Experimental MicroMod "T3.2" board

I've seen JLPCB offers SMT assembly for free (one side on the PCB) .. and only 72Hrs extra .. wow. Have to look how this works!
less than 9€ for 5pcs 100 x 80mm boards..shipping incl.
Just ordered a few for my "living room audio mixer"
 
Last edited:
I asked them about soldering that MicroMod connector. Unfortunately they will only assemble what's in their parts library. This library is suprisingly large and they have a lot of NXP processors (e.g the MK20DX256) but not the correct IMXRT.

I'm already doing a motor driver carrier board for my MM32 board from above. Want to use that offer as well :) (need to hurry up, its only valid until Christmas I think...).

Screenshot 2020-12-22 115950.jpg
 
Ah, ok.
But they have the ESP in their library. Nice. I'll look for other WiFi chips, too...

The motor dirver looks great. The floating parts are most nice;)
*kidding*
 
The motor dirver looks great. The floating parts are most nice
*kidding*
Yes, they are a bit fragile but I want to check the assembly capabilities of JLCPC. Lets see how they manage that :)
 
I've seen JLPCB offers SMT assembly for free (one side on the PCB) .. and only 72Hrs extra .. wow. Have to look how this works!
less than 9€ for 5pcs 100 x 80mm boards..shipping incl.
Just ordered a few for my "living room audio mixer"

Well, for free is a bit of a come-on. There will be engineering fees, a lot of parts have a min order count and smaller SMD parts have "attrition" (think of it as the number of parts wasted loading the tape on their PnP machines. Also, the inventory of parts in their library is quite variable, suffering from outages relative frequently. I generally avoid using parts that have a low instock count.

But still, their assembly service works well and I've used it for quite a number of boards. The Wifi modules are a recent addition.

The web interface to their library is a bit clunky. I use this instead.

Also, if you use Kicad you can add the LCSC numbers to your symbols and then produce an exact BoM with a script (JLCPCB points you to it).
 
I meanwhile finished the design of a MicroMode compatible stepper driver carrier. So, I'm ready for the MM Version of the T4 :) (or, at least I will have something to play with my (hopefully) T3.2 compatible MM board from above)

3d_top.jpg 3d_bottom.jpg

Features:
  • Up to 4 stepper driver modules (DRV8825, A4988 etc)
  • V_In can be choosen from USB (standard) or an external supply
  • V_mot 8-35V
  • Microstep selection and enable done via a I2C multiplexer (MCP23017)
  • Step/Dir signals directly connected to the GPIOs
  • A couple of signals are broken out on to 100mil pads to connect to the real world (see silk screen on the bottom layer)
  • Two layer, 80x50mm
  • USB-C connector

GitHub: https://github.com/luni64/mmStep
 
Yay... after some delay due to the Christmas holidays the boards for my first DIY Teensy (compatible to the SparkFun MicroMod carriers) arrived today...

MicroMod3.2_empty.jpg

IMG_0108.jpg IMG_0116 (2).jpg

.. and I tested my rusty 0603 soldering capabilities. They definitely need some brush up but after all the board actually is alive. Here the inescapable blinking evidence: https://youtu.be/BuMABkxE_Ro

Unfortunately the Bootloader chip is somehow dead, so I uploaded with a JTAG/SWD probe (J-Link EDU Mini) connected to the debug header on the SparkFun carrier. Need to have a look what I did wrong with that chip. Soldering seems to be OK. Does anyone know if the pad on the bottom of the chip needs to actually be connected to GND or is this just for cooling and the other GND pins are sufficient?

Anyway, this gave me a chance to test JTAG debugging using visTe (formally known as VisualTeensy) and CortexDebug. Here a quick debugging test (video is a bit chaotic but you'll see how it works...): https://youtu.be/3MQKpvlO7OM. Need to get hold of the correct *.svd file to get access to the peripheral registers as well.

So, now find the bootloader issue...
 
Actually I very much hope that Paul is working on it and that the rumors are not just rumors... :cool:

IMHO, this form factor is really intriguing for integrating processors into custom carrier boards. I'm currently working on a test board implementing the STUSB4500 to get (up to) 100W over the USB-C line to such a carrier. That would be perfect to control and power motorized devices over USB without dedicated power supply. Lets see if that works out. Currently PD capable USB PC cards are a bit rare but I think that will change.
 
Looks nice. I am sort of waiting for the T4.x version. Not sure when/if that might happen, but I did order a Micromod board and an ESP32 board for it that shipped today.
 
Thanks for the feedback. Meanwhile, I did some measurements and found that the boot input pin is permanently low. I assume some shortcut under the bootloader or it simply is broken...

Anyway, I built a second board which worked without any issue :). Good thing with those carrier boards is that they provide all the required infrastructure (power, reset and boot buttons...). So, just to make sure the design is correct, I only soldered the bootloader on the board. With this minimal setup I checked if it generates some signals on the JTAG/SWD lines after pressing the boot button which this time it did. The rest of the parts was quickly soldered and the board seems to work now. The RTC is nicely fed by the battery on the SparkFun carrier and keeps the time after power cycling.

Next I'll try the 8bit data bus and the rest of the signals.

@Paul/Robin: Do you have an European distributor for the bootloaders? Would be much easier than buying them in the US...
 
I'll build one or two of these boards, too., as you know ;) Thanks for the boards etc.
Have you thought of "upgrading" to a Teensy 3.6 "MM"? Is there room enough for the non-bga version on the board?
 
Has anyone verified that grounding PTA1 (or more likely PTA3) on the MKL02 disables it properly?

To follow that up: Grounding PTA1 works, the SWD pins get tri stated and an external debug probe can connect. Grounding PTA3 does disable the bootloader (it doesn't react to the boot button anymore) but doesn't tri-state the SWD pins.
=> PTA1 (~RESET) of the MKL02 is the correct pin to ground if you want to attach a hardware debugger.
 
I got the boards. Oh..they are SMALL ! Thank you, Luni!

Oh my.. my eyes.
I have an oven. Never really used so far. Maybe i try it.
Does anybody have any tips for me, how to use it? I have no stencil. Does it work without?
I own a simple hot air station, too. Is it better? I've used it some times, but needs a bit experience, too.
With an soldering iron, I had never luck with this pitch.


Have you thought of "upgrading" to a Teensy 3.6 "MM"? Is there room enough for the non-bga version on the board?
Hm, this chip is bigger than the whole board.
 
I got the boards. Oh..they are SMALL ! Thank you, Luni!

Oh my.. my eyes.
I have an oven. Never really used so far. Maybe i try it.
Does anybody have any tips for me, how to use it? I have no stencil. Does it work without?



Hm, this chip is bigger than the whole board.

I have not had success without a stencil, but I haven't tried it more than once. Maybe someone more experienced at freehanding it can comment...

Regarding the reflow, I recommend the following temperatures, which are based on using Kester EP256 Leaded solder paste. I've done lots of prototypes this way with good success:
1. 6 minutes at 250F. Timing isn't very critical, this stage is to just pre-heat all the components so they are more likely to heat evenly during the next step. The long duration is because some parts will heat faster than others and this gives everything time to stabilize at the desired temperature.
2. Raise the heat to 450F. Once the oven reaches 450F (you'll know because the heating element will turn off and then cycle off and on to maintain temperature), I set a timer for 2 minutes.
3. At the end of 2 minutes, turn the oven off and open the oven door to drop the temperature.

Obviously, don't use the oven for food again because of the lead fumes. Also do this in a well-ventilated area. Lead-free solder paste will require a different reflow schedule.
 
brtaylor, thank you.
It is a real soldering oven, very heavy, and it is possible to select different temperature profiles. Never used :)
The Kester solder paste is not available here. Do you have a link with a description? Or @ EU or German users: Is there a good alternative?
 
Actually, hand soldering it is not _that_ difficult. If I can do it using my 40 year old DIY soldering station anyone can :). Some cheap loupe glasses help a lot...

Good thing is that my daughter is currently experimenting with video cutting, so I asked her making one showing the assembly of the board: https://youtu.be/CCVqATbAmtM

 
My experience is that paste is better, even when you don't have a stencil. But I save time and frustration - unless JLCPCB puts most of the parts on, I always order a stencil. More advice - use a syringe + tubing to make a mouth powered pick-and-place.
 
brtaylor, thank you.
It is a real soldering oven, very heavy, and it is possible to select different temperature profiles. Never used :)
The Kester solder paste is not available here. Do you have a link with a description? Or @ EU or German users: Is there a good alternative?

It's 63/37 tin to lead ratio. The full description is:
https://www.kester.com/products/product/ep256-solder-paste

I typically order through OSH Stencils, but looks also available on Amazon. I would check with the vendor if you go that route - Solder Paste does have an expiration date and gets noticeably harder to work with if it gets too expired. Keep the paste in the fridge. I typically remove when I start assembling my jigs and stencil and organizing my parts. It gives the paste about 20 minutes to warm up while still leaving me a lot of working time left. I use some high quality tweezers from EMS (https://www.emsdiasum.com/microscopy/products/tweezers/ultra_fine.aspx) for placement; although, I don't recall which ones and it's probably overkill. With the Teensy 3.2, you can always rework the pins with some solder wick to remove bridges after assembly too.
 
Obviously, don't use the oven for food again because of the lead fumes. Also do this in a well-ventilated area.

Because we occasionally do solder-based repair/rework (cables, components, etc.) in our development lab at work, we are required to take a Lead Safety class annually. One of the biggest surprises that came out of this class is that, at normal soldering temperatures, lead fumes are not created and therefore, lead fumes pose absolutely no hazard under normal conditions of soldering. Don't misunderstand: lead is still very hazardous, but never in the gaseous form when soldering. For this reason, eating, drinking, smoking, and chewing gum should be avoided when working with lead solder. Thorough hand washing is highly recommended upon completion of the soldering activities as well. Fumes from solder paste are the much bigger hazard. See <this> as just one reference for more information. Avoiding using your spouse's breakfast toaster oven for occasional solder (re)flow work is still highly recommended !!

Mark J Culross
KD5RXT
 
Back
Top