I would definitely recommend using 0603 (imperial size) rather than 0805. I consider 0603 a good general purpose size for at lot of things, and yet they are still easy to handle with manual soldering, that be iron or hot air. Just my comments, although I know I did not add much more information here to all the other replies 
1 Like
It also seems to be the case that it is easier to get 0603 compared to 0805. Was especially noticable during the MLCC crisis.
1 Like
Yes, 0805 and larger is verging on obsolete/specialist.
The snag with the smaller parts in some of my projects is low voltage rating even when the power rating is no problem. Most volume consumer products are built with 0402 or 0603
1 Like
For me, 0603 is comfortable to hand-solder and all my designs use 0603 parts (if possible).
I generally use paste and a reflow oven most of the time but for prototypes I’ll often hand-solder first ( unless there BGA parts involved).
I use Kester 959T liquid flux in a squeeze bottle with a needle for SMD work. Can’t stress how good this stuff is for hand-soldering or touching up after a bad reflow. Also, high-quality solder is extremely important. I use Chip Quik SMD291AX paste for reflow and Kester 24-6337-8806 for hand soldering. Kester solder is the best you can buy in my opinion. I also have a Hakko 915F solder station with a collection of different tips. I find for soldering 0603 caps and resistors a chisel tip works best and for QFP and SOIC parts a bent conical (or flat) tip with a reservoir works best.
I’ve been able to successfully hand-solder QFN but I don’t recommend it.
With 0402 I’ve hand-soldered a few times but avoid caffeine because hand-shaking can be disastrous. 0402 parts often get stuck to the iron tip or go flying across the room never to be found again. Definitely need a 10x microscope or you can get +7 reading glasses on eBay which work well.
0805 looks enormous to me now and I rarely use this size in my work unless it’s a large value capacitor
1 Like
What I have seen is that 0603 and smaller is still quite common for “everyday values” and 1206 / 1210 are common for parts that are currently impossible to cram into an 0603 (very large capacitors, high voltage ratings, high power resistors, etc). 0805 I guess doesn’t offer enough of a useful middle ground
1 Like
they suggest me to use 0804 SMD Packages.
Here are I think important points to consider before applying a one size fits all rule (which usually is a bad idea):
-
It all depends on what your approach to soldering these SMD components is: I use DIP only for quick prototyping, but found 0603 easier to solder than bigger ones like 0805,1206 when using solder paste with my SMD rework station because you can cover more components with your hot air in a given area. Now when using a soldering iron, of course 0805 or even 1206 can be a better choice…
-
You should be careful not to use always your preferred size (i.e. 0603) but instead, always check the datasheet of your component during your design phase, as many smaller components have different characteristics.
The most common mistake being using a small resistor ref. that is 1/10W when your design needs at least 1/8W …
Same for capacitors, always check the max voltage of your particular package (should be at least twice what your design is supposed to use by default).
- Make sure you use hand soldering special footprints when you use a soldering iron
SMD soldering is both fun and efficient with a bit of practice, good luck!
-Fab
1 Like
I found 0603 letting me to get smaller PCB still enough easy to hand solder.
It is popular rule used for electrolytic capacitors. They are typically used at supply input and you should expect some overvoltage there during surge.
But for ceramic capacitors in the circuit where you don’t expect overvoltages I have seen few times the info that you can use them up to their specified voltage. Not sure at that moment as many years passed but as remember I sow in the TaiyoYuden specification that ceramic capacitors were tested for 2,5 times the nominal voltage.
So if you need 10 uF at 5V regulator output I think you can use ceramic 0603 10u/6.3V X5R capacitor.
1 Like
You can always try, but finding shorted capacitors is a real *&^%$#@!
I have no trouble at all with 0603, or probably even smaller, as I have a decent stereo optical microscope. I’ve heard several others complaining (mainly on youtube) that 0603 is already getting hard to solder because they are too small to see properly. This is especially true if do do not have a stereo microscope (Around EUR400) and your eyesight is getting less (often worsening with age, after about 30 years it seems to start slowly). A cheap partial solution is to use a lot of light. And I mean an stupendous large amount of light. When there is a lot of light, your iris contracts, and a smaller iris results in a “pinhole camera” effect on your retina and this gives a sharper picture.
To give a reference for a stupendous large amount of light
Bright sunlight is around 1000W per square meter (perpendicular to the sunbeams), and efficiency of LED’s is no better then 20%.
Your brain fools you. It does not reveal to you that your eyesight is getting fuzzy or a bit vague. It takes a conscious effort to focus on that.
1 Like
Since at least 10 years I need glasses to read.
I don’t have a microscope - I am using the table lamp with increasing glass.
I have just done experiment by looking at PCB directly and through this glass - the PCB looks being 1,5 times bigger.
Normally I don’t assemble our PCBs, but when I had to select elements for RFID antenna circuit I desoldered and soldered 0603 capacitors more than 200 times a day.
1 Like
These were just examples to incite beginners (or beginners with SMD) to always check the datasheet when using a package.
Now if I can understand that sometimes it can help your design a lot to be minimum with your maximum voltage specifications, it is not something I would recommend to someone starting.
Especially for high voltages, people should be even more careful (I saw a cap exploding and traversing a lab wall and the guys that dimensioned this cap did not do a great job and could have killed or caused a lot of harm to one of us…)
Finally, I would avoid as much as possible to make my design sensitive to a particular manufacturer for which you know about their testing, because I like to be able to provide component substitutions.
But again, there is not a one size fits all rules and reading you, I am sure that you know what you are doing when you apply a more relaxed rule …
1 Like
Many ceramic capacitors lose capacitance as the voltage across them approaches the maximum rated voltage. It’s often suggested that if the capacitance value is at all important you should select one rated at twice the expected working voltage.
When I made my first SMD board I was very hesitant to use 0603s. My last design used a mix of 0603 and 0402 parts and I didn’t find the 0402s that much harder to mount.
However, I have a stereo microscope designed for board repair and a cheap knockoff hot-air soldering system. I really can’t imagine placing these smaller components with a soldering iron.
1 Like
To tell you the truth I don’t use 6.3V capacitors at 5V
I use them at 3.3V. But I believe that ceramic capacitors can be used up to their nominal voltage.
When tantals were gone (I think it was 2002) I switched to ceramic uF. Those time I sow that TaiyoYuden testing information. And may be for higher capacitances the overvoltege was smaller.
I believe others also test for some higher then nominal voltage.
Long time ago when tube technology was standard I suppose ceramic capacitors were used up to their nominal voltage because such voltages were in circuits. Now as most pF and nF capacitors have nominal voltage of 50V and we work typically with up to 5V then we are used to think that capacitor voltage have to be higher then the voltage in circuit, but I believe it is not true.
With uF ceramic capacitors I look at prices as the same uF/V combination typically have higher prices when case is bigger and higher when smaller then some ‘optimal’ case. I assume these case is still easy to manufacture so if I select it I will have no problem to get many sources.
I also assume that electrolytic capacitor nominal voltage need not to be higher than the maximum expected voltage in circuit. But for example when I have 12V powered device I place at input 18V (in other method of specification the 15V) transil. That transil during 25A overcurrent limits voltage to 24V. So I use 25V capacitor which is in accordance with rule ‘x 2’.
1 Like
But when you are limited with space and you have the place for 0603 and then (at 5V) you can use 10uF/6.3V or 4.7uF/16V the 10uF will have higher capacity.
According to Hitano datasheed I see that with 5/6.3=80% of nominal voltage the capacitance of 10uF capacitor will be about -33% so 6.7uF, and 4.7uF capacitor with 5/16=30% of voltage will be about -5% that is 4,5uF. And it is your decision if you wont to have more uF or more voltage (just for sure).
1 Like
TVS are great devices when used properly but again, one should read datasheet characteristics carefully: as an example they typically fail short quite badly so it is great as specified for clamping in the case of transient but I hope you also have other circuit protection design for short circuits.
1 Like
Limited current supplies and PTC.
1 Like
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.