I'll start with a picture...



The two red circles are magnets embedded in a 3D printed pulley that I marked before printing to make sure I had the right side towards the sensor. The sensor itself is glued on the inside of the post with the wires coming out the back with roughly 3 mm between the sensor and the magnets as they pass by. Both magnets are triggering the sensor as expected and everything appears correct when turned by hand.

What is happening is as the pulley rotates at higher speeds the sensor does not seem to register the magnets correctly. I measured the pulley RPM with a separate tachometer and it showed around 3000 RPM, but the Hall sensor was only triggering around 110Hz which would equal out to around 55 RPM.



I watched the scope as long as I could handle the noise (it's a siren project) and as far as I could tell there was no erratic signalling, every edge looked nice and crisp, and no other oddities where seen. I tried some different triggering and timing to catch anything out of the ordinary saw nothing different. I tried everything again with a separate power supply and temporary circuit with only the sensor and pull up connected (what is actually in the above picture) and nothing changed.

I did buy the sensors off Amazon and they don't seem to have any manufacturer's logo on them at all so maybe it's just a matter of buying proper parts, but I wanted to see if anyone has any ideas as to what could be wrong. They are your typical transistor looking device https://www.amazon.com/dp/B0CHDSD62C?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1. As far as I can tell from other datasheets they should be fast enough to register at 3000 RPM and even still good at the 5120 RPM I need to hit assuming they are to have similar specifications.

 

The A3144 is obsolete.
40 units for 10 bucks, so typical of Amazon.
Find a newer substitute from vendors like Digikey or Mouser.

 

Probably semiconductor manufacturing rejects, that instead of being crushed were “salvaged” by a far east trader and conveniently resold by Amazon.
If something is too good to be true……

 

Probably semiconductor manufacturing rejects, that instead of being crushed were “salvaged” by a far east trader and conveniently resold by Amazon.
If something is too good to be true……

I'll admit that was my initial fear, but I wanted to see if anyone knew of anything else I may have missed or didn't understand. I have never attempted anything with Hall Effect or magnets yet. I was hoping for "clearance" type stock instead of rejects. Time to poke around digikey I guess.

 

I'll start with a picture...

View attachment 363108

The two red circles are magnets embedded in a 3D printed pulley that I marked before printing to make sure I had the right side towards the sensor. The sensor itself is glued on the inside of the post with the wires coming out the back with roughly 3 mm between the sensor and the magnets as they pass by. Both magnets are triggering the sensor as expected and everything appears correct when turned by hand.

What is happening is as the pulley rotates at higher speeds the sensor does not seem to register the magnets correctly. I measured the pulley RPM with a separate tachometer and it showed around 3000 RPM, but the Hall sensor was only triggering around 110Hz which would equal out to around 55 RPM.

View attachment 363109

I watched the scope as long as I could handle the noise (it's a siren project) and as far as I could tell there was no erratic signalling, every edge looked nice and crisp, and no other oddities where seen. I tried some different triggering and timing to catch anything out of the ordinary saw nothing different. I tried everything again with a separate power supply and temporary circuit with only the sensor and pull up connected (what is actually in the above picture) and nothing changed.

I did buy the sensors off Amazon and they don't seem to have any manufacturer's logo on them at all so maybe it's just a matter of buying proper parts, but I wanted to see if anyone has any ideas as to what could be wrong. They are your typical transistor looking device https://www.amazon.com/dp/B0CHDSD62C?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1. As far as I can tell from other datasheets they should be fast enough to register at 3000 RPM and even still good at the 5120 RPM I need to hit assuming they are to have similar specifications.

You're misinterpreting Revs per second (rps) for RPM, your 110/2 =55 rps x 60 =3300 RPM!

 

You're misinterpreting Revs per second (rps) for RPM, your 110/2 =55 rps x 60 =3300 RPM!

I can't believe I was that blind!! I think I have a bit of project fatigue at this point. It looked way too consistent for part issues and it even worked out the same in programming on the pic microcontroller which is where it would have really showed any problems.
Thank you a thousand times.

 

I made a little video of this contraption and thought I would share a link

There's a bit more to the video before this, but decided to skip to the relevant part. The left side of the screen is a screen recording of the app I use to control it.

 

where I am from we have plenty of those, full scale!

 

I can't believe I was that blind!! I think I have a bit of project fatigue at this point. It looked way too consistent for part issues and it even worked out the same in programming on the pic microcontroller which is where it would have really showed any problems.
Thank you a thousand times.

Been there and done that.

Ron

 

I'll admit that was my initial fear, but I wanted to see if anyone knew of anything else I may have missed or didn't understand. I have never attempted anything with Hall Effect or magnets yet. I was hoping for "clearance" type stock instead of rejects. Time to poke around digikey I guess.

Then I read post #5 and it seems to have been a data analysis error. Errors do happen, and reading RPM from a frequency counter that displays events per second happens to be a common goof.

The first step is to look at the data sheets of those hall effect sensors, and see what their response time relative to the magnetic field strengthis. Certainly it is not instant! BUT that was not the problem, as we see!!!