This weekend I decided to look into optos on Williams machines. Optos for TZ's clock are infamous, but they are also used all over the place. For registering shots going up ramps, around orbits, for flippers, drop targets, and various other "toys." I know others must have lots of knowledge in this area. Maybe we can share some of it here?
Most people reading this thread probably have read about the QVE11233.0086 opto. This is the magical "more sensitive" opto that is needed for TZ's clock. Note: I quoted "more sensitive" because that's a generalization of the part's unique behavior. I was wondering why the TZ clock needed this unique opto and whether all optos in Williams games required these characteristics (ex: flipper optos, etc.) and, if not, why?
Let's move onto the switch matrix and summarize it's behavior ever so briefly. Ignoring all the row/column behavior, which is interesting in itself, the switch matrix operates on 12v. The WPC design determine whether a switch is "active" by sensing whether the voltage on a switch line is above or below 5v. So we start off with 12v using a pull up resistor and if that voltage gets below 5v then the CPU will read a different state for the switch.
When the CPU "strobes" (check the activity level) of an opto, it connects the switch circuit to a ground path. This is what allows the switch circuit to read below 5v --- assuming the switch itself is closed.
That's my understanding at least. I welcome comments / corrections.
So what's unique about optos vs regular switches? It comes down to how much current the switch can pass. A physical switch allows a good amount of current to pass. Some optos, no so much. Remember, we are starting with 12v and need to get it down to less than 5v in order to change the state. Using the trusty V=IR (voltage = current * resistance), we can quickly see that the switch needs to support greater than 7 mA to get the voltage down below 5v. How did I calculate this? To go from 12v to 5v you need to drop 7v. There is a 1KOhm resistor inline with the 12v supply, so using I=V/R i got I=7v/1KOhm=7mA.
And 7mA is just to get down to 5v. You really want to get down further, so this could easily need to be more. 12 mA if you wanted to go to the edge.
Again, my understanding... happy to hear comments / corrections!
So what's the problem? Optos available today have specs that state they only allow a significantly less current flow than needed! While it appears to vary, 0.5 mA isn't uncommon.
What problem does this cause? Ignoring any a part may enter odd states of behavior when run out of spec and could potentially do many unwanted things, perhaps the two most obvious symptoms are:
1) When the opto activates, it doesn't allow enough current to pass, and the 12v doesn't get below 5v. The opto has "triggered" but because of the WPC because, the CPU will not sense it as it needs the voltage to go down below 5v. If you are lucky and get a part that allows a significantly higher current, it might work all the time. However, you might get a part that simply meets spec and never works in this circuit. Or, you might get a part that works sometimes and causes the WPC to read the opto togging on and off.
2) Pushing too much current through the opto can cause the opto to prematurely fail. If you are lucky to get na opto that appears to work, it may not work as long as expected.
And for the 3rd time, that's my understanding... love to hear comments.
I've come across info that implies that only the TZ clock optos need to behave like the "0086" optos but not others, such as drop targets and so forth. However, if my info above is right, I cannot see why all optos in a Williams design wouldn't require supporting these amount of current.
Which leads me wondering...
1) Do all optos in a Williams machine need to support this or is there something special about the TZ circuit?
2) The new flipper boards that exist, do their optos support these high current ratings?
3) What's a good source for optos that support Williams games (say other than Marco)?
I'd really like to hear more details on this topic. Please chime in if you think the above is factually correct or if it needs modifications, as well as if you have answers to my pondering...