5 ms·

There is nothing wrong with bipolar transistors. They aren't antiquated. FETs have definitely overtaken them in power applications (for good reason) but it seem

There is nothing wrong with bipolar transistors. They aren't antiquated. FETs have definitely overtaken them in power applications (for good reason) but it seems hard to displace them for small signal applications. Small signal FETs just aren't that plentiful or cheap. There are many instances where a small signal FET makes sense (level translators, etc) but the cost doesn't justify its use. Using a 2n2222 and the like is appropriate. On top of that, BJTs excel in many analog applications. There is a good reason why many analog IC companies make parts using a BiCMOS process. If you are doing a discrete analog design (generally for performance reasons) you often need the performance of BJTs or JFETs.

spott 10y ago
2n7000, BS270 are both pretty cheap for small signal FET applications. For analog applications, BJTs are still very much important, but for digital stuff (and especially power stuff), it is usually easier to grab a cheap FET than use a BJT.
Florin_Andrei 10y ago
The article seems mostly right when it's about interfacing logic circuits with loads that require even moderate currents. FET is king in this field. I've a little aquarium and I've noticed how fish always get scared when lights turn suddenly on or off. So I took an Arduino and wrote a timer app for it, using PWM to slowly drive the output up over 60 minutes (sunrise), keep it at max for a few hours, then drive it down slowly over 60 minutes (sunset). I used a MOS-FET to drive the power LEDs directly from the Arduino. Works fantastic, and the FET is cold at all times. A 2N3055 would have been pretty hot most of the time. I've been playing with electronics for a few decades now, so I can't be accused of change for the sake of change. The old stuff still works, sure, but there are new toys which are better in some cases.
- mrob 10y ago
  What frequency are you PWMing at, and are you applying any analog filtering? I frequently get annoyed by slow PWMed lighting, and I wonder how other animals with possibly faster vision perceive it.
  - Florin_Andrei 10y ago
    No analog filtering - that's how you keep things cold all the time. The frequency is either 500 Hz or 1000 Hz, I forget which (depends on Arduino type). I doubt there are any critters that can tell it's not continuous.
    - mrob 10y ago
      If I count as a critter, I can easily tell the difference by moving my eyes and seeing the phantom array effect. If you're already using resistors for current limiting then putting a big enough capacitor across the LEDs will filter them without much difference in power consumption. I have no idea if the fish will care.
      - paulannesley 10y ago
        That effect drives me nuts too; but it seems to bother a minority of people. I've always assumed/guessed it's in the order of 50–100Hz; the 500–1000Hz mentioned may be fast enough to appear continuous to sensitive humans and maybe fish...
        throwanem 10y ago
        The most usual source of that flicker these days is driving LEDs (usually in a string of Christmas lights) with half-wave rectified AC, which can be considered PWM with a 50% duty cycle over the same period as the input signal - 60Hz in the US and I think Japan, 50Hz most other places. That's slow enough to see easily; it would surprise me to learn that PWM at ten times the rate is slow enough to see at all. Looking at the code I wrote a while back to drive a four-digit common-cathode LED display (you know, a TV bomb timer!) as a clock, I find that after all the tuning I did to get rid of flicker, I ended up activating each digit for 1ms at a time. Granted, running under Raspbian this is going to flicker anyway from time to time because multitasking, but that aside, I found no PWM flicker perceptible, even at what is effectively 20% duty cycle over a 5ms period. (The decimals are implemented as a fifth "digit".) So I'm going to guess that .5KHz and up are fine for most humans. Fish I don't know about; (1) cites a flicker fusion rate between 2 and 40Hz for swordfish, but for goldfish I've no idea, other than to guess that .5KHz and up is probably fine for them too. (1) https://www.google.com/amp/amp.livescience.com/3799-fish-eyes-adjust-shutter-speed-track-prey.html?client=safari https://www.google.com/amp/amp.livescience.com/3799-fish-eye... (Sorry for AMP, but taking it off the front nets me a timeout from livescience.com. Something broken there, I'm guessing.)
- jacquesm 10y ago
  If you properly drive a 2N3055 (so all the way to 'on' or all the way to 'off') you'd have to put a pretty good current through it to get it to get hot. The trick is to use the highest voltage you can get away with (most power for the least amount of current). .6V * quite a few amps before you overcome the housings capacity to get rid of the heat. You'll need to properly drive it though and not at too high a frequency (and for obvious reasons, not too low either because you'll notice the flicker).
  - cr0sh 10y ago
    > before you overcome the housings capacity to get rid of the heat. The thing is, most of these larger BJT parts were designed to be used with heatsinks, not usually standalone. IMHO, there's nothing wrong with dropping some voltage and dumping some heat. Sure, it isn't efficient, but if you size everything properly, and your project can accommodate the heatsink - then where's the problem? Heck, a long time ago you could get tiny heatsinks for some of the small round metal-can transistors that existed - I've got an old 16K minicomputer core-memory board from the 1970s that has drive electronics set up like this. The author of this "rant" mentions not using L298 (and presumably the L293) either - but there's nothing wrong with these parts for hobbyists (outside of the L298's weird footprint, and the fact that heatsinks for them are almost unobtanium) - provided that the limitations are kept in mind (finding a low-cost mosfet 4A h-bridge in a thru-hole design is virtually impossible). These parts were originally meant for automotive applications, where having a 12-14V large amperage battery supply was the norm (people trying to use them for circuits with 6V or less always wonder why their small motors won't turn).
    - jacquesm 10y ago
      > outside of the L298's weird footprint, and the fact that heatsinks for them are almost unobtanium In a pinch: any piece of aluminum U profile will do, just drill a hole in it and use some heat conducting paste.
      - Florin_Andrei 10y ago
        It could not possibly be easier. It's literally a flat piece of metal with a hole in it. Add a drop of heat goo in between for extra coolness.
- blackguardx 10y ago
  The problem with the article is that there is no nuance. Some beginner is going to read it and think all BJTs are bad, but that isn't even remotely true. It all depends on application. I would certainly use MOSFETs for lighting (a power application). I probably wouldn't design an analog front-end for a specialized sensor with them though.
- aexaey 10y ago
  There is a purpose-designed Chinese chip exactly for your situation - control brightness of a LED strip based on PWM input. Compared to a bare MOSFET, this chip gives you higher efficiency, current-controlled (i.e. flicker-free) output and ability to drive LEDs directly (without current-limiting resistors). PT4115, pretty cheap on ebay and other usual places.
  - Florin_Andrei 10y ago
    Well, the LED strip I use is a black box (from an electrical perspective). It just needs 12V, no external resistors. So I just hooked it up to 12V via the MOS-FET. The PWM signal from the Arduino (plugged into the MOS-FET) does the job of changing the apparent brightness. No need for magic chips in this context.
    - pareidolia 10y ago
      Your ledstrip probably also consumes about 15 watts per meter since integrated current limiting resistors dissipate most of the energy; a black box with the efficiency of a lightbulb.
julian_1 10y ago
Higher gate capacitance of FETs is also a drawback. Often you need a bjt or even dedicated gate-driver IC to drive the gates quickly. The extra pole can also make negative feedback more complicated. And there's less component choice when you want more control in the ohmic region versus switching - for a non emitter/source follower circuit. Another issue is the higher gate voltages required - compared with bjts even using logic-level FETs. The article would have been much better with a different title - "How to use FETs effectively and when to use them", not a binary oppositioning of "antique" versus "modern".

spott 10y ago

2n7000, BS270 are both pretty cheap for small signal FET applications. For analog applications, BJTs are still very much important, but for digital stuff (and especially power stuff), it is usually easier to grab a cheap FET than use a BJT.

Florin_Andrei 10y ago

The article seems mostly right when it's about interfacing logic circuits with loads that require even moderate currents. FET is king in this field. I've a little aquarium and I've noticed how fish always get scared when lights turn suddenly on or off. So I took an Arduino and wrote a timer app for it, using PWM to slowly drive the output up over 60 minutes (sunrise), keep it at max for a few hours, then drive it down slowly over 60 minutes (sunset). I used a MOS-FET to drive the power LEDs directly from the Arduino. Works fantastic, and the FET is cold at all times. A 2N3055 would have been pretty hot most of the time. I've been playing with electronics for a few decades now, so I can't be accused of change for the sake of change. The old stuff still works, sure, but there are new toys which are better in some cases.

mrob 10y ago
What frequency are you PWMing at, and are you applying any analog filtering? I frequently get annoyed by slow PWMed lighting, and I wonder how other animals with possibly faster vision perceive it.
- Florin_Andrei 10y ago
  No analog filtering - that's how you keep things cold all the time. The frequency is either 500 Hz or 1000 Hz, I forget which (depends on Arduino type). I doubt there are any critters that can tell it's not continuous.
  - mrob 10y ago
    If I count as a critter, I can easily tell the difference by moving my eyes and seeing the phantom array effect. If you're already using resistors for current limiting then putting a big enough capacitor across the LEDs will filter them without much difference in power consumption. I have no idea if the fish will care.
    - paulannesley 10y ago
      That effect drives me nuts too; but it seems to bother a minority of people. I've always assumed/guessed it's in the order of 50–100Hz; the 500–1000Hz mentioned may be fast enough to appear continuous to sensitive humans and maybe fish...
      - throwanem 10y ago
        The most usual source of that flicker these days is driving LEDs (usually in a string of Christmas lights) with half-wave rectified AC, which can be considered PWM with a 50% duty cycle over the same period as the input signal - 60Hz in the US and I think Japan, 50Hz most other places. That's slow enough to see easily; it would surprise me to learn that PWM at ten times the rate is slow enough to see at all. Looking at the code I wrote a while back to drive a four-digit common-cathode LED display (you know, a TV bomb timer!) as a clock, I find that after all the tuning I did to get rid of flicker, I ended up activating each digit for 1ms at a time. Granted, running under Raspbian this is going to flicker anyway from time to time because multitasking, but that aside, I found no PWM flicker perceptible, even at what is effectively 20% duty cycle over a 5ms period. (The decimals are implemented as a fifth "digit".) So I'm going to guess that .5KHz and up are fine for most humans. Fish I don't know about; (1) cites a flicker fusion rate between 2 and 40Hz for swordfish, but for goldfish I've no idea, other than to guess that .5KHz and up is probably fine for them too. (1) https://www.google.com/amp/amp.livescience.com/3799-fish-eyes-adjust-shutter-speed-track-prey.html?client=safari https://www.google.com/amp/amp.livescience.com/3799-fish-eye... (Sorry for AMP, but taking it off the front nets me a timeout from livescience.com. Something broken there, I'm guessing.)
jacquesm 10y ago
If you properly drive a 2N3055 (so all the way to 'on' or all the way to 'off') you'd have to put a pretty good current through it to get it to get hot. The trick is to use the highest voltage you can get away with (most power for the least amount of current). .6V * quite a few amps before you overcome the housings capacity to get rid of the heat. You'll need to properly drive it though and not at too high a frequency (and for obvious reasons, not too low either because you'll notice the flicker).
- cr0sh 10y ago
  > before you overcome the housings capacity to get rid of the heat. The thing is, most of these larger BJT parts were designed to be used with heatsinks, not usually standalone. IMHO, there's nothing wrong with dropping some voltage and dumping some heat. Sure, it isn't efficient, but if you size everything properly, and your project can accommodate the heatsink - then where's the problem? Heck, a long time ago you could get tiny heatsinks for some of the small round metal-can transistors that existed - I've got an old 16K minicomputer core-memory board from the 1970s that has drive electronics set up like this. The author of this "rant" mentions not using L298 (and presumably the L293) either - but there's nothing wrong with these parts for hobbyists (outside of the L298's weird footprint, and the fact that heatsinks for them are almost unobtanium) - provided that the limitations are kept in mind (finding a low-cost mosfet 4A h-bridge in a thru-hole design is virtually impossible). These parts were originally meant for automotive applications, where having a 12-14V large amperage battery supply was the norm (people trying to use them for circuits with 6V or less always wonder why their small motors won't turn).
  - jacquesm 10y ago
    > outside of the L298's weird footprint, and the fact that heatsinks for them are almost unobtanium In a pinch: any piece of aluminum U profile will do, just drill a hole in it and use some heat conducting paste.
    - Florin_Andrei 10y ago
      It could not possibly be easier. It's literally a flat piece of metal with a hole in it. Add a drop of heat goo in between for extra coolness.
blackguardx 10y ago
The problem with the article is that there is no nuance. Some beginner is going to read it and think all BJTs are bad, but that isn't even remotely true. It all depends on application. I would certainly use MOSFETs for lighting (a power application). I probably wouldn't design an analog front-end for a specialized sensor with them though.
aexaey 10y ago
There is a purpose-designed Chinese chip exactly for your situation - control brightness of a LED strip based on PWM input. Compared to a bare MOSFET, this chip gives you higher efficiency, current-controlled (i.e. flicker-free) output and ability to drive LEDs directly (without current-limiting resistors). PT4115, pretty cheap on ebay and other usual places.
- Florin_Andrei 10y ago
  Well, the LED strip I use is a black box (from an electrical perspective). It just needs 12V, no external resistors. So I just hooked it up to 12V via the MOS-FET. The PWM signal from the Arduino (plugged into the MOS-FET) does the job of changing the apparent brightness. No need for magic chips in this context.
  - pareidolia 10y ago
    Your ledstrip probably also consumes about 15 watts per meter since integrated current limiting resistors dissipate most of the energy; a black box with the efficiency of a lightbulb.

julian_1 10y ago

Higher gate capacitance of FETs is also a drawback. Often you need a bjt or even dedicated gate-driver IC to drive the gates quickly. The extra pole can also make negative feedback more complicated. And there's less component choice when you want more control in the ohmic region versus switching - for a non emitter/source follower circuit. Another issue is the higher gate voltages required - compared with bjts even using logic-level FETs. The article would have been much better with a different title - "How to use FETs effectively and when to use them", not a binary oppositioning of "antique" versus "modern".