Creating a Atomic Chess Board...

[Multiwizard]

Hi,

a few years ago I had a new hardware idea for the Atom...

It is very time-consuming to make and when it's finished it's totally useless...

Unless someone decides to (re)write a program for it...

So these are only the pics of phase one...


After the meeting yesterday I am gonna modify it again, phase two...

Kees told me if I wanted to make it work I had to make some confirm buttons for every move and a RESET button,, so I will...


Greetings, Wim...

[Multiwizard]

#2...

[Multiwizard]

#3

[Multiwizard]

#4

[IanS]

Phase 2 needs diodes to avoid ghosting.
http://pcbheaven.com/wikipages/How_Key_Matrices_Works/

[Multiwizard]

Phase 2 needs diodes to avoid ghosting.
http://pcbheaven.com/wikipages/How_Key_Matrices_Works/

Hi,

thanks for the advice...


Greetings, Wim...

[Multiwizard]

Hi,

Kees has made me another schematic with these diodes...

Looks better this way IanS...


Greetings, Wim...

[IanS]

Diodes look the wrong way round.

Each column is pulled high normally via the resistors. When each row output of the 7445 goes low, you want to be able to pull the row low via a switch and a diode. So the diode wants to pointing away from the column wire towards the row output signal.

Have I misunderstood how it works?

[roland]

You are right. With the diodes this way around the switches can't pull the inputs to a low level. The diodes will block the current.

Btw ... I am still not convinced why we need the diodes here. There is only one active row (low level) when scanning the switches so imho the other rows cannot pull the inputs low whether the switches are opened or closed.

The Atom keyboard works the same way and I have never noticed some ghosting.

Edit: I reread the theory about ghosting and I am now convinced In fact, now I am worried because with the right key presses you can damage the 7445 because you can "short circuit" two outputs. In real life it will survive because the short circuit won't last very long, but still.....

[hoglet]

Edit: I reread the theory about ghosting and I am now convinced In fact, now I am worried because with the right key presses you can damage the 7445 because you can "short circuit" two outputs. In real life it will survive because the short circuit won't last very long, but still.....

It isn't a problem with the 7445 because the output stage of that device is open collector, so it doesn't matter if two output are shorted. Not sure the same is true of the 74LS45 (or if it even exists?)

What did you end up using in your New Atom?

[Multiwizard]

Ok thanks, so I have to turn round all the diodes ...

Now I have to wait for Alie , (100pcs for € 0,72 how do they do it)


Greetings, Wim...

[roland]

It isn't a problem with the 7445 because the output stage of that device is open collector, so it doesn't matter if two output are shorted. Not sure the same is true of the 74LS45 (or if it even exists?)
What did you end up using in your New Atom?

Thanks, now I'm not worried any more I used a 7445 from an old Atom.

[daveejhitchins]

Thanks, now I'm not worried any more I used a 7445 from an old Atom.

There's lot's on eBay, if you get stuck

And Dave's correct, there's no 74LS45 listed - except someone on eBay is selling one

Dave H

[Prime]

Being open collector, don't you also need pullups? Just checked the Atom keyboard circuit and that does indeed have pullups, unless the 6522 has them internally, I know one of the ports on the 6821 do so might be the case here too

It's also worth noting that 74xx145 are also compatible.

Cheers,

Phill.

[oss003]

I always have trouble with positioning diodes in the right direction that's why I told Wim to do a test before soldering all 67 of them ..........

Ian was right about ghosting which will probably never happen with a keyboard but is more likely to happen with the chess board.
The problem occurs without diodes when there are pieces at reed relais A1, A2 and B1.
When O6 is going low, it looks like there is a piece on reed relais B2 but there isn't.....

Greetings
Kees

[hoglet]

Being open collector, don't you also need pullups? Just checked the Atom keyboard circuit and that does indeed have pullups, unless the 6522 has them internally, I know one of the ports on the 6821 do so might be the case here too

Yes, indeed pull ups should be added to each "column".

[oss003]

Dave,

are you sure the 7445 has open collector outputs because the function table states Low and High outputs ......

Greetings
Kees

[hoglet]

are you sure the 7445 has open collector outputs because the function table states Low and High outputs ......

I'm pretty sure, yes.

In the TI data sheet there is a picture of the output stage, and it looks like an open collector:
http://www.cnic.ro/ed/ic/SN7445.pdf

But even if it wasn't, the pull ups would still be needed for the case where no key is pressed.

Dave

[oss003]

You are right, the 8255 has internal pullup resistors, but why aren't there pullup resistors in the Atom?

Just wondering if Port B of the 6522 also has pullup resistors?

Greetings
Kees

[roland]

The resistors R17 ... R25 (at the bottom of the picture) are the pull up resistors.

[oss003]

R17-R25 are pullups for the colums to the 8255 but there are no pullups on the rows from the 7445??
Because these are open collector outputs, do they have a stable state if the row is not selected (H)?

Greetings
Kees

[Prime]

As the datasheet says "all CMOS designs", since the 7445 (or LS45) isn't maybe Acorn found they needed the pullups.

Cheers.

Phill.

[MartinB]

Just wondering if Port B of the 6522 also has pullup resistors?

I predicated my I2C interface on the assumption that there are internal pull-ups present on the 6522 outputs. The datasheet shows them schematically and I measured the values to be around 1k ohm. (I2C won't work unless there are pull-ups present on the bus lines and I don't use external components on any slave devices.)

[roland]

R17-R25 are pullups for the columns to the 8255 but there are no pullups on the rows from the 7445??
Because these are open collector outputs, do they have a stable state if the row is not selected (H)?

They don't have to be pulled up separately. The Atom selects a row and that row goes "low". The other rows are not important at that moment. The resistor of the selected column acts as the pull up also for a non-selected row when a key is pressed on that row. That prevents an occidentally "low" read.

[oss003]

Ok, so just turning all diodes should be ok for Wim's Atomic Chess schematic.
Thanks ....

Greetings
Kees

[hoglet]

I predicated my I2C interface on the assumption that there are internal pull-ups present on the 6522 outputs. The datasheet shows them schematically and I measured the values to be around 1k ohm. (I2C won't work unless there are pull-ups present on the bus lines and I don't use external components on any slave devices.)

Just in case you aren't aware, the port B pull ups are not present on the W65C22.
http://archive.6502.org/datasheets/wdc_ ... 3_2010.pdf

Dave

[MartinB]

Ah, cheers, now you mention it, I did know because this has cropped up before but I had also completely forgotten! If I'd thought, I would have put a note in the manual but no worries, I'll add something to the thread. It's not a problem because if someone wants to use a WDC device, it's easy enough to hang a couple of resistors off the bus somewhere and many slave devices have options for the use of resistors. Ta Dave...