Senior project - Power Saving Adapter

[fender7802]

Hi guys and girls,

New forum user here, but a big fan of the show. In fact I got my idea for this project from a video Dave had about the 6W standby power on the InfiniiVision scopes. I guess I'm another one of those people who waits to make an account until the very moment they need a life raft. 

Well, I'm a senior majoring in electronic engineering technology which means I've been working on a final design project. I thought I had the project going smoothly until I hit a bit of a road block earlier this week. I've come to you all to ask for any help which might get me back on track.

My project is called the power saving adapter. The idea is to take any appliance with high standby power (also called phantom power) and reduce it as much as possible. I designed a circuit which would do this by switching in 12Vac instead of 120Vac when the appliance is turned off. That way the voltage would be reduced by a factor of 10, the current by a factor of 10, and the power (hopefully) by a factor of 100.

Of course theory never matches up to reality and I'm now realizing that this idea does not seem hopeful.  Let me just elaborate on the functioning of the circuit a bit to make my problem more clear. An appliance (let's say an electric heater) is plugged into my project, and then my project is plugged into the wall. The heater is then turned off. Now there is a current sensor in series with the heater. A push button is connected to a microcontroller which stores the value from the current sensor into the memory of the micro. So the "off" current of the heater is now stored in memory. Whenever the microcontroller is reading the value of the heater current to be equal to the value of the "off" current, 12V is switched in, and power is now being saved in the heater's standby mode. Now when the user turns the heater on, there will be an increase in current. The microcontroller sees this increase in current which then switches 120V back in so the heater functions normally. Make sense?

Well, I stupidly waited several months to test my idea of feeding an appliance 12V. As I should have guessed, the appliance makes no attempt to turn on when it is only given 12V. There is no increase in current at all. Without that increase, there is no way to tell the microcontroller to switch back to 120V.

If anyone's curious here's what the circuit looks like:

So, I'm not sure what to do next really. Shit-canning the whole project isn't an option. I was wondering if there's some kind of chip which I could put in series with the output of the transformer which would somehow limit the current and not dissipate much power. That way I could still send 120V to the load, but at a much lower current. Maybe some kind of transistor configuration?

Anyway, thanks to anyone who reads this and has any kind of suggestions at all. All replies are welcome. If you would like me to elaborate on anything just let me know.

Cheers!

[c4757p]

I'm about to be a senior in BSEE (this fall) and frankly I'm baffled that this project was approved.

You're going to have to change a lot. The idea of reducing power consumption by doing something when the appliance is switched off, by sensing a drop in current, is reasonably sound. I just can't imagine that being anything short of "switch it off".

So, I'm not sure what to do next really. Shit-canning the whole project isn't an option. I was wondering if there's some kind of chip which I could put in series with the output of the transformer which would somehow limit the current and not dissipate much power. That way I could still send 120V to the load, but at a much lower current. Maybe some kind of transistor configuration?

Voltage and current are not exclusive; one is always a function of the other.

[fender7802]

To be honest I'm surprised my project was approved as well. But it was and I need to stick with it.

About the voltage and current not being interchangeable, my idea was coming from something along the lines of a power supply with a constant current mode. In that way you can change the maximum current delivered by a source.

[c4757p]

A constant-current power supply is just a constant-voltage power supply with a separate control loop: if the current goes too high, it drops the voltage until it's not too high. Unfortunately, there really is no way to vary one without varying the other. If you choke off the current supply to the appliance, the voltage will have to drop too.

[fender7802]

Makes sense, c4757p. Thanks for the explanation on that.

[c4757p]

No problem dude. I'm sure you can figure something out. What about if you could shut the appliance down completely, and then have some other way to know when it's time to turn back on?

[fender7802]

It would be preferred to shut the appliance down all the way. But then we're kind of in the territory of designing a plain old switch. It's silly because that's really the best solution to the problem - pull the plug. We're basically trying to simulate pulling the plug without actually doing it. That's why we kept some standby power - so we could sense when it was being turned back on.

I'm trying to think of a solution based on what you said, without getting overly difficult, either. We don't have the option of tearing open the appliance. It needs to be something you can just plug the appliance into. I was thinking maybe if there was some kind of a wireless sensor that you could stick on the appliance. Maybe a pressure sensor that could send a signal to the MSP that it's time to turn the appliance on. You press the sensor instead of pressing the heater's on/off switch. So it would sort of be a wireless mains switch. I don't know, is something like this on the right track?

Thanks.

[tom66]

Here's a possible suggestion.

One thing TV technicians and other electronics technicians do when bringing up a PSU after doing a repair is to use a series lightbulb jig in place of the fuse. The idea is that the bulb will limit fault current considerably in the case of a short.  Plus you can get cheap 100W resistors this way - granted they have a tolerance of about 50% and a huge TC but that really doesn't matter.

What do most devices do when they have a bulb in series with them? They tend to cycle on and off, fairly rapidly, because as soon as current is drawn the bulb acts as a resistor, dropping the supply voltage often to the point of the device "browning out".

The idea here is to replace your 1:10 transformer (which by the way will consume a couple of watts idle, many times over that of most standby power, which in the EU is now limited to about ~0.5W and due to come down further) with something which will actively limit current. The bulb acts as a resistor but you don't want that - it's just wasting all that power that you'd otherwise save.  An alternative is to use a large ballast inductor, which will also limit current but won't waste it as heat (at least not as much.) Such an inductor will need to be very large (many henries), and the running losses are likely to exceed that of any standby power usage, but....

One thing to be aware of if you open circuit that inductor you're going to get a large voltage spike so you need a way to catch that and prevent it from nuking attached appliances.

Ultimately if anything this device is simply going to waste more power than it ever saved. Most power supplies are more efficient at high line voltage, especially simple standby switchers. Keeping the supply voltage low will put the converter well out of its desired operating range, which will mean it will be consuming a lot of power.

Most modern LCD, LED & plasma TVs I've tested use about 0.1W to 0.2W in standby. And this is lighting an LED, running a processor, keeping the time & waiting for a remote control signal. I've seen a recent power supply with 0.00W standby (i.e. <10mW) when the power indicator was turned off via a push button on the side of the unit (although it would still respond to remote commands.) Scoping the output showed the converter switched just three times a second to maintain the output. If your TV uses 100W switched on, then leaving it on for just 1 minute more than necessary uses the same power as 1,000 minutes of standby (that's almost 2/3rds of a day.)

[fender7802]

Thanks Tom. I think that's an interesting suggestion about the inductor in series.

From my experience I've seen much high standby power levels. For example, this electric heater I have is 12W standby. Agilent has a signal generator, the E4432B which is also 12W standby. Where I live that's $20 a year, so it is significant. I've also tested oscilloscopes, multimeters, dehumidifiers, all of which were over 6W. I don't think it would be out of the question to make a device which reduces that standby power, even taking into account that device. I had the two relays attached to the transformer so that it would be off completely when not in use, so I am conscious of this problem.

If anyone can comment on my idea about using a wireless sensor, I'm starting to think this wouldn't be a bad idea. Basically I just need a transmitter which when I press it, the receiver at the MSP gets a "1", and when I press it again, the MSP gets a "0". I would think bluetooth would be the best choice. Can anyone provide a suggestion?

[tom66]

How does an electric heater have standby power... surely it's just a thermostat and heating element?

If it has logic/timer/clock/etc, remember it probably has a capacitive voltage dropper. The real power usage (watts) of such a dropper can be less than 1W but the power factor is very poor so it can fool a lot of meters.

Your measurements are the exception rather than the norm. I believe it's a legal requirement now to meet some Energy Star standards such as less than 2 watts in standby power usage for US products. There might be exceptions for test equipment or certain industries. Average consumer gear now regularly achieves less than 0.5W due to "Blue Angel" in European products, which bleeds over into US market because it's easier to make one PSU for all models.
http://www.eceee.org/policy-areas/EEES/public_sector/GermanyAppendixPROST.pdf

For a transmitter you can use those simple 433MHz modules - data goes in and comes out the other device. Available for a few bucks each. No channel separation, and quite vulnerable to interference so you might want to transmit a code rather than an on/off state.

[fender7802]

Thanks Tom. I've been thinking about what you said about it being a reactive load. I took out my old textbook but I'm kind of having a hard time jogging my memory on how this stuff works.

But as for the fan, we put an ammeter in series with the wall outlet and the fan. All I can say is that the multimeter reads 0.1A. I just multiplied this by 120V to get 12W. I realize this could be wrong because the actual formula is P=cos(theta)*V*I, and I'm assuming voltage and current are in phase. If they weren't in phase, that would mean that we're not getting an accurate reading of the standby power, correct? Does the 12W mean anything? Would I need a wattmeter?

Thanks.

[tom66]

You're also assuming the offset current of your meter is zero. It isn't.

If something is using 12W and not doing something useful (e.g. spinning) then it's probably going to get quite warm. Perform a sanity check. Is anything warm when it's off? If not, then there might be an error in your measurement.

[Stonent]

I always thought it would be a good idea to put a small button inside a USB based charger, so wen the cord is plugged in, the button is pressed and turns on the charger, without that, the charger is completely off.

[fender7802]

Good point. But I wouldn't think the offset current could be that high. It's a Fluke 87-V (not that that means it doesn't have an offset current). But thinking back to Dave's video about the scope, we are getting a similar figure for that same scope so I don't think I'm too far off.

Any other ideas? I realized that that sensor idea may not work. When the sensor sends 120V back to the appliance, the appliance is still turned off, so you'd have to press the power button on the appliance too.

The only way that could work is if I put the sensor directly on top of the power button of the appliance. That way when you turn the heater off, you're cutting all power to it. When you turn it back on, you're switching 120V back to it and turning it on at the same time. The problem is finding a transmitter with a form factor that I can fit over the power switch.

[Jay_Diddy_B]

Fender7802 and the group,

Here are couple of ideas for 'saving' your project.

1) You could study, and document, the standby power drawn by the devices as a function of line voltage. This would require a variable transform and a power meter.

What you will probably find is that below a certain voltage, the Switching supplies (SMPS) will draw nothing. They typically have an under-voltage lockout circuit. The UVLO is to prevent excessive current draw at low input voltages.

You will also find that above the UVLO threshold the device consume constant power. If you assume that the SMPS is 80% efficient. The input power will be the output power +20%. If the output power is fixed the input power will also be fixed (Give or take changes in efficiency with line voltage).

If you do a teardown, you will probably find devices like this: http://www.powerint.com/en/products/tinyswitch-family

Although this is a negative result, it is a result.

The second thing to do, is make a power strip with one of the outlets the master. Plug the most important device into the master outlet. If the current consumption on the master outlet increases, turn on the other outlets.

For example I could use the TV to turn on the power to my DVD player. The problem with this scheme is that my 'pesky' Digital Receiver looses its menu if it is powered down. But there are some applications that work, turning on a computer will turn the printer into the standby state.

Just a few ideas for you...

Regards,

Jay_Diddy_B