Author Topic: Output protection  (Read 6940 times)

0 Members and 1 Guest are viewing this topic.

Offline abbtechTopic starter

  • Contributor
  • Posts: 39
    • Alan's Electronic Projects
Output protection
« on: September 02, 2010, 01:40:10 am »
Hey guys,

I have a project that will have 16 12 volt outputs which is expected to deliver up to 250mA each.

The circuitry is capable of providing about 600mA so it could pop a fuse but I would like to keep the cost down to the bare minimum. How is everyone else providing short circuit protection? I was thinking of going with a bunch of PTC devices but even at 25 cents each this would add $4.00 to each unit.

Personal projects http://alan-parekh.com
Other cool project ideas http://hackedgadgets.com
 

Offline DJPhil

  • Frequent Contributor
  • **
  • Posts: 511
  • Country: 00
Re: Output protection
« Reply #1 on: September 02, 2010, 02:45:36 am »
I think it would depend a lot on how your circuit is set up and what degree of protection you're after.
(I don't know exactly what you've got, so I'm thinking of something like a computer power supply.)

If speed isn't really a factor and you're only concerned about preventing a hard short you might consider protecting the primary 12V rail further back in your circuit. A hard short on any of a bunch of parallel outputs would drive up the total current draw on the supply rail well above nominal output and trigger any protection you choose: fold back limiting, fuse, whatever. You can implement a much more thorough protection scheme on one rail for the money than attempting to protect each separately. The only downside is that a fault drawing >250mA and less than your protection scheme (4A plus margin) isn't handled, and your power supply will happily incinerate whatever's attached if it can't take the load.

Adding circuitry to protect each output would make the supply more robust, but more costly (as you mention). At some point it may be cheaper to regulate each output separately with a three terminal regulator and take advantage of the inbuilt protections they offer. Of course, this doesn't really qualify as cheap, but I don't think there's an elegant method to have it both ways.  :-\

Hope that helps. :)
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9330
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Output protection
« Reply #2 on: September 02, 2010, 03:23:12 am »
How about use PTCs to protect a group of outputs?
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline abbtechTopic starter

  • Contributor
  • Posts: 39
    • Alan's Electronic Projects
Re: Output protection
« Reply #3 on: September 02, 2010, 04:55:29 am »
Thanks for the feedback DJ Phil (I thought it was DR Phil at first  :) ) and Mike

The outputs will be used to power LED lighting, in our existing product we have biased the output transistor to power the expected 20 to 40mA LED load and on a short circuit it draws something like 80mA, no damage to the circuit results from a short circuit.

On the new design there is a need to drive much more current, I am thinking that it will still be a simple transistor driven output but since it will be biased to drive hard I am sure that it will be damaged if the output is shorted. That being said I am not sure if a fuse or a PCT would act quick enough for the 2N4403 transistor to be saved anyway.

The power supply will be a 12V 5A supply with an on board 5 volt regulator for the low voltage electronics.  

I was thinking of having a current sense resistor that could bias a transistor to kill the circuit when desired. This circuit looks promising.
http://www.vidisonic.com/2008/07/10/current-limiting-circuit/

I would still need to flip the circuit since I will be using a low voltage MCU to turn on and off the output.

Any thoughts?


Personal projects http://alan-parekh.com
Other cool project ideas http://hackedgadgets.com
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9330
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Output protection
« Reply #4 on: September 02, 2010, 04:59:53 am »
Don't include overload protection and warn that overloading the unit will damage it.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

alm

  • Guest
Re: Output protection
« Reply #5 on: September 02, 2010, 07:04:55 am »
I doubt that a current sense resistor and some transistors is going to be much cheaper than a fuse. Neither is a PTC or polyfuse. Their advantage is that you don't have to replace the fuse after an overload.
 

Offline DJPhil

  • Frequent Contributor
  • **
  • Posts: 511
  • Country: 00
Re: Output protection
« Reply #6 on: September 02, 2010, 07:56:09 am »
I made a horrible mess in LTSpice trying to sort this out. I've attached a picture of the circuit and the .asc file in case you find them useful.

I think I've got it to the point where it'll put ~30mA across the LED and if you short the LED the fault current is limited well below the transistor's dissipation rating. Most of the load is taken up by the 330? resistor, and totals a bit over 400mW. I haven't figured out if there's an easy way to switch the circuit without another component, I'm exhausted and also a noob.

It might not be cheaper, I'm not sure. The dissipation of Q3 are especially low, and I don't think I managed to get more than about 70mW out of Q4, so you might be able to use transistor array ICs.

It might be worth playing with, though in the end I think Alm's probably right. PTCs are likely to be the sweet spot between ease and cost.

Hope that helps. :)
 

Offline abbtechTopic starter

  • Contributor
  • Posts: 39
    • Alan's Electronic Projects
Re: Output protection
« Reply #7 on: September 02, 2010, 08:30:23 am »
I made a horrible mess in LTSpice trying to sort this out. I've attached a picture of the circuit and the .asc file in case you find them useful.

I think I've got it to the point where it'll put ~30mA across the LED and if you short the LED the fault current is limited well below the transistor's dissipation rating. Most of the load is taken up by the 330? resistor, and totals a bit over 400mW. I haven't figured out if there's an easy way to switch the circuit without another component, I'm exhausted and also a noob.

It might not be cheaper, I'm not sure. The dissipation of Q3 are especially low, and I don't think I managed to get more than about 70mW out of Q4, so you might be able to use transistor array ICs.

It might be worth playing with, though in the end I think Alm's probably right. PTCs are likely to be the sweet spot between ease and cost.

Hope that helps. :)

Wow thanks for taking the time to look into this and draw this up. I have to apologize about the transistor I will be using to do the switching. I have the old version on my desk and read it from there. The existing version has a 5 volt micro and 5 volt outputs, it is providing a switched positive output using the 2N4403. The next version will have a 5 volt micro, 12 volt outputs and will be providing a switched negative to the LEDs using 2N4401 transistors.

As far as cost goes I think using a transistor based design could go a long way to keep the cost down since the PTC I was looking at was around 25 cents and the 2N4401 transistors are ones that we stock and they cost us just under 2 cents each. A fuse and solder in fuse holders would be around 45 cents.
Personal projects http://alan-parekh.com
Other cool project ideas http://hackedgadgets.com
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14129
  • Country: gb
    • Mike's Electric Stuff
Re: Output protection
« Reply #8 on: September 02, 2010, 09:04:43 am »
This can be a surprisingly difficult problem, more so as numbers of outputs increase.
The problem with  PTCs is that the output MOSFET has to withstand the output current while the PTC heats up. If your supply has a high current capability, this can easily be tens of amps.
The simple answer that is to use a bug chunky MOSFET - in practicce you need at least a TO-252 style package, however this costs money and PCB space, and the increased gate capacitance may cause problems in PWM applications.

I recently did a 48 channel 24V LED dimmer with 1 amp per channel output,so you potentially have a 50 amp PSU waiting to vapourise the FETs in a fault condition.


For space reasons I wanted to use SO-8 dual mosfets, and I found that protecting these with PTCs wasn't possible - the FET always blows first.

The answer I found was these nice internally-protected FETs for automotive applications  
. OnSemi also do some similar devices - I don't recall the part mumbers.
These switch off in a few microseconds when shorted, making them indestructible.
The only downsides are that the internal current-sensing resistor makes them dissipate more power than a  standard FET, and being aimed at the automotive market the leadtime can be quite long.

You could make a similar arrangement with standard FETs and a current sensing circuits,  and you could use comparators to keep the drop on the sense resistor down.
« Last Edit: September 02, 2010, 09:38:46 am by mikeselectricstuff »
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9330
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Output protection
« Reply #9 on: September 03, 2010, 12:57:45 am »
How about a bunch of diodes from the low side of each output to a comparator? An overload will cause the voltage to rise and trip the comparator. You might even be able to use two resistors and a transistor or sensitive gate SCR to shut off the power supply using the feedback circuit.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline iainc

  • Newbie
  • Posts: 1
Re: Output protection
« Reply #10 on: September 23, 2010, 02:44:00 pm »
I'm a bit late to this, but in addition to the IR parts suggested by Mike you could also consider an ST VN800S or VN800PT (depending which package you're after) or an Infineon BTS4140N.  These are also Automotive parts, but if you're in the UK then Arrow, for example, usually carry stock of the BTS family.

Good luck!
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14129
  • Country: gb
    • Mike's Electric Stuff
Re: Output protection
« Reply #11 on: September 23, 2010, 04:58:18 pm »
I'm a bit late to this, but in addition to the IR parts suggested by Mike you could also consider an ST VN800S or VN800PT (depending which package you're after) or an Infineon BTS4140N.  These are also Automotive parts, but if you're in the UK then Arrow, for example, usually carry stock of the BTS family.

Good luck!

There are a few, ONSemi also do some,and these are all fine for on/off control, but if you want to do PWM be aware that many  of these devices have significant turn on/off delays (10s of uS) so read the datasheet carefully. Some of them also need some series resistance in the gate for the protection to work properly.
 
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf