Another Teensy 4.0 Breakout Board

I fear already that mine will look the same when they’ll arrive here in France in a few days... Why don’t they use a precise laser to do the cut-outs?

The edge cuts are generally milled out mechanically. I think that a laser to cut the boards would need to be pretty powerful.

I know that the flex boards that OSHPark does are cut with a laser, but that laser won't cut traces, just substrate, so castellated connections can't be done directly with them.
 
Re thicker boards - you can purchase male headers with longer pins that will make the soldering easier on the top. I have a few with long pins on both sides. And it's relatively easy to move the pins up and down in the plastic rail - which I frequently do.

I still need the the modded board with extra spacing between the rails, but seeing as I have 8mm ones already at the fab, I'll wait 'til I check those out before reordering thicker ones.
 
My boards came with the second version of the layout (0.14" depth), and I'm puzzling out how to set up my Teensy.

Normally I use large female stacking headers cut down to size:

I like using female stacking headers, because it allows me to either use a standard breadboard or to wire up a prototype board that has various connections using the standard pinouts. I can go Arduino shield style and stack things within reason, with each stack bring out something different (SPI, I2S, etc.). Also on the top stack, I can use those rows for using jumper wires to add more things (or to try things out during debugging).

I realized I can't use female stacking headers like I'm used to because in order to use the board, it has be next to the Teensy, so you can make solder bridges to connect the two boards. I imagine I could rig up a jig when soldering so the stacking header isn't sitting directly on top of the Teensy, and I carefully solder all of the connections.

Or I can use extra long male pins with half of the pin below the Teensy, and half above. I would solder the Teensy on the top, flip it over and solder the underneath pins. I would then have to rework my boards to put the female headers underneath instead of on top.

As has been noted, having the extra pins on the row outside of the two rows of 14 pins is problematical for standard breadboards.
 
I've gotten the SD card sidecar and the USB Host sidecar in .8mm and the ESP32 sidecar in 1.6mm. Assembled and all working fine. I haven't tested the 3.3v regulator on the ESP32 board yet, running off the built-in 3.3v from Teensy seems to be ok, even while doing some basic wifi.

I have a bit of software to write to get the firmware updated on the ESP32 module, but with the factory firmware I'm able to associate my phone with the module.

For the USB, SD and ESP32 sidecars, I don't think it matters what thickness. If you are ordering from OSHPark, the thicker versions are likely to ship faster.

For the Teensy 4.0 breakout board, folks have had poor success with the 0.8mm process, my single order of 1.6mm came back fine.

So, unless you have a specific need, I'd go with the 1.6mm for everything now.

Ordered in regular 1.6mm PCB.

I'll keep this thread updated once I receive them to let you know how they turn out.
 
I did receive the thicker ones and I am trying to decide if I will try one of them out or not. They do look exciting to try to solder in, but maybe not any worse than the other ways to try to get these pins.

I guess for me, the question is, how/where would I use this board? With the Tall Dog version, all of the pins are setup that I can plug them into breadboard, which is how I use all of the ones I have from Tall dog. (I think I have 2 of them with 3.6 and another with 3.5, and another one I have not used yet...
Tall dogs board is a .1" wide than this one, which is good and bad. Bad is it is .1" wider... But the good thing is that with this, It has has labels for each of the pins on board, which comes in handy when I am experimenting.

Where this board could come in handy is if I design another board to plug this one into which is a possibility. Still deciding if I want to take that route or not.
Maybe it might work pretty well if that board had something like: https://www.amazon.com/uxcell-2-54mm-2x14-Pin-Straight-Connector/dp/B07F3RJBFS/
On it to plug these into...

Maybe I will try a quick and dirty one. Maybe my RA8875 test board I was thinking of hacking up...
 
Received my three 0.8mm breakouts yesterday. Unusable, went directly to trash. The small castellation for the bottom pads 24,26,...,32 simply missing, the big castellation totally misplaced. First time I have such a bad experience with OSH Park. :(
 
Well, I had a very pleasant and unexpected surprise today! I had ordered my modified versions of Blackketter's breakout board in 0.8mm thickness from OSHpark before we learned of the woes with those boards. They arrived today, and I was ready to toss the whole bunch, BUT, when I opened the package I was amazed to find that they were actually 1.6mm and perfect in all the castellations!!! No tearing of the copper anywhere.

So I set out to assemble one, and it went very smoothly. Soldering the castellations was no problem, even the small 1mm ones. I've tested all the pins with a blink function, and they all work except one - pin 34 (on one end of the 1mm castellations) - which is stuck high! I've resoldered it, examined it under a loupe, checked the KiCad drawings for the layout, measured the resistance to adjacent pins etc, and while all looks good, it is still high!

Anyway it's usable for prototyping, even with that nonfunctional pin, it looks good, and I really like the form-factor. I'll build another one tomorrow morning, and post photos. I'm letting the first one "soak" overnight.
 
Well, I had a very pleasant and unexpected surprise today! I had ordered my modified versions of Blackketter's breakout board in 0.8mm thickness from OSHpark before we learned of the woes with those boards. They arrived today, and I was ready to toss the whole bunch, BUT, when I opened the package I was amazed to find that they were actually 1.6mm and perfect in all the castellations!!! No tearing of the copper anywhere.

So I set out to assemble one, and it went very smoothly. Soldering the castellations was no problem, even the small 1mm ones. I've tested all the pins with a blink function, and they all work except one - pin 34 (on one end of the 1mm castellations) - which is stuck high! I've resoldered it, examined it under a loupe, checked the KiCad drawings for the layout, measured the resistance to adjacent pins etc, and while all looks good, it is still high!

Anyway it's usable for prototyping, even with that nonfunctional pin, it looks good, and I really like the form-factor. I'll build another one tomorrow morning, and post photos. I'm letting the first one "soak" overnight.

Looking forward to an update! Got 3 each of the @blackketter USB and SD boards today and no base board to use them with given the thin board fail. Perhaps should have already ordered the thick like KurtE.

Odd you have issue with #34 - on the @loglow beta board I found issues with #36 IIRC perhaps another so ignoring those for first rev and it works USB and otherwise so far. But was working on a sketch to ring those out as they seemed odd.
 
Well, I had a very pleasant and unexpected surprise today! I had ordered my modified versions of Blackketter's breakout board in 0.8mm thickness from OSHpark before we learned of the woes with those boards. They arrived today, and I was ready to toss the whole bunch, BUT, when I opened the package I was amazed to find that they were actually 1.6mm and perfect in all the castellations!!! No tearing of the copper anywhere.

So I set out to assemble one, and it went very smoothly. Soldering the castellations was no problem, even the small 1mm ones. I've tested all the pins with a blink function, and they all work except one - pin 34 (on one end of the 1mm castellations) - which is stuck high! I've resoldered it, examined it under a loupe, checked the KiCad drawings for the layout, measured the resistance to adjacent pins etc, and while all looks good, it is still high!

Anyway it's usable for prototyping, even with that nonfunctional pin, it looks good, and I really like the form-factor. I'll build another one tomorrow morning, and post photos. I'm letting the first one "soak" overnight.

Glad to hear you got a (mostly) working board. I just checked my pin 34 (ie. the top left pin on the breakout board and it seems to work fine by toggling it on and off in software. I don't see an obvious way for a mechanical short to 3.3v, what was the measured resistance to the 3.3v power pin?
 
My bad, in fact my very, very bad today:( :( :( So, I was all very pleased with myself this morning when I wired up breakout #2. Everything went swimmingly, I soldered all the castellations, installed the pin headers, remembered to cut the USB power jumper.... Fantastic! Just one thing left to do - solder the USB power connecting wire through the two boards. Funny - I can't seem to poke it through! Then I looked, and said a very naughty swear word because - wait for it - I had soldered the breakout to the T4 UPSIDE DOWN. It was flipped right to left - and of course all cut-outs and castellations are symmetrical about the midline! Silly me, silly me, silly me...

And that was the sad end of my very last unused T4. There was no way to take it apart and salvage anything. At least I have board #1 to play with.
 
My bad, in fact my very, very bad today:( :( :( So, I was all very pleased with myself this morning when I wired up breakout #2. Everything went swimmingly, I soldered all the castellations, installed the pin headers, remembered to cut the USB power jumper.... Fantastic! Just one thing left to do - solder the USB power connecting wire through the two boards. Funny - I can't seem to poke it through! Then I looked, and said a very naughty swear word because - wait for it - I had soldered the breakout to the T4 UPSIDE DOWN. It was flipped right to left - and of course all cut-outs and castellations are symmetrical about the midline! Silly me, silly me, silly me...

And that was the sad end of my very last unused T4. There was no way to take it apart and salvage anything. At least I have board #1 to play with.
DOH! So sorry, I nearly did that myself. I'll update the design with a silkscreen note that says "This side faces Teensy!"
 
My bad, in fact my very, very bad today:( :( :( ...

And that was the sad end of my very last unused T4. There was no way to take it apart and salvage anything. At least I have board #1 to play with.

Very sad for that T4 - Still work on topside pins right?

And we lost a chance to hear if board works :(
 
So I got the sub-boards USB/SD - but haven't RE-ordered a main breakout again yet. Is there a right one to order?
I just updated the design with the new silkscreen warning, and iit also brings out ground to the pins for USB. I haven't ordered one yet but I think it should be ok.
 
My bad, in fact my very, very bad today:( :( :( So, I was all very pleased with myself this morning when I wired up breakout #2. Everything went swimmingly, I soldered all the castellations, installed the pin headers, remembered to cut the USB power jumper.... Fantastic! Just one thing left to do - solder the USB power connecting wire through the two boards. Funny - I can't seem to poke it through! Then I looked, and said a very naughty swear word because - wait for it - I had soldered the breakout to the T4 UPSIDE DOWN. It was flipped right to left - and of course all cut-outs and castellations are symmetrical about the midline! Silly me, silly me, silly me...

And that was the sad end of my very last unused T4. There was no way to take it apart and salvage anything. At least I have board #1 to play with.

One more thought: I don't see why the teensy itself shouldn't continue to work and can't think of a reason that the castellated pads that you soldered wouldn't work, though on the wrong side of the board.

If you ordered the updated version with the grounds connected to the 4th and 5th pins from the top, they wouldn't be connected properly (and instead be connected to digital pin 0.)

What I think this means is that all the breakout pins should still be available and the SD sidecar would work if installed upside down on the other side! The USB and ESP32 sidecars would be more difficult to rework.
 
My bad, in fact my very, very bad today:( :( :( So, I was all very pleased with myself this morning when I wired up breakout #2. Everything went swimmingly, I soldered all the castellations, installed the pin headers, remembered to cut the USB power jumper.... Fantastic! Just one thing left to do - solder the USB power connecting wire through the two boards. Funny - I can't seem to poke it through! Then I looked, and said a very naughty swear word because - wait for it - I had soldered the breakout to the T4 UPSIDE DOWN. It was flipped right to left - and of course all cut-outs and castellations are symmetrical about the midline! Silly me, silly me, silly me...

And that was the sad end of my very last unused T4. There was no way to take it apart and salvage anything. At least I have board #1 to play with.

well since you seem to have written off that t4 you might as well try to take apart. nothing to lose. heating each pin soldered to teensy until you can pull it. gently, slightly wedge a toothpick or similar between the 2 boards at castellations then with tiny bit of flux heat each one , pressure between the 2 boards might be enough to slightly lift and break connection. none of that might work but at least something was tried.
 
@dkryder,
Don't think I gave up without trying :rolleyes: I sent a couple of hours trying to prise them apart... It was just not possible to heat the pins through both boards and extract the pins. In the end I even tried cutting through the breakout with a fine Dremel cut-off wheel to try to make it easier. As you might imagine that didn't go so well :mad:

It just wasn't my day for Teensies yesterday. I ordered 3 more T4's from PJRC (at least I thought I had) - went through the whole rigamarole - put in credit card info etc, hit the order button and went away for a cup of coffee for 30 mins or so. When I came back there was an error on the screen saying I had not filled out the shipping address page, which I know I deliberately left blank as instructed. I shot off a quick note to Robin asking if the order had come through, I haven't heard back yet, and since there hasn't been an email order confirmation I'm assuming that it fell into a black hole along the way.

Sooo, at 3am this morning I discovered you can buy T4s on Amazon (for more money but free shipping with Prime). So I ordered 2 more. The order was confirmed and all was good for a few hours, and wouldn't you know it - I got a message from them saying they didn't know when they might deliver them...

So together with one at OSHpark I ordered a couple of weeks ago I now have 6 T4s out somewhere in the ether in various stages of being lost - and not a single one to work with.

Somebody just doesn't want me to work with Teensies any more :confused:
 
@dkryder,
Don't think I gave up without trying :rolleyes: I sent a couple of hours trying to prise them apart... It was just not possible to heat the pins through both boards and extract the pins. In the end I even tried cutting through the breakout with a fine Dremel cut-off wheel to try to make it easier. As you might imagine that didn't go so well :mad:

It just wasn't my day for Teensies yesterday. I ordered 3 more T4's from PJRC (at least I thought I had) - went through the whole rigamarole - put in credit card info etc, hit the order button and went away for a cup of coffee for 30 mins or so. When I came back there was an error on the screen saying I had not filled out the shipping address page, which I know I deliberately left blank as instructed. I shot off a quick note to Robin asking if the order had come through, I haven't heard back yet, and since there hasn't been an email order confirmation I'm assuming that it fell into a black hole along the way.

Sooo, at 3am this morning I discovered you can buy T4s on Amazon (for more money but free shipping with Prime). So I ordered 2 more. The order was confirmed and all was good for a few hours, and wouldn't you know it - I got a message from them saying they didn't know when they might deliver them...

So together with one at OSHpark I ordered a couple of weeks ago I now have 6 T4s out somewhere in the ether in various stages of being lost - and not a single one to work with.

Somebody just doesn't want me to work with Teensies any more :confused:

ok. but for future reference i can make a suggestion that might help. i have no idea if it would help in this case. it sounds like it would not. but ive been in similar situations when a secondary pcb was soldered close to a primary and i had a tool in my kit that was kinda shot in dark last hope. it was one of these.

https://amzn.to/2VHQZht

the tip metal can be shaped to fit snuggly over the pin and pin slides into tip. i ended up also shortening the pin so there was not a lot heat sink mass, pin length just long enough to slide a bit into shaped tip of iron and leave some for needle nose. thing is after a period of time you want to back off pin just enough to get a good pair of needle nose on pin to yank it out. i have found this is best way i know so far to direct heat to a pin other than heated air, but heated air wouldn't work here.
 
Bummer on the board loss - and restocking issues. OshP got me a T4 okay with PCB's in expected time, Never trusted to order Teensy from Amazon. AdaFruit and Sparkfun got a boatload of batch #3 and both are showing dropping stock levels.

ok. but for future reference i can make a suggestion that might help. i have no idea if it would help in this case. it sounds like it would not. but ive been in similar situations when a secondary pcb was soldered close to a primary and i had a tool in my kit that was kinda shot in dark last hope. it was one of these.

https://amzn.to/2VHQZht

the tip metal can be shaped to fit snuggly over the pin and pin slides into tip. i ended up also shortening the pin so there was not a lot heat sink mass, pin length just long enough to slide a bit into shaped tip of iron and leave some for needle nose. thing is after a period of time you want to back off pin just enough to get a good pair of needle nose on pin to yank it out. i have found this is best way i know so far to direct heat to a pin other than heated air, but heated air wouldn't work here.

Indeed that "Velleman VTDESOL3U Vacuum Desoldering Pump with Heater 30W " can help ... used it a couple of times to good effect for pin extraction and board re-use - done just in last weeks.

Avoid a yank or pressure on any parts you hope to keep though as nearly melted solder might hold better than the pad/metal laminated to the board at that same temp. That applies more to under pads perhaps than the through hole rings, but seems there have been notes about such loss.

Though all those under pads in this case getting all cleaned and released even with the boards not pinned/soldered together without spacers to cut would be difficult.
 
Another datapoint: I got my 0.8mm boards back today, finally, and only one of the three of them is usable. The cuts were all made on all the boards but two of the three boards each had the plating on a castellation torn out.

By comparison, the 1.6mm boards are quite clean.

I do think that with some the addition of some area to the top and bottom annular rings of the castellated pads and some tweaking to the position of the milling lines for the castellations the yield could go up at OSHPark for the 0.8mm boards, but I'm hesitant to put in any effort if folks are satisfied with the 1.6mm boards. I know I am.
 
1 of 3 better than none. I place a 1.6mm order that should arrive soon.

Soldered one of the FRDM4236 boards with staggered castellation SDIO pattern and it went well - looking now it seems to be a thinner PCB though he had done not at OSH. Also going well were the larger box cuts on the 2x5 area. It left room for my iron to get the solder in while isolating the underside SMD components from damage. And skipping the GND and 3.3V on the SDIO with the staggering likewise left buffer area to prevent bridging the needed pads.
 
this thread has been an eye opener for me. i would have never thought that OSH would just continue churning out defective boards once they were made aware of the issue. .8mm & castellation just don't work with this pcb and they know it yet they keep making and shipping them. i'm glad i've read the thread though cause now i'm aware of the .8mm problem in case i am looking at any other pcb in that thickness.
 
It looks like this time they routed for the 2x5's at least - that was a fail on their part if the 'gerber' was clear for removing that area.

As far as the 1mm castellation cuts - that is close tolerance - the FRDM4236 layout might lead to a better direction

oshpark.com/profiles/trainer4edu's Newest Ver3 with pickup of USB Host pins :: A third Teensy 4.0 breakout board - V3 with USB host

Without nearby metal to better hold the 'tiny' though hole plating top and bottom seeing it tear off on cutting isn't a surprise - and at double copper it will cut harder too.
 
Last edited:
this thread has been an eye opener for me. i would have never thought that OSH would just continue churning out defective boards once they were made aware of the issue. .8mm & castellation just don't work with this pcb and they know it yet they keep making and shipping them. i'm glad i've read the thread though cause now i'm aware of the .8mm problem in case i am looking at any other pcb in that thickness.

They were pretty clear with me that the castellated designs were not guaranteed. Their tips are here: https://docs.oshpark.com/tips+tricks/castellation/

My guess is that the thinness of the board and the thickness of the traces (2oz vs 1oz) conspire together to make this a little tricky to get right reliably.

While I'm disappointed that the 0.8mm design didn't work first time, I suspect that it could work reliably with a few iterations.

For production volumes it would make more sense to work with a board house to dial in the design, production and inspection process to get it right reliably.
 
Back
Top