Building Rolands new Atom design...

discuss both original and modern hardware for the bbc micro/electron
PhilYoung
Posts: 204
Joined: Sun Jun 12, 2011 5:55 pm
Contact:

Re: Building Rolands new Atom design...

Post by PhilYoung » Wed Feb 04, 2015 7:48 pm

multiwizard wrote:
Hi Phil, I took some close-up pictures... :)
Sneaky! Very creative. Mine isn't going to have to fit in an Atom case so I'll be able to use something more conventional,

Good luck when you power it up,

Cheers,

Phil

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Wed Feb 04, 2015 8:04 pm

multiwizard wrote: I will power it up tomorrow...
How can you be so hard to yourself (and us)? Why power it up tomorrow if you've everything ready today? :shock:
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:


User avatar
hoglet
Posts: 9091
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Building Rolands new Atom design...

Post by hoglet » Thu Feb 05, 2015 6:51 am

Hmm, the blue lines don't look good.

Can I just check you have re-programmed the GODIL with the design specific to Roland's board as a few pins changed:
https://github.com/hoglet67/AtomGodilVi ... Roland.bit

The lines look like they are 1 us intervals, so it's possible that it is crosstalk from a clock signal. If you change the speed of the CPU clock, does number of them change?

Dave

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 7:09 am

hoglet wrote:Hmm, the blue lines don't look good.

Can I just check you have re-programmed the GODIL with the design specific to Roland's board as a few pins changed:
https://github.com/hoglet67/AtomGodilVi ... Roland.bit

The lines look like they are 1 us intervals, so it's possible that it is crosstalk from a clock signal. If you change the speed of the CPU clock, does number of them change?

Dave
Hi Dave, Roland has programmed this GODIL so I expect that's ok... :?

Only keys that do work are shift+break and control+break and break so I can't switch CPU speed.

I also have measured the keyboard ribbon cable and that's ok, all the signals do come trough... :D

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 7:15 am

hoglet wrote: Can I just check you have re-programmed the GODIL with the design specific to Roland's board as a few pins changed:
https://github.com/hoglet67/AtomGodilVi ... Roland.bit
Hi Dave,

I can confirm he has the right bit-file. I programmed his GODIL myself. And otherwise I don't expect the GODIL to display any text at all.

Wim hasn't used shielded wires for the VGA signal. Can that cause the blue lines?

Greetings,
Roland
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 7:24 am

multiwizard wrote: I also have measured the keyboard ribbon cable and that's ok, all the signals do come trough... :D
Hi Wim,

What happens when you remove the keyboard connector and connect some input lines to ground with a piece of wire?

Maybe your 8255 or 7445 are broken.

Greetings,
Roland
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
hoglet
Posts: 9091
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Building Rolands new Atom design...

Post by hoglet » Thu Feb 05, 2015 7:26 am

roland wrote: Wim hasn't used shielded wires for the VGA signal. Can that cause the blue lines?
On the original GODIL, the VGA Blue signal was right next to the SID audio signal. If this is still the case, then I expect that is the problem. Separating the SID Audio signal in the ribbon cable might help.

Dave

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 7:42 am

roland wrote:
multiwizard wrote: I also have measured the keyboard ribbon cable and that's ok, all the signals do come trough... :D
Hi Wim,

What happens when you remove the keyboard connector and connect some input lines to ground with a piece of wire?

Maybe your 8255 or 7445 are broken.

Greetings,
Roland
I indeed get random characters... :?

I have double checked the keyboard with very bright light, non of the small drilled holes have damaged any of the copper tracks on the back side. :-k

The only thing to do is making a other ribbon cable and connect it to one of my other Atom keyboards...

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 7:53 am

multiwizard wrote:The only thing to do is making a other ribbon cable and connect it to one of my other Atom keyboards...
Please don't split another Atom [-X

You can also try to find out where it goes wrong. You can use the schema of the original Atom because the keyboard sections are exactly the same. So if you press a key you can check if the corresponding pin on the 8255 goes to 0V.

And of course, you're welcome at out monthly meeting in Sittard tomorrow :lol:
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 8:06 am

roland wrote:
multiwizard wrote:The only thing to do is making a other ribbon cable and connect it to one of my other Atom keyboards...
Please don't split another Atom [-X

:lol:
Of course I don't... :lol:
roland wrote: And of course, you're welcome at out monthly meeting in Sittard tomorrow
I was thinking of it already... :)

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 8:08 am

hoglet wrote:
roland wrote: Wim hasn't used shielded wires for the VGA signal. Can that cause the blue lines?
On the original GODIL, the VGA Blue signal was right next to the SID audio signal. If this is still the case, then I expect that is the problem. Separating the SID Audio signal in the ribbon cable might help.

Dave

Is this what you meant Dave, a shielded cable on E7 and E8? ( see picture)
Attachments
P1090038.JPG
Last edited by Multiwizard on Thu Feb 05, 2015 8:15 am, edited 1 time in total.

User avatar
hoglet
Posts: 9091
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Building Rolands new Atom design...

Post by hoglet » Thu Feb 05, 2015 8:14 am

multiwizard wrote:Is this what you mean Dave, a shielded cable on E7 and E8? ( see picture)
It was :(

One more thing.

Try adding a jumper to the GODIL as indicated in this diagram for the "PL4 VGA Jumper".
https://github.com/hoglet67/AtomGodilVi ... stallation

On Roland's design, this completely stops the SID audio.

Dave

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 8:19 am

hoglet wrote:
multiwizard wrote:Is this what you mean Dave, a shielded cable on E7 and E8? ( see picture)
It was :(

One more thing.

Try adding a jumper to the GODIL as indicated in this diagram for the "PL4 VGA Jumper".
https://github.com/hoglet67/AtomGodilVi ... stallation

On Roland's design, this completely stops the SID audio.

Dave
This page won't be displayed. (time out) :(

User avatar
hoglet
Posts: 9091
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Building Rolands new Atom design...

Post by hoglet » Thu Feb 05, 2015 8:22 am

multiwizard wrote:[This page won't be displayed. (time out) :(
Working for me, but here's the picture:
godil_jumper_and_connections.png
Dave

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 8:32 am

hoglet wrote:
multiwizard wrote:[This page won't be displayed. (time out) :(
Working for me, but here's the picture:
The attachment godil_jumper_and_connections.png is no longer available
Dave
Indeed, the blue lines are gone... :D
Attachments
P1090039.JPG

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 8:59 am

I made a wire to the ground and tried every pin on P5, this is what I get on the screen... :(
Attachments
P5 Connector.PNG

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 9:13 am

Your connector image is not correct. Pin 1 is not ground:
Schermafbeelding 2015-02-05 om 10.08.06 (2).png
According what you write I suspect either the 7445 or the 8255. Can you check if the pins 12, 13, 14 and 15 of the 7445 are stable or alternating? Same for the output pins 1 to 11 (8 is gnd).
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 11:00 am

roland wrote:Your connector image is not correct. Pin 1 is not ground:

According what you write I suspect either the 7445 or the 8255. Can you check if the pins 12, 13, 14 and 15 of the 7445 are stable or alternating? Same for the output pins 1 to 11 (8 is gnd).
All the lines underneath the green line are connected with each-other .

The 7445:

pin 1 = 2,64 Volt
pin11 = 0,18 Volt
pin12 = 0,93 Volt
pin13 = 1,23 Volt
pin14 = 1,59 Volt
pin15 = 1,52 Volt

Stable
Attachments
Green is connected.PNG

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 11:44 am

It looks like the 8255 is controlling the 7445. So far so good.
multiwizard wrote:I made a wire to the ground and tried every pin on P5, this is what I get on the screen... :(
Do the keys 3, G, Q and - work?

IIRC you have build my simple logic probe. With that probe you can check if the keys pull a line down. You will see the H led on and the L led should show very short (weak light) pulses. Can you try that?
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

AndyGarton
Posts: 191
Joined: Tue May 21, 2013 3:47 pm
Contact:

Re: Building Rolands new Atom design...

Post by AndyGarton » Thu Feb 05, 2015 12:23 pm

Sorry to interrupt the problem solving discussion, just a useless comment to say I absolutely love this thread and what you guys are doing. Thanks for documenting it all so clearly and entertainingly!

As you were :)

User avatar
roland
Posts: 3761
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Building Rolands new Atom design...

Post by roland » Thu Feb 05, 2015 12:42 pm

AndyGarton wrote:I absolutely love this thread and what you guys are doing. Thanks for documenting it all so clearly and entertainingly!
That's why we do this in this thread and not by email. This way everybody can enjoy it and learn from it.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 1:17 pm

multiwizard wrote:
hoglet wrote:Hmm, the blue lines don't look good.

Can I just check you have re-programmed the GODIL with the design specific to Roland's board as a few pins changed:
https://github.com/hoglet67/AtomGodilVi ... Roland.bit

The lines look like they are 1 us intervals, so it's possible that it is crosstalk from a clock signal. If you change the speed of the CPU clock, does number of them change?

Dave
Hi Dave, Roland has programmed this GODIL so I expect that's ok... :?

Only keys that do work are shift+break and control+break and break so I can't switch CPU speed.

I also have measured the keyboard ribbon cable and that's ok, all the signals do come through... :D

User avatar
oss003
Posts: 3249
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Building Rolands new Atom design...

Post by oss003 » Thu Feb 05, 2015 2:33 pm

Hi Wim,

I think you have to test the keyboard seperately.

Disconnect if from the Atom and put a header in the connector for measuring.
Then switch your meter to resistance and connect one pen to the red marked pin of P5 in the table.
Measure with the other pen the green pins on P5 while pressing the corresponding key. If the key is pressed, you should measure 0 ohm.

This way you can check if your keyboard switches are working ok and that no track/wire is broken.

Greetings
Kees
Attachments
matrix.pdf
(38.08 KiB) Downloaded 43 times

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 6:01 pm

Hi Roland and Kees,

many thanks for all the advice and trying to help me to solve these problems... :D

But this is why I didn't finish my M.T.S. school in the Eighties, to much measuring things (for me) and therefore losing total control of measuring.

Perhaps I can do some nice things with hardware, but that's it... :(

I am absolute not that smart like a lot of you guys, I wish I did, I also have dyslexia and when I have to read things it al gets mixed up when I look at it. :(

User avatar
1024MAK
Posts: 9904
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Building Rolands new Atom design...

Post by 1024MAK » Thu Feb 05, 2015 7:33 pm

The key is to take things step by step. If you are working from a diagram and you can print it out, do so. Then you can mark off each bit as you test it. Believe me, us clever guys sometimes get brain ache and have to do this as well.

Do keep going, you have made great progress so far =D>

Mark

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 7:50 pm

Hi guys, I managed to picked up myself. :wink:

I did the test Kees suggested, and surprisingly some keys did not come through... :shock:

See picture... :?:
Attachments
P5 key Check.PNG

User avatar
Multiwizard
Posts: 1865
Joined: Wed Jan 11, 2012 9:03 pm
Contact:

Re: Building Rolands new Atom design...

Post by Multiwizard » Thu Feb 05, 2015 7:52 pm

1024MAK wrote:The key is to take things step by step.

Mark
Hi Mark, I my case have to crawl for it... :lol:

User avatar
oss003
Posts: 3249
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Building Rolands new Atom design...

Post by oss003 » Thu Feb 05, 2015 8:36 pm

Ok Wim, one step taken :wink:

Please check if you did not drill any tracks!

1. ESC key not working:
Check if one pin of the ESC key is connected to P5-1 and the other one to P5-12 (probably broken)

2. DEL key beeping:
Looks like there is a shortcut around the DEL key between P5-9 and P5-4.

3. The 9,C,M,W keys are not working:
There also is a track broken on P5-9 between the DEL key and the 9 key.

4. The , key is not working:
Check if one pin of the , key is connected to P5-3 and the other one to P5-6 (probably broken)

Greetings
Kees

User avatar
sirmorris
Posts: 783
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: Building Rolands new Atom design...

Post by sirmorris » Thu Feb 05, 2015 8:47 pm

=D> Great work Wim! =D>

Can I suggest checking the sawn edge of the PCB? Perhaps there are short circuits there.

Post Reply

Return to “8-bit acorn hardware”