USB Power

[daneck]

I am working on a small commercial device that needs 9v @ 200mA that would only operate for 1 minute each day.  My choices seem to be a 9v battery, 9v AC adapter, and a USB AC adapter.  Considerations are:

   9v battery - may not provide the 1 year life that I would like

   9v AC adapter - OK, but there is such confusion on plugs and sockets for these that I want to avoid them

   USB AC adapter - I see plenty with 2A and even more capability, but USB power is apparently only 5v.  Is that correct?  What is the practicality of using USB power other than for charging?  (I use it for a higher current keyboard which works fine).  Can you leave out input polarity protection when using USB charging?  I could redesign with 5v components (mainly a relay) if necessary.
   
Thanks!

[nickn4]

So you would need 0,0033 Wh per day.
can you use reachable battery's as well? for example 3 Li-Po Cell. it will last for a couple of weeks

if device needs to be mobile + recharge, i'd say just AC adapter.
i never liked the adapters where you can also unplug the adapter and just power from USB (that's what you mean by USB ac adapter correct?).

can you tell us more about the device?
edit; you mentioned you could redesign for 5v. will power consumption be the same?

nick

[mariush]

One thing it's not clear is how stable the voltage has to be. Any kind of battery will slowly discharge and as a consequence the voltage will also drop.

9v batteries aren't designed for such high currents, they will discharge fast if you pull 200mA for even only one minute a day.  Also, a new 9v battery will be close to 10v but as it discharges it will go down to 6-7v quite fast. If you're going to use a 9v battery you'd have to design your product accordingly.

Adapters... you say AC adapter but you probably mean DC adapters... basically the difference is the first ones output 9v AC vs 9v DC so if you want to make your product support both types you'd have to add a bridge rectifier and a capacitor in your product.
AC or DC adapters will be more regulated but may still output more than 9v when your product idles and go down to 9v during that one minute, under the 200mA load.

USB adapters are basically ac to 5v dc converters, and the majority of them will output 4.5v-5.5v and are capable of at least 400mA. Quality ones can do up to 1-2A.  You can redesign your product to work directly with 4.5-5.5v or you could simply add a boost dc converter to boost the low voltage to your 9v. DC-DC boost regulators can do over 85% efficiency, so to get 9v @ 200mA, you'd pull less than 450mA from USB.

If you are going to introduce a dc-dc converter to boost voltage, you'd be better off resorting to using 2-4 AA alkaline batteries or a ~3.7v lithium battery and boost the low voltage to 9v when needed. Optionally, you could also implement charging the internal battery from usb, such charging/management chips cost under 1$ and are very simple to implement.

[daneck]

Thanks for the feedback!

Yes, there seems to be lack of consistency in describing AC to DC adapters.
Most I have seen say something like "9v AC Adapter" and mean "9v AC-to-DC Adapter".

The device basically has a 12v DPDT relay operating a 12v solenoid valve.
Operating threshold is about 8.4v to actuate and draws about 200mA at 9v.
It can actually use between 9v and 12v DC.
Voltage stability is not an issue, just has to be between the threshold and 12v.

Switching to 5v parts seems more expensive with lower availability than 12v parts.

I think I will stay with a simple AC-to-DC adapter and include one with matching plug/socket with the product!

Thanks again!

[PeterFW]

The device basically has a 12v DPDT relay operating a 12v solenoid valve.

So.... you are using one electromagnet to operate another electomagnet?
Does it have to be a relay that operates the valve?

There are many solutions that are more reliable then a relay and need less power.

[MLXXXp]

You should also consider marking the product itself with the voltage (including whether it's AC or DC), maximum current, and the polarity of the plug. If the voltage is unregulated, specify the accepted range. That way if the adapter is misplaced or fails and needs to be replaced, you know what to look for if it's not available or difficult to get from the product manufacturer.

For any product that I buy that doesn't have such markings, I mark it myself from the information on the adapter, but it's nice to have it there in the first place.

[nickn4]

to me it seems also seems your idea is not optimized that great yet. just like peterFW said: "There are many solutions that are more reliable then a relay and need less power."
If you could give us an functional description and a schematic i would be happy to help you out.

nick

[daneck]

I'm open to suggestions!

The circuit is for maintaining water level in a tank between upper and lower levels determined by two float/reed switches.  I could have simply used a mechanical water level valve but it would run too often to refill small amounts.
I could have used a flip-flop for the latching and a semiconductor driver for the solenoid, but that seemed overly complicated, admittedly at the cost of using more power.

Here is the schematic, where
   J1 is power jack
   J2 1 & 2 are connections to upper water level switch
   J2 3 & 4 are connections to lower water level switch
   D4 is reverse polarity protection Schottky
   K1 is DPDT mini relay
      S1 part of K1 is latch to keep circuit hot after activated by lower level switch until upper level switch shuts off
      S2 part of K1 is power for J3 which connects to water valve solenoid
   D1 and D2 are solenoid back EMF diodes
   D3 LED & R1 are for on-state indication

How would you "optimize"?  Thanks!

[daneck]

You are right, I could use just an upper level switch, but in that case I could just as well use a simple float valve with no electronics.
The problem is that it would run too frequently and I would rather have it operate in fewer large "gulps" than in many small "sips"!

[flynwill]

For this application I would just stick with a 9-12V "wall wart".  If you search any of the major suppliers I'm sure you'll find something suitable for well under $10.  9V battery is probably not a good choice, even if your solenoid valve will pull in at 9V, it probably won't pull in at 6V which is where the 9V battery will be at when it's time to replace it.  If battery operation is desired then I would go with 2-4 AA or AAA cells and a small boost converter.

Your relay could probably be replaced with a couple of transistors, which would like likely save some cost (your first post seems to indicate that you'd like to sell many of these, not just build one for yourself).   If you go battery operated then some care would be needed in the design to make sure that the power draw is zero or near zero until that "fill" reed switch closes.

[daneck]

Thanks, everyone!

Great suggestions.

I think I'll take a shot at replacing the relay with transistor logic and driver.

[mfratus2001]

Another consideration is looking around for existing standards. The security camera industry has a standard 12 Volt supply, and Guitar Effects pedals have a weird-but-standard 9 Volt supply.

[PeterFW]

Another consideration is looking around for existing standards. The security camera industry has a standard 12 Volt supply, and Guitar Effects pedals have a weird-but-standard 9 Volt supply.

If it is powered of the grid you can just put a rectifier on the input, and a DCDC converter behind that has a high input voltage tolerance.
For the jack use a socketed screw terminal and just can put anny type of wall wart on there regardless of voltage and polatiy.
Wich a small cap you can even use AC adapters that makes it "9 to 48V AC/DC@200mA"

If you only use a relay you could even use a bigger shottky and a resistor instead of the DCDC converter.

[Ian.M]

Its a control module that is going to be installed as a fixed installation by plumbers/electricians.  A wallwart PSU is unprofessional and may not meet applicable codes.    I don't know how popular DIN rail mounting systems are over in the states, but over here in the EU, DIN rail mount modules are extremely popular for this sort of application.  PSU modules (both AC and DC) are readily available.  I'd use a Schottky bridge rectifier on the input to your circuit so it can accept 12V to 24V AC or DC, with enough bulk decoupling after it, regulate that down to 9V and use a 9V relay with 240V AC / 24 V DC rated contacts for the output.    That gives the maximum flexibility as the user can simply run it off whatever power supply they need for the valve or off a cheap DIN rail bell transformer if the valve is line powered.  For non-DIN rail applications its easy enough to provide a screw down plastic moulding to DIN rail profile for the module to clip to.

[daneck]

The big advantage of a DC Adapter is that I can get a "UL Listed" adapter and avoid the high cost of having my device undergo UL approval, which I should have when connecting to AC mains.  (Is that what you mean by "rail" mount?)

[Ian.M]

http://en.wikipedia.org/wiki/DIN_rail, usually the 'top hat' profile rail.

Your unit would be built in a DIN rail compatible case, have screw terminals for low voltage power, 24V AC or less and use an UL approved DIN rail PSU which are readily available: http://www.mouser.com/Search/Refine.aspx?N=18357235

Of course there would be nothing to stop a hobby user from powering it with a wallwart if they are too cheap to do the job right.

[flynwill]

I think it depends on who daneck's customers are.  You are correct that if this is targeted at industrial control or part of a large system in a commercial building DIN rail (or DIN rail option) would be good.  If the target is homeowners to keep their fountain or fish tank full then a self-contained box with a US-listed wall wart is probably the way to go. 

[Jeroen3]

With a simple DC/DC module (from Recom for example) you can get 7 to 42 volts to 9 Volts.
With a simple bridge rectifier you have polarity independence.

http://www.recom-international.com/pdf/Innoline/R-78Cxx-1.0.pdf

[Ian.M]

Don't forget the solenoid valve (or pump) also needs power, and if the controller is not a UL listed device, that means the valve (or pump) needs a low voltage PSU for as well (because it isn't rated to directly control a mains powered valve or pump).   Putting a DC jack on the product rather than screw terminals makes it harder to run the valve or pump off the same supply so now you need two wallwarts, which is starting to get clumsy.   It's probably better to offer a  wire ended wall wart with enough current to power the controller and a solenoid valve as an option, or at least offer the module on its own to pro installers.

The other thing that needs looking at is the options for level switches  Float switches can be troublesome and electronic tank level sensors are becoming increasingly more popular. 

You will also double your market if the switch logic can also be arranged for applications that empty the tank.  Avoiding short-cycling backup sump pumps comes to mind.

[daneck]

I think this thread is diverging!

The product is for consumers and the DC adapter will work fine.  The problem of multiple "standard" plugs/sockets is simply solved by providing a mating adapter with the product.  USB 5v power is probably not practical because solenoid valves of interest are over 5v.

By redesigning with transistor latching logic and a transistor solenoid driver in place of the DPDT relay, I cut the power requirement almost in half.

This has been an interesting project, though.

I thought the liquid level detection would be the most problematic and got into various types of sensors (Hall effect, light dependent resistors, reed switches activated by floating magnets, variable resistance "sticks", sonic detectors, etc.) and finally settled on simple float/reed switches.

I still have to consider reliability and failure modes, recognizing that the most failure resistant approach may not necessarily have the least power consumption or lowest cost.

Thanks for all the help!