Can't program Teensy 4.0 after Vin trace is cut and supplyng external 3.3v

Status
Not open for further replies.

ilium007

Well-known member
Hi - I have run into an issue whereby I can't program the T4 via USB after the Vin trace was cut and I supply 3.3v to the 3.3v pin on the board. The serial device does not show up on OSX in Arduino IDE. If I supply external 5.0v to the 5v pin the serial device appears and I can again program via USB.

Once programmed the T4 runs fine on 3.3v. Trying to work out now if I need to supply 5.0v in my project to power the T4 when I have designed everything else on 3.3v.
 
It works as far as can be seen here.

I have a T_4.0 beta board I cut the VIN<>VUSB trace on.

Powered a second T_4 from USB while GND and 3.3V were wired to the above Beta - GND is common and the 3.3V pins is the one on the board end opposite USB - either 3.3V should work.

Upload with TeensyLoader by button or from IDE both work.

GND and 3.3V power must be good for it to work with no USB once programmed?

Is the TeensyDuino 1.51 or 1.52 in use and Teensy loader used for upload?

Does it work with button press?
 
I tried again using a lab power supply at 3.3v on the 3.3v pin (circled) but haven’t tried the 3.3v on the other end of the board.

I noticed that plugging/unplugging the USB cable when powered from my 3.3v will intermittently crash the T4.0 running the default flashing led program.

0B808AC4-BE44-49D6-9C9D-3AD0C12A2D8D.jpg
 
Very odd. There was some half question on this during T_4 beta - but when tested all looked well.

If the T_4 is unplugged - do you see a clean break in the cut VIN<>VUSB with a meter?
 
Just tried on a production T_4 with trace cut on the LogLow Breakout. With switch power set to OFF and powered from 3.3V side pin ( going to be common routed ) - it worked as well.
 
I'm having the same 'problem' with my Teensy 4.0, though I don't think it's a true problem, just a feature. The note on the back of the Teensy 4.0 'Welcome Card' regarding the jumper trace between the VUSB and the VIN pin says, "Cut to separate VIN from VUSB, if using battery or external power." I think this would be more properly worded, "...battery or external power for VIN."

I too am powering my board with 3.3V directly (using both 3.3V pins actually). When I do that alone I cannot program the device. And even after the device is programmed, if I want to use the Serial(0) port for diagnostic data, it is unreliable - working intermittently.

To program the device or use the Serial(0) port I have to either 1) leave the jumper trace intact, or 2) connect VIN to 5V. For my application, since I won't always be using Serial(0) for diagnostics, and I won't be programming that often, I'm considering reconnecting the VIN/VUSB jumper with a blob of solder. This works, but I'm not sure if there are considerations for the long term.

The long-term question for me is about the TLV75733P regulator. What happens if the 3.3V I provide isn't exactly the 3.3V the regulator outputs when I have the USB cable connected? For those times will the regulator be 'unhappy'? TI has some seemingly applicable App Notes about reverse current and the like, but they're over my head.

Is there any reason it would be better to re-cut the jumper trace and use the VIN pin instead? Or, what about removing the 3.3V regulator completely?

Thanks!
 
IMG_1356small.jpg

I obviously haven't followed the instructions correctly. My computer says the USB port is drawing too much power if I put in the battery in the orientation shown in the photo. Everything runs fine with the battery out. I think I cut the trace successfully. With the battery out, I can program and power it, so it seems like the diode at VUSB-VIN must be correct. The one at VBat sure looks like what I see in the (generations old) photo at https://www.pjrc.com/teensy/external_power.html

Do I have to turn the diode on VBat around? If I know I'm going to use a 3.7V battery with a diode at VBat is there any advantage to cutting VUSB from VIN?
 
View attachment 19713

I obviously haven't followed the instructions correctly. My computer says the USB port is drawing too much power if I put in the battery in the orientation shown in the photo. Everything runs fine with the battery out. I think I cut the trace successfully. With the battery out, I can program and power it, so it seems like the diode at VUSB-VIN must be correct. The one at VBat sure looks like what I see in the (generations old) photo at https://www.pjrc.com/teensy/external_power.html

Do I have to turn the diode on VBat around? If I know I'm going to use a 3.7V battery with a diode at VBat is there any advantage to cutting VUSB from VIN?

THE VBat pin is for 3V ( THREE VOLTS ) only to power the RTC hardware! As noted there on the Card - that is intended to connect a '3V coin cell' for RTC running in the power off state. Nothing over 3V should be connected to that VBat pin.
 
Teensy 4.0 is meant to be powered from VUSB or VIN, but not directly from the 3.3V pin.

3.3V can be used without VIN, but only in certain ways. Here's 2 things to try:

1: Apply 3.3V power first to VBAT (a coin cell works great), so the RTC & power management hardware is powered first. Then apply 3.3V to the 3.3V power pin.

2: Use a physical switch or power supply with very fast startup. If your power supply as a "soft start" circuit or a large capacitor, the 3.3V power can rise up too slowly. If the rest of the chip sees power before the RTC & power management, the chip can get "stuck" and not boot up properly.

Again, the normal intended way to power Teensy 4.0 is from the VIN or VUSB pins. When done that way, the chip's power management controls the 3.3V regulator enable pin. That way guarantees proper startup.
 
That does clarify :: This came up the other day ... and it tested it to work ...

when I tested to work with 3.3V direct ( from another Teensy ) on TWO ( Beta @PJRC and @loglow breakout ) T_4's - they BOTH happened to have a 3V cell on RTC so that is why they worked!

Back in Beta I saw it not work once … then it did before I could repro to make a note about it.
 
....the normal intended way to power Teensy 4.0 is from the VIN or VUSB pins. When done that way, the chip's power management controls the 3.3V regulator enable pin. That way guarantees proper startup.

Thank you, Paul. As jrraines said, "That makes it clear."

I didn't find anything in the existing documentation that mentions it. Maybe I missed it. If not, I hope it will be added to the web site to prevent this hiccup for new users.

Take care.

PS: I'm a big Teensy 4.0 fan!!!
 
Do I have to cut the VIN<>VUSB pad if I power it without the usb cable plugged in under normal operating conditions, and when I plug in the usb cable to update code, I remove the external power supply connection? Or does the VIN<>VUSB pad need to be cut no matter what if I am using an external power supply?
 
Do I have to cut the VIN<>VUSB pad if I power it without the usb cable plugged in under normal operating conditions, and when I plug in the usb cable to update code, I remove the external power supply connection? Or does the VIN<>VUSB pad need to be cut no matter what if I am using an external power supply?

No, you don't need to cut VIN<->VUSB in that case. If you don't cut the trace and you connect both the USB and external power, it will try to feed back power to one of your power sources (whichever one has the smallest voltage). I've seen some people recommend a pair of diodes to prevent the reverse power situation.

I tend to not always cut the connection myself, but there is always the time you forget to remove the external power source when you connect the USB. It depends on both the USB connection and external power source whether some magic smoke gets emitted.
 
Status
Not open for further replies.
Back
Top