# Here is an elegant (insane?) way of converting 220V to 110V

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

#### Elvee

Here is a possible answer (the mad scientist version) to this problem:

What is the elegant way to convert 220? to 110V? - diyAudio

This circuit is basically a phase control system allowing the power to flow for 30° on either side of the zero crossing of the 230V input waveform (sin 30° = 0.5).
It uses a MOS rather than an SCR or triac to improve flexibility and to conduct twice per half-cycle.
Doing so has a number of advantages: it reduces the stress on the circuit elements and almost doubles the ripple frequency seen at the output of the transformer.
The conduction is a little larger than 30°, because the V.s product with this waveform is smaller and allows it, and because the leakage inductance of the transformer will cause a larger drop wrt. to the peak voltage than a regular sinusoid.
One screenshot shows the result with a simulated transformer, and the other on a pure resistive load.
The IC can be a LM393. The ground connection is for simulation purposes only.
Note that the circuit also acts as a soft- start.

Disclaimer:

If you choose to build this circuit, it is your sole responsibilty, etc etc

#### Attachments

• 220V110V1.GIF
65.6 KB · Views: 1,322
• 220V110V2.GIF
65.9 KB · Views: 1,298

#### zigzagflux

I suspect phase control will not produce a very happy audio system. Think of having light dimmers nearby, and the trouble they create. Now feed your system off a light dimmer, and get ready for some real buzz.

#### Charles Darwin

Still think the easiest way is to stroll down to your local building supplies merchant and buy one of those fetching yellow step-down transformers they use for 110V power tools.

On ebay they frequently pop up for next to nothing s/h and they come in all conceivable sizes..

#### Elvee

One more thing:

Thanks to the frequency-doubling effect, the peak voltage (and consequently the useful output power from the transformer) can be susbtantially increased:
The conduction angle can be increased to 60° without risk of causing the core material to saturate, leading to a peak primary voltage of 280V in a 230/115V system.

On the other hand, the ripple can be optimized by choosing a 45° angle, thus making 4*F the dominant ripple frequency, yielding 230V peak primary voltage.

The conduction angle can be modified by increasing the value of R9.

#### martin clark

Still think the easiest way is to stroll down to your local building supplies merchant and buy one of those fetching yellow step-down transformers they use for 110V power tools.

If you are tempted to try it, be prepared for odds sounds. Those 240:110Vac site isolation transformers (in the UK) have massive leakage inductance designed-in - it's part of the fault current limiting strategy, and doesn't matter to 50Hz power tools. It's also mostly why they run so damn hot even when unloaded.

#### Elvee

The following example compares the situation of a conventional 115V sinusoid solution with the voltage converter set at a higher conduction angle.

The average output voltage rises from 32V to 48V, a 50% increase, while the ripple voltage goes down by 1V pp (much more as a percentage).
Yet, the transformer's core utilization is still lower, meaning lower iron losses.

#### Attachments

• 220V110V3.GIF
59.8 KB · Views: 1,085

#### Charles Darwin

If you are tempted to try it, be prepared for odds sounds. Those 240:110Vac site isolation transformers (in the UK) have massive leakage inductance designed-in - it's part of the fault current limiting strategy, and doesn't matter to 50Hz power tools. It's also mostly why they run so damn hot even when unloaded.

I haven't tried it personally but whenever I see american musicians here they almost invariably have the instruments amps plugged into those.
Friend of mine uses an US-made SWR bass amp which kept breaking down. He eventually had it back-converted to 110V and is using one of those now without problems. Similarly I know of a few people who bought US studio equipment using them. All without odd noises or problems.

#### Elvee

In addition, the circuit also acts as a line voltage regulator: because the conduction time is decided by comparing the incoming waveform with a fixed reference, the output voltage has a constant peak value.
It is therefore a universal power supply front-end, accepting mains voltages from 100 to 300V~, depending on the MOS rating.

#### SoNic_real_one

And in addition will act like a huge voltage-spike generator on the secondary side of the transformers that are fed with such a voltage shape (due to the switching on non-zero transitions).
I would not tie that output to an equipment that has any kind of transformer or even direct rectifier/switched power supply.

#### Elvee

And in addition will act like a huge voltage-spike generator on the secondary side of the transformers that are fed with such a voltage shape (due to the switching on non-zero transitions).
I would not tie that output to an equipment that has any kind of transformer or even direct rectifier/switched power supply.

Here is an elegant (insane?)

By the way, SMPS based supplies see even larger spikes, and 1,000 times more often too.

#### Elvee

It was such a good idea I couldn't leave it untested.

Result?

It works like a charm, exactly like in the sim.

#### Attachments

• DSCF0003.JPG
279.2 KB · Views: 465
• DSCF0010.JPG
269.4 KB · Views: 453
• DSCF0011.JPG
260.7 KB · Views: 334
• DSCF0013.JPG
246.4 KB · Views: 647
• DSCF0016.JPG
285.7 KB · Views: 342

#### audiofan

Do not use wave shaping circuit with transformer & electronics ....... too many HARMONICS are present ...........

Wave shaping is uasefull with resistors ... heaters..... lightbulb ........ ect not transformer .

#### fireworks

Correct me if I'm wrong, but the only advantage compared to a SMPS is that your solution doesn't need a high frequency transformer.
Because in term of complexity, it's as complex as a SMPS.
As a disadvantage, it puts nasty current spikes into the supply and into the circuit that constitutes its load.

Last edited:

#### Elvee

Correct me if I'm wrong, but the only advantage compared to a SMPS is that your solution doesn't need a high frequency transformer.
When you want to use a 110V appliance on a 220V mains, you don't use a SMPS, do you?

Because in term of complexity, it's as complex as a SMPS.
Do you really think this is as complex as a 1KW SMPS?

As a disadvantage, it puts nasty current spikes into the supply and into the circuit that constitutes its load.
As I said earlier, this has to be taken with a pinch of salt.

The title says:
Here is an elegant (insane?) ....

But even that remains open to the discussion.

What you see looks like a nasty waveform.

But what the rectifiers and reservoir capacitors see is a doubled frequency, with current spikes reduced in amplitude.

I am not sure you can tell a priori how an amplifier will behave, and in some cases, it might actually improve the situation compared to a regular 100Hz ripple

#### thaumaturge

I've always known big hunks of iron (transformers) to be great circuit conditioners. So what does the output of a transformer fed by this circuit look like? I'll bet not bad at all.
Doc

#### Elvee

So what does the output of a transformer fed by this circuit look like? I'll bet not bad at all.
The oscilloscope screenshot is exactly that.

It is unloaded. With a rectifier and filter caps, the tops of the waveform are clipped

#### franzm

In a standard PSU, with bridge rectifier and reservoir capacitors in the secondary of a transformer, the conduction angle of each diode is less than 180 degrees of the sine cycle. So under load, the transformer will see current pulses anyway.
The circuit from post #1 seems to work as an forward converter for the first conductive cycle of the halfsine and as an flyback converter for the last.

#### Elvee

The circuit from post #1 seems to work as an forward converter for the first conductive cycle of the halfsine and as an flyback converter for the last.
There is a small flyback episode, but the main operation is purely forward.

But when the primary is interrupted under load, there is a small parasitic flyback spike caused by the leakage inductance of the transformer.

This spike is handled by the gate supply of the MOS, which also serves as a snubber.

Additionally, the prototype also has a MOV across the MOS, not shown on the schematic

#### EWorkshop1708

This circuit looks badass I may build one too if I can get the parts.

#### CharlieLaub

Paid Member
What about the little issue of the output voltage from your insano circuit being zero for most of the input cycle???

-Charlie

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.