Atari AR3 Paperboy - arggghh!

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
This follows on from a thread jengineer started a few weeks ago about a medium res monitor in a paperboy cab.

He couldn't get the monitor to work but could hear the game play blind etc.

Long story short, two smpsu's had been fitted, supplying the CPU board and video board separately with +5v.

The game was playing blind because the +15v rail to the video board wasn't present, the AR3 was sending 15v to the CPU board but not the video board.

I only know this is the case due to making a bench loom recently for Paperboy and getting caught out with this issue.

Btw, +12v is just not enough, it'll start but weird things happen.

So, i've set about getting the AR3 back in the loop as it should be, i've tidied up the connector to the video board for power, everything nice and snug. I have +15v no problem.

However, i can't get 5v out of the AR3 when loaded.

Unloaded i can get 5v and its adjustable, loaded i get 2.9v and not adjustable.

I've changed Q7/Q8 (the 2n3055's), changed the LM305, tested the TIP32 out of circuit etc, checked every resistor and capacitor and diode.

Originally one of the 2n3055's was getting screaming hot within seconds, i reseated it by removing and resoldering and that stopped, they are getting warm but apparently thats normal but not where a smell was coming out.

I have +15v at its test point and the positive end of C19.
I have 14.6v on the 10.3v line but thats from the brick.
I get 21.6v on the +22v line
I have 14.6 on the collector of Q7/8

I have no control of +5v adjustment when the video board is connected, but have control without.

Someone has attempted a sense mod on J7. Pins 2/3 are linked and 5/8 which looks right but not sure if necessary for an AR3.

I don't have CR8populated - a 1n4002 - is this crucial.

I can change the TIP32 but i don't believe it'll fix it.

The only things i can think of are:

1, the board is overloading the AR3 and causing voltage drop
2, there is a short on the board.
3, the AR3 is still faulty.

Anyone had this issue before?
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Dunno, going to try it today.

The way the power rail is constructed on an AR3 is that you have 10v fighting with the 5v, normally, unloaded, i'd expect to see 10v or more but i'm getting a switchmode type output where you can preset 5v without a load.

The pull-up i think is there to compensate for load, and on an AR3 its got two 2n3055 circuits to provide a higher capacity.

Btw, the board in its original configuration would reset while playing, this was likely due to the video board resetting without 15v and also the psu's had gnd and 5v bridged to each other so uts likely the smpsu's were resetting.

The cpu board waits for the video board to be ready before allowing the T11 to be released.

As the 15v return is shared with the 5v return on the AR3, its not a good idea to supply 5v from one supply and 15v from another because the video board commons all the returns.

The video and cpu board are power isolated as Atari designed the game to need two supplies for 5v.

Note the 15v rail going to the cpu board has its own gnd going back to the ar3 (+WRGND) don't know what the designation is but its isolated on the pcb from the 5v gnd on the cpu board.

The cpu board also requires -15v and again its ground is via +WRGND.

I'll try the other video board, i'm assured its working, i need to know how to construct a proper dummy load for an ar3 for the 5v rail.

Talking all this through, its possible something's not connected to the 5v rail that should be.

Look here at the wiring and ar3 scheme:

http://arcarc.xmission.com/PDF_Arcade_Atari_Kee/Paperboy/Paperboy_SP-275_1st_Printing.pdf
 

Alpha1

Do the Shake and 'VAC
Staff member
vacBacker
Feedback
95 (99%)
Credits
5,653CR
I wonder why two smpsus were fitted to give the boards +5. One should be enough.

See if the PCB set itself is dragging the +5
 

jengineer

Senior Member
vacBacker
Feedback
49 (100%)
Credits
1,205CR
ok James, that all went a bit over my head to be honest but it all sounds bloody good stuff!

I think when the power supply mod was done it was working (a bit hacked and bodged up though)

gareth said the 2nd board set did the same thing, plays blind

so going back to your first post I'd say the AR3 is still faulty (can you swap it with your other paperboy AR3 ? )

or both board sets have the same fault , a short on both boards? possibly caused by the bodged psu set up?
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
I think the set is just super thirsty, i had to split it on a bench rig and use two supplies, both would give 15a apparently on the +5 line but under load of everything on one, the voltage drop was phenominal to about 4.6v, too low for reliable boot, plus 12v doesn't run the gfx properly and you need -15 as well.

We curse atari but it was the parts available at the time.

Also the returns issue above prevent running the odd voltages separately, they are isolated on the cpu board but not the video board.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
The other paperboy doesn't have an AR3, it has the reduced size audio amp found in some gauntlet cabs, i'd have to look at its specs but the other cab is at the unit and gareths is at mine.

I'll not be at the unit for about 2 weeks now due to work and personal commitments so i'm doing as much as possible at home.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Yeah, i could rob one from the roadblasters cab to try i guess.

Gotta be something simple i'm missing, i just don't fully understand what the circuit is doing with those transistors.

We need a modern replacement.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
I've proven its the AR3 at fault, changed boards out and still the same problem on a known working set, used bench rig and boards all come up.

I added a 1n4007 at CR8, my next and probably last ditch attempt at this is to reflow around all the 5W resistors caps etc and change the TIP32A.

I can only see this as a feedback to the boost circuit where the issue can be - a transistor that tests as fine but is failing under load, i've seen that before.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
No joy, gonna have to park this for a bit.

Summary of things done:

TIP32A @ Q6 changed

2N3055 @ Q7 and Q8 changed

LM305 @ Q5 changed

Inserted 1N4007 at CR8 - previously missing

All resistors seem to test ok

Reflowed around R23,25,26,27

All electrolytics check out ok

36vAC coming in at J6 7/8, +/-21 at the correct pads

14vDC coming in at J6 4,5,6 instead of 10.3

Totally stuck now, anybody repair these? I'm out of ideas.

Anyone who lives locally got a working AR3 they could bring over I could test?
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Yeah, ar2 are fairly plentiful, well, easier to come by at least.

If i understood what the circuit was doing then i'd be most of the way there to fixing it.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Ol, i would sincerely appreciate a loan of a worker to test and compare.

I had an email exchange with a guru yesterday, i might revert to the switcher setup using a single 15A psu, however, we are going to bring a proper load tester to this ar3 and see whats going on.

I am just getting info about drawing the +/-15 properly.

I want to get the AR3 fixed now because i've invested time and want to solve the mystery...
 

david_orton_2000

Green Mario
vacBacker
Feedback
8 (100%)
Credits
344CR
Hey James,

Remember some outputs are unregulated, if I remember correctly, the 10.3v one is:-
Quoting you:-
"14vDC coming in at J6 4,5,6 instead of 10.3"

Paperboy would awesome to get going...it was so popular went it was in the Arcade.
Most times people were queuing, and the other times the handlebars were broke! :)

Cheers,
DaveO.
 

Ordyne

Active member
vacBacker
Feedback
5 (100%)
Credits
1,282CR
Sounds to me like your chasing your tail, get a dummy load on it and test it..

I've had similar faults with system 2 which have been all PCB related.. Not that the ARIII couldn't be the culprit.

I've got a super sprint which was running but blowing fuses after a few mins or just booting up with garbage on screen, some how I figured out it would only blow fuses if the coin door was connected. In the end this was due to a short in the audio section of the game board!.. It would run fine with no sound with the coin door lamps disconnected tho! Oh how I laughed.... NOT :(

I'm not getting your theory on the grounds or the load so I cant help there.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Thanks Craig, i've parked this for now until I can get a dummy load on it and we can ramp the load up manually to see when it gives up. If it turns out to be the board then thats a whole different ball game.

As for the grounds, i was worried that the +/-15 rail on the ar3 doesn't have a separated ground from the +5 so if i used a switcher to supply the +5 thhen the ground for that would be on the switcher and i'd be using the gnd on the ar3 for the 15v but on the board all the gnd paths were connected together.

I was concerned that this wasn't electrically correct as the switcher was generating one ground reference and the ar3 another. I realise they are all DC so the gnd should be at 0v but having never done this before I wasn't sure.

I have a pole position upright with two ar2's - when I got it and the previous owner gave me a tour of the innards while swapping in a newly fixed board he said, whatever you do, don't cross the streams with the 5v rails meaning don't plug one 5v from one ar2 into a board already being supplied by another (rhis was the usual pp bullet proof mod to use extra power wires). I assumed thus equally apllied to the ground returns.

I'm cautious not to connect anything in a way that could cause further trouble./damage.

It might seem trivial to the experienced but i am still a learner - a fast learner but a learner all the same.
 
Top