If you want to be triple safe, do some of the additional things you suggested in your original post. Probably not necessary, but I always tend to overengineer anything which takes any amount of heat (having had components fall of boards myself, although usually only surface mount fets and/or regulators in fault conditions). The biggest thing I've discovered is to put as much outer-layer copper as possible at the component leads, as this will help pull heat away from the joint as well. But I'm glad to hear that theory and past experience worked in this case ;) On Thu, Jan 25, 2018 at 10:48 AM, Bob Blick wrote: > Hi Forrest, > Good points, and thanks for the video. I decided to test for myself, so I > made a string of 2W metal oxide resistors connected to a power supply and > cranked it up so the resistors were at 2W each. With full leads, 8mm from > any surface and 21C ambient. > > Using .4mm(0.15") 63/37 solder for my test, solder melted quickly on the > body of the resistor. But I could not get solder to melt on the leads at > all. Attached is a picture with a small piece of solder wrapped tightly > about 1mm from the resistor body. After 20 minutes there wasn't even any > flux escaping from the ends of the solder. > > A magnet sticks to the leads of these resistors. Most power resistors see= m > to have similar leads. I have another (more expensive) brand that are > non-magnetic, and although their datasheet shows the leads as "CuFe" they > are not more specific and I did not do a solder test on them. > > In summary, this makes me feel good. I wanted to rule out failures due to > resistors desoldering and falling out of the board. Using lead-free solde= r > will give me an extra 10-15 degrees of margin for high ambient conditions= .. > Board scorching is another risk, so I will space them off the board as mu= ch > as reasonable. > > As I've mentioned before, I'm only worried about a temporary overload > condition until other safety systems kick in. Under normal conditions the > resistors will be dissipating under 1W. > > Thanks to everyone for your input, this has been fun. > > Bob > > ________________________________________ > From: piclist-bounces@mit.edu on behalf of > Forrest Christian > Sent: Tuesday, January 23, 2018 1:55 PM > To: Microcontroller discussion list - Public. > Subject: Re: [EE] Resistor temperature rise > > Since this thread seems to be still alive, I'll add my $0.02: > > The degrees rise figure is the typically body temperature, not the > temperature of the leads at the board, which be much cooler. Note tha= t > on page 10 of the datasheet you linked it shows recommended lead > termination bending, effectively offseting the resistor off of the board = by > 8mm. I was trying to find a good thermal camera picture to show this, t= he > closest was at https://www.youtube.com/watch?v=3D-e-qIjttIIY&t=3D134s , a= t > 1:57 > in, there is a thermal camera pic which shows the leads basically being > cold, while the resistor body itself being hot. I wouldn't be surprised > to find that when mounted as recommended, the units will not desolder fro= m > the board. > > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > > --=20 *Forrest Christian* *CEO**, PacketFlux Technologies, Inc.* Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602 forrestc@imach.com | http://www.packetflux.com --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .