Teensy 4.1 - USB device connection extension - High Value Mistake!

T

Trouserhams

Member
Hi All - some help/advice would be greatly appreciated!

I've made a mistake by not paying full attention to the datasheet showing the USB pads under the 4.1 board to be USB host only and therefore unable to connect/programme through.

I've got 15 ENIG assembled PCB's that now won't function, along with full CNC'd cases for this PCB layout, and I cannot seem to get the USB connection to play ball.

The board is currently as so;

IMG_0324.jpg

I put the teensy round this way to keep the USB differential pair as short as possible, and used POGO pins underneath the board, but this obviously wont work due to the issue mentioned above with the host port, but this also means there is no space to use the onboard USB port either (I like a racey PCB, but this has obviously backfired here!)

I detatched the external connector pins from the PCB so the host port wasnt connected anymore, removed the original micro connector with a hot air rework wand and carefully soldered two same length wires to the external connector from the board D- and D+ connections.

The power is applied externally though the Vin pin, and I cut the trace under the board for this.

I joined the USB GND and my power supply GND for testing (eventually it will all be powered through the USB connection), but when plugging it into my PC, I get the following error from Device Manager;

Code: 
Windows has stopped this device because it has reported problems. (Code 43)

A request for the USB device descriptor failed.

I've continuity tested from the board solderpad through to the USB A connector going into my PC and cant find any issues. I'm confident the wires are only joined to their restrospective pads on the teensy.

The teeny fires up into 'Blink' when its powered through Vin, and does go into the bootloader when the button is pressed. Teensy Loader does not recognise it when it is plugged in, regardless of if the button is held whilst being plugged in.

From the above, I beleive the teensy is still alive, and I dont think I have somehow damaged it, but I also dont know why my modification hasnt worked - I appreciate USB2.0 is a higher speed protocol, but I have seen some absolute atrocities with people making FE1.1s USB hubs having no issues before, so not sure how my 20mm trace could cause complete USB protocol failure?

Any help would be greatly appreciated to resolve this mess!
 
Could it be that you accidently swapped the D+ and D- connections? I vaguely recall that I ran into a similar issue long time ago.

Paul
 
Hi Paul,

I thought this also might be the issue - especially with some lovely manufacturers not being able to stick to a common code for green/white in cables!

I have tried with them around both ways by swapping the terminals in my test loom to the PC, and get the same fault each time.

I've just verified that it is the exact same reported fault by swapping them again to be sure.

Looking at the board, it might be possible to put a breakout USB micro connection into the board before its soldered to the PCB if the two USB external connector pins were cut higher up. This might be the least destructive way going forwards if this works, but it would still have some soldered wires to the USB cable
 
I know you mentioned GND, but I would really sure you had it.

I am no EE, so I don't recognize your USB connector that you are now using jumpers to the Teensy.
There is obviously more than 4 pins coming out of them. Are any of them possibly also need to be connected and/or interfering in some way?
 
Another thought: could it be that either D+ or D- is not making it up all the way to the PC?
From your photo in message #1 I can't tell what kind of external connector you are actually using and whether that pin numbering is OK.

Paul
 
Hi Both, Sorry I wasnt able to upload a high res image as it kept failing.

The connector system I am using is the Molex MX120 family of automotive sealed connectors.

On the mating connector I have literally terminated a 1m USB cable.

I have continuity checked from the teensy USB pads (after removing the micro USB connector) to the USB-A connector connecting to my PC and cant find any issues, but I will redo this in the morning wiht a fresh mind to be certain, as its likely something silly!

I have attached 5v and GND to the connector (with the USB and board GND attached to avoid floating issues)

The remainder of the pins on the MX120 are all CAN and LIN bus, as this is the intention of the PCB - they dont need to be connected for this initial test, as I cant flash the teensy to make use of them anyway!

I'm making up another PCB (a few IC's werent available at time of assembly) and will not put the MX120 on so I can access the USB port for now, whilst I try to work out whats going wrong with the first one!
 
after some diagnosis of the PCB - I found two diodes placed backwards from the agreed SMD, which were causing power issues on the teensy board (quite rightly!) This issue is now resolved.

For future refference to all - on the Teensy 4.1 the two USB pads underneath ARE directly connected to the device USB port, contrary to other advice on this forum.

You CAN flash through the pads rather than the Micro USB port.

On another note, relating to my earlier comment on some USB cable manufacturers not sticking to convention with D+/D- colours, I have also jsut found a USB cable wiht black/red switched for +5v and GND!!!???
 
Glad to hear you found the reversed diodes!
And about that black/red-swapped USB cable: been there too.

Success with your project,
Paul
 
You must log in or register to reply here.
Back
Top