Turning a 600 mil chip to 300 mil

We’ve seen a few builds featuring NXP’s LPC1114 microcontroller before. This chip – the only breadboard friendly ARM microcontroller available – comes in a ‘still a little too large for prototyping’ 600 mil, 28 pin package. We won’t hazard a guess why NXP chose this rather large package, but the good news is it’s possible to shave this chip down to the more common 300 mil, 28-pin package used by AVRs and PICs.

In the video tutorial of this procedure, the chip is first taped down to a desktop CNC mill. 150 mil on each side of the die are removed, exposing the very cool-looking pattern of leads coming out of the chip. This isn’t enough area to solder, so the chip had to be further milled to expose some of the internal wiring.

After soldering everything to a set of pins, the new 300 mil package is covered in epoxy putty, milled down again into a nice cube shape and painted. Yes, the modified chip does work, and no, we can’t figure out why NXP chose a 600 mil package for this microcontroller over the far more common 300 mil chip.

Video after the break. Tip ‘o the hat to [Ian] for sending this one in.

76 thoughts on “Turning a 600 mil chip to 300 mil

    1. that’s what i thought too.
      cool thing, but why didn’t they just make an adaptor board. 300 mil pins on one side and 600 mil socket on the other.

    2. You are not getting the point here.
      He wanted a smaller package for breadboard use.
      By making an adapter it would even get bigger.

    1. How many hours to modify a 3$ chip?
      How much is one hour of your time?

      A TSSOP28 to DIP28 breadboard adapter can be had for about 5$.

      I’d rather spend 5$ and 10 minutes soldering, and the rest of that time programming the little bugger. But it’s just me…

      1. 5-10 minutes of soldering, plus the 2 days it will take for your adapter to arrive if you don’t have one.

        You guys are a little too critic happy. There are many situations in which a hack like this is very useful. That it is not for you doesn’t mean it is not for someone else!

  1. Man – the DDR music brings back memories.

    Searching for the mill’s name “mini cnc black II” gives several links in Japanese so not sure if available for import to the US or other countries.

    The final product here looks REALLY nice.

    1. The mill looks like an eBay job. Just search for CNC mill kit and you’ll find quite a few made of standard aluminium extrusion that you can mount your own multi tool (dremel) cutter onto.

  2. This is the most stupid hack ever. In the time that took one could use that cnc to make a pcb and solder some pin headers and stick it on the breadboard, without potentially destroying the chip in process.

    1. The point is, that the chip used too much space.
      Meaning that, if you put the chip on a breadboard, you will lose a bunch of those holes, because they would end up being under the chip. Or under your pointless PCB.

      1. If you make a PCB for something like this then use 90 degree headers so the PCB stacks edge-wise and uses vertical space rather than covering all your breadboard holes.

  3. Done that stuff too, but because I had to and the replacement chip wouldn’t otherwise fit into the socket of the scope I was repairing.

    Doing that stuff without really needing it ==>
    pointless

  4. The reason for using 600mil is it’s aimed at low-cost apps like white goods, with single-layer PCBs. PCB space is typically not at a premium, but routability on one layer is.
    600mil width gives more space to route tracks out of the ends.

    1. You mean NXP don’t make stuff the way that hackadays editors/commentards want it because they need to actually design stuff for some target market that is bigger than tens of units? What is the world coming to!

  5. Look at all the negative comments! Since when did hacks have to be practical? Maybe some people would be more at home at an industry website full of practical business applications than at Hackaday?

    Maybe He/She/They did this to:

    Have a challenge

    See if they could

    Make a point that it was dumb for NXP to use such a large package

    Practice soldering to really small things

    Take a small step towards DIY component building

    Practice with the mill

    Make you wonder why…

    Get attention on the internet

    1. >Make a point that it was dumb for NXP to use such a large package

      Look at Mike’s comment above. Do you really think a company that is going to fab massive piles of something hasn’t actually thought about what package to use in great detail?

    2. +1, right on the money. Jeeze there are a bunch of whiny people on here. Nobody is forcing you to read these if you think they are stupid…

  6. That video is off the hook! I love it. Really cool way to go about solving the problem even if it isn’t a very efficient way to do it. I would think turning the original package vertical and running the leads on top down into a socket would be a lot easier. Very cool though – it’s neat to see the internal leads laid bare like that.

      1. Thanks for explanation. I don’t think mils are used outside of the US, and not in Britain anymore either where they have gone almost completely metric.

    1. in english 600mil would be shorthand for 600 millimeters

      thou isn’t good enough for you? you have to steal our units now too?

      Mod edit: That is so wrong I had to delete it.

  7. I am amazed that no one here understands the point of the .6 vs .3 package.

    Management asked, “How do we get people to not think we are an 8-bit processor ?”

    Marketing said, “Put it in a larger package”.

    The story ends with “Now no one will try to plug our 32-bit chip into an Arduino”

  8. i did something like that for a video encoder. but i had the orig psu and power button and heatsink. core 2 duo had it on a table until someone spilled something on it.

  9. I completely do not get WHY he did it. So many more practical solutions. Fact is: I really like HOW he did it; nice movie and the end result (type number engraved, painted black) really gives a customized / retro feel. Unnecessary but cool!

    1. I know. I’m not sure why all the hate; it’s not something I would have the balls to try, nor hope to successfully accomplish.

      Whether or not it’s a waste of his time would be up to him, really. If that’s what he wants to do with his time, then he’s perfectly free to.

  10. To the people complaining about the (stupid, pointless, idiotic) waste of time, money and/or effort:

    When was the last time you played a computer game, played in a sandpit or simply just played?

    Not everything is about optimized efficiency, not everything is about saving money, not everything is about saving time.

    When did you forget to love to do things, simply for the sake of doing them? Who took this love away from you?

    Why do you feel so much irritation toward those who you consider to be sub optimal? Why do you label their work pointless, stupid or idiotic?

    Have you thought about the effect of your quickly written words on the person who thought someone might find what they’ve done to be cool or interesting? What if they stop sharing what they’ve done with the world? What if they stop *doing* because of you? Is *that* hacking to you?

    I enjoy reading Hackaday for the comments as much as the projects the articles feature. There is praise and criticism, but it is constructive, reasoned, and normally doesn’t address the why, instead focusing on the how.

    To the hacker: I learned something from your work, the final chip looks slick, and I think your time and effort were worthwhile.

    1. well for starters the public school system in the good old US of A is rather good crushing passion for life and the things in it you are interested in and as follows drive and curiosity suffers as well, students are advanced when they are defined as ready by the system not when they are ready.

      we say well you have learned this bit of math or english or other basic skill set on to the next level you go but we have so many basics that our students don’t have time to discover let alone dive in to the subjects they are really interested in and that hurts them more down the road than not knowing some of the basics from the get go does in the short term. Once they get so far in whatever field it is they are passionate about they decide well maybe I need more math to further my passion and then all of a sudden that student wants to take a math class instead of dreading it.

      so we get students that after 18 years of dictator ships have learned that life is bullshit and some one else makes all the rules and everything new or different is impossible for them to create so one might as well do whatever they are told to as efficiently as possible so one can move on with watching tv and not caring about anything because they have learned to repress their passions and be good little worker drones. a lot of our community college student’s enroll because that’s what your supposed to do next without the faintest idea of why they might want to be there, they sludge through it and just scrape by they don’t try to do well they just want the shiny piece of paper, not the education that’s supposed to go with it.

      every one is born with the desire to learn with curiosity but we teach that learning is painful and bad when we force feed our students rather than letting them sample all the flavors and find the one that will drive them to excel without out side coercion.

      yeah maybe I have a bone to pick maybe I dropped out just in time to be able to see all the damage that the public school system did to me before it became permanent, sorry about the rant

      1. yes, criticism not constructive criticism aka a waste of everyone’s time. oh no some one spent their time doing something they enjoy for the sake of that enjoyment quick call the thought police. I’m sure he knew there are quicker options but this is about the method and the end result not the efficiency of the result, microwave diners are very time efficient, but I still prefer eating and preparing a home made meal when I have time to cook.

  11. This is absolutely awesome. Hacking is not always going for the easiest solution, sometimes it’s about challenging yourself to try new and funny ways to achieve your goal, just because you CAN :)

  12. …or dead-bug it with some ribbon cable, tin the other ends of the wires, make up your own chip spacing, long enough wires and you could have a 600mmm chip :~) … (Note to self, must stop feeding the Trolls)

    1. A 600mm package would contain how many cores? Given that most current ARM’s are packaged to about 1 cm square (give or take 50%) maybe somewhere around of 3,600? Might be enough to get an OK volume discount, and ooohhhhh the throughput!

      Should it have a hotplate mode? Mmm… ARM-fried bacon, served on ARM-fried half-meter-diameter crepes!

  13. I can’t help but think all that epoxy, primer and paint combined with halving the surface area does nothing good for it’s thermal characteristics, still it’s pretty original!

  14. 1. bend the legs on one side straight out.
    2. solder long headers or wires and normal headers to the other legs.
    3. fit in breadboard.

    This will be pretty tall, but i would rather do this than try to cut down the package.

    1. I did this 20 years ago with a 74154, although I put the(almost) straightened pins into a narrow machined socket. I then wired the pins on the top to the other side of the socket with enameled wire. I then soldered the top side of the socket on both sides, and had a 300mil breadboard-friendly demultiplexer.

      Not to discount this guy’s hack. I have removed the epoxy encasing on several chips in the past to be able to use them after a pin has broken off, and it does work quite well…

  15. don’t be so negative – this hack is awesome in itself. Though I have to agree with some ppl – 600 mil is just twice as much as 300 so why bother ?
    if space is so important for you just use smd

  16. Awesome, I love these impractical hacks where someone stubbornly refuses to accept the way the world turns and pours massive amounts of resources into making it spin his/her way.

    Don’t listen to the haters, they will never get why you do this.

    1. Indeed. Haters would ban mountain climbing because it wastes time, it is pointless, you could fly a plane over it, dig a tunnel, etc. :P

  17. Fantastic Hack!!!
    This has become one of my most favourite hacks, the header image of the article is somewhat deceptive as I didn’t expect the attention to detail that went into the project until I watched the video and could see the finished article – inspiring.

    Love it!

    It makes me think of ideas, where one could take a smaller microcontroller and kit it out with much of the needed electronics – say an accelerometer and shift registers or whatever – wire it together internally and cast a body like you have done here to make a cool little prototyping device with most things already built in that I might use. neat.

  18. Huh, I hadn’t realized these were for sale yet — just placed an order on mouser for a couple of them — anyone know what a simple “blink a led” schematic looks like + toolchain / programmer?

    I did some googling but it’s slow going.

    I wonder how non-trivial porting V-USB to this device would be… the logic levels match (~3.x volts), so it seems like this would be a good candidate for software USB.

    Or I wonder about the .NET Micro Framework…

    This is going to be fun! :D

  19. Very cool!
    I am just wondering why the manufacturer would make a 600mil chip if it is a 300mil with some extra plastic on the sides? It makes no sense to me, but what do I know :)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s