How small can you go?

[kt315]

Hi,

I've made a few boards with SMD components using toner transfer. Worked quite well for me.

I've got a few interesting parts I am willing to fry, unfortunately they came in MSOP package...

How feasible is it to create a board with toner transfer for MSOP?
Any useful tips?

Thanks.

[c4757p]

0.65mm pitch? Absolutely. Just pay special attention to that area when you're transferring.

I wouldn't go much smaller though. I've done 0.5mm-pitch packages before but it was a bit tricky.

[kt315]

Yep, they are .65s. I've read at some forum (do not remember which one), that for MSOP solder mask is a must. Is it?

[c4757p]

Only for reflow soldering.

[kt315]

Ok, thanks a bunch.

[poorchava]

I have successfully done boards with MSOP/TSOP down to 0.5mm. Not a problem. Although uv resist method seems a bit better, because toner will sometimes spill to the sides when heated and pressed, which creates occasional shorts between pins. You have to inspect that carefully and correct with a scalpel.

I think 0.5mm is borderline for homemade pcbs. Never had much success creating 0.4 or 0.45mm patterns (and those parts are rare for a hobby use anyway)

[glatocha]

I remember once I made a PCB with two lines going between DIP 0,1" pins. So I guess it was also around 0.5mm

[Dave]

The minimum width/spacing of traces also depends on the size of areas that need etching. If you design a board with really tiny traces, you better make ground pours, because otherwise your tiny traces will get under-etched before the large copper areas are done etching.

[EEVblog]

Yep, they are .65s. I've read at some forum (do not remember which one), that for MSOP solder mask is a must. Is it?

No, but it helps a lot.
Instead of drag soldering across the pins, you'd wipe away from the pins.

[TerminalJack505]

I use the toner transfer method and have no problems with 0.65mm pitch so I say go for it.

Like Dave says, solder mask isn't necessary.  (Personally, I've always found clean, well fluxed fiberglass to do a great job at repelling solder.)

Here's what I think are the tricks to successfully soldering these types of packages.

• Use a lot of flux.
• Use very thin diameter solder.  I use 0.015"(which is almost too thick) 62%/36%/2% solder.  (Tin/Lead/Silver.)
• Make the pads long enough that you can get the soldering iron tip on them when soldering the pins.  I like to leave 20 to 30 mils (mil=thousands of an inch) of room.
• Make sure the pads are wide enough so that the solder can flow around the pin and get to the 'heel'.
• Use the method Dave describes.  I use a chisel-style tip wide enough that I can solder two pins at once.  Just lay the solder on top of the 'foot' of the pin, bring the tip down and pull it out and be sure to make contact with the pad.  You may need to pause briefly with the tip on the pad for best results.

The attached photo is a board I did recently.  That's a SSOP package with 0.65mm pin pitch.

Edit: Added a few more photos taken with my microscope.  The soldering isn't the best (it looks great with the naked eye  ) but is perfectly usable.

[Psi]

Side etching can become a problem when you have tiny traces.
You may find you have to increase the pad size and decrease the gap between pads in order for them to come out correct after etching.

[kt315]

Awesome, thanks for the tips.

Jack, does it look like you tined the pads before soldering?  Does not it make the board a bit uneven? Did you have any issues with getting the chip laying flat?

[TerminalJack505]

Yes, I always tin my boards shortly after etching them so the copper doesn't tarnish.  It makes soldering components to the board easier too.  "Tinning" in my case is nothing more than fluxing the board and running a soldering iron (that's had some solder applied to it) along the traces and trying to keep the solder as thin as practical.

As you can see, it's hard to create a completely level surface this way so, just before I solder on a part, I make sure all the pads are fairly level by wiping away any excess with some flux and a clean tip.  (I use solder wick on pads this small only as a last resort to avoid lifting the pad.)  The pads with a little extra solder can be handy for tacking-down the part too.

So, to answer you question, no, I don't have any problems.  This is the smallest pitch that I use for my homemade boards, though.

[DavidDLC]

Ask Frantone, from his website:

"To create the CuteUino prototypes I had to push the absolute limits of DIY printed circuit board density..."

Check it out:

http://www.frantone.com/designwritings/design_writings.html

David.

[kt315]

David, thanks.

I've seen her post, it is indeed quite remarkable ... and cute.

I believe it is 0.8 32 TQFP package though.

[nukie]

This is for people living in Australia, I was never able to produce super thin tracks due to side etching as mention by Psi, etching with Ammonium Persulphate. Then I went to a swimming pool store and got myself 5L(too much!!) Hydrochloric acid 32% and I went to the hair stylist and got me self a bottle of vol40 hair bleach 12% Peroxide. Mix them 1:2 ratio and I made 0.1mm tracks no problem at all. This eats the copper in seconds! I am guessing a 1oz board 40x40mm took around 30-40 seconds in 30mL ? I don't know I was dipping and taking it out, I couldn't believe how powerful it is.

For some reason, the etching process doesn't foam much but due to the cocktail mix of hair bleach formula I got produce a layer of thick creame which I scoop out and drained. Other than that it's just perfect. I looked under the magnifier its pretty good. I believe 0.05mm is possible I will try it next time.

[poorchava]

Wow,  as if 0.1 wasn't hardcore enough :-) That's 4mils. And 0.05 is 2 mils or 50um... I guess that starts to count as HDI board...

[Rerouter]

etching in seconds, wowza, any chance of taking some close up photos of some of your boards? each process has its little quirks and curious how this one reacts.

[nukie]

My printer is 600x600 DPI. That's 23.62 dots per mm. 2.3 dots for 0.1mm. And roughly 1 dots for 0.05mm. That's pushing it to the limit. I wonder if it will actually print. I will try it soon with photos after etching. I still can't get away with large copper areas where pitting occurs have to manually fill in with marker pen.

[TerminalJack505]

You might try humidifying your paper and see if that takes care of the pitting.  I have to put my paper in a air-tight container with some hot water before I print on it to ensure the toner sticks completely.  I use the paper that has a coating of corn starch, though, so this trick might not work for other types paper.

[nukie]

I thought you need static on the paper to attract the toner from the drum ?

Sent from my Nexus 10 using Tapatalk 2

[TerminalJack505]

I only know what works for me.  Like I said, this trick might work just with the corn starch treated paper.  I read about the trick here.