|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
30th Nov 2021, 1:17 pm | #1 |
Octode
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
|
SMD failures
Hi peeps...., Is it me ?? am I missing something ? I am developing a 2 metre front end...nothing unusual in that..however.. I designed the front end pcb with SMD 0603 devices in mind,...... the circuit is a copy of the Mutek preamp, using original coils from a Mutek module, upgrading the "leaded" caps to SMD. However, I find that the capacitors are very fragile, not liking being recycled, particularly the 10 nF and 100 nF devices. Yesterday .. I was on Ver 14, which looked very promising, as I had found I had missed a 1.8Pf cap from the original "copy"... My "M O" is to test the circuits, one at a time...with a 10nf coupling cap to a SMB socket for convenience...Gain was intermittent.... again it was traced to the 10N that was outputting...unsoldering.. the end came off...
replacing cured that...... Resistors seem to be much more "reworkable"...am I being too "tight" re using caps ??
__________________
Should get out more. Regards Wendy G8BZY |
30th Nov 2021, 2:29 pm | #2 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: SMD failures
Yes, for SMT multilayer capacitors. Thermal shock of soldering them by hand is risky, they are easily cracked and cracks don't necessarily show.
The little devils are supposed to be stored in low humidity environments before use. Dried in ovens if necessary, placed on two dods of solder paste, and then heated to soldering temperature and back down using carefully controlled time/temp profiles which can be quite complex. This way, they're pretty reliable. Modern production plans crash in flames if there is any significant DOA rate. Done with a hand soldering iron, warm the pad on the board by partial contact with the iron and with the part touching the pad. Use its thermal mass to filter the rise rate. Sneaky soldering! David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
30th Nov 2021, 3:27 pm | #3 |
Nonode
Join Date: Jan 2016
Location: Stafford, Staffs. UK.
Posts: 2,529
|
Re: SMD failures
SMT capacitors are fine when processed correctly.
If they are stressed in any way (heat, strain) they crack very easily, as already said. This becomes very important in certain circuits when there's a DC element as that can cause the metal layers to creep through the cracks and short the capacitor. So these failures take a while to show. And can be spectacular if the DC is for example the main DC supply to something. |
30th Nov 2021, 3:34 pm | #4 |
Octode
Join Date: Mar 2016
Location: London, UK.
Posts: 1,046
|
Re: SMD failures
I have been wrestling with this issue recently, and got some responses (thank you!) from Duncan and David here
All of the research I've performed suggests that unless you really need the package size or are making sufficient production volume commercially, through-hole MLCCs make life a lot easier. |
30th Nov 2021, 4:47 pm | #5 |
Guest
Posts: n/a
|
Re: SMD failures
I have no problem with hand moving SMTs, as long as the bit is big enough to cover both ends (NOT a tweezer). I let surface tension move the component and plop it into place, this works best for the weeny ones and with proper leaded solder and some flux. I never bother to "tidy up" the job, as long as both ends are connected that will do. MLCC capacitors are in themselves remarkably resisient, it's the ends that drop off. I used to work at Vitramon, we made multilayer ceramic capacitors just down the road from me. The body is subjected to 1000C plus, the ends are not!
It is a wipe off, wipe on, technique. Last edited by Guest; 30th Nov 2021 at 4:48 pm. Reason: Added a bit |
30th Nov 2021, 9:54 pm | #6 |
Octode
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
|
Re: SMD failures
Hi gents.. I guessed as much...I tried to veer away from leaded as in leeeeded components to keep the footprint smallish. I always used leaded.."ledded" solder... dinasaur that I am. Obviously as I am "amateur" the processes as described by David are not practical...... I should order a few more 10 and 100 nan 0603's. and not recycle.
__________________
Should get out more. Regards Wendy G8BZY |
30th Nov 2021, 10:08 pm | #7 |
Nonode
Join Date: Oct 2010
Location: Tintinara, South Australia, Australia
Posts: 2,324
|
Re: SMD failures
Buy the 10n & 100n by the reel - it works out to a fraction of a penny each (plus you won't run out in a hurry )
|
30th Nov 2021, 10:42 pm | #8 |
Heptode
Join Date: May 2015
Location: Chatham, Kent, UK.
Posts: 947
|
Re: SMD failures
Hi i am building the USDX 8 band hf sandwich rig , its the size of a credit card all smd 0805 all hand soldered inc the Si5351A-B-GT MSO10_3x3mm_P0.5mm clock chip. The pa is run in class E gives around 5 watt out on ssb ? from a tiny
IRLML2060 (sot23) fet . If any one was going to build this rig i have some of the chips spare (Farnell quoted back order 14-11 2024) any way back on track yes smd easly damaged by heat , Mick |
1st Dec 2021, 12:21 am | #9 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: SMD failures
I do a lot of work with SMD capacitors and resistors and if removing them (say from a scrap PCB) I always use two fine tipped irons to melt the solder at both ends and 'tweezer' the component off as soon as the solder melts. You can get specialised soldering irons which are like a pair of heated tweezers, but I don't have one of hose.
Using only one iron, rapidly zipping from one end to the other while applying a little bit of lateral pressure to the component so that it moves when the solder melts is not the right way to do it, there is quite a good chance of breaking one of the end contacts off that way. Another little stunt I use to either hold an SMD so I don't misplace it, or hold it still while I test it, is to temporarily solder only one end or leg (in the case of a transistor) to any handy solder joint or blob on the PCB. As we all know, once you lose track of an SMD you may never find it again. |
1st Dec 2021, 10:53 am | #10 |
Hexode
Join Date: Dec 2013
Location: Torrington, Devon UK.
Posts: 445
|
Re: SMD failures
I would not reuse them unless it was just temporary. I’m no expert but I’ve started replacing smd electrolytics with smd tantalums. I did a bit of research as I had heard they are tricky to hand solder. I came across this document from NASA which I found very interesting especially the section on the ideal way to solder them https://nepp.nasa.gov/files/16600/06...Capacitors.pdf
|
1st Dec 2021, 11:53 pm | #11 |
Octode
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
|
Re: SMD failures
Very interesting . No wonder the space programme is so expensive.. I can hold my head high, knowing that if its good enough for NASA.. "soldering SMD's" that is.. I should be ok.
__________________
Should get out more. Regards Wendy G8BZY |
2nd Dec 2021, 1:50 am | #12 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: SMD failures
I find that there is a huge variation in ruggedness when it comes to MLCC parts.
The old Murata GRM42 100nF 50V from the mid 1990s was fairly bulletproof and these were in 1206 package. Normally 1206 is one of the most fragile packages. The Murata 1nF COG 0805 parts from the 1990s were equally robust. These Murata parts had tiny codes stamped on them 1A5 and 1A3. They could be reused and chased around a PCB with a hot soldering iron with very little risk of damage. I keep meaning to buy more of the GRM42 100nF caps from Murata, the particular part number went obsolete many years ago but they are still available from some traders. I'd also like to buy more of their old 0805 1nF parts. By comparison, the modern equivalents from the other manufacturers are very fragile. Note that it really isn't good to place an MLCC DC blocking cap near an RF connector at the edge of a PCB as the MLCC will eventually crack after several connection cycles of the RF connector. The old Murata caps and some ATC cap types can survive in these locations but a modern MLCC won't last long if stressed like this.
__________________
Regards, Jeremy G0HZU |