'Think I fried 2 DC board channels, can I replace the triacs?

[TJ Hvasta]

Last year I bought the QC(spade lugs) version of the CMD16 DC board.. has been working well all year (landscape lighting).. Well, 'couple weeks ago, setting up for Halloween, I mismatched the RFlood and RSpots cat5 sockets to the lamps, plugged in a couple of Rainbow Floods (some hardwired, some socketed..) well, the hardwired ones, I think burned the DC gate on channels 2 and 5.. I noticed it got reeeeally hot right away.. I shut off the board, pulled out the cat5 cable.. checked the plugs.. yeah, lil melted plastic on the cat5 sockets.. got it sorted out (and each properly labeled, thank you very much), but now with the board powered, those two channels stay on.. and the two heat sink tabs are darker than the rest.. Has anyone replaced a DC board triac or is it something Dan and his elves have to do?

[-klb-]

I believe they are MOSFETs, not triacs. If you can find the right part numbers, it should be no different than changing a triac.

[TJ Hvasta]

yep, mosfets.. knew that I'll dig out the manual for part #'s.. Thanks!

[-klb-]

I know at least 2 different parts have been used. You might want to verify any part numbers you find against the actual labels.

[TJ Hvasta]

I have the controller in the stack of others from clearing the Halloween display, so best is to look at whats stamped on the case. Thanks.. was just watching Tim's Sandstorm bouncing house video..

[TJ Hvasta]

Same year and a half old CMB-16D-QC board, but a new problem... Dan sent three mosfets to replace the bad ones.. thank you Dan!

These mosfets (original, not the replacments).. has anyone else found they get very brittle, to the point the body breaks off the legs? I have 5 that have broken off at the slightest touch with 2 more about to break off if I look at them crossly.. I understand these are DC isolated (no common metal heat sink allowed) but isnt there some way to secure the tabs together so they dont flex? The dongles (Cat5 socket pins to quick-connect spades) are zip-tied together, but it seems any flexing at all just does these in.. I have a ticket into LOR to see if they can send me replacements or some suggrestion how to secure them, otherwise, I'll prob have to look to Mouser for 16 replacments

[WilliamS]

now i know this is a tad dated but in terms of where you were in error in connections Im lost in what exactly happened? Did you create power wires out of Cat5 or am I just too noob to read between the lines? Or am I seeing it as some where wired differntly using cat5 for ease of connections but you shorted something out having the wrong lamp plugged in?

[-klb-]

Nothing shorted, and this is a DC low voltage controller, so nothing intrinsically wrong with cat5 for wiring, as long as the loads are calculated to work with it.

But it sounds like the mechanical flexing of the cables is brushing against the MOSFETs. Not sure why those would be so brittle though..

[TJ Hvasta]

My first two RFloods I hard-wired (soldered) Cat5 pigtails to the boards, the others after that I used the Cat5 sockets Greg supplied, and Cat dongles suggested by Steve last year.. so the wire pairs I used in the first two were different than the rest, plugged into the cat5 socket dongles which were plugged onto the lugs.. well, one of the pairs were + and - so that shorted the mosfet.. the second, the wires twisted, shorting, took out the second one.. as to why the others got brittle, I have no idea, not much flexing at all, but without any type of support structure (no heat sink tying them together) the dongles moving in the case(as little as it was) were enough to get them to bend/snap.. I have since secured the dongles to the inside if the case, so even if the cable gets kicked/pulled it wont move inside.