Wednesday, November 28, 2012

Something bad happens, something good happens.

All right, I've also done a few minor fixes such as attaching switches to the playfield, a right loop spinner, protection for sharp edges etc. In some of the older pictures you could see that the center scoop (or Vent-entrance) had the metal edges unprotected. Being so close to flippers and a device screaming for attention, it basically means that over time both the scoop itself and the balls will be damaged.

1) Scoop protection and the right loop spinner.
I'm thinking about making this spinner a "cash counter" in order to unlock new upgrades,
but its final feature is not yet written in stone. 
I designed a protection part for the scoop which allows me to attach rubber pads to the side without touching the inside of the scoop. This part was laser cut alongside the rest of the pieces and is visible in a previous post.





















2) The broken screw next to the left bumper and "the secret lane".
Note that I'm using super-bright LED's in the bumpers.
Hopefully they'll do the trick and require little or no maintenance, since they're not that easy to
replace once the cap and upper level is in place...

I also found a sharp edge around the bumpers that needed to be taken care of, since a stray ball could trash that area (and itself) pretty fast. I figured the easiest way and probably the only way was to fix a small mini-post just next to the edge.

But the screw was too short.
And it broke in half when I tried to unscrew it.

First I tried to unscrew it from the back, which of course didn't work. I started a mini excavation in order to grab the screw with a pair of pliers, but since there's VERY little room to maneuver that didn't go that well either. It may look like there's space in the picture, but believe me - there's not.

During night it came to me -
I'll simply screw the screw through the playfield using a pair of pliers.
Brilliant!

The only downside is that I had to remove the entire rail in that area in order to reach the screw.

But by doing so I also got my thumb out in fixing the blue rubber pad on said rail, since there was a gap between the rail, pad and ramp.

A quick bend and it looks perfect!

3) Left ramp entry with the blue rubber pad in its correct place. 

Back to the mini-post.

With the screw out, I could use a better one to securely fasten the mini post in position.
It came out great and with the bumper cap on it looks rather designed that way, actually...

4) Bumper area with attached mini-post. Also visible is the up-post, which needs to be raised slightly esthetically, but it works 100% as is. It's just outside ball-reach so there's no risk of it getting jagged at the edges or interfering with a ball so I'll leave it be for now.   

Other than that I've also applied mylar to most areas that are exposed and in need of protection.
Hopefully it will keep the machine looking fresh a little longer!

Eject! Eject! Eject!

I've finally completed the ball-through.

I really took a long ball on this one, since I started out using an old ball-through from an old Bally game called "Night Rider". What I didn't know was that this game was a single ball one. Bummer.
Once I got the parts I was thinking about different ways of getting it to work like a newer machine's ball through, thinking I'd get away cheaper than investing in a new ball through.

I had to precisely tweak the solenoid strength of the drain kicker so that the balls are shot just hard enough to reliably land on the other end of the "hill", without getting kicked back by Mr Newton on the other end. This turned out to be a bit tricky since it could work just fine, but then be too weak in case there's multiple balls waiting to "drain". This would simply block the shot and the ball would require a couple of shots. But I got it right in the end.

1) The fully loaded 4-ball through, including the roof which will further direct the balls.  
For the entire assembly to function I needed to put a kicker solenoid in an angle (visible in the previous post) so that the balls would be shot out correctly.


2) Shooter lane. Note the missing groove for the ball. This was because I didn't have all measurements at hand when it
was time to CNC the playfield. Better safe than sorry... but now I really wish I'd grooved it.

In the end, it works and functions reliably - but if it breaks I'll probably buy a new ball through instead.
To sum it up:

Cheaper, yes.
Easier, no.

Too Much POWER!

It's been a while since my last post, which is mainly because my focus has been partly elsewhere, and partly because I've been busy doing stuff on the machine.

First off -
I've received the parts from the welder and they look awesome!
From a user perspective it's almost impossible to detect any seams, which I was a little bit worried about. It turns out that 'tigging' is a technique used to melt two materials together using no external materials, meaning the weld will be 100% the same strength as the rest of the piece.

This would not have been the case if I would to pursue my blazing thing I had going there for a while...

Behold!

1) Habitrails welded together. I guess that was obvious by looking at the picture thou.
I had to modify the pieces slightly, most noticeably the VUK rail. The metal stop at the end was not wanted and had to go, and the loop had to be evened out to smooth out the trajectory of the ball.

I've also changed some of the solenoid coils since a couple of them were too weak. I may have to switch around the center VUK and top VUK since the top one has a bit too much power now.
Speaking of power...

I've replaced the autolaunch coil with a beefier one since the old one couldn't quite push the ball all the way up the ramp. I studied the manuals of the game I've shamefully borrowed the ramp from and found out that, as I already knew, I was using a too weak coil.

I did a couple of test shots and for some reason it didn't shoot the ball all the way up anyhow.
After a little detective work I noticed that the ball bounced ever so slightly against the edges of the shooter lane and lost momentum. So I tried to tweak the positions a little and all of a sudden the autolaunch simply gave up.

Tap. That's all the thunder it could bring.
Tap. I saw the kicker arm move but very weakly.
Crap. Did I just burn my new shiny coil?

I looked under the playfield and saw this...

2) Autolaunch...has become neverlaunch. Just behind it is the new ball eject solenoid, which is rather inconveniently placed. 
As you can see, the reason for it not working very well is obvious.
The link had simply cracked...

This wasn't unexpected thou since the link was from an older game that I've modified to fit. Luckily for me I bought the correct link in the last shipment of parts so it's all just a matter of pulling the roll-pin, inserting the link and reinserting the roll-pin.

Phew.

Saturday, November 10, 2012

Solenoid Madness!

Finally - the solenoids are in place and everyone of them working properly!

I've also remade the control cables for the power boards since several of the pins I planned on using for control didn't work since they were shared with other functions, such as the SD card loading etc. This is a drawback on the Chipkit-board since the actual PIC32 chip has a lot more I/O's than the board has, so some of them are shared.

I programmed a simple but effective diagnostics page as well in order to test each of the solenoids which has already proven very useful (I can't believe I originally thought I wouldn't need a diagnostics mode...) since I can properly diagnose how much power is needed for each coil, and of course - verify that they work.

Unfortunately I noticed that some of the coils are a bit too weak, while some are far too strong. So I'll probably switch around some of the solenoids to get the best power usage possible. Worst case scenario, I'll buy a couple of new ones.

Here's a short clip of the adjustments menu and, of course, the coil diagnostics page -



Feels good to see some movement from the machine after all this time!

p.s
Dramatic music is dramatic!

Friday, November 9, 2012

Fight The Power!

Yes!
Got the magnet working.

When attempting to fasten the diode I noticed that I may have inserted the ground cable in a non-correct way (i.e tightening the screws around the plastic exterior instead of the copper directly), basically creating a current loop causing all the 48V to pass through the little tiny resistor instead of through the much beefier MOSFET. I say 'may' because I'm not sure how else the magnet would work, but a theory could be that the ground cable just barely made contact causing the resistor to become the least cost path.

I've added a diode now for safety as well, and hopefully the magnet will continue to work in production. In any case, I am not touching this now.

Another thought springs to mind - I haven't extensively tested the powerboards with 48V.
I've just noticed that they in fact work and that there has been zero problems shooting around for a while. But how will they stand running game after game?

Thursday, November 8, 2012

Resistance is futile...

It was a bad resistor.

Or rather -
When voltage was applied to the control pin of the magnet MOSFET, the resistor was fried immediately. This caused the MOSFET to be always on.

I've tried moving the magnet cable around and so far I've fried two MOSFET's and four resistors.
The two regular solenoid's on the same board works perfect.

I have checked the resistance which is fine, and the magnet seems to function both ways, i.e there's no polarity or diode in there. I think, unless it's been fried as well. I'll try putting one of my spare N1001 diodes on the magnet to see if that makes any difference - that's the one difference I can see at least.

I'm running out of spare MOSFET's here...

Electric magnetism...It burns.

So I hooked up the third and final powerboard today.
Then I finished creating the third powercord mini-harness for the three extra solenoids.
Finally I turned the power on and tested the coils - first one ok, second one ok and we got smoke. Smoke?

Turns out something wasn't right with the third (actually first) MOSFET so it was in a always-on state.
This of course caused the magnet it was connected to to be always on, causing it to get EXTREMELY hot! I mean, surface of the sun hot.

And then I accidentally touched it, just to be sure it really was hot. It was.

Tomorrow's another day and then I'll troubleshoot the board to see what's wrong with it. Most probably I'll switch the faulty MOSFET and the accompanying resistor to make sure it works.
I really, really, really wish I'd tested the board before mounting the harness and the board etc...

Anyway -
Thinking the MOSFET is toast anyway, I decided to videotape its last moment on this planet doing what it should do - grabbing balls.


Edit....
Now that I think of it - the MOSFET is probably fine since it does in fact work. Maybe it's just the resistor that needs changing. Does anyone know if a broken MOSFET leads current...?

Gone International!

Found an online show called The 404 Show talking about a lot of things - and of my Bioshock pinball. Cool and interesting!
You can find the video here, and the part about me and my build is from around 4:44.

To the hosts -
Feel free to contact me in case you want to know more about the machine or anything! We'll have a virtual beer or something. ;)
Cheers!

Wednesday, November 7, 2012

It's ALIVE!

Quickly hooked up the motorized targetbank just now, and here's the first test run:



As you can see, it's not perfect -

The rotation range on the servo was supposed to be a full 180 degrees, but this particular servo only has approximately 170 degrees of motion. This means I need to manufacture a better link between the servo and the sliding bracket in order for the target bank to fully rise.

Side note - 
The clicking noise in the video is caused by one of the servo extensions hitting the frame itself. 
The excess plastic have been cut off and the problem is no longer there. 

Tuesday, November 6, 2012

Metal On Metal

I've left the habitrails for professional welding today, since my brazing didn't improve and the results were inferior. There were two major parts that needed welding, the right habitrail exit from the upper level and the transport from the center VUK onto the upper level.
Can't wait for the finished ones!

I've also gotten in touch with a a metalworker who've helped me "water-cut" the custom steel parts that I needed to have fabricated. Great guy, awesome quality and the results came out great!

1) From top left, going clockwise - Servo brackets, 3-targetbank servo bracket and supports, top VUK transport/ball lock, center VUK transport, upper level transport, top VUK support, spinner bracket, protector for center scoop... 
2) A couple of (near completed) pieces. Folding stainless steel is too easy, that's for sure!
Sure, a little touchups here and there are needed - but I think they look great!

The folding on the VUK bracket took quite some time to get right!
Now I'll have a couple of more hours in front of me bending and folding the pieces into their final shape!

A Helping Hand!

While bolting the legs on with the new plastic protectors, I realized that the bolts were only just long enough to reach into the bracket on the other side. I double checked this with a couple of pinball experts online and they all said that there were two different bolts available - and of course I got the wrong ones...

But then I received a mail from a guy called Purre who was willing to send me a complete set of bolts free of charge! Awesome stuff!

Thank you, Purre!

The bolts made all the difference as you can imagine when compared side by side with a regular bolt...
1) Top - old bolt. Compared to the new bolts there's almost 10 mm difference in length!

Finally the legs are rock solid!

Sunday, November 4, 2012

Status update

The current alpha state of the gameflow and menu system is as follows...



I've done a couple of things that doesn't show in the video, such as - fixing the tilt blob, tidying the cables, soldered an additional powerboard, prepared the upper level so it's more or less complete...

Things are going forward!