I am a beginner in electronics and I need help on my project.
I am working on a project for which it's required to control 272 (8x34 matrix) electromagnets which are self made, the current drawn each electromagnet is 250 mA on 12 v power supply, I am able to control one column using TPIC6B595 shift register IC and using 34 ICs for each column I think I can control all of them individually, but basically these electromagnets are pushing another permanent magnet but I should reverse the polarity of the power supply to electromagnet to reset it to it's original position, I am not sure how to handle this scenario in this application.
In another topic there was discussion about controlling flip-disc displays which matches with my requirement, but the suggested IC in that topic FP2800A Decoder Driver is not available in my country, it will be really helpful if someone suggest me a better approach to accomplish this task.
Also I am not sure if I can use matrix solution for the same to reduce number of shift registers.
This is a lot harder than you think. You cannot simply reverse the polarity of the power supply because then the electronics will not work.
Do you need more than one on at a time? If so then I don't see any option but to have a h-bridge circuit for each magnet.
The electromagnets are 250mA, but the TPIC6B595 can only do 150mA.
Grumpy_Mike is right, the circuit depends on whether you want to individually control the reverse current or not.
An H-bridge is for controlling a motor, to be able to make the motor run in both direction. The L293 can be used for that.
You could consider to have a serial bus with 8 slave circuit board. Those 8 circuit board can be placed near the electromagnets. That way it is easier to build and it requires less wires.
Perhaps using transistors as output stages, and using a serial bus output port (i2c to output pins, shift register, or something like that) is a more common option. In that case you could use very common components.
Do you already have one column running ? Can you upload a photo of the electromagnet with the permanent magnet ?
If you mention another discussion, please provide a link to it (just copy the url in the text).
In this thread http://arduino.cc/forum/index.php/topic,17607.0.html
I can see FP2800A IC h-bridge on each output to change the power direction, but I will try to see how to implement h-bridge circuit on each output.
Thanks for your detailed reply, I am new to electronics, I didn't yet started using TPIC6B595, right now I am able to control one column using 74HC595+ULN2803 combination, these magnets are winded myself by hand, so still the resistance is not consistent, I think I can still increase the number of turns to reduce it's current consumption. (I thought once I find a way to control all these I could wind all with a coil winding machine.)
If you can mechanically reset them, you could continue to make all columns with the 74HC595+ULN2803. They are very common and cheap components.
I don't know if the ULN2803 can turn all outputs on. 8 * 250mA is 2A. That is a lot for the ground pin. Since they are darlington outputs, the ULN2803 also gets warm.
So many electromagnets will cause a lot of spikes on the power supply. Be sure to have good decoupling for the power supply at every column.
Thank you for your suggestions, I will follow the instructions, I think I have to learn a lot to complete this work, now I am little confident on my task.
but the suggested IC in that topic FP2800A Decoder Driver is not available in my country
I am not sure this will help you. It can only drive one coil at any one time and can not switch the polarity of the current. Sourcing and sinking is not the same as switching polarity.
It is a lot of circuitry you are thinking of making here.
By the way I went to Bangalore a few years ago in the Phillips center, I was giving a seminar to Pace India. Fascinating place, I really enjoyed it. I was glad I was not doing any driving. I stayed in The White House, I was astonished to find such places still existed.
It's no help to you in deciding what sort of drivers to use, but if you're planning to drive these via a matrix of rows/columns then the number of drivers you would need would be the total number of rows plus columns. It's worth bearing in mind that the dimensions of your driver matrix do not need to match the dimensions of your display layout, and the closer it is to square the fewer drivers you would need to use. The minimum in this case seems to be 17x17, requiring 34 drivers. It seems to me that the simplest driver would still require 34 H-bridges, or 34 relay+drive transistor plus a single overall H-bridge, so there's going to be be a considerable amount of work and money involved either way.
Thank you for your reply, I actually underestimated this project, I think I can't work on it alone, but the information you have provided very much helpful as it directs me properly.
Seems H-Bridges seems something very critical, since all the electromagnets are going to reset at once, I think I should do it mechanically, so that I can stick to normal 74HC595+ULN2830 driving logic.
It's nice to hear you visited Bangalore, it's a nice place as you have mentioned except heavy traffic on roads, I live very near to Phillips center :).
Hi PeterH,
Thank you for your reply, now I am planning to reduce the matrix size by 8x8 and first try to control it without H-Bridge, then think I can go for matrix circuit.