Shop OBEX P1 Docs P2 Docs Learn Events
Silly Question on power switches for boards w/props — Parallax Forums

Silly Question on power switches for boards w/props

TymkrsTymkrs Posts: 539
edited 2014-06-16 20:56 in Propeller 1
Hey all,

I'm still working on a board design involving a Prop (what else?!) and I'm trying to find a proper power switch. According to the PropBOE BOM, it's using a slide switch that's good for 0.1A at 12V - but should I be using a 0.3A at 12V instead given that the Prop can source 300mA at its max? (ideally maybe even 0.4A). Just wanted to see what others have been using. I've been looking for switches on digikey, mouser, and octopart, and as soon as you hit the 0.4A range the price goes way up, or if you're at the 0.3A mark and try to go from 5V to 12V, the price goes up and availability goes down....

Thanks for any suggestions!
Addie

Comments

  • TymkrsTymkrs Posts: 539
    edited 2014-06-12 05:11
    Thanks Basil, I should have mentioned that I was looking for SMT options :p. I've used the filters on digikey before, but say from your list that you found, smt options are in the 6-9 dollar range per switch *yikes*

    Addie
  • BasilBasil Posts: 380
    edited 2014-06-12 12:57
    Ahhh I see. Maybe use a cheap low current switch to turn on a cheap FET to switch the higher current?
  • TymkrsTymkrs Posts: 539
    edited 2014-06-13 15:31
    Ah yeah, I did that for the solenoid, but seems like a lot of steps for a power switch - ie, it's just as easy for people to unplug it from a wallwart...
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-06-14 22:37
    The simple reality is that these low current slide switches that you might find in an SMT format are not really durable.

    After trying to find something that is both durable and small, I gave up and just use a toggle switch that is on the power cable and eliminate trying to put a tiny power switch on boards.

    I know this is not exactly what you want, but it just might be the best way to build. I even distrust the longevity of the slide switches that Parallax puts on their Propeller Proto Board and expect that eventually I will have remove and bypass the devices when they fail. (We have a lot of corrosive air pollution in Kaohsiung due to a local steel industry and high humidity.)

    I also have trouble with the tiny Reset buttons failing, and the DIP slide switches. On board I tend to stick with jumpers instead of switches.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2014-06-14 23:47
    I'd favor the original low-current switch that grounds the pulled-up gate of a pMOSFET. It's only two more very cheap components.

    -Phil
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-06-15 01:22
    Using a small low current switch with a MOSfet is likely better than tryiing to provde a 400ma switch for a 300ma load.

    The problem with low current mechanical switches the same as with the contacts on a mechanical relay. The type of load might significantly accelerate their failure. Often a relay is de-rated as much as 60% or 80%, to assure the contacts will survive over the long-haul. So a 300ma load, might be better accommodated with at least a 1 amp mechanical switch. And a tiny low current switch has a much better chance at a long life if it is just toggling a MOSfet that handles the real power.

    In a MOSfet situation, even a DIP slide switch might actually survive. Like these ===> http://www.goldmine-elec-products.com/prodinfo.asp?number=G17778
    And if the switch is in a DIP socket, it is easily replaceable.

    The DIP slide switches do come in a single switch unit, AND you can bend the tabs flat and outward to solder as surface mount. Try here +++> http://ex-en.alliedelec.com/search/productdetail.aspx?SKU=70128458

    ++++++++++
    There is the additional issue of miniaturization creating greater sensitivity to ageing and wear.
    So I just prefer using a traditional switch with a proven track record that is very over-rated for both volts and amps. Not only is the longevity more assured, these devices are much cheaper. You can get them at places like Electronic Goldmine. And not everything has to be on the circuit board... a lot of creative and better solutions can be resolved with wire.


    http://www.goldmine-elec-products.com/departments.asp?dept=1180

    Everything that is new and miniature is not always the best solution for durable use.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2014-06-15 14:43
    I'd favor the original low-current switch that grounds the pulled-up gate of a pMOSFET. It's only two more very cheap components.

    -Phil

    This would mean there's always a little bit of current flowing through the pull-up right?

    I don't suppose I could talk you into a link to an appropriate pMOSFET? I confess to not understanding many of the search parameter when attempting to find one at Digi-Key. I got the results down to 38 pages but I really have no idea (or very little idea) of which would work well for this application.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2014-06-15 15:04
    Duane Degn wrote:
    This would mean there's always a little bit of current flowing through the pull-up right?
    No. When the gate is pulled up to the source voltage, there's no differential across the resistor, so no current. There's a tiny bit of current flowing when the gate is grounded, but the resistor can have qute a large value, since the gate itself draws only a minute amount of leakage current.

    -Phil
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-06-15 23:59
    Well it seems a 2n7000 mosfet is rated at about 200ma tops, so it won't quite make it unless you put two in parallel.

    What would you suggest for a good mosfet in this situation that can handle up to 1 amp?
    And I was thinking a 10K or 50K ohm pullup resistor. Is that about right? Any smaller seems to always give me trouble with sensitivity from outside influences (touching the board, inductive loads nearby, and so on).

    Is this good? http://www.onsemi.com/pub_link/Collateral/MMFT3055VL-D.PDF
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2014-06-16 00:22
    The 2N7000 and MMFT3055VL are nMOSFETs, Loopy. The power switch needs to be a pMOSFET, such as this one. A 100K gate pull-up would be more than adequate.

    -Phil
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2014-06-16 00:27
    Tymkrs wrote: »
    Hey all,

    I'm still working on a board design involving a Prop (what else?!) and I'm trying to find a proper power switch. According to the PropBOE BOM, it's using a slide switch that's good for 0.1A at 12V - but should I be using a 0.3A at 12V instead given that the Prop can source 300mA at its max? (ideally maybe even 0.4A). Just wanted to see what others have been using. I've been looking for switches on digikey, mouser, and octopart, and as soon as you hit the 0.4A range the price goes way up, or if you're at the 0.3A mark and try to go from 5V to 12V, the price goes up and availability goes down....

    Thanks for any suggestions!
    Addie

    I think you started off on the wrong foot when you said that the Prop can source 300mA max, that is NOT an operating figure, these relate to ABSOLUTE MAXIMUM RATINGS, a bit like red-lining, something might happen (or stop happening). The datahsheet says in this section "Functional operation of the device is not implied at these or any other conditions in excess....".

    Maximum current that the Prop itself would draw with all 8 cogs @80MHz running flat out would still be under 100ma. A power switch that's rated for 100ma is good for 100ma. The other thing is that if your power supply utilized a switching regulator then the actual current drawn at 12V would probably be more like 50ma.

    BTW, a very compact and cheap switching regulator is this one here.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-06-16 04:38
    Hmmm.
    Peter J always brings something new to the table and sets me thinking.

    One of the quirks of living in Kaohsiung is that the only 3.3VDC linear regulars that I can buy locally are a 4 wire device that has one extra pin used to Enable the device. I had never thought about it before, but the devices come in 5.0V as well as 3.3V. And they certainly can use a 100ma switch to toggle their 1 amp capacity.

    http://www.onsemi.com/PowerSolutions/product.do?id=NCP691

    Here are a couple of tiny ones .. 150ma and 350ma, but I have some 1.0 amp ones amongst my clutter.

    http://www.onsemi.com/PowerSolutions/product.do?id=NCP600

    http://www.onsemi.com/PowerSolutions/product.do?id=NCV8772
  • GenetixGenetix Posts: 1,754
    edited 2014-06-16 09:10
    Addie,

    I would use a through hole version of a slide switch so customers don't complain about issues with it coming loose or off. Also, most switches are rated for higher voltages so they will handle more current at lower voltage. You may also see ratings for peak (highest for a moment), average, and continuous (same for a long time) voltage.
    Many companies, such as Cherry, offer samples and will help to find you what meets your needs.

    A switch is one of those items that gets a lot of use so spend a little more for a quality component.
  • Tracy AllenTracy Allen Posts: 6,664
    edited 2014-06-16 10:21
    I agree with the advice about using a thru-hole switch if possible, for stress resistance.

    I like is the e-switch 200 series toggle switches, and the 200U in particular. Not only is it small, it is sealed to IP67 (dustproof and immersible to 1m depth). That allows it to face up to a lot of environmental challenges. It is a dry-contact switch. That is the kind of low-current switch we have been talking about, and you will see a lot of them rated at 0.4VA. What that means is that you multiply the current the switch will be carrying (especially at the instant it needs to be turned off) times the voltage it will have to sustain as it opens.

    Dry-contact switches have a thin gold layer over other metals, typically gold then nickel over a springy base such as beryllium copper. If the circuit stays within the 0.4VA rating, an arc will not form as the switch opens, and the thin gold layer will stay intact for the ~10000 activations typical of these switches. Switches rated for higher currents may have silver contacts, and oxidation is not so much of a problem due to the high conductivity of silver oxide. Or they use a wiping action to keep the contacts free of corrosion. Some even incorporate a magnet to help suppress the arc.

    For a pmosfet, I like the FDN304, ID=2.4A and Vgs=0.8V.
  • TymkrsTymkrs Posts: 539
    edited 2014-06-16 17:54
    Hi all,

    Thanks for your various suggestions of wisdom! Sounds like a throughhole is the simplest solution, but I will carefully read through everything folks have been suggesting in terms of using pmosfets. That said, I may just need to hide the solderpoints with a lrf!

    Addie
  • GenetixGenetix Posts: 1,754
    edited 2014-06-16 18:41
    I think Kapton tape is non-conductive.
    I like to clinch or bend the leads 90-degrees so they're parallel to the board and trim the lead where the pad ends. Removing the part later though can be a hassle.
    I also prefer eutectic or 63/37 Sn-Pb solder because the joint hardens immediately after the iron is removed.
    Some people bend the leads slightly to keep the part in place while they solder it and then trim the excess later, usually near the end of the solder cone that's formed.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-06-16 20:00
    The 2N7000 and MMFT3055VL are nMOSFETs, Loopy. The power switch needs to be a pMOSFET, such as this one. A 100K gate pull-up would be more than adequate.

    -Phil

    Thanks. In any event, the MMFT3055VL is apparently discontinued. I forgot about high side switching versus low side switching.

    I still think that just using a voltage regulator with Enable is a solid solution.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2014-06-16 20:56

    I still think that just using a voltage regulator with Enable is a solid solution.

    Yes it is, but from my experience solid or simple solutions (IOW best) are rarely followed or applied in this forum :) ... :(

    BTW, many shutdown regs have fairly low absolute maximum input voltage ratings so this needs to be checked.
Sign In or Register to comment.