Very very nice comment! Help understand better...

Decoupling caps: 0.1uF ceramic will be fine.

Ok!

zener:
No, just signal diodes, anything that will handle a few mA

OK! but I'm so ...that I can't even choose one from digi-key...a digi-key link will really help!

Internal PIC resistors:
I think you will find that they are pull-up resistors and small ones at that. It will depend on the Hall device you chose. If it has an open collector output then the answer is probably yes and you would be able to loose the diodes as well. See comments below.

then I will have to decide based on the comment below! and especially when the 'Hall sensor' has been chosen!

I will start with this quote:

The point is you need to DESIGN the functionality first and then source the components. IF you then find that a component cant be found, or more likely found cheep enough, then modify your design so that you don't need the functionality that you are struggling to provide.

I thought everybody already knew the functionality!
What I need to achieve is simple! Just sense/check whether or not a piece is on the board (scanning the board) and when a change has been made on the board (ie a piece has changed location on the board), inform the PC (chess engine) of the change made so it can compute its counter move. That's all!

When starting the programming part there might be a few things to considerbut hardware comes first! And we also know the initial position of each piece...Now I am trying to come up with a simple and good circuit (with the right components) that will do the job.

Digi-key alone list 65 Hall effect sensors that will work at 5V including all the output topologies I have mentioned and several more besides.
You will be able to pick one when you decide what you want to do.
Remember you need a switch or an analogue output NOT a latch!

Switches will cost less but check the initialisation and switching times as was mentioned in an earlier post.
Again you need to define the design, how long a scan time is acceptable.
If one board scan per second is OK then even a slow 125mS device is quick enough.

What I want to do (design definition) has been described in the previous quote! Now everybody knows what it all about...

Ok! 'switch' or 'analogue output' and Not a latch!
@ dyslexic: two types of sensors have been suggested at post #28 and #34 which one do you think would be best to use (knowing the application)?

one think I like about the #34 one is that apparently I won't have to use the A/D converter (less coding! that's great!!) + it has an 'open collector' output...but is unipolar...

And what 'package type' would be best to use given that the circuit will be PCB (it a requirement)? of course I'll have my platform on top! well maybe the 'type' doesn't really matter...huh?

About the initialisation/switching time...
Well I can't really tell right now but I'll assume that max scan time is 'one second' but I believe using multiplexing and fast chip we can have a complete scan for less than one second, correct?

I had a quick look at the #34 post one 'datasheet' but have not seen where they mentioned about the 'initialisation/switching' time...the ideal would be to have a complete scan for less than a second...any help here!?

The hall effect sensors I used as an example require a positive supply and have a bipolar analogue output. If you use a device like this then you need to ensure that when multiple device outputs are connected together that one isn't sourcing current for another to sink, that's wheat the diodes are for.

Now it is most likely that as only one row will be powered on at a time that this wouldn't happen anyway but it doesn't hurt to be safe particularly as you are not an experienced designer. Without the diodes you only get to make a tiny programming mistake once and you will destroy 16 sensors.

However if you look at different solutions the the layout will change ... If your sensor is has a unipolar digital output, IE it is either on or off, then you don't need the diodes because there is nothing to protect.
If you plan to measure field strength to identify pieces or sense when they are just above the board, as feedback for your manipulation system, then you will probably want to use Schottky diodes to minimise voltage drop or even replace the diodes with current limiting resistors but that would depend on how the devices behave when not powered on.

I have combined two paragraphs here...
I think I will go for the solution that uses 'schotty diodes' as like you said being safe doesn't hurt...

Since my post and drawing you have talked about 3 other options, all different.

Now there is nothing wrong with that, in fact I think exploring options is a great plan especially as the example I used was just the first common device I found as opposed to being the best option, technically or cost wise.

Nomore time to explore other solution...I sticking with the Hall sensor ones as we are advanced on this one!

Thanks so so much for this post!

 

Well for the chessboard circuit would connect to MCU using two 8-bit PORTs (input and output)...


the microcontroller will connect to PC using USB cable (male - male) via FTDI IC chip and 3 jumper wires for TX, RX and GND:

https://www.sparkfun.com/products/9873
https://www.sparkfun.com/products/515
https://www.sparkfun.com/products/9385

Are the last two link correct I mean the right component?

I also think on the chess circuit I would need some kind of header to connect the 3 wires, correct? like these ones: https://www.sparkfun.com/products/8096


Now How the arm circuit (I know it not designed yet...) connect to the microcontroller? how many pins will be needed!?

Note that only one microcontroller will be used...the PC (chess engine) will send commands via the same hardware interface for the arm. Once command received...mcu will instruct the arm to do the proper move...

Hopefully in 2 days time I will start the 'arm' design...

thanks!

 

Now How the arm circuit (I know it not designed yet...) connect to the microcontroller? how many pins will be needed!?

I thought you were getting the hang of this .... Please!

You cant work out or ask how do I control when you haven't specified what to control.

I know it's an arm, I know it will have an electro magnet on the end but that is only a small part of the required spec.

This is an order of magnitude harder to do than a simple board ...

How many joints will it have?
What are you going to drive them with?
How accurate will it need to be?
Are you going to program an absolute position or 'look' for the piece when you are in the vicinity?

The above are just the initial fundamental questions ... detail come later.

If you try and buy parts that you like the look of and then cobble his together afterwards it will most likely not work well if at all.

Sorry to be blunt but you really need to understand this before starting to build it.

 

I thought you were getting the hang of this .... Please!

You cant work out or ask how do I control when you haven't specified what to control.

I know it's an arm, I know it will have an electro magnet on the end but that is only a small part of the required spec.

This is an order of magnitude harder to do than a simple board ...

How many joints will it have?
What are you going to drive them with?
How accurate will it need to be?
Are you going to program an absolute position or 'look' for the piece when you are in the vicinity?

The above are just the initial fundamental questions ... detail come later.

If you try and buy parts that you like the look of and then cobble his together afterwards it will most likely not work well if at all.

Sorry to be blunt but you really need to understand this before starting to build it.

You are correct! I am aware of everthing you have mentioned...

I'll get back to the arm very soon...I am trying to design it (I mean address all these issues) with very less degree of liberty etc...

I am working on it... will get back to it very soon...

thanks!

 

@ dyslexicbloke:

As for the choice of the Hall sensor...there was 2 suggestions...

do you think this http://www.digikey.co.uk/scripts/dksearch/dksus.dll?vendor=0&keywords=OHN3120U will work better...(as mention on post #34)

I am asking you coz you seem to be the expert with Hall sensor...
I have not placed the order yet coz I need confirmation with it...

thanks

 

Here's the parts list for order:

- Hall Sensors (64)
- magnets (32)
- decoupling caps: 100nF= 0.1uF (64)
- signal diodes - schotty diodes (64); http://www.digikey.co.uk/product-detail/en/SD103A-TR/SD103AVSCT-ND/3104400
- metal thumb tacks (32)
- solenoid/eletromagnet - for lifting/dropping pieces

Anything missing?

I still can't order coz there hasn't been a total agreement on the hall sensors...I have mentioned several times but not being adressed! I can't really decide on my own but looks like I will have to...the clock is ticking...have to start working on the arm...

The magnet will depends on the hall sensors chosen...so a bit stuck on these two...BUT I will go for Bmorse ones (post #34).

I MUST make my order by tomorrow and move on! will have to make tough decision!

Currently studying/researching/thinking about the arm...will post all my ideas/solutions when done!

thanks!

 

You might already have it but something for switching the solenoid. As usual I'm no expert, but a logic level mosfet might be a good choice for switching 500ma at 12V and a higher current schottky diode than the ones you already have for a flyback diode.

I get my magnets from this guy on Ebay, he's got lots of other magnet sizes, these might be about right, but I don't know how to calculate it.
http://www.ebay.co.uk/itm/50-small-...t=UK_Dolls_Accessories_RL&hash=item20c977b4f3

 

You might already have it but something for switching the solenoid. As usual I'm no expert, but a logic level mosfet might be a good choice for switching 500ma at 12V and a higher current schottky diode than the ones you already have for a flyback diode.

A bit lost here... do you mean as the solenoid is to be used, it needs a 'logic level mosfet' and 'higher current schotty diode' for switchin it?

I really feel stupid right now...

I get my magnets from this guy on Ebay, he's got lots of other magnet sizes, these might be about right, but I don't know how to calculate it.
http://www.ebay.co.uk/itm/50-small-...t=UK_Dolls_Accessories_RL&hash=item20c977b4f3

OK! but these magnets are to work with which 'Hall Sensor'? Im asking coz coz Bmorse said the ones he linked only work with 'S end magnets'...

Also you don't know how to calculate what?

thanks

 

You might already have it but something for switching the solenoid. As usual I'm no expert, but a logic level mosfet might be a good choice for switching 500ma at 12V and a higher current schottky diode than the ones you already have for a flyback diode.

I get my magnets from this guy on Ebay, he's got lots of other magnet sizes, these might be about right, but I don't know how to calculate it.
http://www.ebay.co.uk/itm/50-small-...t=UK_Dolls_Accessories_RL&hash=item20c977b4f3

@ Eric:
Markd77's ebay magnets will work with the hall effect sensors on post #34, (If the sensor does not detect the one side of the magnet, just flip the magnet over to the other side)
6mm N35 Neodymium should work just fine, they will be strong enough to where the magnet will not have to be directly in front of the hall effect sensor which will make piece placement easier since you would not have to be so precise in placing the chess pieces. (Of course the bigger the board, the larger diameter the magnets should be.)

As for a flyback diode for the solenoid, a 1n400x (1n4004, etc.) series diode should be just fine to prevent any back emf....

 

@ Eric:
Markd77's ebay magnets will work with the hall effect sensors on post #34, (If the sensor does not detect the one side of the magnet, just flip the magnet over to the other side)
6mm N35 Neodymium should work just fine, they will be strong enough to where the magnet will not have to be directly in front of the hall effect sensor which will make piece placement easier since you would not have to be so precise in placing the chess pieces. (Of course the bigger the board, the larger diameter the magnets should be.)

Alright! will buy them from ebay Mardk77 link...

As I am reading/researching on the arm (hopefully will post soon) I realized that the size of the board (platform) is VERY important when designing the arm!!! coz we need to know the workspace of the arm...so as to determine the 'length' of each linkage..etc...

Any suggestion for the board size? does all board have the same size? need to be investigated...

As for a flyback diode for the solenoid, a 1n400x (1n4004, etc.) series diode should be just fine to prevent any back emf....

so something like this http://www.digikey.co.uk/product-detail/en/1N4004-E3/54/1N4004-E3/54GITR-ND/754810 would work right? guess already one is needed, huh? how 'bout the mosfet level?

Ok! Hall sensor from post #34 will be bought then...

thanks again!

 

Apart from the weight I am not sure I would be worrying about the solenoid.
Given that you are only going to turn the thing on and off every 30 seconds or so you could just use a relay, simple cheep!

I appreciate that the business end is the natural place to focus but it probably shouldn't be, try and think wider

 

Apart from the weight I am not sure I would be worrying about the solenoid.
Given that you are only going to turn the thing on and off every 30 seconds or so you could just use a relay, simple cheep!

Pardon my ignorance...and understand that my electronics background is I would say poor...I will have to learn everything on my own from scratch...but for now all I have to do is quickly read around and ask...

sounds like you have another option/solution...instead of the solenoid...
Btw, happy with Hall sensor from post #34?

I appreciate that the business end is the natural place to focus but it probably shouldn't be, try and think wider

Huh?

 

What about the power supply? How will the whole circuit be powered? guess a proper one needs to be bought, right?

 

Diode linked in recent post looks fine, this MOSFET:
http://www.digikey.co.uk/product-detail/en/NTD5867NL-1G/NTD5867NL-1GOS-ND/2401422
is massive overkill, but is the cheapest through hole logic level MOSFET that can manage 500mA (actually it can handle 40 times that).

 

What about the power supply? How will the whole circuit be powered? guess a proper one needs to be bought, right?

I would guess a 12V regulated supply with 5A or more would be fine for motors, solenoid, etc. Then at least one 5V regulator to supply all the lower voltage parts. If any motors are 5V or less, get a regulator for them and another for the microprocessor to help avoid voltage spikes to the microprocessor.

 

I would guess a 12V regulated supply with 5A or more would be fine for motors, solenoid, etc. Then at least one 5V regulator to supply all the lower voltage parts. If any motors are 5V or less, get a regulator for them and another for the microprocessor to help avoid voltage spikes to the microprocessor.

Alright! so if I am correct there will be one for motors, solenoids,...(arm circuits...), another one for chessboard circuit (mcu included) and have to mix for mcu if any motors require 5v too

any digi-key link...?

thanks

 

Maybe this:
http://www.digikey.co.uk/product-detail/en/ALS75-12/237-1295-ND/1010424
to supply everything, you can then use linear regulators to get your 5V, 3.3V powered from this 12V supply.
Something like these:
5V
http://www.digikey.co.uk/product-detail/en/MC7805CTG/MC7805CTGOS-ND/919333
3.3V
http://www.digikey.co.uk/product-detail/en/UA78M33CKCS/296-21633-5-ND/1494026

 

What I meant by 'business end' was the end that does the work, it was a poor choice of words, sorry.

BY the way if you are going to use RC servos to build your arm, which simplifies the mechanical build considerably, then you will probably want a 5v supply that is good for a several amps.

Servos are available in a huge range of torques and speeds and therefore supply requirements.
What you need will depend on what you want to achieve but I suspect that you will want slow precise movement and will be dealing with fairly high loads, we in RC servo terms anyway.

Torque is measured in mass/distance ... for example.
A 0.5Kg per Meter servo will lift 0.5kG on a 1m beam which is the same as 1kg @ 500mm
The faster you want to move the load the more inertia you are dealing with, that is force necessary to accelerate and decelerate the load.
This directly translates to power, in your case Amps.

Fortunately RC servos are usually specked this way .. IE torque at some maximum speed in degrees per second.
The maximum current stated will be what is required to move the maximum load at the maximum speed.

Hope that helps.

 

Again thanks for that useful post...I am aware of the use of RC servo motor for the arm...Hopefully will post in a few hours on the initial stuffs I came up with for the arm design...

thanks Dyslexcbloke!

 

First thing, IMO, will be to come up with a free body diagram (FBD) showing all the joints (DOFs) and linkage (actuators). Attached is a scketch of the arm I came up with 4 DOFs.

joint1 at the base for the 'rotation' (moving to the left/right);
joint2 and joint3 for the 'translation' (this would help to reach the 1st and 8th row) and joint4 used to pick/drop pieces at an angle of 90 degree.

I believe this is the minimum DOFs required to do the job! Using less than 4 DOFs as described would not, IMO, help much! And more than 4 DOFs is not necessary adding on complexity.

Each DOF (joint) is a RC servo motor! Now what needs to be determined is which RC motor is to be used at each joint and the length of each linkage (atuator).

As for the length of each linkage we need to determine the workspace (size of chessboard) of the arm and the fixed position of the arm relative to the chessboard. I mean there should be a minimum distance between the linkage1 (see scketch) and the first row for it to work well (especially for the first row).

As for the choice of each RC motor, the following need to be taken into consideration in order to determine the 'moment' (torque applied) and hence choose the motor having the minimum required (computed)
torque:

- weight of each linkage (and which material);
- weight of each joint (motor);
- weight of piece to be lifted;
- length of each linkage;
- maybe I missed other stuffs!

I can write 'moment equation' for joint4,3,2 but not too sure about joint1 (base) and apparently 'Forward kinetics', 'Inverse Kinetics' and ...velocity needs to be computed! Sounds like lots of work...

Guess I am ready to spend some $$$ to anyone willing to come up with the full design... that is just too much (for me alone with my skill level) and no time...I must submit a paper design (all schematics) in 2 weeks time to be precise and will have another 4 weeks for all the implementation (and programming)

BUT I am more worried about the 'programming' part of it!

Or just Bmorse method described below that looks much easier:

.......

Build this arm on a fixed position relative to the board (Meaning the arm should always be in the same place near the board all the time), then work on getting it to move to each grid coordinate first (see attached sample grid), this will all be done with a uC first, with no interaction from PC yet....

View attachment 44856

This will have very basic functionality built into it, basically it should just be able accept a simple command to move to a specified X,Y location and pick up a piece, then another simple function to move piece to another X,Y location and set it down. And also add a "teaching" routine, so you can "teach" it the grid locations (basically, manually move arm to location, press button to save coordinates, repeat for each grid position, etc.) that way, if you change board size, you can easily have the arm learn it.

once this is done, then you can move onto the PC comms, then all the Chess engine has to do is send 2 sets of grid coordinates, one for where to pickup piece, and second to where it needs to go.

Thanks for your comments!