Adjustable AC power supply

An adjustable 30V AC power supply sounded like a cool thing to have on the bench, so I decided to throw together figure 3 on this web page:

formatting link

I used a few junk box substitutions for the BC550B and the BC141, and that may entirely be the source my problem, but I wanted to check if the random- found-on-the-Internet design is sound before I spend the time to source the same parts the author used.

The problem I'm seeing is that when adjusted to approximately half voltage, the output isn't a smooth 15V AC wave, but it's a 30V cycle followed by th ree truncated 10V cycles. I see a sawtooth on C4. It gets charged up on the full cycle, this causes the op-amp to shut down as C4 drains over the nest 3 truncated cycles, the op-amp turns back on, C4 charges to full, repeat.

Thanks, Scott

Reply to
smbaker
Loading thread data ...

e bench, so I decided to throw together figure 3 on this web page:

formatting link

t may entirely be the source my problem, but I wanted to check if the rando m-found-on-the-Internet design is sound before I spend the time to source t he same parts the author used.

e, the output isn't a smooth 15V AC wave, but it's a 30V cycle followed by three truncated 10V cycles. I see a sawtooth on C4. It gets charged up on t he full cycle, this causes the op-amp to shut down as C4 drains over the ne st 3 truncated cycles, the op-amp turns back on, C4 charges to full, repeat .

It doesn't looks well thought out. If you want AC the Baxandall class-D osc illator isn't too bad a way of getting it - the output voltage reflects the DC supply to the oscillator, but if you don't need much power, that may no t be a problem.

formatting link

formatting link
r1.htm

--
Bill Sloman, Sydney
Reply to
Bill Sloman

That's elaborate, ugly, and has no real hope of working. Get a 24V stepdown transformer, and hook it to a variable autotransformer. Or, if you want real flexibility, feed a low AC voltage into an audio amplifier.

Reply to
whit3rd

or use a simple linear regulator where its Vref is just an adjustable fraction of the voltage rather than zenered. x2 for ac.

NT

Reply to
tabbypurr

You could try adding a capacitor across R2, at a guess try 0.1uF or 0.47uF.

piglet

Reply to
piglet

Possibly. It's wired with fairly high gain, suicide biasing (hFE limited, through three transistors, no less!), and has terrible loop dynamics (a peak detector responds instantly and recovers slowly). Without any appearance of intentional compensation (perhaps the limited gain of the op-amps would be helpful -- doing it the bad way, throwing away loop gain instead of fixing the actual dynamic problem), oscillation or chaotic behavior (driven limit cycles?) should be expected.

Besides that, the waveform in general will not be sinusoidal, but rather flat-topped or lopsided (depending on reactance, and how slow the compensation is). This is because, if feedback doesn't change appreciably during a cycle (problem is, it does), the current flow is constant, so for a resistive load, it just clips the peaks.

There's also absolutely no current limiting, so calling it a "3A power supply" is misleading; it could potentially source much, much more, to its own detriment. (Even those awful Chinese (DC) bench supplies know better than to do that. Well, for a little while, until the bottom-shelf fan siezes or something and the whole thing wheezes the smoke away.)

How to fix it? The simplest way is also the best way. Start over from scratch. Build an audio power amplifier, and connect the input to a potentiometer on the power transformer. Voila, adjustable AC output. Boring use of an amplifier? You bet. Solid state control? Current limiting? You bet! (Assuming you added the current limit / protection transistors to the output stage, of course... you did, didn't you?..)

For bonus points, you can build the amplifier DC-coupled, and have a selectable input from three sources: DC (via TL431 reference perhaps?), AC (via power transformer and resistor), and external (via RCA or BNC connector, perhaps?). All of which go into a potentiometer (aka volume control) and the rest of the amp.

That's what I did for a bench supply. Difference being, that one started from tape drive parts. I mean rack mount, spinny reels, computery kind of tape drive. Not the full power, multi-HP kind, but a pretty good one nonetheless. Anyway, that had 20A+ darlingtons in it, and a 500VA (at least) transformer, good for +/-25VDC. So I have a 0-20V adjustable supply capable of short circuit current over 40A (measured!), which doubles for whatever kind of brick-shithouse-driving amplifier I might need (200W+ into one or two ohms, anyone?).

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
 Click to see the full signature
Reply to
Tim Williams

MMMMmmmm grumble...... Use no active parts, just a plain resistor in series; NO distortion; how about that being a hot idea? OR.. Use a variac to the primary of a 30V transformer?

Reply to
Robert Baer

The esiest way to make that work is keep the transformer, and chuck the rest out, and then put an ajustable autotransformer (Variac) on the input side.

You'd only need a little variac to tun it like this "0.231kVA" one:

formatting link

As others have stated, the circuit presented doesn't actually make any sense.

Be sure to fit a fuse and ground the chassis.

--
  \_(?)_
Reply to
Jasen Betts

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.