GE865/GL865 reset line internal pullup

11 thoughts on “GE865/GL865 reset line internal pullup

  1. Cosmin

     

    Can you give me the size (min./typ./max.) of the internal pullup resistor on the RESETinput line, both for GE865 and GL865. And what are the logic levels for low and high conditions, please?

    I am planning to put a capacitor here to extend the pulse width of an external watchdog circuit.

    Thanks, Tom

    1. Hi Tom,

       

      Hardware User Guides give values for high/low signal voltage levels, also values of max current that can be drained from RESET, and advise to use an open collector transistor there, not a simple pull up or totem pole configuration. How do you intend to hook the watchdog here, it won’t be much more simple and efficient to put the logic on the power lines of the module?Fresh and clean startup.

      1. Thanks Cosmin,

         

        you are right. The logic levels for RESET indeed are mentioned in the HUG. But I cannot find the value of the internal pullup resistor on the RESET pad.

         

        My idea is to use Maxim´s MAX6371 watchdog ckt (which has an open drain output) but it produces only a minimum of 100 msec pulse. So therefore I need to add a suitable capacitor between GND and RESET to reach the required minimum of 200 msec.

         

        Assuming (HUG Rev 13, page 23/76) a current of 0.15 mA from the RESET pad, one can calculate roughly the size of the cap: C = (I * dt)/U, where I = 0.15 mA, dt = 100 ms (extra needed RESET low time) and U = 1.8 V (minimum level of RESET high.

         

        This would result in 8.33 uF, rounded off to at least 10 uF, maybe 22 uF. But the value for current is inaccurate, would be better to have a value for internal pullup resistor, to be on the safe side, right?

         

        What do you think, in general, about this idea?

         

        Regards, Tom

        1. Don’t know what to say about a capacitor there, might break things with improper rise/fall edges, maybe a monostable is cleaner. I’ll ask for the resistor value.

           

          Take care:

          The  RESET#  is  internally  controlled  on  start-up  to  achieve  always  a  proper  power-on  reset
          sequence, so there’s no need to control this pin on start-up. It may only be used to reset a device
          already on that is not responding to any command. 

          1. Yes, you have a point. Rise time might be an issue. Ideally, the RESET line should have a Schmitt-trigger inside, with hysteresis. Please check if you can have more information.

             

            As to the boxed remark of yours, I do know that there is no need to control the RESET line as such, but only in case of derailed sw, on a very unmanned spot where nothing can be done from distance, if the unit stops working.

             

            To be on the absolutely safe side, the added capacitor could be isolated from the RESET line by adding two external transistors and keeping the capacitor in between them.

             

            Would be nice to hear if anyone has any practical experience here?

             

            Tom

  2. Cosmin,

     

    any further news here? Did you get some estimate about the internal pullup resistor on RESET line? And how about any rise time issues, if I add a capacitor?

     

    Have you seen/heard of any external watchdog implementations?

     

    Tom

    1. Hi Tom,

      RESET line is internal pulled up to 1,8V with a 1K resistor.

      So I think 20uF doesn’t help!

      Anyway teh rising edge of such a capacitor is very slow, I’m not sure it is a good idea.

      Many years ago I designed a "pulse extender"  but, as well as I can remember I used a circuit with a couple of diode to selsct thw different resistor to charge discharge a capacitor and a gate to sharpen the edge..

      1. Hi Luca

         

        Thanks for your reply. I agree, it seems best to keep the RESET line clean, driving it only with an open collecor setup, as your documentation suggests. I must extend the pulse somehow before applying it to the RESET line.

         

        Tom

         

          1. Hi All,

             

            monostable can work fine, no reason to don’t use it.

            I designed a differnet circuit, in attachment circuit itself and a simulation.

             

            Both can work.