need help identifying a device

[dpmiller]

OK chaps, my google-fu is coming up short here  @Radian @ProDave etc

 

I'm looking for a replacement Hall-effect sensor. It's an SOT-23 device, and the markings are W90B, 65AA. Does anybody know what it really is and/or what it might cross to? It's the valve position sensor in the pump for our treatment plant.

 

ta...

[Radian]

Are you certain it's a Hall-effect sensor? W9 prefixes a lot of sot-23 Zener diodes. No other matches that I can find. Maybe trace out the pin connections for additional clues, Vdd and GND are on pins 1 & 3 and output on 2 for all the ones I've come across but if yours looks different, it might help eliminate some obvious candidates.

[dpmiller]

Yep, definitely Hall. there's an actuating disc- a cam- directly above it and there are two magnets clipped into this. Problem is, I don't know if this is polarity-sensitive, but presume it is. The 2 magnets are 180deg apart, and it has no other way of telling cam-up or cam-down for valve position feedback

 

so something like a DRV5053 I was thinking

[Radian]

DRV5053 is bipolar analog but what circuit does it feed in to? It might have been digital. Biasing resistors and a comparator would be a big hint that it was analog as you suspect.

[dpmiller]

I'll have a look on the control pcb later, but in the interim the US distributor is putting me in touch with someone in mainland Europe who can assist as the wee sensor board is an available spare apparently

[dpmiller]

Well I can't follow the signal trace to anything active on the main board. the little sensor board has capacitors between GND and both Vs and OP. The OP is pulled up by a resistor to Vs. Where the cable lands on the mainboard, there's another cap between GND and OP, and a resistor between Vs and OP.

Supply voltage is 5v

 

 

I'm tempted to bring the OP line high and/or low through a few k Ohms to see what pattern results in the unit being satisfied and starting the blower

[dpmiller]

So with a bit of experimentation, pulling the OP alternately high and low (through 1kR just in case) happily sequenced the pump on. Getting just a couple of the SOT23 devices looks like a pain, so I've ordered a few of the 41F sensors in thin TO92 off Amazon so we should be sorted tomorrow. They're commonly used in E-bike motors and the like, and I've got room on the PCB for the TO92 device once I do a bit of leadout wire origami.

[Radian]

1 hour ago, dpmiller said:

So with a bit of experimentation, pulling the OP alternately high and low (through 1kR just in case) happily sequenced the pump on.

Excellent. Would have been over-engineered if they'd used an analogue sensor for a go/no go switch.

[dpmiller]

yep, I was pretty sure it had to be polarity-sensitive as I've previously watched it boot up and it didn't stop after the first magnet passed by. One end of the valve travel N, the other S.