Arduino In-Circuit Tester: Build Project

[Dave2084]

I'm happy to pitch in with costs.

Don't mind buying a couple extra of sets (and lots of probe heads) if it helps spread the burden. I can always sell them later.

Once the software has evolved more I think a lot of people will want them. As I'm working on 6802, it's only a short step to do System 6/7 Pinball, System 11 Pinball, Squawk and Talk, Data East Pinball and with a little more effort (to get to 6800) System 3/4 and Bally Classic pinball. That encompasses a large proportion of Solid state pinballs = large user base.

 

[Dave2084]

on the dirty PCBs site:
"NEW ORDERS SHIP AFTER FEB. 16TH! IT IS SPRING FESTIVAL, CHINA IS CLOSED."

 

[PaulSwan]

Some Zaccaria pinball & video sound boards also use 6802.
6800 is a bit like 6502 in that some designs rely on using the Clock Phase CPU inputs as "E", necessitating a clock mastering configuration like 6809.

 

[guddler]

Last time it took circa 2 months

WHAAAATTT???

I've only ordered from them once but I don't remember it taking any more than about 2 to 3 weeks, either I got very lucky or they've got very crap!

 

[Judder]

Last time it took circa 2 months

WHAAAATTT???

From the last order

"But our system shows that we scanned the bar code and the boards went out on the 7th October. It can take up to 8 weeks "

When I mailed them on the 6th of November!

 

[PaulSwan]

For those who want to "roll their own" versions, here are he Design Spark source files:

http://www.paulswan.me/arcade/tech/ArduinoMegaICT/ArduinoICTProbe - V1_01 - 03.zip
http://www.paulswan.me/arcade/tech/ArduinoMegaICT/ArduinoICTShield - V1_10 - 03.zip

Paul.

 

[Dave2084]

So after lots of head scratching I now have the 6821's /IRQ latch and clear working.

It has occurred to me there is one more function I have yet to test on the boards ... the 55516 Speech chip.

Does anyone have a deep enough understanding of the kind of data streams needed for CVSDM to make *any* sound through it just to prove it works? It's not a big deal as if the DAC test works with the speech board connected and the ROMs are good then the op amps should be too, so the only other component to fail is the 55516 itself, but it would be nice to test it.

I've uploaded my latest code to github, I'll try it out on my box if dead sound boards to see how it performs before making a pull request. (I also want to try it on the earlier rectangular sound boards).

 

[Dave2084]

Update: I can now play a series of tones from the CVSDM chip

 

[Dave2084]

Hmm, and how I have a Defender boardset with a poorly ROM board on my bench ... time to diversify?

 

[Judder]

Update: I can now play a series of tones from the CVSDM chip

Awesome

Hmm, and how I have a Defender boardset with a poorly ROM board on my bench ... time to diversify?

If diversification is changing from focusing on the 6808 to the 6809P then I'm definitely all for it

I've also got a chunk of Williams motherboards that I've been meaning to go through and test, so that was the driver for wanting to do that, especially with the 6809P core written already by Paul for Star Wars

Guddler recently wrote some Fluke tests to write to the memory (and thus the screen) so my plan was to try and combine the two

Martin - perhaps you can describe the memory routine in pseudo code and Dave and i can translate into objective C (or which ever flavour of C we are writing in)?

 

[Dave2084]

I'm quite adept at using the fluke on Williams boards, I should be able to come up with something.

What I'm not sure about till I get going is how this system will work with the Williams DRAM, I think everything up to now has been SRAM. [Paul???]

I did have a quick look at the Defender schematics and for a later series CPU the clock mastering would probably go onto Pin 9 of 7P (with it socketed and the leg not in the socket as Paul did with star wars).

Maybe not till in the week.

I'm 99% certain that this ROM board has a bad PIA, I could just change it or test it with the fluke, but where's the fun in doing it the old way?

I think I might have to start off with ROM, CMOS RAM and IO tests and come back to the DRAM later.

 

[Dave2084]

(new to) Github question:

I just created a new branch to start working on some Williams 6809 stuff, expecting Github to create a new local repository on my macbook. Instead it resets all the code on my current repository back to master. Luckily all my changes are check in but they now only exist on Github.

Being used to SVN I find this very surprising, where did I go wrong? Is there any concept of repository/vs working copy here? Surely I should be able to work on multiple branches of the same trunk/master at the same time?

 

[Judder]

(new to) Github question:

I just created a new branch to start working on some Williams 6809 stuff, expecting Github to create a new local repository on my macbook.  Instead it resets all the code on my current repository back to master.  Luckily all my changes are check in but they now only exist on Github.

Hi Dave

Common practice is to git clone to a new directory the branch you are working on if you want to maintain two development environments simultaneously

https://git-scm.com/docs/git-clone

Within a single directory you can only checkout one branch at a time, but you can still have a branch, say 'soundboard', and then check back out master, creae a new branch based on this, say 'motherboard' and then you can just switch between the two as you like - both being separate derivatives of a clean master

You can then merge them into a third branch at a later date to be merged into master etc.

We moved from SVN to Git a number of years ago and it really is worth it for collaborative development

git branch -v
(list branches)

git checkout soundboard
(get back your changes for your soundboard branch)

.. Do some soundboard development ..

git checkout master
(switch back to clean master)

git branch motherboard
(create new branch based on clean master)

git checkout motherboard
(checkout motherboard branch)

.. Do some motherboard development ..Judder2016-02-06 23:15:07

 

[Dave2084]

Not looked at the git stuff yet, however, I have coded up a Defender test for the 6809 project with ROM and IO Testing including all the bank switching and IO initialisation logic.

I did add basic code for RAM as well but unsure how DRAM will work.

I've got it as far as compiling, I'll test it tomorrow if I get time

 

[Judder]

Not looked at the git stuff yet, however, I have coded up a Defender test for the 6809 project with ROM and IO Testing including all the bank switching and IO initialisation logic.

I did add basic code for RAM as well but unsure how DRAM will work.

I've got it as far as compiling, I'll test it tomorrow if I get time

Super cool Dave

Here's how I have my local git repository setup which might help [edited - PM me if anyone needs more information]

Judder2016-02-08 00:24:30

 

[Dave2084]

The Williams branch hasn't had any commits yet. I want to make it at least semi functional before that.

Have you tried the sound board test code? Would be interested in feedback!

To test the interrupts you have to ground a sound input before the timeout.

Don't forget the two custom sound creation features (You'll need a speech board for the CVSDM test).

Dave20842016-02-07 20:08:38

 

[Judder]

The Williams branch hasn't had any commits yet. I want to make it at least semi functional before that.

Ah - that makes sense!

Have you tried the sound board test code? Would be interested in feedback!

That's the plan for tonight - as I have one working sound board and a couple of not working ones I can try it out with the Williams test rig (http://www.thedefenderproject.com/?p=329). Will post back with some pictures and feedback

To test the interrupts you have to ground a sound input before the timeout.

On the IOJ3 as per my tests here? http://www.thedefenderproject.com/?p=304

Don't forget the two custom sound creation features (You'll need a speech board for the CVSDM test).

Cool will do - need to find / borrow a speech board to test that last bit out (*if anyone has one kicking around!)

 

[Dave2084]

To test the interrupts you have to ground a sound input before the timeout.

On the IOJ3 as per my tests here? http://www.thedefenderproject.com/?p=304


Don't forget that some jumper configurations move inputs around:

1. W4/9 switches between pin 6 and DIPSW2.

2. W12/13 disconnects pin 8 (or not).

 

[Judder]

Have you tried the sound board test code? Would be interested in feedback!

Dave - tonight I did and I have to say officially, awesome

This thread is going to be a little bit picture heavy but for everyone to appreciate the great work you and Paul have been doing thought it was a good thing!

Here we go (PS it's rigged for Joust at the moment)...

1. Test rig setup - Linear PSU, four wires for +5V, +12V, -5V (does for -12V) and Ground (Martin - note CPU head now the right way around ) - Dining room table borrowed from very understanding wife

2. Close up of the CPU head in the Williams sound board (for orientation purposes - little * to top left for pin 1 and wires - 2 = +12V, 3 = +5V, 5 = GND, 8 = -12v (-5V))

3. Initial Bus Idle Test - OK!

4. Bus Check

5. ROM Check All

6. RAM Check All

7. RAM Check All CS

8. Interupt Check

9. Input Read

10. Output Write

11. ROM Check

12. ROM Read

13. RAM Check

14. RAM Write-Read

15. ROM Read All

16. RAM Read All

Really great to see it all working Dave and now time to try and diagnose the ones with the E:s next to them

 

[guddler]

Nice work!

Reason I wasn't about at the weekend and had every intention of committing Asteroids on my day off was because I helped a fellow collector (and good friend of course) move just under 40 of his cabs. I'd love to say that was all of them but it was all we could get done in two solid days.

Sadly I was broken on Monday (no kidding) so nothing happened and in the morning I am back away for work again. Such is life I guess. I will get to it.