Shop OBEX P1 Docs P2 Docs Learn Events
Looking for a mosfet recommendation (SOLVED) — Parallax Forums

Looking for a mosfet recommendation (SOLVED)

mparkmpark Posts: 1,305
edited 2020-07-22 06:33 in Propeller 1
Hi all. I need my P1 to switch a load on and off, and I thought I'd try using a mosfet for a change. So I don't know from mosfets, but I'm using this as my guide (from learningaboutelectronics.com/Articles/N-channel-MOSFET-switch-circuit.php):
N-channel-MOSFET-switch-circuit.png
I figure I can replace the 3V gate input with a Prop pin and replace the 6V buzzer with my 5V device. Does that sound reasonable so far?
Now the aforementioned web page recommends the 2N7000 mosfet, which IIRTDC has a maximum drain current of 200mA. Unfortunately, the circuit I want to control draws ~250mA, so the 2N7000 is out. The related NDS7002A could work, but I'm only finding surface mount packages.
I'm hoping someone can help me find something like a 2N7000 in a TO-92 that can switch maybe 300mA to be on the safe side.
«1

Comments

  • Try an IRF510, works best with a gate driver, you can use a simple 2N3904 to drive the gate
  • When it comes to TO92 packages and switching light loads at low voltage then I would always use a plain old BJT. There are plenty available that have high gain and can handle 1A with low Vce saturation. One example is the ZTX618 which handles 3.5A continuous with very low Vce(sat) < 100mV and typical current gain of 300.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2020-07-20 19:53
    DigitalBob wrote:
    Try an IRF510, works best with a gate driver, you can use a simple 2N3904 to drive the gate.
    The 2N3904 is an NPN transistor. By itself, it will not adequately drive the gate of an IRF510 to a high enough level. To drive the gate to a voltage higher than 3.3V, you will need both that and a PNP transistor.

    Here's a schematic:

    npn_pnp_mosfet.gif

    Incidently, I don't recommend going this route, unless those are the only parts you have on hand. It would be much better to use a gate driver like the MIC4427-series. But if the load is very light, you could get by with just a ZVN4424A MOSFET without a gate driver.

    BTW, be sure to connect the drive voltage negative supply to ground right at the MOSFET pin. You do not want load current flowing through any logic ground wires or traces!

    -Phil
    586 x 320 - 4K
  • I used the 2N3904 for a gate drive on a IRF510 step driver, but in my situation the suppy voltage was about 12 volts and worked it like a champ. Can't argue with success.
  • In my appication the 2N3904 switched the IRF510 off, the gate was pulled high with a 10K.
  • JonnyMacJonnyMac Posts: 8,912
    edited 2020-07-20 21:53
    If you've got a couple circuits, the TC4427 -- which you can get in a DIP-8 package -- is a nice way to step from 3.3v to your desired gate voltage.
  • DigitalBob wrote:
    I used the 2N3904 for a gate drive on a IRF510 step driver, but in my situation the suppy voltage was about 12 volts and worked it like a champ. Can't argue with success.
    Okay, if you were using the gate pull-up resistor to drive the IRF510 gate, that could actually work, although turn-on time is compromised a bit by the gate capacitance -- as is turn-off time in my circuit, quite frankly.

    -Phil
  • In my stepper driver it wasn't that critical.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2020-07-20 23:37
    What's all this about? Isn't it just a buzzer? Normally I'd just be using a tiny sot-23 digital npn to drive it. There are thousands of bjts in TO92 that will do this job very nicely. One resistor and one transistor. If however you really really want to use a MOSFET for this low voltage low current circuit then just drive it directly from the Prop as there is no need for fast repetitive high-current switching that requires a gate driver.

    P.S. I know the thread title is " looking for a mosfet" but the application is driving a 250ma low voltage buzzer. If I saw a gate-driver + MOSFET in a pcb driving a little old buzzer, I'd laugh.
    There are only hundreds of npn bjts in to92 that you can choose from these days, rather than thousands, however the 2N5550 is cheap and handles 600ma at <100mv Vce(sat). I just remembered that I used to use BC337s for this kind of thing back in the day (even when I used mosfets for other things), they handle 800ma and were a good all-round npn.
  • jmgjmg Posts: 15,140
    mpark wrote: »
    I'm hoping someone can help me find something like a 2N7000 in a TO-92 that can switch maybe 300mA to be on the safe side.

    TO-92 is very much trialing edge, but if you go to Digikey and search for through-hole with VGS <= 2V, you find the cheapest is NTD3055L104-1G in IPAK

    If you look for easy to use SMD packages, the BUK98150-55A/CUF in SOT223 is cheaper than any thru-hold alternative,
    Older parts tended to not target explicit specified 3v3 operation, and FDT439N on SOT223 is well spec'd (RDS(on) = 0.058 Ω @ VGS = 2.5 V) but not so easy to find.

    It may be that packages like SO-8 are more a middle-ground, of easy to solder but modern enough to have 3V drive spec's - here, you may find dual-mosfets are more mainstream.

    eg The DMN2028USS in SO-8 has 2.5V drive spec and includes ESD diodes. FDS9926A & NTMD6N02 are dual fets in SO-8

  • mparkmpark Posts: 1,305
    My high school Latin teacher told us that the old Roman orators would use very sophisticated language when addressing even uneducated audiences because the listeners, though they might not understand what was being said, appreciated the compliment of not being condescended to.

    So thank you all for assuming I would understand your replies! :) But yeah, no. The circuit in my first post appealed to me precisely because of its simplicity, 'cuz I'm pretty darn simple myself.
  • Give this a look

    NUD3124LT1G

    I used it to drive some 120 volt 10 amp relays and It took the place of 5 parts.
  • mpark wrote:
    The circuit in my first post appealed to me precisely because of its simplicity, 'cuz I'm pretty darn simple myself.
    Well, I know better than that! But the transistor I recommend above (ZVN4424A ) will work just fine in your circuit.

    -Phil
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2020-07-21 03:21
    @"Phil Pilgrim (PhiPi)" - that mosfet is only rated for 260ma continuous and has an rds(on) of 5.5 ohms. That means it would be dropping 1.375V and dissipating 340mW which would also affect its rds(on). Are you sure that this is the right part?

    I keep saying that an npn is far simpler, even if you do need one resistor.
  • mparkmpark Posts: 1,305
    I admit that not needing a resistor is why I want to try this mosfet idea, but I may well fall back to an npn :smile:
  • I’dI reach for a bipolar transistor too for that, and my go-to is the Zetex (diodes inc) ZTX10xx where xx is 49, 41 or 53. They are super beta npn transistors, which means it doesn’t take much base current to turn them fully on, and they come in the Zetex “E-line” TO92 size package, which means they can handle an extraordinarily large current without letting out the magic smoke. Complimentary PNP transistors are available in the ZTX11xx series.

    If the buzzer happens to be electromechanical (peobably not?), it will kick out nasty high voltage spikes.

    Diodes Inc has some nice Thru-hole mosfets with low thresholds, like good for 3V gate drive, but I don’t know about TO92 size. (DMT6009). Mosfets may seem simpler at first glance, but there are pitfalls for the unwary, complications.
  • I’dI reach for a bipolar transistor too for that, and my go-to is the Zetex (diodes inc) ZTX10xx where xx is 49, 41 or 53. They are super beta npn transistors, which means it doesn’t take much base current to turn them fully on, and they come in the Zetex “E-line” TO92 size package, which means they can handle an extraordinarily large current without letting out the magic smoke. Complimentary PNP transistors are available in the ZTX11xx series.

    If the buzzer happens to be electromechanical (peobably not?), it will kick out nasty high voltage spikes.

    Diodes Inc has some nice Thru-hole mosfets with low thresholds, like good for 3V gate drive, but I don’t know about TO92 size. (DMT6009). Mosfets may seem simpler at first glance, but there are pitfalls for the unwary, complications.

    +1 for the Zetex - I use these for those special cases for these very reasons. DIODES are now manufacturing ZETEX.

    As for "simpler" I agree, but I'd use the digital transistors in SOT-23 packs that have series resistors built-in. One tiny black dot on the board is all I need to handle 1A or more.
  • Cluso99Cluso99 Posts: 18,066
    edited 2020-07-21 05:37
    Why not try the trusty old 2N2222 / PN2222 ?
    TO-92 case, 600mA max, hFE 100 min at 150mA
    30c at Mouser
  • mparkmpark Posts: 1,305
    ...Mosfets may seem simpler at first glance, but there are pitfalls for the unwary, complications.

    As I'm learning in this thread ;)

    Btw, I'm not using a buzzer. That's just an example diagram I found on the interwebs.
  • mpark wrote:
    Btw, I'm not using a buzzer. That's just an example diagram I found on the interwebs.
    Okay, what is your load, exactly. Enquiring minds want to knowTM -- so we can better help you, of course!

    -Phil
  • Cluso99Cluso99 Posts: 18,066
    Your load is important to what is recommended.
  • mparkmpark Posts: 1,305
    Wait, so when I wrote "replace the 6V buzzer with my 5V device" everyone assumed "device" = "buzzer"?
  • A big ? back because it's all a bit vague and unless it's a national security matter, you should be able to tell us what the "device" is right up-front so that we can steer you in the right direction, instead of all over the place. There were plenty of opportunities to correct any misunderstanding and specify the device.
    However it doesn't matter much because at 5V or 6V you are not going to be switching any high power load. If however you are thinking of connecting inductive loads to the 5V rail then you will have to protect your whole board from inductive spikes which may also be radiated, but that depends upon the load.
    Is it a national security matter? Your call.
  • mparkmpark Posts: 1,305
    I'm mostly curious as to why people thought I was hooking up a buzzer. I didn't explicitly say it wasn't a buzzer, but I thought the implication was pretty strong. Apparently not, though? I'd love to hear people's thought processes.

    Basically, I tried to communicate something and failed, and now I'm doing a little post mortem to see where my judgment was off.
  • Anybody whose read all the words in your first post would know that you wanted to drive something other than a buzzer. However, a lot of people skim though the posts, and might not fully comprehend what you want to do. I have been guilty of that myself.

    It would be useful if you could provide some details about the device you want to drive, such as it's resistance, inductance or capacitance. An RLC model of the device would be useful.
  • Cluso99Cluso99 Posts: 18,066
    We all tried to help, but...
    No you've told us it's not a buzzer a number of posts ago. It really matters but you're not telling us. Do you want help???
  • Okay, time for 20 questions:

    1. Is it an inductive load?

    -Phil
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2020-07-22 00:51
    Ok, see if I can help because I wonder myself when things go askew as they did when everybody started to recommend gate drivers for a tiny 250ma 5V load :)

    In my first post I simply referred to switching "light loads" with a cheap and effective npn since that is the design choice I make for anything like this, especially in to-92 which aren't amenable to the larger die area of mosfets. Basically, the lower the rds(on), the more parallel elements in the mosfet array, the larger the die but to-92 is limited.

    However when I read the thread later on there were posts talking about gate drivers so prompting for more specific information I ask the question "What's all this about? Isn't it just a buzzer?" but got no confirmation or correction. Nonetheless, I wasn't too fussed about the load because what can you be switching other than a light load at 250ma at 5V?

    So all the while if you read the posts you could say "hang on, hang on guys, I'm just trying to switch power to another xyz circuit board, not PWM a motor" or something like that.

    All this because you toyed with the foolish idea of using a mosfet instead of a bjt to save one resistor :smile::smile::smile:

    BTW - you still haven't told us what "the device" is.

    If I ask Google for directions to some place, it normally assumes I am driving and might direct me out of the way onto a freeway or something. But I may be cycling and so I let Google know that too so that it gives me the correct answer. Nonetheless, I decide to take a shortcut through the bush instead.
  • Perhaps it may be RTFpost saying using buzzer circuit as a guide may not suggest a buzzer, but a big as the post schematic with a BUZZER labeled symbol would at a glance say buzzer to the average (context at a glance) reader. High WOT-age thread. Time to go see some celestial objects......
  • jmgjmg Posts: 15,140
    edited 2020-07-22 04:44
    .... especially in to-92 which aren't amenable to the larger die area of mosfets. Basically, the lower the rds(on), the more parallel elements in the mosfet array, the larger the die but to-92 is limited.
    The problem is not so much the package, but more the ancient nature of TO-92, which means no one bothers to place the modern MOSFET die into that package. The volumes are just not there.

    A SOT-23 sized SMD package can deliver down to VDS - 12V 10.7A (Ta) 1.2V, 4.5V Rds = 10mOhm @ 9.7A, 4.5V Vgs(th) = 800mV @ 250µA


    In BJT tech, the device would be something like BC337-40 - that's TO-92, and has good HFE so does not need very low base resistors when used with a ~300mA load.
Sign In or Register to comment.