Auction Find: Bandsaw

I bought a bunch of stuff from an estate sale auction.

The biggest item was this ~ 1976 Craftsman 12" band saw / sander.

Model 113.24350c: I think the c suffix is for Canadian market.

It has a cast iron table!  I think it suffered a fall at some point.  The bolts that retain the bottom bearing have no heads.

Luckily they are 1/4-20 pan head bolts so I can replace them easily.  Might be tricky extracting the old ones, we'll see.

I'll need to pick up the blade guard (part no. 69044) and lens (69025).

Millers Falls 9C Smooth Plane

Today I lucked out at the Re-Store (Habitat for Humanity's thrift store for home supplies) and got this little gem:

It's probably worth about $20 in good condition, so perhaps I overpaid, but they're a good cause.  I also got some clear fir trim for $2.50 so it balances out.

More info after the break!

High Scores, thanks to Logic Probotype

Solar Ride now saves high scores!

I don't need to use a post-it note on the glass any more.

I first got it going a month ago, by replacing the 5101 memory chip with a new one.  Things went fine until two weeks ago, when it reverted to its previous behavior of only accepting one credit and not remembering high scores at all.

I futzed around with the power supply, replacing capacitors in a shotgun approach to hope to luck out and find the problem, but that wasn't working out.

So I made a logic probe;

And I probed.

The answer became clear shortly thereafter - the 5101 chip wasn't being enabled by the MPU.  The enable circuitry is an oscillator (multivibrator) which, combined with a signal from the CPU, sets one of the wires on the RAM chip to logic high when the chip should be used.

The process went thus:

Pin 17 on Z22 was logic low.  That means the Z22 (5101 RAM) isn't enabling.
Examine Z1, a quad-input AND gate (4081).
Z1 pin 6 is high.  That's the CPU signal.  So, the CPU is telling us to enable the RAM.
Z1 pin 5 is low.  It's coming from pin 3 on itself
So, logically, either pin 1 or 2 is also low.  It's pin 2.
Pin 2 is wired to Z2 Pin 9, Q2.

The C2/R2 wire is fired by the CR timing from R158 and C32.  Oddly enough, when I measured the voltage drop across R158, the signal from my multimeter was enough to trigger C2/R2 and cause pin 9 to go high, enabling the RAM.

I replaced C31, C32 and Z2 and now we're getting high scores every time!

Things On Walls

This weekend I put things on the shop walls!

An air hose reel from Princess Auto!

A board! It's got the Cleco fasteners on it.

A fire extinguisher from Costco!

A puzzle

There is one mechanical design flaw with Solar Ride.

The ball can and will get stuck in the playfield, up near the shooter exit.  There's a gate back to the shooter lane with a spring-loaded switch to award 500 points when the ball passes through.  However, sometimes the ball will get caught by the wire gate instead of rolling into the playfield.

I've dreamed up an electronic / electro-mechanical solution:

"If the switch is on for more than a few seconds, activate a solenoid that changes the spring tension, forcing the ball back into the playfield."

It's so simple!  Just one sentence!

However, after some cursory research, I have discovered that delay circuitry is far from simple.  Do I use a timer IC like the NE555 to introduce a delay?  What about an op-amp and a capacitor, can I get a few seconds out of that?

I want to make this component as simple as possible, but am so far unable to come up with anything that doesn't use a timer device.

UPDATE: There ended up being a much simpler fix for this problem.  The upper mounting hole for the wire gate is fairly large and there's a lot of play.  When the ball was getting stuck, I saw that the return spring for the wire gate was rubbing up on the edge of the through hole in the playfield - changing the return force and not letting the wire bail push the ball through.

Loosening the mounting screw to move the upper post away from the binding edge reduced the problem quite well, and now it hardly ever gets stuck.

Futaba Display fixed!

Tonight I replaced the zener diode on the power supply for Solar Ride.

It is a pain to work on the power supply board - the main power transistor is soldered on through the chassis, so you have to desolder it before you can remove the board from its carrier.  All things considered, though, it went fairly smoothly.

A new RottenDog power supply would set us back $100 or more.  This fix cost $12 - and I got lots of spares, if it happens again.

But it works!  The displays no longer share numbers.  Now, I have to sort out the high scores - but for that I will probably just get a 3xAA battery holder and get some cheap NiCAD or NiMH batteries to stay in here forever.

Up to Potential

We've had a Gottlieb Solar Ride pinball for a while now, and it has had consistent problems with the displays.  They have ghost numbers - the score on the first player display will appear on the other displays as the numbers increase.

Yesterday I traced it down to a failed zener diode / resistor causing the +8V reference voltage to go away.  From the Futaba web site:

"If the filament potential is lower than the anode and grid cut off voltage, thermionic electrons can reach the anode and cause illumination of the phosphor. The filament bias voltage should be increased to prevent this problem." 

Naturally.