# 9V AC stabilizer

I've got a power supply:

• IN: 220V AC
• OUT: 9V AC 1A

Note that the output is AC, not the conventional DC output.

It has no regulation systems built in (simple transformer). I'm looking for a circuit to regulate the 9V AC output, because it varies as a result of unstable input (180-230V). I couldn't find a circuit to do this anywhere.

Alternatively, I'll be happy if someone gives an example of simple low-power 220V AV regulation circuit, so I can regulate the input of the transformer.

• I don't see how this would help get a stable 9VAC. It can help get a stable 9VDC, but not AC. – Kellenjb Mar 9 '12 at 22:26
• The way I read this question is the OP has 180VAC to 230VAC in (basically the variation from the power grid) and he wants to output 9VAC. – Kellenjb Mar 9 '12 at 22:29
• Are you really sure you mean AC output? That's possible, but rather unusual. I wrote a answer below not noticing the output is AC instead of DC. I have meanwhile deleted it since it doesn't apply. Why do you need tightly regulated 9 V AC? What are you going to do with it? – Olin Lathrop Mar 10 '12 at 0:00
• @Kellenjb, Yes, you got it right, I've got problems in my power grid. Usually the voltage is much lower then 220V and that's what causes problems. OlinLathrop, Yes, I'm sure. And indeed, this is pretty unusual. I'm using Zyxel ADSL modem which has all the rectification circuits inside. And it hates low input. – dmytro Mar 10 '12 at 10:23
• @OlinLathrop, Yes, I'm sure. And indeed, this is pretty unusual. I'm using Zyxel ADSL modem which has all the rectification circuits inside. And it hates low input. – dmytro Mar 10 '12 at 10:38

"And indeed, this is pretty unusual. I'm using Zyxel ADSL modem which has all the rectification circuits inside."

So, tell the doctor, you don't need the AC really, do you? First thing it meets in the modem is a rectifier, followed by a smoothing capacitor. And then the AC is gone. The modem works on a DC power supply.
The 9V AC has a peak value of 12.7V, the bridge rectifier drops about 2V, so the smoothing capacitor sees 10.7V minus some ripple. If you supply the modem with a 12.7V DC supply the result is exactly the same: 2V drop across the rectifier and 10.7V on the capacitor, sans the ripple then. The absence of ripple means that it's pretty certain it will also work at 12V in.

So you don't need a regulated 9V AC, but a regulated 12V DC. You can use a linear regulator or a switcher if you want more efficiency. Things to look out for:

• Smoothing capacitor capacitance should be sized for 180V on your transformer, but the working voltage for the 230V
• Input voltage of the regulator should be OK at both 180V in and 230V in

There may be a minor issue with the bridge rectifier in the modem. If you supply AC to it diodes conduct current on alternating half cycles, with a DC input only two diodes will conduct all the time, so their power will be double. If the diodes get really hot you may replace them by a more heavy version.

edit
Fake name mentions this rectifier:

and indeed if that's used the DC solution doesn't work. Actually I'm not really worried about this. This design has some disadvantages. First, only half of the input power is available for V+. In a modem you'll need most, if not all, power for V+. Second, this is only a half wave rectifier, which either needs very large smoothing capacitors (expensive and space-consuming), or gives a large ripple, which means an inefficient power regulation. I doubt this will be used.

• The other issue I can think of here is if they are using the AC to produce a split-rail supply. With a 9VAC input, you can produce both + and - ~10V rails. This will not work if you feed it pure DC. – Connor Wolf May 5 '12 at 8:31
• @Fake - I edited my answer. – stevenvh May 5 '12 at 9:21

Even though your device says it needs 9V AC, it actually just rectifies it to just under 9V DC.

If it does so with just a bridge rectifier then you can safely feed it a DC voltage, and it will still work. This assumes that the device doesn't use the AC waveform as a 50/60 Hz timer. (Such as a cheap alarm clock.) The DC voltage will be different than the AC voltage the device was designed for.

Otherwise, the most robust way to power such a device would be similar to a online UPS only without the battery. You would convert the mains power to a fixed DC voltage, and use that to power an inverter that powers your device.

• I had a (probably similar) DSL modem which used a 9V AC PSU on a standard 2.1mm plug. I accidentally ran it for over a year on a 12V DC wall wart without realizing it until I discovered the disconnected AC-outputting transformer below the desk. – Kevin Vermeer Mar 12 '12 at 4:20
• @BradGilbert, Hm, using DC power supply might be a nice idea if it works, since it looks like the device has a diode bridge right next to the 9V AC input. I'll try it and report – dmytro Mar 12 '12 at 8:15
• @KevinVermeer, cool, I should definitely try DC – dmytro Mar 12 '12 at 8:18

The easiest way I can think of to do exactly what you describe yourself as wanting (if not necessarily the best way to solve your problem), is a inexpensive audio-amplifier, and a sine-wave generator.

Basically, you feed the audio-amp a 60-Hz sine-wave (you can make a simple sine-wave oscillator with a couple of op-amps). You then run the modem off the output of the amp. You just need to adjust the amplitude of the input signal so the output is precisely 9VAC RMS.

Since this is fairly low power (wall-wart transformers are rarely more then ~10-15W), a 20-30W amplifier should easily suffice.

A DC supply, as the other answers describe, is really a far more elegant solution. I'm posting this as more of a FYI, and as a precaution against the modem manufacturer doing something unusual, like using the AC input to produce a split-rail output, or multiplying it up to get a higher voltage, etc...

If you take the modem apart, you should be able to ascertain what they are doing with the input AC without too much trouble.