Status as of 9/18/2013

Last week was spent moving pinball machines from VT to MA, and doing some ordering for Disaster.  I bought a bunch of cheap 48V power supplies off ebay.  They are low current but should be good for testing the solenoid driver card at 48V instead of the 5V/12V that I used for the initial testing.

I bought a PC power supply that is beefy enough to be used for a boost converter from 12V to 48V.  I’m pretty certain that I’m going to power most of the low voltage solenoids/lighting/motors with a PC power supply.  They are inexpensive and ubiquitous.  The one question that I have is should use the PC power supply and a boost converter to switch the 12V supply up to 48V.

My current guess is that I need approximately 15A at 48V to power the high power solenoids.  The 15A is complete conjecture on my part, and I won’t really know what the real requirements are until I take some current measurements using an oscilloscope.  (Tried to win a free oscilloscope at work today and failed.  Guess I have to continue to scour ebay looking for a good deal.)

There are currently two different designs for the 48V source.  The first design using the 12V supply from the PC power supply and then a boost converter to up the voltage to 48V.  One difficulty in the design is that an inductor with a very high saturation current is needed.  (about 100A).  The second design would use a simple full bridge and smoothing capacitor to convert the AC into DC, and then a P channel FET driven by a microcontroller to act as a voltage voltage regulator.   Best case would be finding an inexpensive 48V power supply, but I haven’t located that yet.

I bought a couple pieces of acrylic to act as a playfield surface.  I bought a small piece to make the mini-pinball machine which will prove out the driver, input and lighting cards.  The second piece of acrylic is for the Disaster playfield.  I have to go through my extra parts box and see how many things I’m missing to make the mini-pinball playfield.

Last thing is the new Taxi machine.  To say it doesn’t work is an understatement.  At this point it doesn’t boot.  I need it to be up and running by Thanksgiving.  I just ordered a bunch of basic soldering supplies, supplies for surface mount, and a bunch of general things.  Most of this stuff Mark had in his basement, and we just used his supply.  (While Pb may be a big no, no in new electronics, it still reigns supreme in the Taxi machine.)  I bought a good amount of lead solder for re-wetting connectors, etc.

Since so many things are not working on it (i.e. not booting), I need to visually inspect the whole machine and see what is the problem.  Just a quick look last night was Q73 and Q75 were toasted.  One of the pop bumpers is seized.  One of the kickers switches was always closed.  Guess it’s time to start with testing all the voltages.

Advertisements

7 responses to “Status as of 9/18/2013

  1. I could be wrong here, but I believe solenoids don’t draw more than 2-3amps at 48v. If they really pulled 15amps, you’d be at half your AC current input (720watts) from flippers alone (never mind the logic circuitry, lamps, etc. Most people can run 3-5 pinball machines off one AC circuit. Also I wouldn’t upconvert 12v up to 48v, you’ll get horrible efficiency. I’ve seen people take $20 ATX computer power supplies, and tie them together to get 48V (you have to do something funny with the ground wire). That way you get 3.3v, 5v, 12v, 24v, 36v, 48v:
    http://hildstrom.com/projects/atxbenchpowersupply/index.html

  2. yeah my sega and stersn have like 4a fuses on the flippers. dont forget solenoids fire for like 50ms ‘ON’ and in stern/williams wpc style they get pwm’d so they are only then on for like 10ms and off for 30ms type deal, and they mostly dont all fire concurrently, so you wont be drawing uber power.

  3. Here was my thinking for 15A. Flippers 2 x 3A, pop bumpers 4 x 2A, slingshots 2 x 2A. Total for all that stuff is 18A but of course most of that isn’t firing all at once so you get some statistical multiplexing of the current needed, so I said, WAG, 15A. I can’t imagine that I would need more than that. Since all of my drivers are using PWMs as BloodyCactus mentioned, it is even a little better.

    I had the same thought about tying the PC power supplies together before, but it seemed too expensive with what I had budgeted. (I also need to make a board to tie the power supplies together which would run about $10-$15 for boards/connectors). Then I remembered, that I had a 2nd line item for a PC power supply for running general illumination lighting. That makes the whole thing much more possible.

    Just a note if you happen to tie PC power supplies together. I believe that the case of the power supply is tied to ground. That means if you are putting four power supplies together to get 48V, the last power supply’s chassis (metal case around it) is at 36V. While not considered “dangerous” voltage, you can definitely feel it.

    Other bonus is I would have 24V and 36V rails that I could use to power my less aggressive solenoids. It would make the pinball machine a little more energy efficient. (Nice but certainly not a requirement).

    Thanks for all the suggestions. Now if I could only locate somebody that has a CNC router table I’d be all set. (moving from VT to MA, I lost access to the CNC machine. It wasn’t a router table, but I definitely could have made my mini-pinball using it.)

  4. I went through a similar decision making process in my build. but ultimately I chose to just go with a 50/24v (AnTek PS-5N50N24 ) dedicated supply for the solenoids.

    Even at 12v the solenoids would produce non-trivial dips across all rails in the PC switching power supplies I was using. It was quick to recover, but that’s noise I simply don’t want in the logic side.

    Maybe you’ll have better luck, but I’d do some serious testing with a scope before committing to a single supply design around a PC PSU.

    • All of the solenoids are driven on dedicated paths, so I’m not really worried about my logic levels drooping. That being said, I’m still slowly working on my “test” pinball machine to test all of my drivers. I was planning on using a BAC to try and filter out some of the quick current draws from the PC supplies if necessary.

      I’m not 100% sure that it will work, and it is tough to say until I get a good chance to fire multiple solenoids at the same time. One of the interesting parts of having a distributed system is that you can’t see that you need to fire another solenoid, so if the solenoid isn’t important (say a VUK or kicking out another ball for multiball) you could hold off firing those solenoids until the supply recovers.

      While it might be possible to do that sort of statistical multiplexing of the power supply, it seems like it would just be better to make sure that you size the power supply appropriately.

      • You should still be able to detect this in software, I assume you have some central point that acts as a control interface for the solenoids in general. A map at that level indicating solenoid groups that shouldn’t be fired together would probably do the job just fine. You’ll want to monitor flippers as well if you can afford to (A fun thing to do on old Bally pins is hit both flippers right as the outhole kicks, it will almost always fail to kick hard enough if you time it right)

      • Nope. It is a completely distributed system, so there is really no central location that you could do something like that. I could centralize the high power solenoids on a single driver card so I could do that, but it would defeat the purpose of a distributed system. I’m trying to minimize the wiring. We will see if that is possible when I make the first prototype. It would be fantastic if I did the first prototype and none of the wires were longer than 5 inches long.

        All the real times aspects of the drivers are done on the remote cards. The central controller only grabs edges from the remote drivers so it can be used for scoring and such.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s