The Beginner Vs. Star Wars: Ep II 'The Board Wars'

[Griffin]

As part of my larger goal of scratch building a complete Star Wars cockpit I'm having to take on various sub-projects. The first was building a 25" XY monitor from scratch....

Episode 1 is here The Beginner Vs. Star Wars: Ep I 'Building a 25" AMP XY Monitor

As part of my ongoing documentation of this project now I have to take on my non-booting Star Wars board set and see if I can get that running. The set has been assembled from 5 separate sources so it's a real mish-mash of spare parts but my hope is to create a fully working game from the sad forgotten remnants of machines long gone.

For this part of the project I have collected the following pieces...

Atari Color XY Power Brick - Working (Thanks Komodo)
Game boards - Non tested (Thanks eBay seller)
Game board interconnect - Working (Thanks Komodo)
RFI Board interconnect - Working (Thanks Komodo)
AR2 - Working (Thanks Komodo)
RFI Cage/Shield (Thanks Komodo)
Full uncut Star Wars Cockpit Wiring Harness (Thanks eBay Seller)
Star Wars Service Panel - Working (Thanks MitchO)
Original game Yoke - Rough, but solid (Thanks eBay Seller)

A special thanks to DezBaz, EclipseEye and Komodo for general ongoing support and Talon2000 who is helping me via PM with some expert guidance and advice on board repair to keep me trucking on as I take on this task.

[Griffin]

What I know so far.

The XY Power supply is good. Using this guide I have checked all voltage across the pins in the J5 power connector block without the supply under load and got the following...

1. 15.59v dc
2. 15.6v dc
3. 15.6v dc
4. Ground
5. Ground
6. 19.5v ac
7. 19.5v ac
8. 13.1v ac
9. 7.8v ac
10. 15.4v ac
11. 7.85v ac
12. No pin
13. 31.5v ac
14. No pin
15. No pin

As I understand it the voltages on pins 1-3 show higher than they should be as there was no load on them while testing, under load they should be around 10.6VDC

Pins 6 & 7 are good as they combine to create the 36VAC which is really 18/0/18 CT (Center tapped) so 19.5 is OK compared to 18. You should get 36VAC between pins 6 and 7

Measure 8 to 9 directly (Not to GND) See if you get 6.3VAC or so

Measure pins 10, 11 & 13 directly also - Use pin 11 as GND for testing this winding. It's another CT winding, so you should get 25/0/25VAC, and 50VAC between 10 & 13

(Thanks to DezBaz for walking me through this initial power testing)

[Griffin]

The AR2 works properly.

Using this guide I checked the voltages on the pins of the connector blocks on the AR2 board

With the power supply plug J5 back in place (if correct voltages are present at the power brick)...

Label, then Unplug J7 and J8 from the AR2 ** There are 2 plugs on an AR2 that are the same. (IF you get it wrong, you will BLOW UP Game boards) So label!! **** / end blunt warning

With the power turned on I checked the AR2 voltages for +5VDC, +12VDC, +22VDC, -22VDC, +15VDC, -15VDC, -5VDC, all to GND

(Thanks to DezBaz for walking me through this initial power testing)

[Griffin]

Confident that my Power supply and AR2 are sending nice friendly voltages out to the wiring harness I am now moving on to plugging in the game boards and powering them up to see if I get anything out of them.

I have my test speaker grounded to the AR2 and attached to the SP1 capacitor leg (it's labeled on the AR2 board) so I can at least hear the sound coming directly off the AR2.

I've checked all the mating pins in my two interconnect blocks on my wiring harness to make sure that none of them are broken, burnt or bent out of whack. Also for now I am going to leave the RFI interconnect board out of the equation and just connect my wiring harness directly to the game boards...

The RFI Interconnect in place


Direct connection to the game boards with the RFI Interconnect removed.
(Thanks EclipseEye for these images)

I've hooked up my XY monitor, have my service panel plugged in and switch set to test mode and I have my coin door and yoke plugged in too.

Time to power on and see what happens...

Well, a whole load of nothing, but we knew that or I wouldn't be writing this thread

Apparently I should expect to hear at least a series of 16 beeps from the speaker when powering up the game boards in test mode, but nothing. I also have no picture on my monitor and I've noticed that the spot killer light is on at the monitor deflection board meaning the is no XY signal being sent to the monitor from the game boards. I have also noticed that I have solid red lights on all three game boards (which is apparently a good thing).

[Griffin]

At this point it's time to start basic testing on the game boards, so I have to do the following.

- Check the voltages at the various voltage test tabs on the main game board.
- Check the clock crystal and circuit.
- Check or swap out the CPU

So this is where the absolute beginner meets the challenge, I have no idea what a clock crystal is or what the circuit does, so this is where the real learning is going to start.

I do know I need a logic probe and a oscilloscope at 100mhz or above... how do I know these things??? Because Talon2000 told me, and that's good enough for me.

So I have my logic probe on order from Amazon ($13) and I'm currently combing the local classifieds and looking on eBay to try an O-Scope good enough that I can use it for these kind of jobs long into the future without completely breaking the bank.

[Griffin]

I'm starting to read may way through the infamous Star Wars troubleshooting guide found here... http://falz.net/files/starwars/troub...guide/#mainavg


As recommended I'm going to to have to check the Main PCB as follows:

1. Visually inspect the board for obvious problems:
Loose, missing, or improperly oriented integrated circuits.
Solder shorts.

2. Check for the following voltages:
+ 5 V
+ 12 V
+ 5 V EAROM (there is a separate + 5 V for the nonvolatile RAM)
+ 10.3 V unregulated.

3. Check for the following clocks:
12 MHz
6 MHz
3 MHz
1.5 MHz
E
Q


4. Check that the microprocessor HALT, NMI, and FIRQ lines are always high.
5. Check that the microprocessor RESET line is high except when the RESET test point is grounded. If necessary, disable the watchdog circuit by grounding the WDDIS test point on the Main PCB. Refer to Troubleshooting the Watchdog Circuit if the watchdog circuit is not working.
6. Check that the microprocessor IRQ line is not permanently low. (An intermittent low on the IRQ line is acceptable.)
7. Perform Hardware Diagnostic Test 1 to check that the Main PCB RAM and ROM is working. If the RAM or ROM is faulty, refer to Troubleshooting With the CAT Box for checking the Main PCB.

Ok so testing voltages, that's easy enough to do with the test tabs on the game boards using my multi-meter.

Check the clocks - I have no idea what this is or how to do it, so this is my first homework, how do I check the clocks, what am I looking for, where do I even find the points to test?

[EclipseEye]

Griffin - I will help where I can, but I am no expert at repairing Star Wars PCBs and have only ever repaired two sets in the past.

Things to consider, the PCB set consists of three parts, the Sound Board (Top smaller board), the CPU Board (The middle section) and the AVG Board (Bottom section).

The PCB set can be troubleshot by removing the Sound Board, although obviously it means you will have no sound, but full access to the CPU Board.

I think you have already tested voltages on-load, so if you have not already done so, as a starting point I would remove all socketed ICs, clean any tarnished legs and refit them.

Big problem area's on a SW PCB set are the ROMs and their sockets, RAMs and their sockets. The AVG chip is also prone to failure.

It would probably be a good idea to see what the CPU is doing, so locate the 6809 CPU on the CPU Board, it will look like this;



Place a probe on pin 37 (RESET) to see if the PCB is 'Watchdogging', if it is you will see the signal flip from LO to HI repeating constantly.

I have to warn you though, learning to repair PCBs on a SW is an incredibly steep learning curve. However, if you can tackle this then you can tackle anything. Hopefully others with more experience will pip in and maybe I will learn something too.

[Griffin]

Excellent!! Thanks.

I don't mind the steep learning curve, the way I see it, if I can do this and others can follow along then it will at least show that you don't have to be an expert electronics engineer to attempt board repairs.

This guide to repairing PCBs for Beginners by Womble is proving to be an excellent introduction. I like the way he writes, keeps everything in very easy to understand terms and really seems to cover some very important first steps... http://forums.arcade-museum.com/showthread.php?t=177192

I think many new users get very intimidated by the verbiage and seeming complexity of it all and give up. The hobby can only benefit by more people feeling encouraged enough to give things a try themselves and experience their own victories.

[PL1]

Originally Posted by Griffin

Check the clocks - I have no idea what this is or how to do it, so this is my first homework, how do I check the clocks, what am I looking for, where do I even find the points to test?

Here's a little theory to help you picture and measure what is going on with the clocks, Joel.

In musical terms, the crystal (indicated as Y1, Y2, etc. on the schematic) that generates the clock signal acts like a violin string.

Thicker strings vibrate slower (lower frequency) than the thinner strings.

Shorter strings vibrate faster (higher frequency) than longer strings.

The combination of string thickness and length determines the resonant frequency of that string.

When a voltage is applied to the crystal it vibrates at its resonant frequency, just like drawing a bow across the violin string causes the string to vibrate at its resonant frequency.

The measurements you mentioned are frequencies (except for E and Q which are covered in the guide you just linked under "== Clock =="), but frequency counters can be expensive and you can measure it another way.

Since frequency and time are inverse functions, you can use an oscilloscope to measure the frequency by measuring the time each cycle takes.

Example:
50 Hz = 50 cycles/second ==> 1 cycle = 1/50th of a second = .02 Sec = 20 mSec

On the oscope, with time/div set to 5 mSec, the signal will take up 4 divisions or squares (5 msec/div * 4 div = 20 mSec)

Also check the amplitude to be sure that it will be "loud" enough for the circuit to use -- usually close to 5V rather than in the mV range, but I'm not sure what the correct readings would be in this case.


Scott

EDIT: Derp. Accidently put "X" as schematic symbol for oscillator instead of "Y" -- corrected above.

You can find the 12 Mhz oscillator Y1 on the parts breakdown in the SW Cockpit PDF on page 72 (pg. 5-33), upper right corner, 3rd from right.

Originally Posted by dezbaz

That was great, thank you

Thanks, Dez. Always glad to help.

[dezbaz]

Originally Posted by PL1

Here's a little theory to help you picture and measure what is going on with the clocks, Joel.

In musical terms, the crystal (indicated as X1, X2, etc. on the schematic) that generates the clock signal acts like a violin string.

Thicker strings vibrate slower (lower frequency) than the thinner strings.

Shorter strings vibrate faster (higher frequency) than longer strings.

The combination of string thickness and length determines the resonant frequency of that string.

When a voltage is applied to the crystal it vibrates at its resonant frequency, just like drawing a bow across the violin string causes the string to vibrate at its resonant frequency.

The measurements you mentioned are frequencies (except for E and Q which are covered in the guide you just linked under "== Clock =="), but frequency counters can be expensive and you can measure it another way.

Since frequency and time are inverse functions, you can use an oscilloscope to measure the frequency by measuring the time each cycle takes.

Example:
50 Hz = 50 cycles/second ==> 1 cycle = 1/50th of a second = .02 Sec = 20 mSec

On the oscope, with time/div set to 5 mSec, the signal will take up 4 divisions or squares (5 msec/div * 4 div = 20 mSec)

Also check the amplitude to be sure that it will be "loud" enough for the circuit to use -- usually close to 5V rather than in the mV range, but I'm not sure what the correct readings would be in this case.


Scott

That was great, thank you