Testing the new reduced-sized ICSP connector

Posted on Monday, December 17th, 2012 in #liveupdates, demonstrations, Development, manufacturing, programmers, Prototypes, testing by Sjaak


This week we received a batch of PCBs and amongst them was the latest revision of our USB TFT breakoutboard. This board is the first board that features a new ICSP footprint, which occupies less space on the PCB and should improve testing and programming times at Seeed.


The footprint consist of a key hole and 5 small pads. Each pad carries an ICSP signal (/MCLR, V+, GND, PGC and PGD). One of those pads has a small hole, which helps to align the probe better. The spacing between the pads is the bare minimum required so the pogopins wouldn’t touch.


The design is based on little pogopins we found on our geek tour in Shenzhen, China. Due to the lack of a datasheet, we measured the pogopins precisely and made our own.

We were eager to while away the winter doldrums by hacking on this board, but after soldering it we realized there’s no adapter for the new programming header yet. We needed a way to program it quickly.


We hacked this programming adapter together with a couple of pogopins. Here’s the final result alongside the regular .1″ programing header we normally use. The first person that guesses the datasheet on which this photo was taken will receive a free PCB of the USB TFT breakout!

To make the adapter:

  1. First we used nail polish to make the outside of each pin less conductive
  2. Next we carefully aligned the pins on a piece of tape and glued them together with hot-glue
  3. Hot-glue is fantastic stuff, but it cooled too quickly and everything got misaligned No worries, we just reworked it into submission with our trusty hotair station.

Not the most repeatable process, but a PCB and laser cut acrylic programming probe is on the way. We expect it will be much more elegant.

After the glue set we tried out the new programmer. Amazingly it worked right away the first time and without any hassle!

This entry was posted on Monday, December 17th, 2012 at 9:00 pm and is filed under #liveupdates, demonstrations, Development, manufacturing, programmers, Prototypes, testing. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

16 Responses to “Testing the new reduced-sized ICSP connector”

  1. Superfro says:

    According to google, the datasheet is a TMP512 OR TMP513 from TI. :)

  2. Dan Collins says:

    It is page 6 of the TMP512 datasheet, as found by Superfro above.

    • BlueSkeye says:

      Take a closer look at the datasheet. You’ll see this is the Texas Instruments Temperature and Power Supply System Monitor TMP513 model due to “negative differential shunt voltage”. The TMP512 has a “positive differential shunt voltage” spec.

  3. Louis Mamakos says:

    If you’ve not seen it, you might consider this product: if you don’t want to have to fabricate your own connectors/plugs.

    • Sjaak says:

      we seen those, except they are vey bulky. Also we had a couple of pogopins at disposal :)

      • ken says:

        I’d agree w/ Louis. Tag-Connect has their 2030 cable which is fairly small (meant for low pin count connections like SBW or SWD) with an RJ11 jack on the other end of it.
        The benefit is it’s easily purchasable (i.e., a standard) so any hobbyist can use it. Building your own connector is great for yourself, but still IMHO a waste of time…my time is worth more than a $30 cable :-)

    • kmel says:

      Nice find!
      Will order a TC2030-MCP-10 asap… no more RJ12 jacks anymore!

      Why inventing the wheel again?

  4. Markus Gritsch says:

    I have a bit mixed feelings about a custom made programming connector. Makes it a bit harder for tinkerers to program such a board by themselfes.

    • Sjaak says:

      All our boards come preprogrammed with a bootloader, so not much an issue for board purchased through ‘official’ channels. I tried to make that point into the post.

      I agree on people that get a free pcb. If we ever move over to a custom made programming connector, we’ll sell them with Seeed off course

      • Ian says:

        I think you’ll see the standard 5 pin header on our boards whereever possible. We’re a long way from using this as our only ICSP connection. This is part of an experiment for designing for a standard test and programming rig. We’ll play for a while before making any changes to projects.

  5. BlueSkeye says:

    Take a closer look at the datasheet. You’ll see this is the Texas Instruments Temperature and Power Supply System Monitor TMP513 model due to “negative differential shunt voltage”. The TMP512 has a “positive differential shunt voltage” spec.

  6. Simon says:

    I’ve been using the Tag Connect cable for some time now, and imho there’s a big flaw with it:

    You can’t route signals between the pins!

    I think you might have this problem with your pogo pin connector too. Even though you have designed it for minimum pcb size, the fact that you need to route around it, you have effectively enlarged the area it requires.

    But hey – you guys HAVE MADE YOUR OWN. Well done!

  7. Brx says:

    I’ve looked at using the TagConnect TC2030 or other models for years for JTAG, I2C, or other ICSP, but always ended up turning away from it because we only needed such a connection for proto bring-up & debug, while production programming would be through a ‘bed of nails’ test fixture — and for eng debug, a 1.25mm or 1.27mm pitch thru-hole header took up less space – and they can be made to self-retain pretty well by carefully bending the pins into to a common ‘wavy pin’ pattern.

    A real advantage of the TagConnect TC2030 and similar models is that the guide pin pattern is designed to protect the rather delicate pogo pins from the common abuse of getting smacked around on a table-top or floor by assembly techs and they make it easier to hold a decent connection together with one hand. I’ve finally found a good app for the TC2030 in designs for medium-volume boards where the producer doesn’t want to bother with a bed-of-nails test fixture. I think it will be worth the 30$ ea on the production line.

Leave a Reply

Notify me of followup comments via e-mail. You can also subscribe without commenting.

Recent Comments

  • Edward Mallon: A lot of us have ended up at this sensor / screen combination. But I couldn't afford the extravagance of six dedicated control lines on...
  • Andy: Me!!!
  • Bill F: Taking my place in line........
  • Mikee: 11
  • Clara: Thanks, Cinco!