Go Down

Topic: Want to make an Arduino Mini-kinda Clone? (Read 13930 times) previous topic - next topic


It would appear that I have alot more tip options with the Hakko 936.

I don't know what crazy tip types you plan on needing, but I bought half a dozen tips for mine, and all but two of them still have the factory tinning. Any soldering iron will come with a tip that is ideal soldering PTH components and adequate for 99% of the soldering work you may need to do. A sharper tip will be helpful for some surface-mound components, but what else are you going to need?

Remember, the tip and the shaft that it connects to get very hot when you are soldering, and it takes a while to cool down, so you don't want to be changing tips all the time. It's best if you can complete a job with only one, and for most jobs, the one that comes with the iron is the best choice.

Buy whichever one you want, but unless there's a specific style of tip that is only available for one iron that you think you'll need, don't let tip selection be the deciding factor.



I don't expect to be changing the tips very often or ever.  At least not when they are hot.  You only need to grab a hot iron once to learn that lesson.

As for the current use I see for the soldering iron.

SMD components.  As talked about above.

Sub-cell battery packs.

Misc other sodlering tasks solutions.

i.e. wire to connectors etc.

I just thought it was neat, that if the task required.  There were additional tips around.

Also, the Hakko appears to be a tad bit cheaper.  The one that I'm currently looking at is new and ~70 bucks shipped with 10 tips of my choice.  Which was roughly 30-50 bucks cheaper than the Wellers that I could find.

If there is a huge difference between the two, then I'll opt for the better.  However, for the amount of soldering that I will do.  I'm mainly looking for the iron that has the proper tip and steady voltage to do the job.

*back on topic*

I think I finally started to beat eagle into submission.  I have a PCB that is laid out roughly the way I want it.  In a fairly compact size(7"x1.4").  I used the reference schematic from the website.

Although, I'm confused.  Since on the of the connectors is name something, but plugs into something else.  I don't know if this is by design of a misprint.

On the FTDI Basic connector.

TXO connected to RXI
RXI connected to TXO

I left it as it was, but seems a little strange.


That's correct. TX is transmit and RX is receive, so the atmega's TX line has to connect to the FTDI's RX, and vice versa. It can be confusing, you just have to remember which device's TX/RX you are working with.


Looks like I'm going to have to bring out the programming pins also, to burn the bootloader.  

I thought I could buy the chip preloaded, but I guess not.

I'm thinking about using just pads, and a little pogopin programming station.  Since the goal is to make this thing small.

On the schematic, it has the SCK pin pulled to ground through a resistor.  Is this required?

If I'm reading the info page right.  I need the following pics for the in circuit programming.


Am I missing anything else?

I brought out the 6 pins for the FTDI programmer board, then I bought out the pins I want.  Now I need to program it with the


Aug 06, 2009, 10:21 pm Last Edit: Aug 06, 2009, 10:23 pm by whosawhatsis Reason: 1
No, you shouldn't need a pulldown on SCK. I think the resistor you're looking at is a current-limiting resistor for the pin 13 LED.

If you design your board so that it uses digital pins 11-13 for I/O, they will already be exposed. If you also break out the reset pin, you can wire up a custom cable to temporarily connect to the pins you're already using. Then you don't need a separate ICSP connector.

You could also pre-bootload the chips yourself. I recently saw a thread that mentioned doing this by holding chip down on a footprint connected to the programmer without soldering it in place.


I was reading the schematic right, but in arduino land.  Pin17 aka port 17 of the chip maps to Arduino pin 13.


That is a great idea about bringing the multi function ports out to be abused for the programmer.

Is there anything I should be concerned about?  Or any other ideas for multi use on the pins.

I can see this causing me more pain in dragging around the parts for a pretty layout.


Is there any specific settings for the GRID in eagle that I should us?

I don't know if you use eagle or not.  I read on the sparkfun tutorial that it wanted me to is .1" grid.  

Is there a real reason to do this.  By that, I mean does the manufacturer need this?  I thought the boards where printed or cut out with a cnc router.  I can understand minimal trace width and clearance.  But I didn't know if there were other requirements I needed to deal with.

Is Batchpcb the best spot for my initial, this probably won't work prototypes?


You can build on any grid you want, but I find that it's best to lay out at least your PTH components (headers and such) on at least a .05" grid. You'll probably then want to switch to a smaller grid to route the board. Read the Batchpcb faq for their requirements, even if you don't use them to print boards, they're probably good guidelines.

I use Batchpcb for my boards. It takes the better part of a month to get the boards back, but it's nearly two orders of magnitude cheaper than any comparable service I've seen, so it's really nice if you're not confident that the design will work. If you find another site with comparable prices that doesn't require you to get a whole panel made, let me know.

I will suggest that you order more than one set of boards and parts for your first prototype. Especially if you're working with SMDs for the first time, you don't want to ruin a board and have to wait for batchpcb's turnaround time again. Plus, with their $10 handling fee, it's not very economical to just get one board (most of mine are around $3 each). My last 3 or 4 orders, however, have all been doubled (I received two of each board I ordered), probably due to the slight soldermask defects I have found on a few of them causing the manufacturer to re-print the panels.



I read about the slow turn over, but I guess that is to be expected.

I figure that I will use them at least for the first couple.   I'm sure that I will screw something up.  Either in the design or the assembly.

If I actually get to the point that I need alot of the same part, I'll look at other options.

Currently I'm trying to relayout the board with your pin out suggestions.  

Is it standard that you place all of your components on the top layer?

Or is there a way to send some of the smaller parts to the back side?

Like resistors etc...


Components on one side is generally preferred, but as long as you're assembling the board by hand, you can put stuff on the bottom. The board I showed had about half of its components on the bottom. You can flip them to the bottom with the mirror tool. It's right under the move tool, and has an icon that looks like a letter E next to its reflection.



Is there any problem with using the D0/D1 Pins... which are the RX/TX.

I guess that goes for any of the pins... Are there anything that I shouldn't try to use for secondary purposes?   I'm assuming reset would be one.

Thanks for all the help...  


The arduino fuse settings disable the use of the reset and xtal pins for I/O, and I believe ADC pins 6 and 7 are ADC only (analog 0-5 are also digital 14-19). You can use digital 0-1 for non-serial functions as long as nothing will be connected to them when you are using the serial port. I like to avoid using them so that I can use the serial port for debugging while my programs are running, but if you're looking to minimize the number of pins broken out, especially if you can test your code on a full arduino board and just change the pin numbers for your board, there's no reason not to use them.


Sounds good.

I'm getting a little confused with the documentation on the site.

From what I showed earlier... as pin/port 17 = pin13/SCK = led

However, when I look at this link.


Pin/Port 19 = Pin13/SCK

It's trying to make my head hurt.

I'm going to assume that the TQFP has the same pins as the DIP?  Or is there a TQFP pinout cross reference page.  I grabbed the mapping above, off the mini hardware page.  I believe that the pin count is different between the DIP ad the TQFP  32 vs 28?

That confusion aside.  For programming purposes.... I was thinking...

That I could place test points on the bottom side, and use pogopin.  And/or press the programmer against those.  VS bringing the pins out to the connector that I'm using.  Is the "test point" in eagle a physical device that has to soldered down.  Or a glorified soldermask/point?

That way, I have access to all the pins, but don't have to see them.


The pin numbers are different on the TQFP package because there are 4 extra pins. If you count counterclockwise starting from reset (pin 29 on the TQFP), you get all of the pins in the same order, just with pins 3, 6, 19 and 22 added (gnd, vcc, A6 and A7, respectively). I did a quick-and-dirty pinmapping image for the TQFP that I'll try to get cleaned-up and posted.


Aug 07, 2009, 01:21 pm Last Edit: Aug 07, 2009, 01:22 pm by Deviant Reason: 1

Thanks for the help.    I pulled up the pro and did a side by side as I labeled the pins in eagle.

Submited my little design to there robot last night.   Passed all the little rules.

But when I got the files back from them. The top_lg had little copper colored pieces sticking out from some of the traces on the board.

I believe that it is just where the parts and the traces overlap.  But I don't know why it would be doing that.   By design or is this something I should attempt to correct?

Go Up