12v clean power distribution

[qwertypo]

On my mobile sound cart I have several wireless RF receivers that run on 12v.  I'd like to build a small distribution box from my main 12v powersupply.  Mainly to prevent shorts from loose plugs, as I am currently using Y cables and W cables to multiply my power... loose coax jacks tend to short out on any grounded metal. 

Two questions.  I'd like to use a polyfuse on each output so if it does short, it will interrupt flow.

And, I'd like to consider filtering any potential noise on the power that could be shared between the receivers.

Its gonna be a 12 port box, with one power input.  The line that feeds its can handle 3A at 12v.  Each RF receiver has about 200ma demand at 12v.

Currently I was planning on running a rectifier on the input and then feeding that directly to each output, with a polyfuse just before each output.   And LED for  status of the polyfuse would be nice if easy to incorporate.   Ideally only on when the fuse is blown...

The addition of capacitors to filter out possible RF bleed into the powerline is the second question.  Typical RF range would be 500-900mhz.  I'd assume that the receivers are designed to filter the power themselves... but I had some recommendations to consider this in the design.  Any thoughts?

Thank you for any help.

[joaquins]

For radio recivers I think I'd use a 7812 in each output, but for that you should have at least 2V exceding power supply(recomend 3V), with 200ma and 15V PSU each one will be dissipating about 0.6W with only the TO-220 case without heat sink is about 30ºC+Ta, so you wont need more than that. If you wont more noise rejection you can add a couple of caps at outputs and maybe a bigger cap at the input for better regulation.

It has a short circuit current limit, but it will drop more than 1A so will protect but also will drop the others voltage, with normal load in 11 outputs.

If you use the polyfuse you could put a led across it with a resistor and if the output shorts it should turn on, but you should try it, I'd never use one.

JS

[Speedskater]

Is your main 12V power supply a battery or a line voltage AC to 12 V DC supply?  If it's the latter, does at have any output adjustment range?

[JohnRoberts]

PTC fuses are useful resettable fuses.

You can pretty simply wire a led in series with a resistor across the PTC. As long s the PTC is cold and low impedance the LED is off, if the PTC switches to high impedance the LED will turn on.

I did something similar in an old console power supply I did years ago... I used bi-color leds on each output, Green meant that the PS rail was up and fuse was good. Red meant that the rail voltage was up, but fuse was open. Dark meant PS rail was down, no voltage.

Time to get it repaired, instead of an easy fuse replacement the customer could handle.

JR

[millzners]

Yeah I'd also go with a resetable fuse and a separate  ldo regulator for each output.  Seems you have ac on the input so after the rectifier you could filter with caps then regulate.  For rf filtering use a ceramic cap on the output of the regulator.  You could use an ldo like an LT1764 and maybe even get away with regulating the signal twice, one ldo after the rectifier feeds the separate ldo's for each channel.  Maybe you're not into overkill, but I tend to use this arrangement all the time.

[CJ]

watch out for the polyswitch ratings, if it says 100 ma, it is good to use on a 50 ma circuit,

the reason is that if things get a little hot inside the box, the self heating mechanism of the polyswitch can lead to a false shutdown,

so for 200 ma, i would use a 400 or 500 ma polyswitch.

you can tell if their tripped, the get pretty warm,

and they do not work well in parallel, one will get hot and shutdown before the other,

[ricardo]

watch out for the polyswitch ratings, if it says 100 ma, it is good to use on a 50 ma circuit,

CJ, while I understand your concerns, if the breaking current was critical, I'd do it the other way round.

For a 50mA 'never exceed' current, I'd use a 25mA Polyswitch.  I've tested many Polyswitches in da old days and they all switched at about twice their rating.

If the current rating was just to prevent the shebang bursting into flames, this isn't that important.  But I don't think you can find Polyswitches of 50mA, let alone 25mA.

[CJ]

i had a bad time with those switches shutting down the product i was shipping,

it is very similar to thermal runaway in a transistor, in fact, that is about the same mechanism they use to make those things,

you can put an ammeter in series with the switch and watch the current sloooooowly creep down, then it creeps down faster, and then real fast till it gets warm, the I^2R losses will then hold the switch off, till what ever it is fusing is either turned off, or blows up, which is unlikely since the circuit is now current limited.

so yeah, i went back and forth with digikey on this till i got something that would work, the tolerances on those things are pretty loose, so you have to protect against the occasional lemon.



and they do not like high voltage, if they shut down across a lot of volts, they tend to explode like a tant cap hooked up backwards,

[JohnRoberts]

Fuses are not very precision either, but when properly applied PTC fuses work and reset when the fault is removed. 

While a little higher tech you could design a more precision current limit circuit using active devices. An opportunity to learn about fold-back current limiting, where the current output is reduced for dead shorts (based on actual output voltage), while you could size the current limit pass devices to handle the dissipation for full current into a dead short circuit.

A practical current limit circuit will need a few tenths of a volt to detect current draw, while you might make that even smaller with even more added complexity.

JR

PS: I used PTC parts in a trick power amp, back in the '80s (PMA70+) where I used the PTC to power limit the average power of a boosted rail amp. The 35W amp would make over 100W on transient peaks, 60W limited by the PTC (for roughly 15 secs), and 35W all day long.  With dynamic music it sounded like a 100W amp.

[qwertypo]

Is your main 12V power supply a battery or a line voltage AC to 12 V DC supply?  If it's the latter, does at have any output adjustment range?

2 12v Lead Acid batteries supplying a power distribution system that allows AC input to charge the batteries.  Basically a high end UPS meant to be clean for audio use

[qwertypo]

Yeah I'd also go with a resetable fuse and a separate  ldo regulator for each output.  Seems you have ac on the input so after the rectifier you could filter with caps then regulate.  For rf filtering use a ceramic cap on the output of the regulator.  You could use an ldo like an LT1764 and maybe even get away with regulating the signal twice, one ldo after the rectifier feeds the separate ldo's for each channel.  Maybe you're not into overkill, but I tend to use this arrangement all the time.

I got a couple resetable fuses... 3A with 1s trip, and .5A with .1s fuse... gonna test for timing on a short, as well as stability over 16 hours use at my typical capacity... I want a fast trip in the event of a trip.. but I'd love 3A per capacity if possible.   Might add an extra 3A at the main as that is my limit per line from my main supply.

The 1764 is a voltage regulator?  Will that filter noise?  Or just protect against fluctuation?

[qwertypo]

watch out for the polyswitch ratings, if it says 100 ma, it is good to use on a 50 ma circuit,

CJ, while I understand your concerns, if the breaking current was critical, I'd do it the other way round.

For a 50mA 'never exceed' current, I'd use a 25mA Polyswitch.  I've tested many Polyswitches in da old days and they all switched at about twice their rating.

If the current rating was just to prevent the shebang bursting into flames, this isn't that important.  But I don't think you can find Polyswitches of 50mA, let alone 25mA.

My main concern on the individual outs is shutting off in case of short circuits.  The equipment is on a metal cart, and its all grounded, so its easy to make the entire cart hot with a misplaced cable.

[qwertypo]

Thank you for all the replys.

Any suggestions on an LED to determine status?  I am in debate between a green "powered" led or a red "tripped" led or both if easy enough

[abbey road d enfer]

Thank you for all the replys.

Any suggestions on an LED to determine status?  I am in debate between a green "powered" led or a red "tripped" led or both if easy enough

There are bi-color LED's, green and red in the same envelope. Unfortunately, I think they are available only in common-cathode 3-wire configuration or back-to-back 2-wire, none of these being adequate for your application (or it would require too much glue to make it work). Do you have any practical reason not to use two separate LED's? These are so unexpensive, they are a natural choice.

[qwertypo]

Thank you for all the replys.

Any suggestions on an LED to determine status?  I am in debate between a green "powered" led or a red "tripped" led or both if easy enough

There are bi-color LED's, green and red in the same envelope. Unfortunately, I think they are available only in common-cathode 3-wire configuration or back-to-back 2-wire, none of these being adequate for your application (or it would require too much glue to make it work). Do you have any practical reason not to use two separate LED's? These are so unexpensive, they are a natural choice.

no problem with using two at all, I am sure I have them lying around somewhere anyway.  thanks for the drawing, I give that a try and see how it works.

[millzners]

Yeah I'd also go with a resetable fuse and a separate  ldo regulator for each output.  Seems you have ac on the input so after the rectifier you could filter with caps then regulate.  For rf filtering use a ceramic cap on the output of the regulator.  You could use an ldo like an LT1764 and maybe even get away with regulating the signal twice, one ldo after the rectifier feeds the separate ldo's for each channel.  Maybe you're not into overkill, but I tend to use this arrangement all the time.

I got a couple resetable fuses... 3A with 1s trip, and .5A with .1s fuse... gonna test for timing on a short, as well as stability over 16 hours use at my typical capacity... I want a fast trip in the event of a trip.. but I'd love 3A per capacity if possible.   Might add an extra 3A at the main as that is my limit per line from my main supply.

The 1764 is a voltage regulator?  Will that filter noise?  Or just protect against fluctuation?

The LT1764 is a low drop out (LDO) voltage regulator, it's function is to take a DC voltage and to provide a low noise, low ripple DC output; it's a low drop out regulator because it only requires a few hundred millivolts drop across the regulator to do this as opposed to 2 volts like an LM317 for example.  It will help to reduce/isolate noise from the source DC voltage getting into the output of the regulator, and since you don't need much of a voltage drop, you can put them in series to help isolate power supply rails from each other.

For example, if you have 12 independent supply lines you need to make, you could just feed all of them off the output of one regulator.  But if one of those 12 supply lines has a noisy part on it, that noise is going to infect everything else, the regulator won't be any help at all.  However, if you were to feed 12 different regulators from one central regulator, you'd prevent that since each line would then have significant isolation.

Filtering the noise is different, but also important, you'll want a lot of capacitance on the input and output of the regulators, and you'll want a mix of ceramic/film and tantalum/electrolytic.

[qwertypo]

So I've got this system prototyped on a socket board.

Both my .5a and my 3A polyfuses trip very quickly on a short.  This is good.  The 3A fuse.. gets very hot in a shorted state, but does seem to protect everything else from the load.   I'm a little nervous about if it was shorted for a long period if it would melt for burn.  Can these things hold this load for a long time? 

The other issue I am curious about, it I was hoping the fuse would stay tripped until the voltage is removed... but it resets instantly when the load is removed.  Anyway to achieve my desired effect?  As a result if I short the circuit, things will power up immediately after the short is removed, and if I overload a circuit, it results in the device constantly power cycling.  I guess preferable if the fuse was tripped, it would act more like a tradition fuse, and actually switch off.... afraid this thing will smoke up.

Thanks for any advice.

[qwertypo]

Thanks for all the feedback... the resulting unit came out beautiful and works great... I plan to build two more!

cheers