Request for carrier board

dgustavus

Member
Hi Paul,

Following your suggestion, I am posting a request for a carrier board for the Teensy 3.1. Suggestions for the carrier board are that in addition to a socket for the Teensy, it would have the leads brought out in a fashion useful for the connection of external devices. Something like ground/power pins for every one or maybe two connections (for connecting pwm to motors or maybe servos). There is an example from KurtE of a carrier board he is making in the post at:
http://forum.pjrc.com/threads/24649...an-amazing-price?p=42063&viewfull=1#post42063.

This looks like a good starting place. I believe such a carrier is really necessary and am stopped from using the Teensy in projects by the knowledge that I will have to design and build something similar before I can actually use the Teensy. As an example, for a robot, I would like to see four pwm connections with pwr/gnd and four encoder connections with two digital pins, pwr and gnd on each. One might even consider some voltage level shifting, especially on analog to handle 5v sensors. Some good options on power connection would also be nice (batteries?).

This carrier board would be more expensive than the Teensy, of course, and would preferably have the surface mount components already installed. I believe it would greatly enhance the Teensy, however, by reducing the effort to employ it in a new project. I suppose it might even be possible to provide processor upgrades to the carrier if the Teensy 4+ maintained the pinouts.

Thanks for reading my suggestion,
Duane
 
rather difficult to have a one size fits all (applications) carrier board. For some a universal board would have many N/A pins interfaces.
Not easy to get a Goldilocks board! So there would need to be: 2? 3? 4? kinds.\
 
No, 2, 3, or 4 would not cut it. In the Arduino world - and this is one of the key elements that has made it so successful - a carrier board is called a shield.

As someone who has designed my own "carriers" HERE and a "shield" ( see sig) I think I have come to the following conclusion specifically in resprect to Teensy:
The more generic an adapter/carrier will be the bigger it will get. This would be somewhat counter productive as he teensy has its name for a reason, it's small.
A purpose and project specific custom board also likely will be smaller and will likely perform better from a signal standpoint.

The tools to create (small) custom boards are free and easy to use e.g. Eagle, Diptrace, KiCad etc. there are loads of Tutorials and component libraries available. Getting professional quality boards actually manufactured also is very easy and cheap. OshPark, Hackvana, Itead, Seedstudio etc.
There is really no reason not to roll your own!

Also for one-off deployments there are plenty of prototyping solutions that would allow one easily to create a form of adapter/carrier with the needed connectors without having to go the custom board route. There are several breakout boards available for Teensy, there is possibly a very nice Arduino adapter board coming soon. There are the adafruit Perma Proto boards etc.....

Then again, making custom boards really is exceedingly easy.
 
Last edited:
Soon PJRC will be releasing its first carrier board, aimed at LED projects using OctoWS2811.

Well, I suppose the audio adaptor and wiz820 adaptor could be consider carrier boards too.

Odds are good we'll make more. Maybe a more generic board makes sense, perhaps with high power drivers like H-bridges and screw terminals? But so far, this specific request isn't really clear to me, both in terms of what ought to be on the board, and why. I really need to understand more about what uses the board will really have.
 
Everyone has their own opinion on what an ideal setup would be. My main interests are for using on robots. My thoughts on this are very much influenced by the earlier boards I used from Lynxmotion. First the ABB that then was replaced by the Bot Board II (http://www.lynxmotion.com/p-252-bot-board-ii.aspx). Later these same ideas went into our design of the Botboarduino (lynxmotion Arduino 328 clone)

Things that I like and as such included on my own carrier boards include:
a) Screw terminal power - On my first carrier board I have a 2 terminal, on 2nd I have a 4 terminal (maybe different power for AX servos Lipo 3s, RC Servos 6V)
b) Power Buss
c) Selectable power to groups of pins. - For some projects and/or some pins - you want 3.3/5V, on others you want VS(servo) to run servos or the like.
c1) Lots of power and ground pins - In my case I want/have these as 3 pin servo headers for lots of the IO pins.
d) Some feedback - I like having a little speaker, that for the most part just makes a little noise. I want some way to know I pressed a button on a remote, or the robot is now on... Likewise one or more LEDS. I have not added any additional over D13, but they are handy.
e) Maybe one or more buttons - First ones I did here did not have any, but in new versions I added one.
f) i like to use XBees - So my main one has an XBee.
g) Some like Arduino headers (like Linuxguy/Geekguy here ) so I played around with a design that has them. On this one I traded not having XBee
h) Size: Obviously small is good. On my first board, I kept to the original Lynxmotion/Basic Micro sizes (BB2/SSC-32/Arc-32) of 3"x2.3", but for 2nd board went bigger as was not going to fit everything including the Arduino headers.

i) Try to avoid really small parts. In earlier boards I have built I was using 0805 and 0603 SMD components, which I can do. For these boards I went up to the next larger sizes 1206 - For me now I probably would go back to 0805 on things like capacitors...

Things I have not included, that I wondered about was things like SD card support. Things I have not considered include motor controllers. So far I prefer using ones like the Roboclaw that have encoder support and the like, but that is me.

I know I have posted pictures and the like of the two boards. I have also posted up on Lynxmotion and Trossen zip files that include the Gerber/drill files that you can upload to OSHPark or Seeedstudio. They also include a sub-directory that contains the DipTrace design files, as well as parts list. Also I know that in the one with XBee I included PDF files showing schematic/board don't remember if I added these as well to the one with Arduino headers (but can). Let me know if you would like me to upload to here as well.

So far I have built two with XBees (made slight changes for the 2nd one which I have shown the pictures with the OSH Park board with OSH Park Teensy). This morning I received email notification from Seed Studio that my order for the one with Arduino headers has shipped, so hopefully in another week or so I will receive them from China

Kurt
 
IMO I would be more interested in a full pinout Teensy3 than any particular carrier board.

The surface mount pads are difficult to utilize and unless you use some kind of breakout, there is no good way to breadboard it. Soldered wires sort of work (for a few connections), but then there is the problem of ripping pads off the board as soon as it gets a little stress on the wire.
 
IMO I would be more interested in a full pinout Teensy3 than any particular carrier board.
Would sure make doing a carrier board easier ;)

I forgot to mention, I was also playing around with a version of the 2nd design that used as many of the Standard Seeedstduio parts (http://www.seeedstudio.com/wiki/Open_parts_library). When you use their parts you can pay to have them solder up a limited number of prototype boards. But this list is pretty limited. You only have so many resisters to choose from. If you want a resister it will mostly have to be 0603... Also limited on Voltage Regulators. Most of them need at least 7V, which is not good when you might wish to use with 6V NIMH. For larger batches (100+), you can have them use other parts. But personally I don't need 100 of them.

Also I wonder that if someone were to build a production version, would you still want to mount a Teensy on it or would instead want to directly mount the parts of the teensy on the main board. But for me, doing one offs, it is just fine using the Teensy 3.1 as a breakout board.
 
IMO I would be more interested in a full pinout Teensy3 than any particular carrier board.
That would be nice as well, particularly if there was an option for soldering all of the pins in it.

But I can also see the need for a carrier board that the current Teensy plugs into, which might not give you access to the bottom pins, but would provide better power supplies to deal with lots of WS2812B lights and grumpy strips.

The third might be a way to plug into Arduino shields and xbee sockets.
 
I notice that Tindie has breakout boards with all the pins if that is what you need. The Teensy is small, and that is an advantage in some applications where few connections need to be made to the board; I am not suggesting making the Teensy larger if minimum size is the feature you require. Along with the breakout boards is the possibility of using a prootbaord and designing your own carrier that is specific to your requirements. How easy that is depends on your experience and talents, but I think we can agree it is not equally easy for everyone. Clearly, no carrier designed by someone else will be perfect for what you are building, but it might be very close, as is the carrier being built by KurtE to my needs since I am also building a robot. That is why I put some of my ideas, and pointed to KurtE's project, in the request bucket. I believe most robot projects will need two to four channels of pwm (in general I would not worry about the motor driver as there are lots of options for that). More advanced robots will want PID speed controllers which will require decoding quadrature encoders, plus some motor direction digital signals. Digital sensors, at a minimum, would be bumpers, tilt, a configuration DIP switch and probably a big red kill switch. Analog sensors would be ranging (ultrasonic and optical) as well as power consumption; these are where 5v input would be most useful. A minimal control interface is useful on-board (LCD and buttons), and communication could be anywhere from a USB port to an XBee or bluetooth.

That is a bunch of stuff, however the carrier would only have the surface mount components installed, and places for the connector pins; not all of that would be required (for instance a separate Teensy for communication/interface tasks might make sense). Which connectors are populated, and for what would be determined by the project. I think for lots of projects, especially for those feeling their way along, that would be a significant step forward. As the project develops, issues would arise that render the carrier less-than-perfect for that application, and feedback to that effect might improve future versions or engender a new carrier better adapted to that application. I think such a board would make the Teensy appeal to a broader range of potential users.

Duane
 
FYI - Just received the shipment from Seeedstudio of my 2nd carrier design. One with Arduino headers. Here is a picture showing one of them that I put some Arduino headers in and test putting in a shield to verify that I did not screw up any of the spacings. Also next to it is the stack of the other 4 boards. You can also see the most recently assembled of the other design shields (without Arduino headers but with XBee). Also I show the other shields I happen to have.
Teensy-3.1-with-shields.jpg

It may be several days/week+ before I get a chance to assemble one of these. Also not sure how much I will try out the different shields. The servo shield (V1) was sort of interesting when I first started working with Arduino 328, as it allowed you to reasonably control 18 servos (Hex 3dof). I had to slightly modify the board as by default they only supported 16. Also back then I made a few fixes to their library.

Kurt
 
Hi Paul,
I think that a teensy 3.x board that had the built in xbee socket and lipo charging circuit like the arduino FIO V3 board would be a great addition to the teensy product line.

It may wind up being a little longer than a teensy 3.1 but that could be good because a larger carrier board could mount up with it.

It appears that the arduino FIO V1 thru V3 have enough issues that it is not practical to even experiment with. It also only has 2K of SRAM and runs at 8 Mhz.

I think a teensy version with its 32bit processor, on board 1 cell lipo charger and built in xbee connectors would form the basis for many many applications. I hope you agree.
 
The XBee S1 is wildly popular due to its plugin form factor and good documentation. But it's old and big. Digi International has no surface mount non-Zigbee radio. Their small surface mount is Zigbee only and uses the Ember chipset whereas XBee S1 uses Freescale - and an 8051 MAC processor.

I wonder if there would be interest in a non-Zigbee (802.15.4 like the XBee S1s), version of the Anaren module that uses TI's 802.15.4 radio. I have some, got them working with a special firmware pushed onto the radio. This firmware is owned/supported by T.I.
The Teensy could add to the firmware the remote sensor/DIO, A/D that Xbee S1s have. the MCU interface to the Anaren is either/or SPI or UART. Both are in place.
They are amazingly smaller than the XBee, and much lower power consumption. They are much smaller than it seems from the photos.

Below, the link is the ZigBee version. The 802.15.4 version that I've worked with is the same hardware, but with the Zigbee firmware replaced with the 802.15.4 only, via SPI/UART.
http://www.anaren.com/air/products/air-for-zigbee-standard-apps
Mouser and others sell these.

One could also program the teensy to which the Anaren module so that either (a) there is a new library that has the same calls as the XBee S1 library or (b) emulate the AT commands and Xbee binary API. It gets a lot easier with 64KB of RAM.


I do note that Anaren has changed ownership and that's some risk.
 
Last edited:
But so far, this specific request isn't really clear to me, both in terms of what ought to be on the board, and why. I really need to understand more about what uses the board will really have.

In my experience most of the uses for micro-controller boards like teensy fall into two primary categories. First, breadboarding various experiments, trying new things and developing software. For this current teensy format is OK, although the popularity of breakout boards certainly tells one thing - having powerful processor and not being able to use all its resources is really bad idea.

Second category is actual use of the device in personal or commercial projects. For several years I've been working with a company using Atom board and then Bot Board II carriers for custom-designed robotic equipment. Interesting, that they often went with less powerful controllers only to keep the convenience of wiring.

Those people saying that it is impossible to suit all the needs are correct. That is why I think the simplest (and coincidentally cheapest) carrier board is a way to go. Most of the devices typically connected to the controller need just three wires - potentiometers, servos, serial motor controllers, some displays. Having standard servo headers in groups with jumpers to select between 3.3 and external power will cover a lot of requirements. Additional power regulator providing at least 1A of 3.3V is a bonus. Header for remote reset button is nice to have. The rest of buttons, buzzers, LEDs and proprietary headers are overkill, in my opinion.

So, here are my two suggestions (most of this was already suggested above, so I am only adding my vote to those):

1. Make next version of teensy with full pinout. It will not increase board size much and you can use additional space for CAN transceiver, which will sure be appreciated.
2. Make a small carrier board routing each pin to servo-compatible 3-pin headers connected to selectable power rails.
 
Hi wwwoholic,

Speaking as someone who spent quite a few years working with the Lynxmotion Bot Board 2 (by Jim Frye) with the Basic Atom Pro by Basic Micro and/or the Arc32 board, I know what you mean.

Although personally, I believe having a few of those other features like an LED or two, maybe a button and maybe a buzzer do come in handy as in most of my experiments/projects I want some form of user feedback.

So over the last several years, I have made several different Teensy boards (again mostly for my own education and the fun of it.) Also for me personally these days I seldom do much with RC servos, but do use Robotis Dynamixel servos, so I now typically include a setup to run them.

What I have found is for me, there is no one design that fits for everyone (even myself from day to day). My first boards had XBee connectors on them as at the time, I was doing a lot of stuff with XBees (Back to Lynxmotion days with Remote control with XBee...), then others suggested making a version with Arduino shield connectors on it, which I have now done a few versions of. I then did a simpler version somewhat geared around the Teensy LC, and the most recent is one that I am playing with to emulate an Trossen Robotics Arbotix PRo controller (clone of Robotis ...730),

Which board works best? Again depends on what you need... Also with many of these boards, since I have to populate them myself, If I don't need a feature, I don't populate it on the board...

The three boards, that I still play around with and at times make fixes for include:

Simple Teensy LC (Works good on 3.2 as well)
Teensy-ax-board-bld.jpg

The last one I did with the Arduino headers was a through hole version, to make it easier for those of us who are all thumbs to solder. If I were going to make some of these I would make changes, like move the Reset jumper like I did on some others so it aligns with 3.2 and use Pongo(sp) pin... Also probably update AX Buss handling depending on how experiments with the other two boards...
Teensyduino-TH-3d.jpg

Current board I am playing with, is the one to emulate Arbotix Pro (Currently waiting for a batch from OSHPark), the empty looking area is setup to plug in an Adafruit BSO055 IMU:
Teensy-Arbotix-Pro-V03.jpg

Note: I am showing these different versions of my own play time, not because they are great, but to say that there are sure lots of ways one could go with these. But if PJRC had a version that worked like the old BB2, I sure would have been having a lot of fun playing with them!

Kurt

P.S - Actually there is a 4th Teensy board, I have recently been having fun with as well. But is is specific to playing with Adafruit Neopixels and are setup to Plug into a Bioloid servo chain (Pardon the pro trinket at the bottom of the picture):
Teensy-and-Trinket-Neopixel.jpg
 
After some thinking, I'd like to clarify my suggestion for making full pinout teensy. To keep in line with the goal of having tiny board, I think the best option is to:
stretch the board just enough to accommodate 3 more rows of 7 holes on a side opposite from USB. That's 21 more pins with only 7.62 mm longer board.
Those who don't need them will not notice, but those who need all the processor power will be ecstatic.
 
You can order this board that FrankB designed from OSHPARK that you solder into place to attach the bottom pins into a 2x7 extension to bring out the pins. The current OSHPARK price is $3.45 for 3 boards. It does say you need his permission if you are going to use it in a commercial setting. I've ordered some, but I haven't soldered them onto the boards right now: https://oshpark.com/shared_projects/cYQjHdCu.
 
Yes, that is exactly what I had in mind, Michael. Except with 21 holes (14 do not cover all necessary connections) and without turning neat teensy board into another Frankenstein's monster as all these breakouts do.

I have to admit that I am looking at this from the position of a user incorporating the controller into small-scale commercial development where buy-plug-run approach is preferred to any tweaking of the board just to get started, which is good mostly for hobby projects. While hobby enthusiasts can support the business for sure, only appealing to commercial customers can bring real revenue.
 
Last edited:
After some thinking, I'd like to clarify my suggestion for making full pinout teensy. To keep in line with the goal of having tiny board, I think the best option is to:
stretch the board just enough to accommodate 3 more rows of 7 holes on a side opposite from USB. That's 21 more pins with only 7.62 mm longer board.
Those who don't need them will not notice, but those who need all the processor power will be ecstatic.
yes, good compromise
 
Yes, that is exactly what I had in mind, Michael. Except with 21 holes (14 do not cover all necessary connections) and without turning neat teensy board into another Frankenstein's monster as all these breakouts do.

Hi, which additional connections do you need ?
 
I asked for 21 because breakout board adds 20. Hmm... actually 14 might be OK, it covers 12 pads and reset. So only 2x7 extra holes, 5.1 mm extra length. Even better!

Oh, and I am removing a request for on-board CAN transceiver. They vary a lot depending on the environment and industrial use high-voltage fault protected transceivers will be better mounted separately.

So, in the end... exactly the thing MichaelMeissner suggested above only as part of Teensy itself. Getting back on topic, it will make proposed carrier board much simpler.
 
Last edited:
I asked for 21 because breakout board adds 20. Hmm... actually 14 might be OK, it covers 12 pads and reset. So only 2x7 extra holes, 5.1 mm extra length. Even better!

Oh, and I am removing a request for on-board CAN transceiver. They vary a lot depending on the environment and industrial use high-voltage fault protected transceivers will be better mounted separately.

So, in the end... exactly the thing MichaelMeissner suggested above only as part of Teensy itself. Getting back on topic, it will make proposed carrier board much simpler.

Yes, CAN is of little use for most of the users.
There's a new Teensy planned... nobody knows the pinout yet, but i will have more connections.

Or just try my board :) It's only ~1,15$ incl. shipping...

(p.s. i earn nothing with it, not a cent)
 
Last edited:
Hi Frank,
Your board is nifty but as I mentioned before, I do not work with hardware directly, I am writing software for it. The guys I am doing it for do not bother with soldering, they are focused on mechanical part of things. The controller for them is just something they stuff into a box, plug their wires and forget. That is why they still using BBII carrier with inferior processor, and they buying bunches of those. I am trying hard to persuade them to switch to Teensy, but they won't do it unless they have pre-assembled carrier with servo-style 3-pin connectors as drop in no-soldering-required replacement. That's what landed me on this forum to begin with.
 
Ah, i see. Maybe the users Onehorse (US, sells on tindie) or Pensive (UK, own shop) are interested to make it ?
 
Back
Top