Magnetic Suspension Circuit: Troubleshooting Levitation

[Franklin Lopez]

I am building a magnetic suspension circuit. Basically to hold up a ball with a a piece of metal inside using a coil with an electromagnetic current. The circuit has already been set up. The ball is attracted to the coil but does not stay levitated. It should stay below the led sensor and the photo sensor. We believe the current should drop but it doesn’t and adjusting the pots does nothing to the current nor the voltage.

 

[berkeman]

We can probably help you, if you post a lot more information. How are we supposed to help you given so little information?

How about up use the Upload button to show us:

• Schematic
• Photo of the setup
• Specs on the weight/size/etc of the ball
• Your analysis of the Feedback Loop (Gain and Phase Margin, Transfer Function, etc.)

Thanks.

EDIT/ADD -- Make sure your files are in PDF and JPEG formats for the Uploads. Thanks.

 

[Franklin Lopez]

I have tried to add a schematic, and actual pictures of set up. At the moment the two 100K pots are not functioning the way we believe they should be. When adjusted the voltage nor the current seems to change.

 

[Tom.G]

Could you post another copy of the schematic with a higher JPEG resolution? Neither the numbers nor the pin names on the opamps are readable. The maximum file size here is 3MB, which should be way more than enough to get a readable schematic to us.

Is the ball expected to end up above or below the sensor? (that is the position it should start at, so it just partially blocks the photosensor)

To figure out what is going on, please tell us what the supply voltage is, then:
1) Set both 100k pots to minimum resistance.
2) With the LED shining on the photosensor, take voltage readings at each TP and the supply voltage
3) With the photosensor blocked, take voltage readings at each TP and the supply voltage
4) Set the lower 100k pot to maximum resistance, leave the upper one at minimum
5) Repeat steps 2) and 3)
6) Post results

I designed and built a similar device many years ago and they can be tricky to set up. The optical system in this one going to make it especially challenging due to the very small sensitive area of the photosensor. Also, the metal inside the ball must be attached to the ball. If that is a hollow ball with the metal rattling around inside you have an almost impossible problem. You can also try it with just the metal, without the ball.

Tom

 

[Franklin Lopez]

The ball is supposed to be right below the sensor.
Will be working on project again tomorrow and will post results

 

[Tom.G]

Much better schematic, thanks.
I can't quite make out the value of R3 (I think it's R3, hard to read, the one from the 50k pot to Ground).
What are the part numbers for the LED, the photosensor and the opamps?
When the ball is at the electromagnet, is the light to the photosensor blocked?
If you have an Ohmmeter that will read that low, I would also like to know the resistance of the coil.
The breadboard photos seem to show a resistor at hole F34 which is unconnected.

 

[wirenut]

Much better schematic, thanks.
I can't quite make out the value of R3 (I think it's R3, hard to read, the one from the 50k pot to Ground).
What are the part numbers for the LED, the photosensor and the opamps?
When the ball is at the electromagnet, is the light to the photosensor blocked?
If you have an Ohmmeter that will read that low, I would also like to know the resistance of the coil.
The breadboard photos seem to show a resistor at hole F34 which is unconnected.

That is r8 and I think it is 8k. As far as I can see r3 doesn't exist.

 

[dlgoff]

I think it is 8k

We need a higher resolution image but playing around with my window size, it looks like 2k to me.

 

[wirenut]

We need a higher resolution image but playing around with my window size, it looks like 2k to me.

I enlarged my screen AND had to use a magnifying glass! I don't know if it's I'm getting old or what, but I've been using the zoom function a lot lately. I tried the same on the pictures and it looks like you are correct on the resistor isn't connected to anything at F34. [edit] On second look the lead might just be long and looped back under the 2 green jumper wires to E35, its tough to tell.@dlgoff As for the R8 8k looking like 2k the font he is using has the top and bottom of the 8 flattened, and his 2's look like Z

 

[dlgoff]

As for the R8 8k looking like 2k the font he is using has the top and bottom of the 8 flattened, and his 2's look like Z

The more I look the more I can't tell.

 

[Tom.G]

I wouldn't worry about it too much at this point. That resistor just sets the maximum gain in conjunction with the 50k pot. There are other more significant problems with the circuit, which can only be resolved when the OP posts the data previously requested.

 

[wirenut]

It would be helpful if he either labeled the parts in the pics, or cleaned up the breadboard a little, and listed the parts used.
I am unable to make out the color bands on the resistors, so I can't tell where they are supposed to go, and don't know what the white rectangles are supposed to be. Also the leads going out of frame need to be labeled.