diode replacement.

[deth502]

here goes the long story. for the short version, skip to the end....

i have a cheap chinese plasma cuter- a cut50, forget the brnd, that i have been using for a while. 110v machine and it works great on sheet metal, with the dial usually at about 1/3 up. today i had to cut some 5/16" plate and i thought id give it a try to make things easier. fired it up and it only gouged about half way trough. so i turn the amperage dial up to about 3/4 of the way, and have a go at it. it cut for about 1.5-2 sec and a saw a flash from the vents on the front of the machine, and that was that.

i opened it up and saw a resistor that was burnt to a crisp, it still read at 25.23k so i going to take a huge leap of faith and assume this is a 24k resistor. there are also 2- 330k resistors that apparently arced and welded themselves together. i dodnt test them at first, but after i separated them they both still read within spec. these are just like giant 1/4 resistors, i thiknk they are 5w, but they are not ceramic, ive never seen anything like them before. but to get back on track..  before the resistors on the output side (where all the power comes out) there is ...kind of... a set of 4 diodes, only one has been doubled up on. i checked these and several were open. im assuming this is why the resistor caught fire. searching the internet, i found a video of a guy with apparently the EXACT same problem i had. you can see in his video the one doubled up diode i spoke of also. the board layout is a bit different, but...



so, it would seem this is a common problem. i dont want to do it again. searching the components, the one i removed seems like a very plain jane (other than being rated to 2kv@.2A) diode.

http://www.semicon.sanken-ele.co.jp/sk_content/rp%201h_ds_en.pdf

now, i CAN get this diode to replace them, but again, i dont want to come back to this again. so im thinking of throwing in some hvm12 to beef it up

https://www.mouser.com/datasheet/2/345/hvm-1201156.pdf


----------------------------------------------------

short story..

so, would a hvm12 be a good replacement for the rp1h in this instance?

[keymaster]

so, would a hvm12 be a good replacement for the rp1h in this instance?

No , rp1h is a Fast Recovery Diode ,hvm12 isn't.

[deth502]

ok, thanks. see, thats why i ask, im no good at dissecting all of the subtleties of a data sheet.

is there any beefier part (that will not break the bank) that is suitable to put in in their place?   

[keymaster]

it is surprising difficult diode to match .
1. GAP3SLT33-214
is faster ,it can handle digger voltages , more current . but is in smd package .You can use the leads from the old diode .
(https://www.mouser.com/ProductDetail/GeneSiC-Semiconductor/GAP3SLT33-214?qs=sGAEpiMZZMtQ8nqTKtFS%2fMVO5fHW5uUQrQbtp%2f9MuOSn8pflXoqULg%3d%3d)

2.BY203-20STAP
is 3 times slower(worse than the original but still way better than hvm12) it can handle more current and in through Hole package . the cheapest of all .
(https://www.mouser.com/ProductDetail/Vishay-Semiconductors/BY203-20STAP?qs=%2fha2pyFadujR63Y9vK37TLjHex%252bXgi0gaHMaPtzwGlH1gvElXe77Yg%3d%3d)

3.DH40-18A
same speed ,it can handle a lot more current(with the proper heatsink) but is 1800V (vs 2000V of the original)
(https://www.mouser.com/ProductDetail/IXYS/DH40-18A?qs=sGAEpiMZZMtoHjESLttvkpfqOErzsE88yk8kJN%252bq6B4%3d)

I personal would go for the no.1

[deth502]

thanks! looks like ill be ordering some of #1. being smt i could always build a little breakout board to mount them too. hopefully the bigger specs will have them lasting the long haul (or close to it) over the stock parts that seem a bit puny for the application. the $10-13 seems a bit steep, but a lot better than some of the diodes ive seen searching ($30+!!! wtf?!?)

[Yansi]

Why not to get the RP1H directly?

[deth502]

again, because mine broke, and in searching the net, apparently a lot of other peoples broke. makes me think the part is under spec'ed and im not looking to make this repair a regular thing.

[Yansi]

Okay, understood!