditions. Many commercial switchers use custom integrated circuit controllers and with creative arrangement of choke and caps, boost regulators, buck regulators, and doublers, etc are possible.

Like Xgamer said, very high efficiencies and consequently reduced size for a given power output (cell phone chargers, etc) are the selling points. Since they can operate at 100s of kHz, transformers and coils can be made smaller (inductive reactance ~ freq)

Noise can be a HUGE issue. For my money, not worth it for tube guitar amps. Large and lossy is the arena that we play in.

cheers,

rob

[Cliff Schecht]

If you don't know how to design and apply a switching power supply then noise won't be your only issue.. If it's switching at over twice the frequency that you can hear then, theoretically you shouldn't run into any noise issues as far as heterodyning is concerned. As far as device noise is concerned, you're chopping the MOSFET so the RMS noise is reduced greatly compared to a saturated FET. Switching noise should not be audible past maybe 25 kHz, but this leaves the chance for heterodyning so again it helps to be switching past the Nyquist limit. EMI issues can be subdued with layout techniques and slew-rate limiting on the switches (high dv/dt will spew out a lot of broadband crap). It's when people don't design switching power supplies correctly for audio circuits that they get a bad rep like this!

The big issue with switching power supplies, especially when involving tube amps, is getting those damn filaments going properly. The inrush surge on tube filaments can be 5x or more higher than the quiescent filament current draw and even though this is for a very short period of time, it can cause issues with sub-harmonic oscillations, lock-up or the PWM just shutting off completely (fail-safe mode). The PWM on SMPS's are usually optimized for a pretty short range of loads, usually maybe 10:1 to 100:1 depending on the current levels. Many switching power supplies won't even work properly without some sort of load. So really to run any amount of filaments on a switching power supply, they most likely will need their own control section and magnetics as opposed to using a flyback to generate heater and HV windings like in a computer power supply. A simple dedicated offline buck converter would probably do the trick and the HV could come from a separate converter, but it would be very wise to sync the oscillators on these two converters to prevent audible beat frequencies.

And FWIW the attractiveness of switching power applied to tube amps is not to boost efficiency, tubes are inefficient devices anyways, but to shed the weight that the power transformer adds. This to me is where the attractiveness of SMPS's are when thinking tube amps.

Switching power is one of those things that involves a little bit of everything that a EE does. You get your general purpose analog circuits, device physics, feedback, magnetics, filters, EMI/noise, strong knowledge of passive components and even a good amount of digital. I'm getting started on the IC side of switching power supplies design at work and with my research so I get lots of practice in all of those fields..

[markr14850]

Just for clarity, are you asking about Switch Mode Power Supplies, or simply about building a power supply for some channel switching relays?

[Cliff Schecht]

Oh shit, you're right, he's talking about relay power supplies!

Damn, I always get excited when we discuss SMPS's around here...

[ToneMerc]

[/quote="Colossal] You could use your 50VAC supply but you might consider dropping the voltage down to something more reasonable like 12VAC before rectifying it. If you plan to regulate 5VDC for your relays, you wouldn't want a regulator like say a 7805 to have to dissipate much power if you didn't have to. It's better to get the voltages closer to say 12VDC before regulating. A small, standalone power transformer might be a better fit. In a pinch, you could just use brute force by rectifying your 50VAC tap to 70VDC with a bridge, then filter and drop it down to 12VDC across a fat resistor, then regulate to 5VDC with a 7805. A 7805 doesn't want to see any more than 35VDC on the input.[/quote]

Ditto, and makes sure that you use a heatsink on the regulator. Any voltage greater than 3-4 volts above the LM regulator target output is wasted and dissipated as heat.

TM

[ToneMerc]

[quote="Colossal] You could use your 50VAC supply but you might consider dropping the voltage down to something more reasonable like 12VAC before rectifying it. If you plan to regulate 5VDC for your relays, you wouldn't want a regulator like say a 7805 to have to dissipate much power if you didn't have to. It's better to get the voltages closer to say 12VDC before regulating. A small, standalone power transformer might be a better fit. In a pinch, you could just use brute force by rectifying your 50VAC tap to 70VDC with a bridge, then filter and drop it down to 12VDC across a fat resistor, then regulate to 5VDC with a 7805. A 7805 doesn't want to see any more than 35VDC on the input.

Ditto, and make sure that you use a heatsink on the regulator. Any voltage greater than 3-4 volts above the LM regulator target output is wasted and dissipated as heat.

TM

[Cliff Schecht]

One more try!

You could use your 50VAC supply but you might consider dropping the voltage down to something more reasonable like 12VAC before rectifying it. If you plan to regulate 5VDC for your relays, you wouldn't want a regulator like say a 7805 to have to dissipate much power if you didn't have to. It's better to get the voltages closer to say 12VDC before regulating. A small, standalone power transformer might be a better fit. In a pinch, you could just use brute force by rectifying your 50VAC tap to 70VDC with a bridge, then filter and drop it down to 12VDC across a fat resistor, then regulate to 5VDC with a 7805. A 7805 doesn't want to see any more than 35VDC on the input.

The more voltage you have to drop, the more heat your regulator creates. While they do have built in thermal protection, trying to ask them to dissipate even 30W (30V dropped from 35V at 1A) without a LOT of heatsinking will make them turn off a few seconds after they turn on. Like other suggested it's better to get a voltage only a few Volts above regulation, although it helps to stay a few Volts above the required drop (usually at or above 2V for 78xx series!) which means you need a minimum of maybe 7.3V for the regulator to even work properly and a few Volts on this for even more margin..

[Colossal]

Not to plug my own stuff, but I had a board made for a 5VDC regulated power supply to run 5V relays. It uses a 7805 regulator and is intended to be bolted to the chassis as a heatsink. As Cliff mentioned, they have built in cutoff/thermal protection. They are best operated with a small differential voltage between input and output to minimize dumping power as heat. The 7805 can provide up to 1A and a little more with good heatsinking.
[img:640:480]http://dl.dropbox.com/u/7561678/Boards/Boards.png[/img]