I am designing a device for some laboratory work where I need to apply different temperatures to dozens of different samples. I am planning to use WS2811s (1 wire LED drivers) to simplify the wiring. The idea is to use one WS2811 (which sinks up to 18mA of current per channel) per heater, to switch on a PNP and run approximately 100mA of current through R3 (220 Ohms). This resistor is used as a heating element. (See Diagram). (R1 and C1 are just recommended by the WS2811 data sheet).

The WS2811 is designed to be used at maximum of 12V - but I’m assuming that this is for direct connection to LEDs which are only dropping 3V. I would like to use a 24V supply for the heating - as I can more easily find a cheap switching power supply that can handle the 3 or 4 Amps I will need for all the heaters. If I put a current limiting resistor at the base of the PNP, so that there can never be more than 10mA sinking into the WS2811 - I think I should be able to still switch on the PNP and drive approximately 100mA through R2.

Does this seem OK? I only have a few through hole WS2811s to breadboard so wanted to check before prototyping it !

 

This is the first time I’ve come across the WS2811 but I’m wondering why you need it for this application? Also, I think your transistor should be an NPN rather than PNP with the resistor on the 24V side of the transistor. A bipolar transistor is fine, or you could maybe use a logic level MOSFET.
How about using a surplus laptop power supply, typically around 19V and 3A or more? And you could turn channels on and off with a processor like an Arduino. The UNO has 13 outputs but you can drive more by using serial to parallel ICs. In this case, the speed should be fast enough to allow you to do a form of PWM to adjust the average current in each resistor

 

Thanks for replying ! I am using WS2811 as it allows me control the 64 odd heaters with a single logic wire. The WS2811 sinks current into the LED channels (even though they are called "OUT") - so that is why I went with a PNP. (The transistor is the wrong way round in my diagram but I think the idea is clear). Yes the reason I wanted to go with more than 12V is because I have lots of 18-24V relatively high power supplies from old laptops.

 

Welcome to AAC.

The Dallas Semiconductor 1-Wire chips are widely used in certain applications but are pretty “old school” at this point. It certainly doesn’t make them a them a bad choice, though. You might check out the DS2413 which will sink 20mA and handles up to 27V.

 

I think using the WS2811 is a brilliant idea as it gives you a as many PWM channels as you want with minimal wiring. Each WS2811 gives you three PWM channels (RGB).

You can use a 24V supply for the heaters by using the WS2811 to switch a logic inverter pulled up to any voltage up to 12V then using that to drive an NPN low side to switch to the heater at 24V.

 

I am designing a device for some laboratory work where I need to apply different temperatures to dozens of different samples. I am planning to use WS2811s (1 wire LED drivers) to simplify the wiring. The idea is to use one WS2811 (which sinks up to 18mA of current per channel) per heater, to switch on a PNP and run approximately 100mA of current through R3 (220 Ohms). This resistor is used as a heating element. (See Diagram). (R1 and C1 are just recommended by the WS2811 data sheet).

The WS2811 is designed to be used at maximum of 12V - but I’m assuming that this is for direct connection to LEDs which are only dropping 3V. I would like to use a 24V supply for the heating - as I can more easily find a cheap switching power supply that can handle the 3 or 4 Amps I will need for all the heaters. If I put a current limiting resistor at the base of the PNP, so that there can never be more than 10mA sinking into the WS2811 - I think I should be able to still switch on the PNP and drive approximately 100mA through R2.

Does this seem OK? I only have a few through hole WS2811s to breadboard so wanted to check before prototyping it !

The WS2811 can sink up to 18mA per channel, so limiting the base current to 10mA with a resistor is a good idea. For your setup, consider using the 2N2907A, PNP (suitable for high-side switching, handles upto 600 mA, 60 V) or as already suggested, if you plan to use NPN, then PN2222ATF (perfect for low-side switching, handle 600 mA and 40 V) is a good choice.

 

Welcome to AAC.

The Dallas Semiconductor 1-Wire chips are widely used in certain applications but are pretty “old school” at this point. It certainly doesn’t make them a them a bad choice, though. You might check out the DS2413 which will sink 20mA and handles up to 27V.

Thanks for this pointer. This could be a nice solution for me. I had my maths slightly wrong in my first post and with a 24V supply I only need to sink 10mA - not 100 ! So in fact I could use this without a transistor. I like the fact that this chip uses parasitic power and also that it uses the official 1 wire protocol. (The WS2811 is proprietary). I will be using it in conjuntion with a 1 wire temp sensor and this will simplify code also. Thanks for the tip. Do you know by any chance if there are more recent versions?

 

Thanks for this pointer. This could be a nice solution for me. I had my maths slightly wrong in my first post and with a 24V supply I only need to sink 10mA - not 100 ! So in fact I could use this without a transistor. I like the fact that this chip uses parasitic power and also that it uses the official 1 wire protocol. (The WS2811 is proprietary). I will be using it in conjuntion with a 1 wire temp sensor and this will simplify code also. Thanks for the tip. Do you know by any chance if there are more recent versions?

I glad it was helpful, I don’t it’s been a long time since I had to source the 1-wire stuff. I think it is still considered current production at least. Also, the temperature sensors are a perfect thing for you. I hadn’t thought of that.