2
\$\begingroup\$

I have a BLE module. The reset is pulled up to 1.8V. The IOVDD is 3.3V. I need to use a programmer that can reset the module but the programmer is 3.3V. Is it ok to use the circuit below to make the programmer reset the module? (The MOSFET has Rds 0.15@Vgs=2.5V, Vgs(th)=0.85V)

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ Write down your expectation and simulate. \$\endgroup\$
    – greybeard
    Jun 7 at 9:20
  • \$\begingroup\$ Looks fine to me. Remember to choose a MOSFET so that Vgs=-1.8V is enough to turn it on (as much as you need it to be turned on). The Vgs when NRST goes low is -1.8V and not -3.3V. \$\endgroup\$
    – user253751
    Jun 7 at 9:23
  • 1
    \$\begingroup\$ When NRST=0V, the PMOS will pull down the voltage close to VTH i.e., 0.85V. It will not go to 0V. To get the exact voltage, you need to simulate the condition. With this voltage, you then need to check if the BLE module can be reset properly. I would not take this risk, I would use an NMOS instead and invert the gate control. \$\endgroup\$
    – sai
    Jun 7 at 9:51
  • \$\begingroup\$ @sai , good catch, reset VIL is 0.3*DVDD (0.3*1.8=0.54 max). Could you provide the NMOS alternative? \$\endgroup\$
    – alsaleem
    Jun 7 at 10:57

2 Answers 2

Reset to default
2
\$\begingroup\$

When NRST=0V, the PMOS will pull down the voltage close to VTH i.e., 0.85V. It will not go to 0V. To get the exact voltage, you need to simulate the condition. With this voltage, you then need to check if the BLE module can be reset properly. Since you confirmed that VILmax is 0.54V max, it won't work as you rightly concluded.

Edit: Here's a possible solution using NMOS. However, do look into the diode solution proposed by Spehro; that seems to be much easier to implement.

enter image description here

\$\endgroup\$
7
  • \$\begingroup\$ Do you recommend to use open-drain logic gate, instead? \$\endgroup\$
    – alsaleem
    Jun 7 at 12:07
  • 1
    \$\begingroup\$ I updated my answer showing the NMOS option. \$\endgroup\$
    – sai
    Jun 7 at 15:38
  • \$\begingroup\$ 74LVC1G07 will do just that. Thanks for suggestion. \$\endgroup\$
    – alsaleem
    Jun 7 at 19:55
  • \$\begingroup\$ Yes, sure, you can use that. BTW, is NRST in your diagram an open-drain output? \$\endgroup\$
    – sai
    Jun 8 at 1:26
  • 1
    \$\begingroup\$ Then, there is a possibility to connect that to the BLE's reset without any circuitry in-between. Probably the only point to check would be whether the J-Link loads the reset pin when it's supply is absent. \$\endgroup\$
    – sai
    Jun 16 at 1:32
3
\$\begingroup\$

You'd be better off to use a Schottky diode. For example, at 1mA a 1N5819 would pull the /RESET input down to the NRST voltage + less than 200mV.

The p-channel MOSFET will pull it down to something like Vgs(th). Vgs(th) is typically specified at a very low current such as 250uA so if R? draws more than 250uA it won't even get that low.

Incidentally, the classic solution (which potentially has less voltage drop than a diode) is to use a single "1.8V" n-channel MOSFET with gate tied to +1.8, source to the module and drain to the 0/3.3V signal.

\$\endgroup\$

Your Answer

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

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