I am using a standard NPN transistor-as-a-switch set-up to use a Teensy 4.0 to switch a bunch of shape memory valves (valves controlled by a wire that changes shape when heated by a current). I have previously got some advice here on the use of a buck converter set-up to run this, but this did not work out well so I decided to go this simpler route and accept the ~0.35W power waste in the 8.2 Ohm resistor (2W type).


simulate this circuit – Schematic created using CircuitLab

When I connected the "Teensy DIO" point directly to the 3.3V DC power source, the transistor switches fine, giving the desired .25A current required to the SMV. When I instead set the Teensy DIO to High as in the circuit, I only get a 0.05A current, and the voltage at the Teensy channel is only .8V, with a current flow from the Teensy at 0.5mA. when disconnected from the circuit the channel goes to 3.3V when set to high, so the Teensy seems fine. Does anyone know why this set-up does not work? Thanks in advance!


1 Answer 1


(3.3V-0.6(=Vbe))/1k = 2.7mA
Output = 50mA Ic/Ib= 18.5 which is expected as hFE reduces near 10% at saturation but is typ. rated at hFE=10.

Design for worst-case Ic/Ib=10 thus Ib = 25mA, Rb ~ 100~120 Ohms


5.5V logic is typically 50 OHms +25% here operating at 3.3V, ESR or RdsOn may be on the higher tolerance side. PIO = 25 mA Abs. max so if tolerances are slim, reduce to 22mA (3.3-0.65V)/22mA = use 120 only

The RdsOn is standard for all 5.5V max logic but given as ESR=60 max on p35 of 49 but uncertain if that is for 5V so even if +50% or 75 Ohms is used, 25mA generates only 50mW.

for a given Vdd, RdsOn= Vol/Iol , (Vdd-Voh)/Ioh

2nd update

On 2nd thought you can use a 2nd bigger transistor (to reduce T) in a Darlington configuration and reduce the waste power of the 8.2 to allow 0.7V extra drop of the Darlington 240mA to 5.6 Ohms with two 10 to 12 Ohm R’s in parallel as it dumps over 300mWto 350mW.

  • \$\begingroup\$ Ib 25mA might be borderline for a T4. Maybe a darlington configuration might be better? \$\endgroup\$
    – luni64
    Jan 18, 2021 at 8:35
  • \$\begingroup\$ Ib= 50mA is the rating for Vce(sat) =0.7max @ 500mA Borderlne? explain your thinking. (D'ton=worse with only 3.3V \$\endgroup\$ Jan 18, 2021 at 11:49
  • \$\begingroup\$ Sorry, my comment was unclear indeed. I might be wrong, but Ib needs to be supplied by the Teensy pin. The IMXRT datasheet is very vague, but sourcing/sinking 25mA seems to be on the high side. \$\endgroup\$
    – luni64
    Jan 18, 2021 at 12:13
  • \$\begingroup\$ What is the (Vdd-Voh)/Ioh=RdsOn and Pd of CMOS driver? 25mA is rated for =Vdd/2 logic levels, but here only needs to be >>25mA * RdsOn \$\endgroup\$ Jan 18, 2021 at 12:44
  • 1
    \$\begingroup\$ Thanks Tony, the Darlington configuration works! I am still a bit at a loss why the original solution does not work, since the current from the teensy is well below the limit. \$\endgroup\$
    – Priegel
    Jan 18, 2021 at 20:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.