Kurt's DIY Teensy Breakout boards with Arduino Shield headers

I thought I should create a thread specific to these boards as to to not hijack other threads. Yesterday I received two sets of fabricated boards from Seeedstudio. One set is some simple modifications of my earlier board, which uses lots of surface mount parts. I thought I would also try a version of it with all (one optional exception) though hole parts. Here is a picture showing the two boards:


Since I have not done a version of these boards with through hole parts, I thought I would start to assemble one of them first. So far the one issue I am running into was my initial first choice for 5v VR, which I knew was big, but I thought I could mount it with the leads going all the way in. Turns out the leads get wider toward the main body and do not go into the holes and so it sits really high. I probably should have set it up to mount horizontally. Other options include ordering smaller version. This one can output 2.5a which is overkill. Another option is to use a switching regulator, like the ones that Trossen Robotics use on their Arbotix-m boards. I purchased a couple of these from digikey(http://www.digikey.com/product-search/en?x=0&y=0&lang=en&site=us&keywords=811-2196-5-nd). They are more efficient but require 7v, which is great for Robotis type servos (3s Lipo), and probably some other servos which are setup to work with 2S lipo, but won't work with 6v NIMH or 6.6v 2s LifePo4 batteries which work with normal RC servos... Here is a picture showing some of the parts populated.


I will continue to populate this board and see what other issues come up. May also look for some other LDO 5v VRs to try.

In some other threads, I mentioned that I would like to use some taller Arduino shield headers. I knew there were some out there as for example Seeedstudio ship some with their XBee shield. Unfortunately they don't sell the taller headers. So I started searching around and found that Samtec sell some different height headers that may work with their ESW series of headers(http://www.samtec.com/documents/webfiles/pdf/ESW_TH.PDF). The following picture show a few different sizes:

The one on the right is the standard size part that I purchased from digikey (S7008-NO), The one in the middle is the shortest of their tall ESW series (ESW-106-12-T-S), which they sent me a couple of sets of these headers as samples. Earlier I purchased a set of their taller ones on the left(ESW-110-33-T-S), which again are of this same series. I decided those were too tall. They also sell versions whose height is halfway between those two.... For this board will use the ones in the middle.

There is more information up on the thread:
http://forum.pjrc.com/threads/25620-Teensy-3-1-to-Arduno-Shield-DIY-Adaptor/page2.

Design files are up on github: https://github.com/KurtE/Teensy3.1-Breakout-Boards

Lots more soldering to do! Should get one of these done in a day or two.

Kurt

Congratulations, looks awesome to me, can't wait to see the final results!

Thanks, still playing with VR. I can fold it over I think to make it work. I have also edited my Diptrace layout, as to make even more room for this. (Moved it closer to the edge and up and the transister for speaker down, so it should be able to bend down flat I think. Will verify it more later. Need to finish the board first and see if anything else needs fixing.

Assuming everything else works fine, I personally probably won't order any new ones fabricated, but will update files up on github so if others want, they will have better version. But need to get back to building it and do some initial tests.

Kurt

Also figured out I did an oops with choosing the 5v VR. As the order of the pins on the LM1085 is different than I thought I verified. I thought it was the same as the LM2937, and the switching regulator I picked to try out. That is on those devices the Pins are: 1=Input, 2=Gnd, 3=Output, but with the LM1085 it is 1=Inupt, 2=Output, 3=GND. So the current part won't work! So I unsoldered it and put in one of the switching units. If I assemble another one, will probably just order an LM2937, or maybe one of these: http://www.digikey.com/product-detail/en/MIC29150-5.0WT/576-1106-ND/771575

Will update parts list soon, showing issue and possible parts...

I finished assembling the one:


I have not done a lot of testing yet, did verify I could program it, which I first tried my Teensy Test sketch. I did verify the speaker outputs...

As I was concerned about the height of the VR, I put on two shields. The bottom one is a Seeedstudio XBee adapter and the top one is the Adafruit 2.8" tft display... As you can see in the picture, that the slightly higher headers does help add additional room under it.


Also I ran my start of a remote control, that I hacked on earlier to work with the Adafruit 2.8" tft with touch and it runs :0

As you can see the SPI output is working and it also responds to the touch input, so I know at least some of the card is working... I will do some more testing over the next few days.

Kurt

Blue Smoke

Probably soldering issue when I replaced VR. Saw puff in middle of CPU when I plugged in 12v wallwart... Will rebuild second one, start with voltage section and verify voltages before I go farther...

Edit: You can see where the chip blew part of it's top...


Kurt

Update: for some reason it does not like when you install the 5v VR backwards As you can see in the picture in the previous post, the main processor sizzled and you can see a piece of it missing. Now trying to decide if I should attempt to remove fried Teensy or just build a new board. It would be nice not to have to resolder all of those IO pins and the like. But trying to figure out the best way to proceed. Obviously having a hot air rework station would probably make it a lot easier, but don't have one.

I am thinking of doing some teensy mutilation, and maybe try to use nippers and cut the chip into pieces, such that maybe I could one or two pins at a time...

Suggestions?

Nip the pins on the MK20 and then desolder whats left with an iron and sucky desolder tool. Then clean up any residue, put in a new MK20 and use lots of Flux.

The really question is, have you also fried the Mini54? I would try and test that before you put the MK20 back in. IE. Power 3v3 to the board without the MK20 in. Hit the prog button and check if the reset line toggles.

Thanks for the suggestion. I could definitely try that.

What I am trying to decide, is it worth the try? In the past I was not very good at soldering these types of components. However a friend did send me a small bottle with flux in it as he also said I was trying to do it without using enough flux.

Obviously the first thing I need to do is to fix what fried it in the first place.
Then my choices are:
a) Punt: I have 4 more of these boards, plus 5 of the ones with SM parts. I believe I have the parts to assemble at least one of each of these...
b) Replace the whole Teensy 3.1 - $17-$20 - probably surer bet
c) Try to replace the MK20 assuming the rest is not fried $7.40 - Most economical, maybe I should first try to remove the chip to see if feasible.

Thanks again

There is no doubt about it, it will be tricky and fiddly. I personally, if I had spare Teensy boards, would unsolder the Teensy and then Socket solder a new board so that in future if it happens again I can quickly change it.

But if you want to try the replace the MK20 and you have one spare then it's a good excersie to try it and experience SMD high pin count soldering.

Thought I would give a quick update.

I purchased the Sparkfun Hot Air rework station (https://www.sparkfun.com/products/10706), which arrived last week. I was able to use it to easily remove the VR, and also decided to use it to unsolder the whole Teensy 3.1 from the board, instead of trying to simply remove the processor. I did this as I have 3-4 Teensy 3.1s currently in reserve. After replacing the VR, I figured out the 3.3v VR was also toast, so I replaced it as well. This time I verified that if I connect up 12v to the board (I used a powered AX servo Hub) and powered it through an AX servo wire... I verified I had 12V at the VIN, I also had 5V where I expected it and 3.3v where I expected it as well. I then soldered in a new Teensy 3.1. I only issue I had was when I soldered on a wire to the reset pad and moved it, it tore the pad off the Teensy, so my reset buttons won't work now.

But I have now verified that the 2.8" tft display is working including my test program and the ILI9341_t3 spitftbitmap test program works, which reads file from SD card and shows it on the display.

So I think this board is now in a workable state

Kurt

But your missing out on the joy's of replacing the processor! And the hours of trying to figure out if you've managed to solder all the pins or not or that you cooked the processor or not

Good to hear that your board is now in a workable state.