How much testing to be "full working"?

DarrenF

Well-known member
Joined
Jan 7, 2011
Messages
5,855
Reaction score
526
Location
Florida
[can't edit the title... should be "fully," not "full"]

So I only worked on my own stuff for quite a while, but now I'm starting to work on the occational PCB (game PCBs, not really monitors or power supplies) that belongs to someone else, or is intended for sale.

So where do most people stop testing and consider a game "fully working," or "working all the way," or "100%" or whatever other terms you see in the FS section?

The following list starts with things that I'm pretty sure should be done, and generally proceeds into things I expect hardly anyone does:

- hook up power and video: powers up/boots & has proper-looking video output.
- verify proper function, no errors, in test mode (if it exists)
- hook up speakers: enable attract mode audio or jumper credit/start and audio sounds good.
- hook up basic controls: able to coin up, start, and play a bit w/o any glitches/probs.
- short soak test: leave on for an hour or two to make sure it doesn't wig out or hang.
- hook up all controls, including any cocktail/P2 controls and obscure inputs (tilt, service, advance etc.) and fully test them as well.
- verify proper function of every dipswitch position.
- play test game extensively, verifying function at upper levels, or to the end if it allows continues.
 
Last edited:
I consider fully working to mean running properly with no errors or glitches and no obvious physical hacks or repairs to the board (or at least making sure they're mentioned beforehand). Sound and controls should be a given for "working." Ideally running it for a few hours, but I know people don't have the time for that. However, if it were a very expensive or rare game I would insist on it.
 
Well for me, all known weak spots are addressed.
Check for errors, run self tests.
Check all inputs and outputs.
Repair/Replace anything that screams hack.
Run a Soak test for 48-72 hours.

Even then failures are bound to happen.
 
I do everything on your list, except maybe the last two. Instead I will usually hook up the Fluke to verify the DIPs (looping read @ the DIP address while flipping the switches) and to test the RAM and ROM.

I had one board that worked fine in upright mode, but would constantly reset when switched to cocktail mode, so I make a point to always check that.
 
So true.. I recently had one the failed the cocktail mode as well. It's something you don't think about until someone puts in their cocktail and it doesn't work.

You test it and it's fine. Then they tell you it's for a cocktail.. DOH..
 
You've got to draw the line somewhere. It's no good spending longer on the testing than the repair.

Personally I wouldn't bother with DIP Switches or getting too far into the game.
 
Yes, definitely have to draw the line somewhere. That's what I'm trying to get a good feeling for... where others typically draw that line, and what buyers expectations are (and how those two differ).

Anybody test outputs such as?:
-leds / lamps (like for the Atari cone buttons)
-coin lockout coil drivers
-coin counter drivers
 
Last edited:
If I bought a board described as fully working, I expect that all inputs and outputs work (including dips and coins, lit leds, etc.). I expect it to coin up and play. I expect it to play properly (Video looks reasonable, controls all seem to work, and I don't expect that you've played it for hours on end). A little burn in testing is nice, and appreciated, but not a requirement.

I expect that all repairs are done reasonably well - if there's long wires, they should be tacked down somehow, etc. No chips hanging off of wires and flapping in the breeze.
 
Back
Top Bottom