Teensy 4.0 seen as SE BLANK RT Family

Status
Not open for further replies.

xdog12

Active member
I have a teensy 4.0 that I started uploading example codes to. I was having issues receiving serial communication so I kept trying different example codes. Then I noticed that it did not show as a teensy device in windows device manager. Now I cannot enter manual program mode or see the teensy on Arduino IDE.

I have tried a different computer that did not see teensy 4.0 at all.
I have ~3.3V output from teensy
I have 4.98V from power regulator to Vin
I tried holding the reset button for 15 seconds but the Teensy only flashes red once
Currently windows sees it as SE BLANK RT Family
 
For the 15 sec Restore the Red Flash is when the Button should be released to complete the restore. Did that happen then followed by about a minute of Red LED on? If held seconds too long after the flash the Restore is aborted.

That "SE BLANK RT Family" appears when the Teensy doesn't start with good USB.

Perhaps unplug the T_4.0 and then hold the Button during plugging it in, and a half second after, before releasing.

If that doesn't work then try the 15 sec Restore again with Button release when the RED LED flashes.
 
To clarify, plugging in the teensy should be followed by powering on the teensy? I have a separate power supply so I want to make sure that i have your instructions correct.

After I release the button for the 15 second reset, the LED flashes once and then the LED never turns back on.
 
To clarify, plugging in the teensy should be followed by powering on the teensy? I have a separate power supply so I want to make sure that i have your instructions correct.

After I release the button for the 15 second reset, the LED flashes once and then the LED never turns back on.

If externally powered - hold the button until just after the Teensy gets power to make sure any code on the Teensy does not run, but stays in bootloader.

If the timing is right for the 15 sec Restore LED flash and Button release, it should then go full RED for the duration of the FLASH Formatting and device Restore and then return to 1 sec Blink with NO USB when 'Factory code' is Restored.
> if that doesn't happen then something isn't working right.

Is the VIN<>VUSB trace cut on this Teensy?

If running under external power that is 'up to spec' then USB need not be plugged in for the Restore to work.
 
I have just tried your suggestions and the teensy will only flash red once. I can not get it to boot into blink. When I plugged the USB I cannot enter programming mode.

I checked that the trace is cut and has no connectivity.

I still have 3.3v from the teensy
 
Can a clear pic of Top and Bottom be shown to show soldering and attachments made?

Perhaps if VIN were Reconnected to VUSB and the external power were not applied it would show something different under just USB power?
 
After I release the button for the 15 second reset, the LED flashes once and then the LED never turns back on.

Just to confirm, when you see that brief flash on the red LED, you're supposed to release the button. If you do release within a few seconds of that brief red blink, the red LED is supposed to turn on bright for about 30 seconds. You're not seeing that? It's just remaining off?
 
Yes, one red blink and then nothing.

That's definitely damaged hardware!

Look for any damage or stray solder in the area near pins 22 & 23. Especially check the tiny pins on the chip facing pin 23. Signals shorted together in that part of the board would be the most likely explanation for the problem you're seeing (red LED does quick blink, but flash wipe with red LED on 30 seconds doesn't happen).
 
I have nothing on pin 22 but a pwm out to headers for servo motor on pin 23. I checked the board and did not find any stray conductive materials.

Are you saying that this board is probably dead? Should I bother with desoldering the 5V and pwm outputs. I already accidentally fried a board due to exposed wires. This was the backup board that I'm slowly adding the components to.
 
Are you saying that this board is probably dead?

I'm saying it's definitely damaged.

The word "dead" normally means so badly damaged that it shows no response, or responds in a way that suggests repair is impossible. That's not the case here.

The good news is we do have signs of life from all the parts. The main processor seems to be working, since it's able to appear as NXP's USB ID. The bootloader appears to be running, since it does detect you're pressed the button and blink the red LED after 15 seconds. With those results, it's very likely all the power supply and clock stuff is good.

Since we know both chips are still up and running, the most likely explanation is something went wrong with the signals which connect between the 2 chips. Admittedly this does involve some guesswork from only the words written on this forum thread, but I have a pretty strong hunch this Teensy 4.0 might be salvageable if you can find whatever caused the 2 still-working chips to stop communicating.
 
Can a clear pic of Top and Bottom be shown to show soldering and attachments made?

Perhaps if VIN were Reconnected to VUSB and the external power were not applied it would show something different under just USB power?


I took pictures of the top and bottom of the circuit https://imgur.com/a/vePtQUx

I disconnected everything but the I2C connections and I soldered VIN to VUSB. Still nothing but a single red blink.

For reference here is me forgetting rule 1 and frying the first teensy 4 https://www.reddit.com/r/arduino/comments/nsi7rm/i_learned_a_valuable_lesson_about_exposed_wires/

So the circuit was working until I made that mistake. What should I do now? Should I just get another board?

Thank you for your quick responses.
 
Maybe remove that yellow wire and look for any stray solder.

wire.jpg
 
That's definitely damaged hardware!

Look for any damage or stray solder in the area near pins 22 & 23. Especially check the tiny pins on the chip facing pin 23. Signals shorted together in that part of the board would be the most likely explanation for the problem you're seeing (red LED does quick blink, but flash wipe with red LED on 30 seconds doesn't happen).



Thank you for all of the help, I realized what happened. I didn't think that the PWM input wire from my electronic speed controller would send 5 volts as it is PWM input. So pin 23 is fried and I have to invest in proper level shifting for PWM going to the teensy AND PWM leaving teensy.

With that lesson learned, can you take a look at my schematic https://imgur.com/a/OPM3Yi6 and let me know if this level shifter will work? SN74LVC4245A. I have 5 of 6 channels from my receiver 5V and 3 PWM outputs for motors. So will this SN74LVC4245A level shifter work or should I pick a different level shifter.
 
I believe SN74LVC4245A should work. But the connection in your schematic seems strange to me.

Looks like it's meant to have Teensy drive the B inputs with 3.3V signals and the A side act as 5V outputs.

On page 10, the datasheet says the DIR input needs to be low for the B to A direction, but on your schematic I see it's connected through a 10K resistor to 5V.

The OE' pin also needs to be low, as high gives "isolation" according to page 10. I see a 10K resistor connected to 5V.


I also see what looks like 2 DCDC switching power supply modules. You should really add capacitors at the input and output of both. That Traco part in particular caused a lot of problems years ago. The words on Adafruit's web page about a 10uF capacitor came from my suggestion, as the Traco datasheet gives very wrong advice. It absolutely does require a 10uF capacitor located close to its VIN+ & GND pins. Other people have been burned by that part. It works fine if you add the capacitor, and works fine if the power supply has a low impedance, but will give you nothing but grief if used without a capacitor when you have a battery which changes impedance as it discharges.
 
Thank you for the information, I decided not to use those chips and instead I used a MC14504B for 5v -> 3v and txs0104E for 3v->5v. However I have to figure out why most of the outputs of the MC14504B are not correct. (between 1000 and 2000 PWM signal). I used .1 uf capacitors on both power lines but I did not use any resistors. Let me know if you see any red flags on the level shifters that I decided to use, I also added the 10uf capacitor.

My real question right now is when I try to run any I2C codes the teensy gives me this error. Board at usb:0/140000/0/9 is not available. I used the example code for Arduino_LSM6DSOX and I could not get serial from the teensy 4. I then used a simple blink example code with serial and slowly added I2C to that. And I once again got the same error and I have to manually press the button on the teensy to upload different code. What do you think is causing the teensy to behave badly once I call i2C.
 
I updated the software and now the teensy is communicating via serial that it cannot find the i2c device. I replaced the broken teensy 4 (i2c worked) with the new teensy 4 and now i2c is not working. I verified the connections between the devices and I am slowly desoldering everything that is attached. I also tried adding a delay. What should I try next.
 
Maybe it would help to see new pictures of what is hooked up.

From your first pictures, it also looked like maybe you you connected +5v to pin 2
screenshot.jpg
screenshot2.jpg
But I can not tell if that connection was cut or not?
 
I tried using SDA1/SCL1 using the i2c scanner and that also did not detect any devices.

What would be your recommendation on connecting the teensy to solderless breadboard? I2C was working beforehand so I didn't think a new board would disrupt communication. Here is a picture of the board before I started desoldering everything again.

https://imgur.com/a/SGxkppH
 
What would be your recommendation on connecting the teensy to solderless breadboard?

You're dealing with so many unknowns, so many ways things can go wrong. So far at least damaged Teensy, maybe due to soldering, maybe from 5V power touching the wrong place, maybe something else. Use of known-problematic Traco switching power supplies, when used without additional capacitors. Now I2C devices not responding. But apparently no longer connected to the primary I2C port as in the prior photos and schematic. Why, who knows?

I see in your latest photo you've also trimmed away a substantial part of the Teensy 4.0 PCB material. That's quite risky, as it's a 6 layer PCB. There are 4 inner layers you can't see, where 2 of them are GND and 3.3V and the other 2 have lots of signal wires. That may or may not have an impact. The point is it's adding a lot more uncertainty!

The point of building a physically large solderless breadboard prototype is you can reduce the number of unknowns. You can check whether the hardware all works as designed, without the extra uncertainty of construction problems from soldering and aggressive modifications like carving away part of the PCB.

You can also get known-good software tested on the breadboard. For example, a month or two ago we had someone else report I2C wasn't working on Wire1. Turned out (after many messages begging for the full code to be shown) the problem was the Scanner example had not been properly changed, so it was using Wire1.begin() but then trying to actually use Wire somewhere else in the code. That may be your problem here. Or it could be those wires aren't connected, or 1 or both are shored to each other or something else, or maybe the hardware has been damaged some way during this construction. It's extremely difficult to follow where those wires go in the 1 photo of your latest attempt, and whether the Scanner program is correct is also blind guessing. With a huge number of unknowns, it's very hard to diagnose what's causing a project to not work. For those of us trying to help you over the internet, the harder it is for us to see what you've done, the less effectively we can try to help.

With a solderless breadboard, you could quickly & easily remove the wires connecting other I2C chips and check again until it works. If it's actually a software problem, you can quickly get to a test with only 1 brand new part connected and then focus your attention on the software side. Debugging is so much faster & easier with a solderless breadboard. Photos from a few different angles can also let us see how things are really connected and try to help you more.

As a general approach to experimental & prototype electronics building, getting each piece working and building up incrementally upon each prior success is a much better path. When something goes wrong, if you've added only a small amount, figuring out what has gone wrong is usually much easier.


Of course, if everything worked perfectly on the first try, there'd be no need to troubleshoot, no need to whittle down the huge number of unknowns. But that's clearly not the way things are going. So my recommendation it to consider changing your overall approach, because clearly this way of building is not giving you a path to success. Using a solderless breadboard as a first step may seem unnecessary, and if this were working well it would be. But unless you luck suddenly changes, I really believe you should consider using a different approach that lets you test as much as possible as you go and leverage each small success to chart a more reliable path toward ultimately making a successful project.
 
Maybe you misunderstood my question, I asked What would be your recommendation on connecting the teensy to solderless breadboard? The teensy has no headers to attach it to a breadboard. I assume you want me to attach headers to the teensy. Attaching headers via soldering sounds the same as using a Adafruit Perma-Proto board. I understand that they are not the same, but what makes a breadboard which I would have to solder headers better? None of the components that I need to test will fit on the final version with headers attached. So if everything was working properly I would have to desolder the headers. So I would rather solder to the Adafruit Perma-Proto board instead.

Can you explain where you see damage on the teensy 4? I think you are misinterpreting the picture and the damaged Adafruit Perma-Proto board.


The reason I connected everything together is because the teensy was working with I2C before I gave 5v to pin 23. So I removed that teensy and soldered a new one. While soldering that board I soldered the level shifters and the capacitors. Currently only the capacitor that you recommended is connected to the i2c devices and the teensy. I desoldered everything else.
 
Status
Not open for further replies.
Back
Top