I've mentioned this in passing before, but hoped to get more advice as
this is my first highish-current A/C device.
I've designed a board that uses TRIACs to control A/C, probably at 120
VAC. Since there are no outlets on this device, I opted to run only one
leg of the A/C on the board. IMHO, this SHOULD reduce chances of Bad
Stuff happening, as a short on the A/C traces should only force a
circuit on.
I'm not putting fuses on the PCB either, as the case these are mounted
in will have a fuse. Is this likely to cause any problems? My
conscious tells me I should fuse the inputs and outputs on the PCB, but
my mind tells me I should be fine with an external breaker, once again
because a short across the A/C lines cannot be between AC-line and
AC-neutral. The outputs are also on breakers to limit the current to 5
A/output.
The minimum trace width on this thing is about .250 inches wide, with
2oz copper, so my quick lookup of this means I should have no problems
with the 5 A/circuit, with 15 A max per set of four outputs I need.
Should this be acceptable for a small number of these devices, not mass
production?
Hummm... I'm sure more knowledgeable people on piclist will chime in,
but as far as shorts and fuses ... what if a grounded metal object
contacts the pcb? kerpow. Or one of the AC devices fails as a short?
Are triacs guaranteed to fail open-circuit?
Personally, fuses are cheap and relatively small, why not use them.
The digikey part# for the fuses and fuse holder aren't on the page,
email me and I'll look them up if you'd like. It took a little while to
find a suitable fuse (readily available, small, 15A, and holders rated
for 15A as well).
> I've mentioned this in passing before, but hoped to get more advice as
> this is my first highish-current A/C device.
>
> I've designed a board that uses TRIACs to control A/C, probably at 120
> VAC. Since there are no outlets on this device, I opted to run only one
> leg of the A/C on the board. IMHO, this SHOULD reduce chances of Bad
> Stuff happening, as a short on the A/C traces should only force a
> circuit on.
>
> I'm not putting fuses on the PCB either, as the case these are mounted
> in will have a fuse. Is this likely to cause any problems? My
> conscious tells me I should fuse the inputs and outputs on the PCB, but
> my mind tells me I should be fine with an external breaker, once again
> because a short across the A/C lines cannot be between AC-line and
> AC-neutral. The outputs are also on breakers to limit the current to 5
> A/output.
>
> The minimum trace width on this thing is about .250 inches wide, with
> 2oz copper, so my quick lookup of this means I should have no problems
> with the 5 A/circuit, with 15 A max per set of four outputs I need.
>
> Should this be acceptable for a small number of these devices, not mass
> production?
>
> --Michael
Jesse Lackey wrote:
> Hummm... I'm sure more knowledgeable people on piclist will chime in,
> but as far as shorts and fuses ... what if a grounded metal object
> contacts the pcb? kerpow. Or one of the AC devices fails as a short?
> Are triacs guaranteed to fail open-circuit?
If a triac fails closed circuit, all it means is the power for that
circuit is on. I only have one leg of the A/C on this board.
> Personally, fuses are cheap and relatively small, why not use them.
Because an internal fuse requires someone to open up the case, rather
than pressing the externally pressable breaker button.
I think I'll add a fuse on the input to limit the current across the
traces to something I know they are safe for, but not fure each output
to guarantee its limit. I'll recommend the user add a breaker per
output channel to limit to a safe value, or limit the input to be a max
per-channel safe value.
At 04:48 PM 12/16/2005 -0600, you wrote:
>I've mentioned this in passing before, but hoped to get more advice as
>this is my first highish-current A/C device.
>
>I've designed a board that uses TRIACs to control A/C, probably at 120
>VAC. Since there are no outlets on this device, I opted to run only one
>leg of the A/C on the board. IMHO, this SHOULD reduce chances of Bad
>Stuff happening, as a short on the A/C traces should only force a circuit on.
>
>I'm not putting fuses on the PCB either, as the case these are mounted in
>will have a fuse. Is this likely to cause any problems? My conscious
>tells me I should fuse the inputs and outputs on the PCB, but my mind
>tells me I should be fine with an external breaker, once again because a
>short across the A/C lines cannot be between AC-line and AC-neutral. The
>outputs are also on breakers to limit the current to 5 A/output.
A typical glass or ceramic tubular fuse will protect the traces, but not
the semiconductors.
A typical breaker will protect the wiring, but not the traces because they
allow very
high surges for more than enough time to vaporize your traces.
Is the AC line polarized? A ground fault could take out either half of the
circuit
if the plug isn't polarized or if the outlet is wired wrong. If you keep the
neutral half the circuit off the PCB (in relatively thick wires- like in
the line
cord or bigger) you should be able to get away with a single glass fuse in
the hot
(fusing the neutral is not kosher).
>The minimum trace width on this thing is about .250 inches wide, with 2oz
>copper, so my quick lookup of this means I should have no problems with
>the 5 A/circuit, with 15 A max per set of four outputs I need.
Are you sure you actually have 2oz copper? Is the 0.25" for the common line or
for the 5A circuits?
>Should this be acceptable for a small number of these devices, not mass
>production?
>
>--Michael
Watch the clearances, especially to ground and double especially to anything
a user might touch. You can get away with less in the US/Canada in some cases,
but creepage (clearance over a surface) of 8mm+ is good to have. If the
board could have condensation or other moisture on the surface, you'll need
more distance and probably conformal coating or even slots machined into the
board. Avoid having any floating metal in the housing etc.
Nothing connected to the AC line should be possible to touch, or should
be possible to come into contact with anything that can be touched. If in
doubt, have someone knowledgeable look it over (perhaps in mock-up form)
before you have boards made.
I am already using varistors for Tyco but I cannot find the PPTCs they
use for primary protection easily. Has anyone seen these (I can only
find PPTCs to 60V, I need 300V) ?
>On Sat, 17 Dec 2005, Peter wrote:
>
>>On this theme, what is the list's opinion on *inexpensive* thermal fuses ?
>
>Continuing on this theme, polyfuse type protection:
>
>http://www.electronicproducts.com/ShowPage.asp?SECTION=3700&PRIMID=&FileName=marray1.mar2003
>
>I am already using varistors for Tyco but I cannot find the PPTCs they use
>for primary protection easily. Has anyone seen these (I can only find
>PPTCs to 60V, I need 300V) ?
>
>thanks,
>Peter
There are some rated at 600V, but beware, their interrupting capacity is
EXTREMELY limited, a sad joke really.
They are NOT suitable for general use on mains circuits.
An inexpensive 1.25" 3AG glass fuse is typically rated for at least 10,000A
interrupting capacity at 125VAC.
Fuses are HUGE, especially if you want to use common replaceable types.
get depressed every time I try to "harden" one of my PCB designs by
adding fuses...
On 12/18/05, William Chops Westfield <.....westfwKILLspam.....mac.com> wrote:
>
> >> fuses are cheap and relatively small
>
> Fuses are HUGE, especially if you want to use common replaceable types.
> get depressed every time I try to "harden" one of my PCB designs by
> adding fuses...
Fuses are really HUGE but I have to put it in for UL approval. ;-(
Of couse big and samll are always relative to the board size. ;-)
I was using three fuses for one EEx ia product (Intrinsic Safety product
used in explosive environment) and the brand of the fuses to choose for
Ex use was really limited for PTB approval. But it is of course necessary
to put them in so that extra reliability/safety can be achieved.
I haven't read this whole thread yet, but I try to avoid fuses in my
products. As pointed out they are big. They also nuisance trip. When the
line current is low enough, I've designed in PTC thermistors for
overcurrent protection. BC Components has a series of them rated at
260VAC. I believe they are a UL recognized component. Our test lab has
listed our products as compliant with UL508 when we use these parts.
The other pain with line voltage on a circuit board is creepage distances.
It's real hard to run any traces when you need to leave so much space
between them!
> On 12/18/05, William Chops Westfield <EraseMEwestfwspam_OUTTakeThisOuTmac.com> wrote:
>>
>> >> fuses are cheap and relatively small
>>
>> Fuses are HUGE, especially if you want to use common replaceable types.
>> get depressed every time I try to "harden" one of my PCB designs by
>> adding fuses...
>
> Fuses are really HUGE but I have to put it in for UL approval. ;-(
> Of couse big and samll are always relative to the board size. ;-)
> I was using three fuses for one EEx ia product (Intrinsic Safety product
> used in explosive environment) and the brand of the fuses to choose for
> Ex use was really limited for PTB approval. But it is of course necessary
> to put them in so that extra reliability/safety can be achieved.
>
> Regards,
> Xiaofan
>
> At 11:26 PM 12/17/2005 +0200, you wrote:
>
>> On this theme, what is the list's opinion on *inexpensive* thermal fuses ?
>>
>> Peter
>
> The type used in many appliances (they open permanently when subjected to
> excessive temperature) seem to be fine. What's your concern?
Assembly on printed circuit board by soldering (by careless personnel).
One too long/too close soldering and you have an oopsed product. I would
also like it to be able to take 15A. The ones used in heftier appliances
seem to be right (ex: microwave oven etc). These are bimetal type afaik.
>> www.electronicproducts.com/ShowPage.asp?SECTION=3700&PRIMID=&FileName=marray1.mar2003
>>
>> I am already using varistors for Tyco but I cannot find the PPTCs they use
>> for primary protection easily. Has anyone seen these (I can only find PPTCs
>> to 60V, I need 300V) ?
>>
> There are some rated at 600V, but beware, their interrupting capacity is
> EXTREMELY limited, a sad joke really.
>
> They are NOT suitable for general use on mains circuits.
Contrary to what the article above says ?
> An inexpensive 1.25" 3AG glass fuse is typically rated for at least 10,000A
> interrupting capacity at 125VAC.
The biggest issue with fuses....is replacement. I have used PICO fuses for years, generally they never blow, just there...in case....
If you feel like the fuse may blow, and the board is hard to get to replace, then just the simple AGC 1-1/4 with an external holder. If the board is going to be in a moist enviroment, then decide....is this a throw away ? Should you pot the assembly? Maybe a combination...pot the assembly but stick a couple of sockets out of it for the fuse. It all depends on the application, and if you think a fuse will blow. Generally, I have found that a fuse will blow if they wire it up wrong...the first time....and after that, they generally never blow UNLESS something really bad happens....and thats what its there for.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
interlog.com Info for manufacturers: http://www.trexon.com
