hall sensors instead of pots !!!

Hi all,

I changed my pots to hall sensors (SS495). After connecting they work within a range from 43 till 1002. I also used hardisk magnets and can't tune my amc to give a reading of 0 to 1023. Only if i trim the blue trimmer about 6 turns is will be 1023. But i want the full ranges without trimming. What can i due to make it happen?

Also started testing by playing r factor and must say its not to bad, but like i said before i miss a few degrease of movement because my motors won't turn the full 180 degrees.
Some other issue is that with pots my h-bridge stays very cool even if i play four an hour, but with the hall sensors i can only play for like 20 minutes and then my h-bridge mosfets are getting too hot.
The connection of the sensors to the amc is ok (looked at pdf of ss495).

Does anyone have ideas to solve this issues !!!!

thanks very much,


Pim

I recommend you read the datasheet for your sensor to find an answer to your range. For some reason the minimum and maximum voltages don't reach 0V and Vref. I don't think this is a big issue since you are only loosing about 6% from the ends.

It is very interesting that the mosfets heat up with the hall sensor, but not with the pots. The only explanation I can give is the sensor is too noisy, making the motors shake rapidly which causes huge current spikes. Try an RC low pass filter on the analog signal.

If the hall-sensors are placed near the motors, the pwm noise may interferr the hall sensor signal.
Perhaps some hall sensor shielding would be of benefit.

Regards

Hey guys thanks for the fast replies, My motors aren't jumpy they behave the same way as with the pots. But my sensor are near the motors, I will try a different position later.

Because you don't notice it, the motors can still move. If there is an error of 4-5 out of 1023, you are not going to notice it, but the motors are still going to correct their position. The noise is very fast changing, the motors probably don't even start to move that fast. Try to listen to them very closely and you will most likely hear them whine.

You should be able to compensate this effect by adjusting the kp value in the AMC via hyper terminal.

Regards

I will play with the kp settings, now they are both x=3 y=3. What kp settings is used with hall sensors ?
See position sensors in the picture is this to close by?
Because i don't have really jumpy readings.

I have been using such sensors without any problems. KP settings I set to 10. No problems with HBridge. Maybe you should try another magnet and equipped with sensors as close as possible. AMC shows the values of 0, but at a small angle. To avoid the overshooting I was a bit limiting in the X-Sim software with percent limiter.
Here are some photos of my installation.






And video
http://www.youtube.com/watch?v=YBUW4A3zhuE

Hi all,

I tested with KP settings x=10 y=10. The results are more wining noise from motors, and mosfet are getting hot in 5 minutes.
So with a kp of 3 i have less noise. I know this sounds strange because with higher setting it should be less.
I'm going to make a new wire from the sensor to the amc board, and also will check again that the sensor is right connected.
I can't find any other reasons why my mosfet are getting hot so quickly. :?

Like i said before no heating problems when using potmeters !!!!
Thanks all again for replying so fast.

Will keep you guess posted.

So with a kp of 3 i have less noise. I know this sounds strange because with higher setting it should be less.
Can you explain why? This is exactly what should happen...

Put the hall sensor further away from the motors, new wires are not going to change anything.

hey Frakk,

Your right forget that sentence. Kp on 3 gives less noise as it suposed to be, my mistake.
Motors move faster when x y is higher i understand. Tried these settings because ru.null was using these.
Also never played with Kp settings before.

But your remarks Try an RC low pass filter on the analog signal you mean between output pin (sensor) and input pin (amc)?

Will do that tomorrow oh yeah moved away sensor, but this doesn't make any differens in noise.

Yes, just put a series resistor and a capacitor to ground on the AMC side of the resistor. 1K and 0.1uF should work.

Hi All!
Pim, you can take pictures of the motor and the magnet at the side?

The requested photo's

With such a magnet, the angle of displacement is not more than 30-40 degrees. This is clearly not enough. You need a magnet, some like me in the photo, with opposite poles at the edges. Also in front of him, too, magnet? If yes, then it can affect the sensor in certain positions.





If you have problems with overheating HBridge, you can try to connect GND, as in this figure. And replace IRFZ44N to IRLZ44N if you use IRF.

I wanted to install magnets like yours but i can't find those type of magnets.
But with the construction i made you get more degrees of displacement than 30 / 40 degree. In the middel it 512 and a quarter turned counter clockwise 998 and clockwise a quarter gives 40.
Also i think this setup has a to weak magnetic field. Otherwise my readings should be 0 and 1023.
If i use an old hdd magnent the readings are 0 and 1023.
But because of the shape of the magnet (same as thanos uses) i can't use it and will look for strong magnets like your construction.

Your tip about other mosfets and the ground wiring i will do for sure. And Frakk's idea i will follow also.
Just didn't had the time to hobby with it today.

I will keep you all updated.

thanks for all the good info guys

@ ru.null , can you give me your .rn2 file so i can see settings u used.

update findings today:

Today wired both connection (ground) like in the picture the blue wires. After testing only one motor (Motor 2) is working on when i move sensor to the magnets. After disconnecting blue wire (ground) from the (motor 1) half site of the h-bridge, both motors are responding again. So this is a bit strange there should be a bad connection of some sort in motor 1 site of the print.

Bin measuring all parts and also bad soldering, but i can't find directly what should be wrong. To me the voltage on mosfets etc.. looks ok. Did not change the irz44 for the other type yet because i still believe must be something else because it worked fine with pots.

For now i have disconnected both ground wires. I checked without the h-bridge connected the amc and measured the voltage of the hall sensors. when in mid position the reading is 128 128 509 509 and in voltage 2.47 V and when reading = 1002 then voltage = 4,87v and when 0 = 0v.
So the hall sensors are putting out the right values and are ok.

When sensors are connected but laying on the ground the motors don't wine. when i use a old hdd magnet and move the sensor to mid position (509) the motor does whine a bit.

Tomorrow i will try the condensators etc... (frakk's idea) to get rid of the wining noise. Didn't head the time to go to electronics store to arrange the parts.

So and conclusion the noise should be solved quite easily, but the heating : i don't know just yet what to think !!!!!

Hi all,

It is working :yippiee:

Installed frakk's (low pass filter) idea, that did the trick for the noise. And the heating of the mosfets is also ok with this. But they are a bit warmer then they where with pots.
I ordered the other mosfets and will also replace those to bring the temperature down just a little bit.

Thanks again guys for all help.

Good to hear. Try to put it as far from the motors as you can because I'm pretty sure they affect the noise as well.

@ ru.null,

I managed to solve issue with ground (blue wires in picture) that you told
me, but did not change mosfets yet.
Mosfets irlz are ordered.
But i also saw a noter upgrade that you made see quoted text:

We use Thanos H-Bridge with IRLZ44N mosfets. Now hbridge working properly.
Another minor modification is that the motors are connected to the mosfet
coolers to reduce electrical losses.

Now i'm wondering how did you wire this, just put a wire from cooler to
ground of the motor?
Can you explain this how to do?


thank you,

Pim