Good morning, all arduino nanos and pro minis have one thing in common. The microcontroller is mounted at 45 degrees. Now. Is that a style thing is it necessary for routing or routing you might say, or is it something to do with soldering? I did look this up on google and there was one comment about wave, soldering and uh.

Here are some more pro minis and yes, they're all at 45 degrees, this one's quite cute. Actually, this nano it's got. This is actually a 168 uh, not a 328. So it's got less memory, but they've managed to cram every component on the top side.

There's absolutely nothing. On the bottom side there's the ch340. The microcontroller is in a smaller quad flat pack. So i thought i'd set myself a challenge.

I'm going to start designing a nano or it might end up being a pro mini. All i'm going to do is put the 328p on a design and put two headers at this. I think it's 0.6 inch spacing and just see what the issues are. If i have the chip at 0 degrees or 90 degrees, or if i have the chip at 45 degrees, so a quick inspection of this nano on yellow a0, a1.

A2 a3 - are the four pins around this corner. Now, there's pin one over there, so these are some quite high numbered pins, but that would seem to make a lot of sense to put that corner adjacent to these four um connectors on the outside of the board up here. However, it's not quite so obvious. The benefit because, okay, yes, we want to root out d 12.

I believe that is. I think this is d 13 down here. 11. 10.

9. 8. 7. 6.

5. But the rooting here is a little bit tricky. There are some quite wiggly rooted tracks. The rooting for this ceramic resonator is really horrible.

It uh this pin comes up to that side, but this one goes all around the outside. Now i can see the logic of having the ceramic resonator ground going to these two ground points. That would appear to be quite uh sensible now, on the pro mini, it's a slightly different situation. Yes, we've got a 0 1, 2 and 3 down on the bottom here around this corner, but at the top it's slightly different because the numbering on the board is different.

This is two three four five, six seven on the um nano. These are all shifted over 10. 11. 12.

13 are all down in this corner again. The ceramic resonator is here: they haven't. Um used the trick of running the ground straight into the two pins on this side, but they have. These are the two resonator pins uh? Clearly, on the chip running to that resonator, however, here's another pro mini and it's completely different, because the resonator is actually over here and they, the two sides of the resonator, go through wires, which, if you look on the underside of the board, run in underneath whoops.

I keep hitting the camera underneath the chip and are connected to these two pins. Clearly, underneath the micro controller. Again, we've got a 0 to a 2 on this corner, but i'm not convinced by looking at this track layout that this 45 degree thing is absolutely necessary, because this brings up other questions. If i do find that 45 degrees is more convenient, will jlc pcb actually solder, my component at 45 degrees - i don't know so i just did a google search for arduino uno and the ones that have a surface mount pack.

This one is at 45 degrees. This one here is at 45 degrees, but then these are both wi-fi boards, so there may be an issue with density of components this one's at zero degrees. This one here is at zero degrees, this one's at zero degrees and this one's zero. So they do seem to be a mixture, so maybe, where there's more space 45 degrees isn't actually necessary.

So here we are in easy eda and i've called this uh giuliano bonanno. I was going to put it on a yellow board, but you can't actually have the smt assembly service on a yellow board. So it's going to have to be a green banana, but that's okay. I don't mind and i've created two printed circuit boards, this one's pcb45, juliano bonanno and i've got pcb 90 the same.

So let's have a look at the 45 banana and you can see as we had on the pcb when i shone the torch through it. The analog pins, a naught to a3 are nice and symmetrical down the bottom here. A4 and a5. Work.

That's fine, but these pins up here, which are digital, pins uh. This one is d12 d13, is down here on the nano. It's all a bit tight in this area here and i've got some um design errors. If i run the design check, you can see that i've got a clearance error here.

You can see that i've got a clearance error here so saying that these tracks are too close together. Um they're, probably i think it says minimum 6 mil. I don't know dlc clearance 6 mil. I don't know whether that's i've got six mil or whether the minimum is six mil, but anyway it's saying this is too small and similarly this one is too small.

Now i could uh use slightly narrower traces, i'm using 10 mil for my trace widths, which is there, but i've already run into a bit of an issue. I'm sure that can be sorted out because that's how the arduino boards that have the 45 degree chip are laid out. Okay, let's go to the 90 degree chip. That's this one and of course this is probably fairly easy to route d12.

All the way down to d5, because they're all on aligning pins - and i can probably root that - but i've got a major problem down here - a0 has to come around to here, and i've run out of space underneath these chip pins and i've got seven chip pins. I need to somehow link up, so i could drag this to the right. Let's do that and slightly offset to the ones at the top. These have all gone nicely.

I have to re-root those i'll do that now and see if this is going to be any easier, but it's already looking like i'm not gaining much with the 90 degree orientation of the atmega328p. So i'm starting to see the problem here if you've got eight chip pins, that pretty much line up with eight header pins um, then you've got a chance of rooting them without getting into this sort of problem, but if nothing lines up having eight chip pins. All in this orientation just makes it impossible to bring tracks out unless you bring them inside and put vias in. So i can see the issue.

It's it's really just about the fact that there's so little space here top and bottom that you just can't run that many tracks with the 45 degree alignment you can squeeze a few of these tracks. I could fix this. I could have slightly narrower tracks, thus creating slightly larger gaps. I'm sure i could fix this and in fact i could move this chip slightly to the right and have this arrangement at the bottom offset a little bit.

In fact, i might try that now and see if i can do it without uh, making narrower tracks here and okay, it's ugly rooting, but i think you can see immediately that um these tracks are able to be rooted onto the eight pads of the chip um. The end pads, actually, of course you can come in sideways on, so you really only have to root six where you have to come out uh norm, not normal, but uh in the same direction as the chip leg, but uh yeah, and i mean i can immediately See that i can root that and this d4 d3 d2 - i do have to fit the crystal in here somewhere and get it attached. I think it's these two pins uh, which might be slightly tricky, but it just looks like that's going to be far less hassle than this one, where i can immediately see that i'm just not going to be able to root those four at all. Really, i suppose one way you could look at this is that in this 45 degree orientation i've got six or possibly, you could say seven pins here, which kind of come out into this void area and another seven down here which come out into this left hand Area, similarly, all of these chip pins come out into the right area, so you can think of it as if there really isn't anything at all coming out into this very small gap between the header, pins and the chip pins.

But if you go to the 90 degree orientation, yes, you've got six pins here which have to come out in this direction. Up towards this header. You've got six pins down here which have to come out downwards towards this header, and the clearance is so tiny that that's a rooting nightmare so immediately. Well, now that i've attempted this, i can see that the 45 degree orientation does make it an awful lot easier.

So i think the issue can be summed up, as these six pads in the center of this row effectively trap your track routing into this very narrow gap, where i can really only realistically fit three tracks now you may get away with it if you've got a One-To-One correlation between the header pins up here and the chip legs down here, you might just manage to make it work, but it it's just much much harder than the 45 degree uh chip, positioning and layout, which actually works really quite well. So i'm satisfied now that this positioning of the chip or the rotation at 45 degrees is actually necessary from a routing or routing perspective. You just can't get uh the six pins on the middle of one of these edges rooted properly um with the chip at 90 degrees, despite the fact that it's slightly further away from the header edge - and this is unique really to the nano and the pro mini. Because of the very limited space between these two header rows, which i think is six tenths of an inch, so it's the same on both of these models yeah.

I think the 45 degree angle is absolutely necessary. So there you go cheerio.