Hall effect end stops

  • I'd like to get a sanity check from the smart folks around here…

    I currently have a Rostock Max V2 running the stock RAMBO board, which is 5V compliant. The Duet WiFi needs 3.3V I/O. Stating the blatantly obvious input parameters. 🙂

    I have retrofitted the end stops at the tops of the towers to use an Allegro 3144 unipolar hall effect switch. Then I printed little magnet holders that ride on the cheapskate carriages and come into proximity of the sensors at the top of the travel. My current 5V setup has these just wired directly to the RAMBO's +5V, signal and ground pins for the X,Y,Z end stops. I know I should probably have a buffer/pull-up resistor on the output and put a filter cap on the power right next to the little sensor package, but it seems to be working really well without any of that. I do have the wires twisted pretty tightly so hopefully I'm getting some noise reduction by doing that.

    My question is how to adapt these sensors (or ones like them) to the 3.3V setup on the Duet WiFi? I can't seem to find any sensors that will work on a 3.3V supply, so I'm stuck with 5V or higher Vcc voltage for the sensor chips.

    I'm thinking of doing it like this: Make a little breadboard for each of the sensors that lives up at the top of each tower. It will be supplied by 5V from somewhere on the Duet WiFi. By the way, where's the best place to get this +5V on the board and how careful do I have to be with current sinking/sourcing levels?

    Also on that breadboard will be small cap between 5V and ground, along with either a pull-up or maybe a buffer on the output pin. So 3 wires coming out of each little breadboard: +5V, Signal (0V to+5V digital), and Ground. The 3 bundles of 3 twisted wires coming from the X,Y,Z towers will terminate on another breadboard near the Duet WiFi. This one will have a level shifter IC on it (like the 74LVC245). Coming from the Duet WiFi, I'll have +5V, +3.3V, GND and the three input pins on the X,Y,Z end stop headers. The level shifter IC will knock the +5V signal levels down to +3.3V to safely interface with the Cortex IO pins. I will also distribute the +5V and Ground to the three sensor cable bundles.

    Am I missing anything? Or making this harder than it should be?

    Thanks in advance.

  • administrators

    Hi Toddimus,

    There are plenty of 3.3V-compatible Hall effect sensors - I have a strip of them on my shelf - but the A3144 isn't one of them. However, if you want to go on using the A3144's then it should be straightforward. You can power them from 5V (which you can get from pin 1 of the expansion connector) and ground (from the endstop connector), and connect the open-collector output directly to the endstop input. No pullup resistor needed because the Duet already provides that. You should add a 0.1uF decoupling capacitor across the A3944 Vcc and ground pins.

  • Thanks dc42.

    Looks like the Allegro A1121 or A1122 chip would do the trick. It has 3V minimum supply voltage and is open drain output. So with that, I should be able to directly wire it to the Duet end stop pins (getting the pin order correct of course) and then I'll add in a 0.1uF cap across the Vcc and ground right at the sensor. That's much simpler and doesn't require adding 5V into the mix.
    Given that it's an open drain rather than an open collector, am I still ok just hooking the output pin direct to the Duet's input pin? I assume so.

  • administrators

    Check that the open drain output can sink enough current. From memory, I think you need about 2.5mA for the Duet WiFi.

  • From the datasheet: …"Each device includes on a single silicon chip a voltage regulator, Hall-voltage generator, small-signal amplifier, chopper stabilization, Schmitt trigger, and a short-circuit protected open-drain output to sink up to 25 mA...."

    So sounds like I am good. Unless you are an order of magnitude off in your memory of the datasheet. 😛

    Thanks David! Really appreciate your inputs on this and your work in general on the project!

    I paid the invoice a couple of days ago. Haven't seen anything from Filastruder on shipment yet. Somewhere I think I saw something from T3P3 that the boards would go out to Filastruder today (presumably leaving England for the US). I won't have time in the next 10 days or so to play with it anyways, so it isn't a huge rush for me.


  • administrators

    A1121/A1122 sounds good then. The reason I asked about the sink current is that many of the newer 3.3V Hall sensors are micropower devices only capable of sinking 1 or 2 mA. We reduced the sink current requirement on the Duet WiFi compared to the Duet 0.8.5, at the cost of making the endstop LEDs less bright.

    I've just measured the endstop input sink current requirement. It's 1.5mA for the Duet WiFi pre-productIon boards, and 2.1mA for the Duet 0.8.5.

  • Are there any real life practical adventages of using hall endstops in delta printer?
    I have some endstops from RRD, they are easy to connect, but I'm not sure if they are really worth it.
    Is there any difference in calibration precision and repeatability that could justify spending extra buck?

  • administrators

    I use ordinary microswitches on my delta (the ones T3P3 supplied with a Mini Kossel kit a couple of years ago), and they are reproducible to 1 microstep. So my view is that with good microswitches, you don't need anything better. In any case, if you have a good Z probe then auto delta calibration will calibrate out endstop error. I even tried running my delta without endstop switches once to prove this.

  • Have you seen these Hall Endstops http://www.reprap.me/hall-e-endstop.html ?

    I use them on 2 machines and they work perfectly.

    Forgot you can actually just about make them out in the pictures of my latest delta on this forum https://www.duet3d.com/forum/thread.php?id=357

  • @ mjhall29 they are the same as the reprap.me ones just more expensive. They are available on aliexpress as well for alot less.

  • You can also use reed sensor like this one :

    I use one those one 2 printer

  • Thanks for the info and links. I was just curious if there is a real advantage in homing precision because I've read about hall being more precise because of no mechanical parts. Definitely one microstep is enough precision, of course if it's a 0.9 deg microstep 😉

  • administrators

    I have found mechanical switched to be repeatable enough for Homing delta axis. I tested the trigger point vs distance over 10 iterations with no difference. I was moving the axis quite slowly because all.my homing files move the axis quickly until the switch is triggered, then back off a bit and fine home slowly (F300).

  • For me its not really the precision that matters as much as the reliability, being non-contact there is nothing to wear out so reliability increases.

  • administrators


    For me its not really the precision that matters as much as the reliability, being non-contact there is nothing to wear out so reliability increases.

    Bear in mind that those sensors with a pot are analog devices, therefore probably sensitive to supply voltage and temperature. Also if the magnets lose their strength, the calibration will change. Finally, depending on how close the magnet goes to the sensor and how you have adjusted the pot, it's possible that the Earth's magnetic field may affect them slightly, which would matter if you turn your printer round. So don't assume that they are automatically more repeatable and reliable than microswitches.

  • Thanks dc, pots are adjusted to min sensitivity setting. I use a larger magnet than the tiny ones supplied. All cables i use are lapp LiCy. Cant do much about the planet i am on but they seem to be ok 🙂

  • @dc42:

    There are plenty of 3.3V-compatible Hall effect sensors - I have a strip of them on my shelf - but the A3144 isn't one of them.

    I'm using a Duet Wifi I got a few months ago and am looking for hall effect end stops. Was looking for 3.3V and noticed a lot of them are 5V. I'm reading this: https://duet3d.com/wiki/Connecting_endstop_switches and it says "The endstop switches each need to be able to sink 1.5mA current for the Duet WiFi" and I don't want to buy the wrong ones.

    Is there a recommended quality hall effect sensor that people use? I'm looking for specific ones that are 3-pin and are known to work with the board without issues. Any recommendations would be helpful.

    edit: Would these work: https://www.digikey.com/product-detail/en/AH1815-P-B/AH1815-P-BDI-ND/5453089

Log in to reply