Monthly Archives: February 2017

02/26/2017 – 24V SMPS with Power Filter

Ooooh, look at the cool things that wordpress does.  If I add three pictures at a time, it will put them in this weird box with all three pictures at a time.

Let’s start with the picture on the bottom left.  That is what the backbox of SS3 used to look like.  You can see the four PC power supplies at the top that are all tied together to produce a mega 48V power supply.  To get that to happen I used the Max Power board to tie it all together.  It worked fantastic, but there was a lot of fan noise (I bought really cheap Diablotek power supplies hoping they would be pretty good quality, but their fans sometimes made a nasty screeching noise.  It was actually so bad, somebody at Pintastic the first year pull the power because they were worried something was going wrong.)

Top right picture is with all the extra stuff removed.  This is getting ready to install the 24V SMPS from China.

The big picture is the final installation.  The 24V SMPS is on the right hand side.  It is connected using a 6 pin Molex connector to the power filter board.  That board makes sure that the cold charging of the capacitors isn’t detected as a short and shut down the power supply.  It works fabulously.  The power filter board also has the ability for a processor to read if the power supply is powered on, can turn the power to the pinball machine on or off by grounding a pin allowing the processor to easily control the high voltage power supply, and a handy dandy LED to tell you when the big caps are charged, and can give you a shock.  The power filter board can be used for two different voltages, but in my configuration, since I only have 24V, I have tied both sides together with jumper wires.  That provides the most amount of bulk capacitance which reduces the instantaneous draw from the power supply when solenoids are fired.  It seems to work really well.

I have talked to the Mezels and they are going to have the power filter boards available for $5.  Currently they don’t have a picture up.  Fast makes a similar board for $120 which is a much better design, but for $5 for the board, plus $15 in parts, you can’t beat the OPP version.  I will eventually add information to the Pinball Makers website when I get time.  I’m busy, busy, busy right now trying to keep stuff moving towards Pintastic 2017.

I’m hoping to do another talk at Pintastic this year.  As I have said before, this is the last Pintastic talk I’m planning to give.  Hopefully it will be a fitting end to this project, and it can live on.

Back to SS3.  So right now everything is working except for the drop target reset.  Since I moved from 48V to 24V, I’ve needed to adjust the initial kicks for all of the different solenoids.  That only took an hour or two to dial each of those in.  The one problem that I have is the drop target.  I can not seem to generate enough power to reset the drop target if all the drops are down.

So when rebuilding that drop target bank originally, most of the drop targets were broken.  I bought new replacements from Steve Young from Pinball Resource, but they were not exact matches.  One issue is that they are slightly taller which forces the springs to be pulled a little bit more taught and requires more force to reset the drops.  Probably wouldn’t be an issue if it was only one or two targets, but since I replaced all seven, each little bit of extra needed force starts to add up.  I also had to replace a couple of the springs that were missing.  I used springs that were as close as possible, but it seems like they might not have been close enough.

At this point, I’m going to take another hour or so trying to adjust the springs to get the reset to work properly.  If that doesn’t work, I think that I’m going down the route of removing some of the turns from the coil to give it a little more umph.  That is a last resort, but I might need to do it to get it to work reliably.

Hopefully in the next week or two I will put up another video of SS3 running at 24V.  It seems to be playing pretty well, but without the drop targets working properly, it didn’t seem useful.

Advertisements

2/15/2017 – Weekend of Suck

Yeah, everybody likes to send out blogs that say how great things are going, and how everything is peachy fine.  Well, last weekend was not that way for me.  Everything that could have gone wrong, did go wrong, and now I’m further back than where I started.

So the big plan was to switch over the power supplies from the four PC power supplies to a single cheap 24V switched mode power supply from China.  The previous week leading up to the weekend, I had soldered everything up, set up all the connectors so I could plug things in, and I was all set to go starting on Friday night.

I plug in the 24V supply, unplug everything else, and start testing.  Flippers work perfectly.  Drop target doesn’t reset, but that is expected because I haven’t changed the timings on the pulses.  Kickout hole, same thing.  Everything is looking great…but here is where things go south.

I think to myself…hey, I have a 36V power supply sitting here.  Why not toss that in and see what happens.  Instead of slowly building things up from a known base, and actually having the machine run on the original 24V that the machine used, I decided that I would add 50% more voltage and with ohms law, yes, 50% more current.  I switch to that power supply hit the flippers a couple of times, and then, all of a sudden, one of the flippers locks on.  Ahhhhhhh.  I pull the power quickly.

OK, guess I will simply desolder that MOSFET and put a new one in.  I “guess” at which MOSFET that I need to replace, and guess incorrectly.  That means I just spent an hour replacing a MOSFET that didn’t need to be replaced, and the board still doesn’t work because there is still one MOSFET that needs to be replaced.  Why didn’t I bother to pull out a schematic?  (Yeah, I knew that it was one of two MOSFETs on that side of the connector, but damn it, I guessed the wrong one!)

After that second failure, I was just mad at myself, mad at pinball, just simply in a bad mood and I couldn’t work on pinball for the rest of the weekend.  I just needed to take a small break from it.

Here is the thing that really irks me.  I knew that throwing 36V through the flippers caused too much current for the MOSFET.  I knew that from when I first brought up the playfield, assumed it was 43V, and only later looked at the manual and found out that it originally ran at 24V.  If I would have simply taken a step back, and thought about what I was doing and not rushed into it, I would have been OK, and could have saved myself a couple hours from replacing the parts.  That’s what angered me the most.  That I just didn’t take the time.  I rushed a decision, and now I am paying for it with extra work.  It teaches me for being a dope.

Luckily there was a ton of snow this weekend, so I could cool off by going out and building a large igloo in the back yard.  Come to think of it, I’m still angry at myself.

2/4/2017 – Automated OPP firmware updates

Previously, to update multiple OPP cards attached together in a chain, each card would need to be updated one at a time.  To do that, a ribbon cable connector would need to be unplugged and a jumper would need to be placed to make it so an individual card could be addressed.  While it only takes about 5 minutes to update four cards, it is a pain in the rump to switch around the connectors.  (Especially when the playfield is installed in a machine).

In the repository, version 0.2.0.1 supports a new RS232 command called upgrade other cards.  That allows the card to pass through all RS232 communication so another card can be updated.  What does that mean?  After updating to this version of the firmware, all future updates of firmware can be done without moving cables.  Yeah!

Here is the new command:  c:\Python27\python.exe Gen2Test.py -port=COM5 -upgrade

You may notice that the version is 0.2.0.1.  I use the following method to name code releases.  Major.minor.engineering.test.  So for this version of code, the major release = 0, minor release = 2, engineering release = 0, and test release = 1.  The important thing to note here is that the test release is non-zero which means that I haven’t done a full regression test on this version of firmware.  After that testing is done, I will change some of the first three numbers, and the test number will go to 0.  Basically, if the last number is 0, a full set of regression tests have been run.  If non-zero, development tests have been run to prove out the new feature, but the validity of the code can’t be guaranteed.

So why didn’t I do a full regression test?  Well it takes me approximately 1/2 hour to run the tests, and it involves using test fixtures which I have to set up.  Since, said test fixtures normally live in a box, out of the way, I don’t want to take the time to run those tests now.  I also know that there are a bunch of new features that I’m going to be adding in the next few weeks.  I will simply hold off running the regression tests until all of the immediate changes are implemented.

The next version of code will not be very interesting to many people because it will be supporting the high side incandescent driver boards.  (I believe I am the only one running those boards at this moment.  Right now, when the Pinball Framework instructs them to turn on, they turn off, and vice-versa).  That is the code needed for SS3 to fix its light displays.