Re: Rollergames magnet -- SOLUTION!

TheKorn <TheKorn@xxxxxxxxxxx> wrote in
news:Xns96AB6AC08FC39qqwwaass@xxxxxxxxxxxxxx:

> "Lloyd Olson" <ltg@xxxxxxxxxxxxxxx> wrote in
> news:LOadnXRP-PcvvmjfRVnyrg@xxxxxxxxxxxx:
>
>> I'd suspect a weak trace on the board. Jumper the transistor legs to
>> the other spots they go to and see if the problem clears up. LTG :)
>
> Hmmm, an angle I hadn't thought of! I'll dig in and see what I get!
> :)

So LllllllLLlLllloyd's comment got me thinking. I decided to start at
the far end of the chain and work my way backwards towards the backbox.

Here's how the circuit works... The pre-driver transistor on the sys11C
board drives the drive transistor (tip 120, or in my case, a tip 102.
:) ). Normally, you'd send this straight to a coil. But the magnet is
one mother of a coil, so they decided to send in a big boy to handle
things, a tip 36C.

The tip36C is located on an under playfield driver board. Basically, it
takes the _voltage_ from the tip 102 (not the _current_), and uses it as
a signal. In other words, there's almost no _power_ flowing from the
tip102 to the tip36C. So you don't have to have a 'great' connection
from the tip102 to the tip 36C in order for the circuit to work, just a
'good enough' connection.

Anyway, since I had already replaced the header pins and IDC connectors
on BOTH sides of the connection, barring a failing wire I knew I had a
more than a damn good enough connection between the two. But when I was
removing the under playfield driver board to check for weak traces, I
noticed something...

See, in order for the tip36C to do its job, WMS had to run three things
to that board: B+ power (i.e. +25V), the output from the tip102 (which
tells it "DON'T FLIP! FLIP!" :) ), and ground. B+ simply goes into
the board, and immediately jumps over to the hot side of the magnet
coil. (i.e. there is power at the coil at ALL TIMES). And B+ also
daisy chains on to another coil. Since I don't have any _other_ weak
coils, I knew that B+ was fine. (And somewhere along the way, I also
measured it. :) )

So that leaves two suspects, the tip102 output and ground. Well, I was
reasonably sure the tip102's output was fine since again, all points
between HERE and THERE had already been changed, and the transistors in
question (pre-driver AND driver) had already been changed.

So that left... the ground. The ground for solenoids does NOT
daisy-chain, since that's the switched part of the circuit. Again,
normally the output from the tip102 would be hooked to the other side of
the magnet. But that's being used to control the tip36C, so they needed
a 'real' ground.

The nice thing is that WMS set up to board as-if it were going to
daisy-chain the ground to another board. So what I did was I crimped a
new trifuricon pin onto some crap wire, inserted it into the molex plug,
and then temporarily hot-wired it to the ground braid. (Hey, ground is
ground. Get there any way you can! Besides, I wanted something I could
undo quickly if that didn't turn out to be the problem.)

IT WORKS!!


I played a couple of games with the glass off, pounding the PISS out of
the magnet... Basically forcing the game to give me sudden death and
then stuffing the ball over and over into the pit. (stuff one, shoot it
to miss, stuff it back in, shoot it to hit to reset the counter, repeat
until you've completed sudden death.) That's roughly 15 magnet
shots/sudden death. Didn't miss a _single_ grab!

I put up the playfield and broke out the IR thermometer. The magnet's
coil sleeve was reading a nice and toasty 145F or so! But now I was
worried that somehow the magnet wasn't turning completely off. 145F on
the _outside_ of a coil is pretty damn hot!! But I _was_ deliberately
jumping up and down on the magnet as hard as possible, far harder than
any normal person could in a real game.

So I left the game in test mode (all solenoids have power, but none
activating) for about 15 minutes. If the transistor was stuck on, or
halfway on, the coil would continue to heat, or at the very least, KEEP
its heat. Luckily, a sandwich and a Coke later the coil was back down
to roughly room temperature. (whew!)

So then it was time for the acid test. I put the "golf tee" core back
in. SheKorn and I played five games in a row, and put up some MONSTER
scores, in the process working the magnet pretty hard. Magnet
temperature? Around 98F.



I'm going to leave the temporary fix in for a few days and play the heck
out of it, but I'm pretty confident that this IS the fix I've been
looking for!




So, for future googlers, here's the steps I'd take (in the order I'd
take them!) if I were trying to fix another flaky *** magnet:

* Replace the tip36C on the under playfield board
* Replace the header pins and IDC connectors on the under playfield
board
* Replace the tip122 (Q79 IIRC) on the cpu board (use a tip102. :) )
* Replace the pre-driver transistor on the cpu board (I think it's Q78)
* Replace the special solenoid header pins and IDC connector on the CPU
board.
* Re-grind the top of the magnet core into a golf tee shape <-- this
helps SOOO much!
* Double up the ground to the under playfield driver board
* Replace the 4N25 opto couplers on the interconnect board






Man, I can't believe it's over. It's _finally_ over!
.