Author Topic: How close to max specification should I come?  (Read 4652 times)

0 Members and 1 Guest are viewing this topic.

Offline docuTopic starter

  • Newbie
  • Posts: 7
  • Country: us
How close to max specification should I come?
« on: August 13, 2016, 11:25:13 pm »
Hello!  I am looking to control an 1800W (nominal) load using a Sharp S216S02.  Conceivably, I could split this load up and control using two of these solid state relays, but would prefer to use one for cost and board space. 

An 1800W load with 120 Vac line voltage yields an RMS current of 15A.  The maximum continuous current specified by the S216S02 is 16A.

Should I use two S216S02 solid state relays or can I use one to control this load?  Is there anything else I am missing here?

Thanks!
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 3341
  • Country: au
Re: How close to max specification should I come?
« Reply #1 on: August 13, 2016, 11:39:14 pm »
Is there any surge current at switch-on like with an electric motor or transformer? Or is it a resistive load that is relatively benign and predictable? How low can you keep the temperature on a heatsink? What is the consequences of failure? Loss of life, fines and jail time, or simply an inconvenience?
 

Offline GBowes

  • Contributor
  • Posts: 33
  • Country: us
Re: How close to max specification should I come?
« Reply #2 on: August 13, 2016, 11:45:11 pm »
If you look at the data sheet, that 16A rating is at an ambient temperature of 25°C with a huge Aluminum heat sink (200 mm x 200 mm x 2mm)
I think you will definitely need to use more than one.
 

Online rstofer

  • Super Contributor
  • ***
  • Posts: 9940
  • Country: us
Re: How close to max specification should I come?
« Reply #3 on: August 13, 2016, 11:57:29 pm »
Buy a 120V 20A Solid State Relay and be done with it:

http://www.mouser.com/ProductDetail/Omron/G3NA-220B-AC100-120/?qs=esLHqywZQsgU1j1%2FBeVOuQ%3D%3D&gclid=CKS10cbLv84CFUSBfgod9mwB9w

Assuming, of course, a resistive load without inrush currents exceeding the capability of the device

 
The following users thanked this post: AndersJ

Offline docuTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How close to max specification should I come?
« Reply #4 on: August 14, 2016, 12:01:55 am »
Is there any surge current at switch-on like with an electric motor or transformer? Or is it a resistive load that is relatively benign and predictable? How low can you keep the temperature on a heatsink? What is the consequences of failure? Loss of life, fines and jail time, or simply an inconvenience?

It is for a home-brew reflow soldering station made from a convection toaster oven.  The load would be a set of 120 Vac resistive heating elements (four total).
 

Online rstofer

  • Super Contributor
  • ***
  • Posts: 9940
  • Country: us
Re: How close to max specification should I come?
« Reply #5 on: August 14, 2016, 12:15:59 am »
I built mine using a Black & Decker InfraWave Toaster Oven.  I buried the solid state relay inside the right cover between the cover and the inner wall.  Plenty of room!

I also filled all of the cavities with high temperature insulating mat from McMaster-Carr.  It helps a lot!  Definitely, do this!

This is the controller I used.  I mounted the display to the front of a metal box and enclosed the controller.  I just mounted the box to the top of the oven.
http://www.rocketscream.com/shop/reflow-oven-controller-shield-arduino-compatible

When I cook boards, I put the thermocouple tip in a via hole in the board.
 

Offline docuTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How close to max specification should I come?
« Reply #6 on: August 14, 2016, 12:24:06 am »
I built mine using a Black & Decker InfraWave Toaster Oven.  I buried the solid state relay inside the right cover between the cover and the inner wall.  Plenty of room!

I also filled all of the cavities with high temperature insulating mat from McMaster-Carr.  It helps a lot!  Definitely, do this!

This is the controller I used.  I mounted the display to the front of a metal box and enclosed the controller.  I just mounted the box to the top of the oven.
http://www.rocketscream.com/shop/reflow-oven-controller-shield-arduino-compatible

When I cook boards, I put the thermocouple tip in a via hole in the board.

Is the SSR you linked earlier the same one you used in your project?

Also, did you end up putting the Arduino inside of the toaster oven housing?  The insulation was on my list - I would like to put the controller and such inside of the housing, if possible, but I do not want it to get too warm.  I was planning on designing my own controller, although I will take a look at the one you linked.
 

Offline zapta

  • Super Contributor
  • ***
  • Posts: 6290
  • Country: 00
Re: How close to max specification should I come?
« Reply #7 on: August 14, 2016, 01:49:22 am »


It is for a home-brew reflow soldering station made from a convection toaster oven.  The load would be a set of 120 Vac resistive heating elements (four total).

Note that the initial current when the elements are cold may be higher.

 

Online rstofer

  • Super Contributor
  • ***
  • Posts: 9940
  • Country: us
Re: How close to max specification should I come?
« Reply #8 on: August 14, 2016, 03:25:22 am »
I built mine using a Black & Decker InfraWave Toaster Oven.  I buried the solid state relay inside the right cover between the cover and the inner wall.  Plenty of room!

I also filled all of the cavities with high temperature insulating mat from McMaster-Carr.  It helps a lot!  Definitely, do this!

This is the controller I used.  I mounted the display to the front of a metal box and enclosed the controller.  I just mounted the box to the top of the oven.
http://www.rocketscream.com/shop/reflow-oven-controller-shield-arduino-compatible

When I cook boards, I put the thermocouple tip in a via hole in the board.

Is the SSR you linked earlier the same one you used in your project?

I really don't know but it is where I would start.  It's just an SSR.and it has been a few years since I built it.  The big thing is to be certain the Arduino can drive the LED in the SSR.  Check LED current...

Quote
Also, did you end up putting the Arduino inside of the toaster oven housing?  The insulation was on my list - I would like to put the controller and such inside of the housing, if possible, but I do not want it to get too warm.  I was planning on designing my own controller, although I will take a look at the one you linked.

No, I mounted the Arduino and the shield in a neat metal enclosure that had removable front and rear panels.  That allowed me to work on the font panel to mount the LCD and the rear panel matched the Arduino connectors.

This isn't the enclosure (I can't find the one I used) but it might work:
http://forum.arduino.cc/index.php?topic=137045.0

I use a wall wart to power the Arduino and the line cord through the SSR for the heater.  So, two line connections, just because it was easy. There's a pair of wires between the Arduino enclosure and the SSR.
« Last Edit: August 14, 2016, 03:27:36 am by rstofer »
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8908
  • Country: fi
Re: How close to max specification should I come?
« Reply #9 on: August 14, 2016, 12:06:39 pm »
Depending on the heating element configuration (how hot do they actually become?), and how the nameplate wattage is specified (cold or hot?), heating elements can be far from nice linear resistive load and may show high inrush currents, possibly double the normal current, lasting for several seconds or even longer, which will be very hard on semiconductors, usually rated for pulses less than a few milliseconds. You could measure the actual current in worst case (switch-on). Even the 20A SSR might not be enough in the worst case. Using two to control the heaters separately has the additional benefit of implementing fancy features later on.
 

Offline grouchobyte

  • Regular Contributor
  • *
  • Posts: 244
  • Country: cn
Re: How close to max specification should I come?
« Reply #10 on: August 14, 2016, 04:54:04 pm »
Hello!  I am looking to control an 1800W (nominal) load using a Sharp S216S02.  Conceivably, I could split this load up and control using two of these solid state relays, but would prefer to use one for cost and board space. 

An 1800W load with 120 Vac line voltage yields an RMS current of 15A.  The maximum continuous current specified by the S216S02 is 16A.

Should I use two S216S02 solid state relays or can I use one to control this load?  Is there anything else I am missing here?

Thanks!

Think of maximum operating current or maximum anything like a road speed limit. Just because your car can do the speed and the posted limit is 100 miles per hour doesn't necessarily imply or suggest that you SHOULD. Perhaps you should consider a little known phenomenon called "margin" Calculate your operating margin under worse case conditions and then you will know exactly how far you can push things.

@grouchobyte
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5031
  • Country: si
Re: How close to max specification should I come?
« Reply #11 on: August 17, 2016, 05:32:06 am »
Its mostly a question of how reliable you want it to be.

Also be sure to check under what conditions the max ratings in the datasheet are. Quite often you see max specs for things like continuous current and dissipated power on transistors specified for a Tcase of 25°C and the internal junction getting to like 120°C. This means you can only reach those specs if you bolt the device to a water cooling block that has bellow room temperature water running trough it to compensate for the blocks thermal resistance. And even then the silicon die inside is almost overheating.

This means how close you can get to max specs heavily depends on your application. Like how cool can you keep the device during operations since your heatsink is going to heat up and in the case of an oven the environment might get hotter than ambient. Also in some cases there might be higher than usual starting inrush currents where you might want to overrate it even more.(tho for a oven the cold heaters shouldn't pull that huge of a current).

In my opinion for this particular example i would rather get a higher rated part since 25A triacs are not all that uncommon or like you said use two to split the load when possible.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8908
  • Country: fi
Re: How close to max specification should I come?
« Reply #12 on: August 17, 2016, 05:41:38 am »
Also be sure to check under what conditions the max ratings in the datasheet are. Quite often you see max specs for things like continuous current and dissipated power on transistors specified for a Tcase of 25°C and the internal junction getting to like 120°C. This means you can only reach those specs if you bolt the device to a water cooling block that has bellow room temperature water running trough it to compensate for the blocks thermal resistance. And even then the silicon die inside is almost overheating.

True for bare semiconductors, but for modules like SSR's, or even IGBT blocks, they are often specified more realistically (for example, they don't talk about die temperature at all, but specify everything for a case temperature of 85 degC, counting in the thermal resistances die-to-semiconductor case and semiconductor case to their product case). So you could find a 30A TRIAC inside a 15A SSR for this reason; they have done the analysis for you and sell it as a product.

Even so, you should always check whether this is true or not; always check the conditions where the rated values apply. Without exact conditions given, the ratings are meaningless. Then add a bit of margin.
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5031
  • Country: si
Re: How close to max specification should I come?
« Reply #13 on: August 17, 2016, 07:31:07 am »
True for bare semiconductors, but for modules like SSR's, or even IGBT blocks, they are often specified more realistically (for example, they don't talk about die temperature at all, but specify everything for a case temperature of 85 degC, counting in the thermal resistances die-to-semiconductor case and semiconductor case to their product case). So you could find a 30A TRIAC inside a 15A SSR for this reason; they have done the analysis for you and sell it as a product.

Even so, you should always check whether this is true or not; always check the conditions where the rated values apply. Without exact conditions given, the ratings are meaningless. Then add a bit of margin.

Good point, the industrial designer shouldn't have to worry about that.

But then again if you buy a solid state relay as a shenzen ebay special for $1.95 with free shipping id imagine if it says 16A then it has a 16A triac inside.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17681
  • Country: lv
Re: How close to max specification should I come?
« Reply #14 on: August 17, 2016, 07:36:29 am »
But then again if you buy a solid state relay as a shenzen ebay special for $1.95 with free shipping id imagine if it says 16A then it has a 16A triac inside.
No, there is 8A triac inside or even worse. Although Fotek is real company, I'm pretty surre that at least 95% of those sold on ebay are fakes.
http://www.ul.com/newsroom/publicnotices/ul-warns-of-solid-state-relay-with-counterfeit-ul-recognition-mark-release-13pn-52/
« Last Edit: August 17, 2016, 07:47:36 am by wraper »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf