Shop OBEX P1 Docs P2 Docs Learn Events
2N3904 Transistor as switch - Page 2 — Parallax Forums

2N3904 Transistor as switch

2»

Comments

  • YanomaniYanomani Posts: 1,524
    edited 2013-11-24 20:11
    Hi amboxer21

    From what I read on previous posts, all you intend to do is related to switch on/off your PS2, that has a 9Vdc @5A power supply, under control of your cpu shield, that I presume, from readings, has a 5Vdc power source.

    Also you intend to exert this control, thru some internet connection, from your phone, that is running a "client" application, gathering command words, such as "on" or "off", then sending those command words, wirelessly thru the web, to your "server" application, that runs in your cpu, whose shield, in turn, is expected to finally do the physical part of the process, controlling BJT's base (or Mosfet's gate), that, in its turn, controls the relay coil (as formerly suggested), that, in its turn, control the relay contacts, that will finally manage the on/off process.

    The (aka) for all of it, is "PS2 remote power button, controlled thru a web interface".

    If the above statements hold true, then the next step will be chose if you want to control the PS2 power on/off process, thru one of the following possible solutions:

    1- Controlling the mains input of your power supply (120/240 Vac @??A), allowing you to totaly shut down the PS2 when not in use, in the intent of preserving energy consumption, but having to deal with potentially hazardous mains voltages within your circuit.

    2- Total energy consumption isn't your main concern, and you only intend to switch on/off the 9Vdc @5A, PS2 power supply output, thru some intervention at its dc power cord or plug, to don't mess with your PS2 case and/or internals.

    3- Another similar approach to (2) above, but dealing with PS2 internals, thru a mod to its own power on/off switch, which I believe (without being sure), could be done without having to deal with a current of 5A at all.

    Waiting for Innn-PUT, as Johnny Five, in 1986 "Short Circuit".:smile:


    Yanomani
  • amboxer21amboxer21 Posts: 15
    edited 2013-11-24 20:35
    Yanomani wrote: »
    Hi amboxer21


    From what I read on previous posts, all you intend to do is related to switch on/off your PS2, that has a 9Vdc @5A power supply, under control of your cpu shield, that I presume, from readings, has a 5Vdc power source.


    Also you intend to exert this control, thru some internet connection, from your phone, that is running a "client" application, gathering command words, such as "on" or "off", then sending those command words, wirelessly thru the web, to your "server" application, that runs in your cpu, whose shield, in turn, is expected to finally do the physical part of the process, controlling BJT's base (or Mosfet's gate), that, in its turn, controls the relay coil (as formerly suggested), that, in its turn, control the relay contacts, that will finally manage the on/off process.
    Exactly!


    I wrote programs to wirelessly interface from android phone to the ethernet sheild on my MCU. If i send the word/string "on", the microcontroller will push 5 volts out to an output pin. Which in turn will saturate the base of the BJT and switch the relay on. Ultimately turning the ps2 on.



    Yanomani wrote: »
    The (aka) for all of it, is "PS2 remote power button, controlled thru a web interface".


    If the above statements hold true, then the next step will be chose if you want to control the PS2 power on/off process, thru one of the following possible solutions:


    1- Controlling the mains input of your power supply (120/240 Vac @??A), allowing you to totaly shut down the PS2 when not in use, in the intent of preserving energy consumption, but having to deal with potentially hazardous mains voltages within your circuit.


    2- Total energy consumption isn't your main concern, and you only intend to switch on/off the 9Vdc @5A, PS2 power supply output, thru some intervention at its dc power cord or plug, to don't mess with your PS2 case and/or internals.


    3- Another similar approach to (2) above, but dealing with PS2 internals, thru a mod to its own power on/off switch, which I believe (without being sure), could be done without having to deal with a current of 5A at all.


    Waiting for Innn-PUT, as Johnny Five, in 1986 "Short Circuit".


    I was going to hook the relay directly up to the PS2 adapter. Which runs at -> OUTPUT: 8.5V and 5.65A, INPUT: 100-240V and 1.5A The PS2 itself has the following ratings -> 8.5V and 5.3A Maybe there is a way to mod the hardware directly to achieve this without interacting with such high ratings?? IDK What do you think?
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-11-24 21:28
    Hi again,
    I know you restated 8.5v at 5.65 amps and I use 9.0v at 5.00 amps below. The truth is that either target is going to have the same issues mentioned herein.
    There is not enouugh data in a pdf to provide exact figures for 8.5v (and you really want 9.0v) or 5.65 amps.

    You jump between a 2n3904 and a TIP3055 like they are the same device..... They certainly are not.

    I guess you just want to explore the devices you have on hand.. fair enouugh. Personally, I'd drive a 12 volt brake light bulb for experiments witha TIP3055 before I would dare involve a PlayStation. Experiments are useful, but they do tend to break things when errors occur.

    The 2n3904 will saturate for 50ma output via 5ma input.... so it works with microcontrollers that have a limited amount of milliamps before damaging and i/o pin.
    The TIP3055 cannot be driven directly from a microcontroller, it requires too much current in a saturated ON condition.

    You can drive the TIP3055 to satuaration with 5 volts, but it will take more than 5 ma to get 5 amps output... maybe a lot more ... a rough guess is 100ma.

    I checked to see if 5amps @ 9VDC is in the safe operation zone (figure 2 of Motorola's TIP3055 pdf).
    There is no point in trying if the part is just going to get too hot and burn up.

    I checked the max current for the Ic and it seems to be only 1.5 amps.. something wrong here as that is not enough for you to drive 5 amps through the CE0.
    /// Seems to be a typo in the Motorola pdf as another pdf says 15 amps.

    How did I get the rough guess?
    The data sheet states minimum gain of 20x at 4amps, dropping to 5x at 10amps.
    I used the 20x to figure 100ma.

    The data sheet doesn't seem to address the situation of 9VDC/5amps switched, so it seems you really need to set up a bench model to properly test this.
    You would need at least a 45 watt resistor (say 50 watt) to dump the 9Volt/5amp load.
    And you would then see very clearly that 9VDC in results in something significantly less that 9VDC avaiable (maybe only 7VDC)... there is a 'diode voltage drop' and in larger hot power transistors it can be more than the 0.7 volts that people often mention.

    I really don't want you to break anything of value or take unnecessary risks. So I'd suggest to build the whole thing as a test bench and control the input to the base from an appropriate pot. You are not going to get anything but "ballpark" numbers from the pdfs.

    SOONER or later you will figure out that a 9VDC power supply switched by a BJT will always deliver 8.3 or less VDC... it is called an internal diode voltage drop that is a minimum of about 0.7 volts. But as the device heats up and the power demands go up, it can be over 1.0 volts... maybe a lot more.

    I previously tried to say that the TIP3055 just wasn't worth the trouble... too much heat AND too much has to be determined via building a proto-type AND you need to build a custom power supply higher than your target output voltage.

    Nonetheless, I learned al this by building a bench model to convince myself. If that will help you, great. You can't always use a pdf, a textbook, and a calculator.
  • Heater.Heater. Posts: 21,230
    edited 2013-11-24 23:31
    Datasheets? We don't need no stupid data sheets. We can measure it!

    Set up a little test with your transistor and the load you want to drive.

    Crank up it's base current until the load is driven on has hard as you would like. You will be measuring load current and voltage as well as base current.

    Measure that. Determin size of resistor required. Done!

    Ah, but "transistors vary widely" you say. I say: a) Never mind "I'm using this one". b) Measure a few, preferably from different batches/manufacturers. c) Double the base current "just to be sure":)

    Of course this kind of characteristics measurement is time consuming and prone to blowing up devices (How do we know this little trany is not going to fry trying to drive that load?)

    But it's interesting and a good way to learn and get a feel for things. Something worth doing at least once.

    Aside: I always have to do this with LEDs. No matter what the spec says it's impossible to calculate that load resistor to get the brightness you want. I always have to try a bunch of values.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-11-25 00:20
    @Heater
    Thanks. that Nordic winter must be setting in about now... a bit of cabin fever maybe?

    1. The Absolute Maximum Ratings are always useful on any pdf.
    2. The safe zone for continous on is an important chart as well as BJTs to experience thermal runaway

    And I have certainly noticed that even Led can be a challenge as every color is different and even the same color Leds have changed over time.
  • Heater.Heater. Posts: 21,230
    edited 2013-11-25 00:36
    Loopy,

    We have not ensconced ourselves in the cabin yet. It's only -3C this morning and the sun is shining :)

    Absolute Maximum Ratings and safe zones are also things you can also determine by experiment :)

    Of course I have my tongue in my cheek a bit when I say all this but I do feel that such experiments are worth while. Plotting out those transistor or other device characteristics for yourself is an education. For sure if you study electronics in college or university you may find yourself doing such lab exercises. Hobbyists might like to try it as well.

    It's nice to have a feel for how all that gobbledy gook in data sheets relates to reality.

    There is a lot to be learned by blowing things up!
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-11-25 03:27
    I get the feeling there are a lot of guys that are just starting out that buy a 2N3904 or a TIP3055 (Radio Shack loves to sell these kind of items) and just want to make it work with their Propeller or whatever.

    For myself, I just wanted to build my own H-bridge from parts regardless of the heat. I thought it was a great way to verify the maths. It was, but the end product was rather useless ... too larger, too hot for most robots. That heat is also wasted battery power, and having 1/3 of your power wasted means more recharges.


    The new users come here and immediately we respond with either trying to help them with a big dose of maths and jargon, or the suggestion that they buy something else. We tend to ignore the 3rd alternative of to really explore what the BJT is doing and why it is not a great device for power switching. We are getting ahead of ourselves when the role should be more helpful.

    When you look at the TIP3055 On characteristics for 10 amps, it seems the base current is something like 5 amps... or a Gain of 2x.
    Sure that is a gain, but 5 amps at 5 volts is 25 watts of heat to get your 10 amp switch.

    Try holding any lit 25 watt light bulb in your hand for 10 seconds... you will begin to comprehend that a nice and cool, but noisy mechanical relay offers some benefits.
    And since there is no voltage drop in a mechanical switch, it pretty much allows you to just plug it in and use it without any voltage reduction issue. It can also use AC or DC, though it doesn't perform at anywhere near 120V with DC.

    Solid-state relays are also easy. They may have a tiny voltage drop. And you have to know beforehand if you want to control AC or DC. The newer ones are small and run cool with direct control from a microcontroller..

    +++++++++++++
    I guess I am saying that just knowing all the best solutions is not always the right answer. Helping a new user define their own project and allowing them to see for themselves is important... it is called teaching.

    The 2n3904 is a tiny bit wasteful, just because of the scale at which it operates, but the TIP3055 is extremely wasteful when you look at the alternatives.

    ++++
    I don't think anyone has clearly explained why saturation is preferred for switching over non-saturation. As I recall, it creates less heat and wastes less power fi you get it right.

    ++++
    I like the car key idea and think it is not that hard to do.
    But what I don't want to do is to explain over the internet to all and everyone how to hot wire a car. I dislike the idea of training a few hundred car theives while helping someone on a project.
    If you really know how, you can by-pass the ignition key and get an engine running in about 5 seconds (at least it was that way before all the ECU junk was added to the auto).
  • amboxer21amboxer21 Posts: 15
    edited 2013-11-27 10:24
    Sorry, I didn't forget about m project or the awesome members helping me!! I really appreciate everyone's patience and answers!! Since work(roofing) is super slow this time of year and I blew my dirt bike up in a race 2 Sundays ago. I have picked up an old programming project. I really dont have money to be venturing out of the house much or spending on electronics lol I am super strapped for cash! Hopefuly next week there will be more work than there has been so I can do things!! But for now, if anyone enjoys programming and wants to help, I am working on a program for linux to download your entire facebook photostream in one shot. I have the base down. Facebook oauth an i can grab the first 25 urls but they are in json format. SO I am working on the parsing part now.

    I will probably be jumping back and forth between this and the NPN transistors.

    LOOPY BYTELOOSE,
    I like the car key idea and think it is not that hard to do.
    But what I don't want to do is to explain over the internet to all and everyone how to hot wire a car. I dislike the idea of training a few hundred car theives while helping someone on a project.
    If you really know how, you can by-pass the ignition key and get an engine running in about 5 seconds (at least it was that way before all the ECU junk was added to the auto).

    Hotwiring a car is super easy. Not because I am a criminal or trouble maker but I have a dirtbike and work a lot on the engine and of course venture of into the automotive world to satisfy my need for mechanics. I have learned a lot over the years and in principle... it is common sense to someone who is handy electronic/mechanic wise. Like you have stated, I wont explain over the internet and aid someone in doing something stupid. I already hooked a toggle switch and SPDT monetary push button up to the ignotion and disabled the steering wheel lock. I also integrated 2 SPDT monetary push buttons under the wheel well to lock and unlock the car. Here is a video of the toggle switch and SPST monetary push button in action ->

    Now how cool would that have been to hook up a key pad so only the possessor of the combo could start the car? That would have been badass!! lol

    But This project has me so bummed... lost and discouraged. I have never worke with transistors and i have learned a lot thanks to the fellow members here but i feel like i have come so close but didnt achieve anything. IDK i would like to do something cool! Worthy of dropping jaws!!

    So if I were to continue this PS2 project, what would I have to do? Migrate to mosfets?
  • kwinnkwinn Posts: 8,697
    edited 2013-11-27 12:03
    If the main purpose of this project is educational you may want to migrate to mosfets. On the other hand if the main purpose is to turn your PS2 on and off remotely I would suggest using an SSR to switch the 120VAC input to the wall wart along the lines of what tonyp12 did in this (http://forums.parallax.com/showthread.php/149577-Dual-SSR-inside-a-PowerStrip) project. Not only does it simplify the project since the SSR only needs about 10mA to turn on, it saves energy and prolongs the life of the wall wart by shutting it off when not needed. An additional benefit is versatility since it can be used to turn on/off any device powered by a wall wart, and any 120VAC device that draws no more current than the SSR can handle.
  • amboxer21amboxer21 Posts: 15
    edited 2013-12-27 15:16
    kwinn wrote: »
    If the main purpose of this project is educational you may want to migrate to mosfets. On the other hand if the main purpose is to turn your PS2 on and off remotely I would suggest using an SSR to switch the 120VAC input to the wall wart along the lines of what tonyp12 did in this (http://forums.parallax.com/showthread.php/149577-Dual-SSR-inside-a-PowerStrip) project. Not only does it simplify the project since the SSR only needs about 10mA to turn on, it saves energy and prolongs the life of the wall wart by shutting it off when not needed. An additional benefit is versatility since it can be used to turn on/off any device powered by a wall wart, and any 120VAC device that draws no more current than the SSR can handle.

    I wanted to have another go at this. I was busy with a program I was writing. Only this time, I want to use SSR's. Where would I start? This has definitely been done and does anyone have a link to a tutorial?
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2013-12-27 17:01
    From the control side of things most SSRs operate just like an LED. You should look at the specifications of course, but most have an LED front end for optical isolation. So with the forward voltage and current you can calculate the resistor needed to turn the SSR on/off quite easily.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-12-27 21:04
    The SSRs come in basically two generations of devices.

    A. The earlier 'hockey puck' devices. which are still being used for higher amp ratings (require heat sinks in many situations)

    B. The newer smaller devices.

    As best as I can understand, both usually include adequate currently limiting with the LED isolation (and are marked for +3 to +32 volt inputs), so you can wire them direct to the Propeller and they work fine as long as you get the polarity correct. But you should make sure that they don't need an external current limiting resistor as there are a few that do.

    The SSRs are not limited to AC, so be sure you get the right kind. You have to chose between DC and AC devices. READ THE SPECIFICATIONS BEFORE BUYING.

    In many cases, a 'hockey puck' device is a much larger device than a mechanical relay. So, if space or size is an issue, consider using mechanical devices. And they do need cooling if running near maximum power ratings.

    http://www.ebay.com/itm/Solid-State-Relay-SSR-25DA-25A-3-32VDC-Output-24-380V-AC-Solid-State-Module-/130905531225
  • amboxer21amboxer21 Posts: 15
    edited 2013-12-28 17:28
    The SSRs come in basically two generations of devices.A. The earlier 'hockey puck' devices. which are still being used for higher amp ratings (require heat sinks in many situations)B. The newer smaller devices.As best as I can understand, both usually include adequate currently limiting with the LED isolation (and are marked for +3 to +32 volt inputs), so you can wire them direct to the Propeller and they work fine as long as you get the polarity correct. But you should make sure that they don't need an external current limiting resistor as there are a few that do.The SSRs are not limited to AC, so be sure you get the right kind. You have to chose between DC and AC devices. READ THE SPECIFICATIONS BEFORE BUYING.In many cases, a 'hockey puck' device is a much larger device than a mechanical relay. So, if space or size is an issue, consider using mechanical devices. And they do need cooling if running near maximum power ratings.http://www.ebay.com/itm/Solid-State-Relay-SSR-25DA-25A-3-32VDC-Output-24-380V-AC-Solid-State-Module-/130905531225
    I plan on using the ssr to directly interface with a wall socket. which is 120VAC. i am going to switch the socket on and off instead of the PS2. I just have no idea where to start. Size is not an issue but money is! Money is kind of tight! I have seen a bunch of power strips with small ssr's integrated into them. Maybe that's a direction i could take??
  • kwinnkwinn Posts: 8,697
    edited 2013-12-28 22:18
    A power strip with integrated ssr's would work fine as long as their current ratings are adequate for what you want to plug in to them and the control terminals are accessible.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-12-29 03:51
    I can understand the desire to go 'all solid-state' as it seems to be modern and better. Still, mechanical relays to have a lot of advantages.

    But with SSRs, there are certain drawbacks

    A. Unlike mechanical relays, they fail to ON. Mechanical relays tend to fail to OFF
    B. A lot of SSRs cost more than mechanical relays.
    C. SSRs are generally single pole, single throw devices; whereas mechanical relays can be gotten in double pole double throw and other configurations that are handy
    D. SSRs have to be seperately purchased for DC or AC control; whereas mechanical relays can handle either -- though a 120VAC relay generally is only useful for 24VDC at the same amps.
    E. SSRs can be larger than mechanical relays for the same task, especially if heak sinks are required.
    F. SSRs may run hot, but quiet; mechanical relays run cool, but click.

    You could just buy a relay board on Ebay. They have these that are both mechanical relays and soldi-state for reasonable prices.

    As for myself, I use 12VDC relays that are rated for 12 amp 240 volt operation and drive 8 of these using a ULN2803 darilington chip. The darlingtons boost the output of the Propeller to 12 volts and 60 mal per pin. Seems adequate and a simpler cleaner build.

    The relays I use have sockets that mount to a DIN rail and all the 120VAC connections are very well protected with insulation and just attach wires by a set screw. I prefer this because a failed relay can be swapped out like a blown light bulb or fuse. I avoid relay boards that have relays soldered on to the board. But the banked relay boards are one of the cheapest options to get started it they will accept 3.3 volt logic control.
  • amboxer21amboxer21 Posts: 15
    edited 2013-12-29 13:48
    I can understand the desire to go 'all solid-state' as it seems to be modern and better. Still, mechanical relays to have a lot of advantages.But with SSRs, there are certain drawbacksA. Unlike mechanical relays, they fail to ON. Mechanical relays tend to fail to OFFB. A lot of SSRs cost more than mechanical relays.C. SSRs are generally single pole, single throw devices; whereas mechanical relays can be gotten in double pole double throw and other configurations that are handyD. SSRs have to be seperately purchased for DC or AC control; whereas mechanical relays can handle either -- though a 120VAC relay generally is only useful for 24VDC at the same amps.E. SSRs can be larger than mechanical relays for the same task, especially if heak sinks are required.F. SSRs may run hot, but quiet; mechanical relays run cool, but click.You could just buy a relay board on Ebay. They have these that are both mechanical relays and soldi-state for reasonable prices.As for myself, I use 12VDC relays that are rated for 12 amp 240 volt operation and drive 8 of these using a ULN2803 darilington chip. The darlingtons boost the output of the Propeller to 12 volts and 60 mal per pin. Seems adequate and a simpler cleaner build.The relays I use have sockets that mount to a DIN rail and all the 120VAC connections are very well protected with insulation and just attach wires by a set screw. I prefer this because a failed relay can be swapped out like a blown light bulb or fuse. I avoid relay boards that have relays soldered on to the board. But the banked relay boards are one of the cheapest options to get started it they will accept 3.3 volt logic control.
    I would rather try on my own. So, to get started, I would need a SPDT 12VDC relay thats rated for 12A and 240VAC??? and a ULN2803 darilington chip? I assume the 12VDC relay will be switched by the transistor amplification? Is the use of a DP for heat purposes?? I also assume the relay will handle the PS2 current/voltage for long periods of time?? Here is a question. In order to start, I would like to get a basis for my foundation. What am I basing everything on? The wall current/voltage? The input of the PS2? The Output of the PS2? Maybe a little of both? What should i be worried about? The output of the PS2 adapter?? Thanks EDIT: PS2 ratings and voltageinput is 100-240 volts rated at 1.5 amps - output is 8.5 volts rated at 5.65 amps - AC adapter - 8.5 volts
  • kwinnkwinn Posts: 8,697
    edited 2013-12-29 19:32
    amboxer21 wrote: »
    I would rather try on my own. So, to get started, I would need a SPDT 12VDC relay thats rated for 12A and 240VAC??? and a ULN2803 darilington chip?

    Yes. The 240VAC rating would be for the contacts and the 12VDC for the coil.
    I assume the 12VDC relay will be switched by the transistor amplification?

    Yes, the power to the relay coil would be switched by a transistor.
    Is the use of a DP for heat purposes??

    No, the DP is typically used to switch 2 circuits. They may be two different circuits or the two legs of a 240VAC circuit.
    I also assume the relay will handle the PS2 current/voltage for long periods of time?? Here is a question. In order to start, I would like to get a basis for my foundation. What am I basing everything on? The wall current/voltage? The input of the PS2? The Output of the PS2? Maybe a little of both? What should i be worried about? The output of the PS2 adapter?? Thanks EDIT: PS2 ratings and voltage input is 100-240 volts rated at 1.5 amps - output is 8.5 volts rated at 5.65 amps - AC adapter - 8.5 volts

    The relay should handle the voltage and current for a long time. The relay contact ratings are based on the voltage and currents you want to switch. For your PS2 adapter connected to a typical north American outlet the minimum contact voltage rating would be 120VAC, although higher ratings are fine. My recommended current rating would be at least 3A if you want the contacts to last a long time. For the coil 12V is a good choice.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-12-30 01:22
    Loopy,

    Nice summary of SSR vs mechanical relays. I personally like the sound of relays clicking but my wife doesn't (guess who wins).
    But with SSRs, there are certain drawbacks

    A. Unlike mechanical relays, they fail to ON. Mechanical relays tend to fail to OFF

    Is this based on your own experience?
    It does seem like a logical assumption but it's opposite of my personal experience. I've had two mechanical relays fail over the last few years and they both failed on. The relays were used on a toaster oven I used to dry ore samples. Apparently the contacts welded together in a closed (on) state.

    I switched our kitchen relays to solid state because of the noise, but I'm switching the relays in my lab to solid state since they seem to be more reliable than the mechanical ones.
  • Peter KG6LSEPeter KG6LSE Posts: 1,383
    edited 2013-12-30 02:29
    Duane Degn wrote: »
    Loopy,

    Nice summary of SSR vs mechanical relays. I personally like the sound of relays clicking but my wife doesn't (guess who wins).



    Is this based on your own experience?
    It does seem like a logical assumption but it's opposite of my personal experience. I've had two mechanical relays fail over the last few years and they both failed on. The relays were used on a toaster oven I used to dry ore samples. Apparently the contacts welded together in a closed (on) state.

    I switched our kitchen relays to solid state because of the noise, but I'm switching the relays in my lab to solid state since they seem to be more reliable than the mechanical ones.

    SSRs are not too happy with low power factors. toaster are ideal for SSRs my large tesla coil OTOH makes them latch on and never turn off ....

    SMPS power bricks may also have a issue too .
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-12-30 03:18
    Duane Degn wrote: »
    Loopy,

    Nice summary of SSR vs mechanical relays. I personally like the sound of relays clicking but my wife doesn't (guess who wins).



    Is this based on your own experience?
    It does seem like a logical assumption but it's opposite of my personal experience. I've had two mechanical relays fail over the last few years and they both failed on. The relays were used on a toaster oven I used to dry ore samples. Apparently the contacts welded together in a closed (on) state.

    I switched our kitchen relays to solid state because of the noise, but I'm switching the relays in my lab to solid state since they seem to be more reliable than the mechanical ones.

    About the failure mode.

    1. I am presuming the failure is on the switching side.
    2. It seems that the semiconductor material in the SSR latches to ON in a stressed failure; while mechanical relays tend to have their contacts burn up and create an OFF mode.

    I do admit that it is possible for mechanical contacts to 'weld' into an on state in some situations. This may be more common in DC switching, where current just runs in one direction. For the purposes of heavy DC loads with induction (such as motors), a mechanical relay is often derated to 20% of its full current value to assure long-life. So a 12A relay is best for a motor demanding about 2.5 Amps! (I am a bit unclear of derating SSRs as I have not found much material on the web about deratings for various loads, and the switch process is so different)

    It seems we are still discovering the oddities of SSR failure modes... latching up is not a good thing.

    Different failure modes imply different requirement in setting up a fail-safe system.

    ~~~
    SPST is the abbreviation for Single Pole, Single Throw.
    DPDT is a Double Pole, Double Pole

    This notation is quite common is complex switches, especially rotary switches that were used in equipement from WWII through to audio hifis of the 1960s-70s.

    In relays the Throw is limited to Single or Double.

    ~~~
    And YES, the draw back with 12VDC relay coils is that you have to provide another power source for the 12VDC. There are 5VDC, 6VDC, and maybe 9VDC relay coils. But I prefer the 12VDC as they can be used with an automotive relay. Automotive relays are set up for wet and dirty environments and temperature extremes -- great for robots in the real world and can switch big loads (like 40 or 60 amp ratings).

    ~~~
    There is a constant struggle between selecting for smallness and selecting for adequate power. One horsepower is roughly equal to 775 Watts. So if you want to drive anything that can lift or go up hills, you need to consider size as well as power. The easiest way to get MORE Power, is to double the voltage -- use a 24 Volt motor instead of a 12 Volt motor.. a 36 Volt motor is even better.

    ~~~~~~~~~~~~~~~~~~
    For the Sony Playstation project, you might find a tiny 5VDC relay that has a switch rated at several amps at 120VAC. If you provide the ULN2803 Darlington with 6 or 7.5 volts, the voltage drop within the Darlington pair is likely going to result in about 5 volts to run the relay coils. In general, relay coils are much more tolerant of inexact coil voltages that are higher than specification than semiconductors. If you go too low on a relay coil voltage, the relay may not fully engage and the switch will bounce.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-12-30 03:23
    Duane Degn wrote: »
    Loopy,

    Nice summary of SSR vs mechanical relays. I personally like the sound of relays clicking but my wife doesn't (guess who wins).



    Is this based on your own experience?
    It does seem like a logical assumption but it's opposite of my personal experience. I've had two mechanical relays fail over the last few years and they both failed on. The relays were used on a toaster oven I used to dry ore samples. Apparently the contacts welded together in a closed (on) state.

    I switched our kitchen relays to solid state because of the noise, but I'm switching the relays in my lab to solid state since they seem to be more reliable than the mechanical ones.

    I suspect the toaster oven relays failed because you used software with a high cycle rate of on and off. Just because the microcontroller can keep the device within a fraction of degree doesn't mean that is going to work with a mechanical relay. If you look at the overall goal of holding temperature and have the on/off cycle in wider range and at a slower rate, the relays will last --- they just need to cease all that neurotic on/off that doesn't improve the quality of the soldering task.
  • kwinnkwinn Posts: 8,697
    edited 2013-12-30 09:21
    A lot of good points made after my previous post. As far as choosing between SSR or mechanical relays, in most cases that ends up being personal preference, although there are applications where one is preferred over the other. A mechanical relay is the better choice for an emergency stop circuit, while a SSR would be better for applications where power cycles on/off frequently.

    For switching receptacles on a power bar/strip my first choice would be a SSR, mainly for the ability to control it with currents low enough that they can come directly from a microcontroller pin. A latching relay would be a close second choice.

    With regard to relay failures, my experience has been that the majority of failures for both mechanical and SSR's have been in the off state, and for mechanical relays it has been the coil that failed. Typically the coil has an internal short that cooks it and burns out the driver circuit.
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2013-12-30 16:16
    kwinn wrote: »
    A lot of good points made after my previous post. As far as choosing between SSR or mechanical relays, in most cases that ends up being personal preference, although there are applications where one is preferred over the other. A mechanical relay is the better choice for an emergency stop circuit, while a SSR would be better for applications where power cycles on/off frequently.

    For switching receptacles on a power bar/strip my first choice would be a SSR, mainly for the ability to control it with currents low enough that they can come directly from a microcontroller pin. A latching relay would be a close second choice.

    With regard to relay failures, my experience has been that the majority of failures for both mechanical and SSR's have been in the off state, and for mechanical relays it has been the coil that failed. Typically the coil has an internal short that cooks it and burns out the driver circuit.

    I suppose if a fuse or breaker were in series from the mains voltage the latching relay could be used in the power system. As for mechanical relay failures of the dozen or so I have seen over the last 20 years or so only one had a burned coil. The rest were mainly bad contacts from arcing and a few had mechanical failures in the contact section. One had bad solder joints on the pins that were switching power.
Sign In or Register to comment.