Page 3 of 6 FirstFirst 123456 LastLast
Results 21 to 30 of 52

Thread: Detecting Water system problems before...

  1. #21

    Re: Detecting Water system problems before...

    To be fair, the 10 uF cap helps bring down the source impedance, so you could probably go with 10k or even 47k for the divider.

    Separating circuits for separate sensors will de-couple them a bit so the readings re more accurage, but with a strong driver (4.7k) that's maybe down in the noise.

  2. #22

    Re: Detecting Water system problems before...

    Quick update, I built the first version of the board and most things are working. I did find a problem on my board where I used the wrong data pin on the LoRa radio for the Interrupt pin. I am able to now jumper it to make that part work.

    Click image for larger version. 

Name:	TGST-Assembled.jpg 
Views:	156 
Size:	38.0 KB 
ID:	6743Click image for larger version. 

Name:	TGST-Top.jpg 
Views:	150 
Size:	67.7 KB 
ID:	6744Click image for larger version. 

Name:	TGTS1-Bot.jpg 
Views:	163 
Size:	49.0 KB 
ID:	6745

    I am thinking of maybe doing another minor rev of the board and some of the things I am probably going to add/change:
    1) The coin battery area had etch running through which could maybe short... So now all ground plane
    2) Add small 3.3v VR (MCP1700) to run the LoRa, Display, ...
    3) Connect the reset pin of the display to reset pin of Teensy, so when reset, the display is as well.

    4) Maybe instead of wiring the display backlight to IO pin instead of direct to 3.3v through resistor. Then maybe can turn off or run though PWM for brightness. My question to myself (and hopefully someone who is better at electronics), is, I assume I should connect the display ( through a transistor. Wondering what transistor to use and a simple circuit... It has been so long since I have had a circuit analyses class, that I forget on how best to use NPN vs PNP...



  3. #23

    Re: Detecting Water system problems before...

    If you want to drive a transistor for PWM or other purposes, you don't want NPN or PNP; you want an N-channel MOSFET.

    The all-around most-convenient transistor for this is the BS170, also known as the 2N7000. It can drive 200 mA or so, and switch up to 60 Volts, while still having a threshold voltage of a little over 2 Volts, so 3.3V MCUs drive it just fine.

    There are plenty of articles on the web about using an N-channel as a switch. The only thing you need to really remember is to hook the gate to the control pin, and hook the transistor to ground, and hook the switched load between +voltage and the transistor (it shuts of ground, not voltage.) You don't need a gate resistor for small N-channel transistors and most microncontroller outputs, but 100 Ohm or so won't hurt and will be "more correct."

    This seems like the first Google hit for how to do it:

  4. #24

    Re: Detecting Water system problems before...

    Thanks jwatte,

    I have looked through a few of these sites I see through google, but the problem I sometimes run into is translating them into what I might need here.

    In most of these examples, the load (example LED) is setup such that the plus pin is connected up to power and the MOSFET controls the negative lead of the power. I am not sure if I am saying that right. But in my case all I have is the positive input to the backlight leds.

    Looking at some tutorials, like at the Sparkfun one:
    I am thinking maybe I need something like a darlington setup, with two transistors?

    Another person up on PJRC has built a board that works using two transistors, that part of the schematic looks like:
    Click image for larger version. 

Name:	Backlight-circuit-for-TFT.jpg 
Views:	178 
Size:	20.4 KB 
ID:	6746
    When I asked about needing two, he mentioned for his next version he may use a FET FDV303 that Paul had mentioned someplace. He mentioned it might not require 2 transistors. Probably my simplest this is to simply copy what he has working now. But still wondering if I should try something different.

  5. #25

    Re: Detecting Water system problems before...

    in my case all I have is the positive input to the backlight leds
    Got it! In that case, you need a P-channel. And, because your LED positive is probably higher than the 3.3V from the MCU, you probably need a small N-channel to drive its gate, and a pull-up resistor. The reason for this is that, both in bipolar transistors, and MOSFETs, the control voltage is related to the voltage you drive, and trying to directly control the driving will blow your MCU output pin.

    Assuming you use BS250 for P-channel, and BS170 for N-channel, and load is < 200 mA. (If load is bigger, you need something bigger for the BS-250)

    EAGLE isn't installed on my current computer, so I can't quickly sketch it up. Basically, it's:
    Positive drive voltage -> P-channel source
    P-channel drain -> positive input to load
    Positive drive voltage -> 10 kOhm resistor -> P-channel gate -> N-channel drain
    N-channel source -> ground
    N-channel gate -> 100 Ohm resistor -> MCU control pin
    N-channel gate -> 10 kOhm resistor -> ground

    You build your own schematic from that :-)

  6. #26

    Re: Detecting Water system problems before...

    Actually, there are better transistors than BS170 and BS250 (run cooler) these days.

    N-channel (replace BS170): TN0601N3 or ZVN4210

    P-channel (replace BS250): TP0606N3

    And, if you need to switch real power, I like the SUP75P03:
    I use this for soft-on-off for all my robot projects.

    Beware, though: MOSFETs really don't like overvolting. If you use more than the rated voltages in your circuit (for example, more than 16V for something rated 20V Vgs) you have to be careful to put in fast Zeners or other circuit design to make sure it doesn't transiently get overloaded. Similarly for the Vds you control.

  7. #27

    Re: Detecting Water system problems before...

    Thanks, I will try these out. Will post later when I finish updating schematic.

    Thanks again

  8. #28

    Re: Detecting Water system problems before...

    I have been playing around.

    Does this look about right?

    Click image for larger version. 

Name:	Backlight--circuit.jpg 
Views:	174 
Size:	53.2 KB 
ID:	6747


  9. #29

    Re: Detecting Water system problems before...

    Assuming JDISP-8 is the panel light power, then that looks okay to me.
    (famous last words :-)
    Depending on how fast you want to PWM the power, you may need to reduce R2 in value ( eg 4.7k, 2.2k, 1.0k)
    The gate charge is not specified for this device so I don't know exactly how fast you need to be switching before this matters.
    The trade-off is one between switching time (which means heating of the transistor) versus losses through the pull-up resistor when on.

  10. #30

    Re: Detecting Water system problems before...

    Thanks again, will play around and see how well it works.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. Detecting surface conditions?
    By Paul V in forum Sensors
    Replies: 9
    Last Post: 02-25-2014, 01:33 PM
  2. Contest Entry Mini Water Cannon
    By tbh726 in forum Project Showcase
    Replies: 9
    Last Post: 11-15-2011, 04:26 PM
  3. Detecting motion in an AX-12+ servo
    By Pi Robot in forum DYNAMIXEL & Robot Actuators
    Replies: 7
    Last Post: 07-08-2010, 10:45 PM
  4. Water Gun
    By shobley in forum Mechanics / Construction
    Replies: 27
    Last Post: 06-24-2009, 01:09 PM
  5. Question(s) Water Gun Sleep Alarm... Need Help
    By polo2014 in forum Robotics General Discussion
    Replies: 29
    Last Post: 05-30-2008, 09:17 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts