The Data East Chase rope lights used in two Data East Pinball machines (Star Trek: 25th) and Hook) are a huge problem on the game. So much so; that my Ebay-purchased game came with no rope lights at all. 🙁 This blog series walks thru my intention of recreating the chase rope lights – but made out of LEDs so they use less current, generate less heat, and can be used for decades without any issues.
My game has no chase lights to repair; so I’m going to have to replicate the chase lights. I decided that I didn’t want to try and re-configure a standard off the shelf rope light because they likely don’t support doing a +12VDC common with three ground leads as discussed in Patofnaud’s post #1. Additionally; I don’t have exact specs of the rope light w/ regards to light spacing, diameter of the rope, ect. Since I’m going to have to recreate the lamps; I have decided to do a conversion to LEDs… specifically, using some “fairy light LEDs” which have recently become available on the market. I started by going to Hobby Lobby and buying a set of their battery operated lights using a 40% off coupon. Amazon has a whole bunch of alternate versions of this product… so you might be better off looking at Amazon if you don’t have a Hobby Lobby near you. B072NH2FQ1 seems to be a nearly identical match to what I got from Hobby Lobby.
These strings appear to be made of discreet 603 Warm White LEDs soldered to a common anode and common cathode. The LEDs are encased in a hot-glue like product to protect the LEDs from mechanical stress and help with water proofing. The fact that all LEDs are in parallel might end up being an issue because you can’t control the current into each LED. LEDs wired this way means that any variance in forward voltage drop (Vf) between the diodes in a series may mean that LEDs with low Vf would get more current that LEDs with higher Vf and could lead to premature failure of those leds. We’ll see long term if this becomes an issue with this project.
Taking apart the battery pack of the fairy LED light uncovered what I thought… A power switch and a 1/4W series current resistor of 15ohms. 3AA batteries supply 4.5V which is current limited by that 15ohm resistor. I did some quick measurements of the entire string and IIRC; the whole string took around 113mA with the 15ohm current limiter. I don’t recall what the series Vf was for the whole string.
Because the LEDs are powered with a 4.5V source; it becomes rather obvious that I can’t use these as a drop in for the 12V incandescent rope light originally in the machine. I’m going to need a conditioning circuit which will drop the 12V down to ~4.5V and provide some current limiting. I started by measuring my machine’s ramp to determine the approximate length of the chase lights. I measured with a piece of string to be about 15inches. The fairy LED lights have spacing of about 3inches which means that my largest segment would be about 15/3 = 6 leds, max. These 6 will become important for the series resistor calculations in the future.
Before I got to far into designing a voltage level shifter and current limiter; I looked at my machine. I couldn’t find the chase light connector shown in post #6 of pat’s thread. I couldn’t find it because the previous owner connected the left and right .156 connectors together with a Z-connector. That may mean there is now an issue with the chase light board (DE #520-5054-00 or #520-5054-01) in my backbox but I’m thinking connecting them together did no future damage since they are wired together in a OR configuration in the schematic manual. Meaning OUT1A is tied to OUT2A via the Orange/Black wire in the cabinet (See post #7, picture 1). I decided that Zconnector is a good place to put my Incandescent to LED converter circuit. My plan is to replace the Zconnector with a board which does the conditioning.
I decided I was going to abuse a LM1117 style +5V regulator to voltage shift the +12V common down to +5V common then use a series resistor to current limit for the LEDs. Since +12V is the common; I’d need to do a wired-OR configuration with a set of fast Recovery SCHOTTKY diodes to the OUT connections on the chase light board. I have no idea if the LM1117 vreg can operate properly with a v+ common connection; but I suspect it will have no issue given the frequency a which the lamps operate. Additionally, given the previous incandescent lamps had issues blowing the NPN Darlington arrays on the lamp chase board; I figured putting in a PTC resettable fuse would be a nice addition. My circuit took shape in eaglecad using Digikey as a part reference.
Here’s my original Fab A circuit, keeping in mind this is untested – but I retain all rights to this circuit for the moment:
Some theory of operation:
D1&D2 provide some polarity protection – for paranoia. D3-D8 are the Schottky diodes which provide wired-OR back from the ground of the +5V regulator (U1) back to the chase light board’s OUT* connections.
C1, C2, C3 provide some filtering for the +5Vreg to help it maintain stability with the wired-OR configuration. D9 provides some additional protection for the Vreg – probably not needed; but extra insurance. R2 – R7 provide the series current-limiting resistors for the various LED strings. Why the different values of 13.3 vs 16.9? Well; my quick napkin calculations shows that there may be 5 or 6LEDs on the longest string and then one less on the shorter strings. I plan on connecting the longest string on the 1A & 1B lines while the shorter strings (with the higher resistors) will be on the others. F1-F6 are the resettable PTC fuses. The bidirectional LEDs at the bottom are monitoring the output of the chase light board to give me an indication the chase board is working properly. They will chase green if working properly. I basically created a small “Z-connector” board to condition these LEDs but also added some small connector boards to help me interface from the LED strings to a simple 4 pin 0.1″ pitch latchable connector. The idea is these smaller connector boards would be fixed to the end of the “rope lights” and allow for a quick connection.
A little background on the Series resistor calculations. Basically; Vf isn’t known for the parallel LED strings. You can’t measure Vf with the diode setting of a Digital volt meter which btw; my favorite is a Fluke 87 series meter. Vf according to my DVM is around 2.4V and we know that no current white leds operate at 2.4V for 20mA. Since I don’t know which LED the chinese put in these strips; I had to make an educated guess. To do this I went to digikey.com and drilled down on warm white 603 LEDs, downloaded the table (using the button near the bottom of the page) and then imported the data into excel so I could get an average of the Voltage – Forward (Vf) (Typ) column. I calculated Vf ~= 3.23V. Armed with Vf, and the knowledge I wanted to operate the LEDs near their 20mA operational current so they would be their brightest I was able to calculate a theoretical series resistor. Here are the “knowns”:
Vf = 3.23.
Each LED should operate at ~18mA.
With the ground diode in the Vreg path… assume Vfdiode = 0.4V. This would cause Vout of Vreg to be 5+0.4V, or 5.4V.
“ground” will be thru the UL2003A darlington on the chase driver board. VCEsat = 0.9 min.
Assume 5 leds needed for longest string and solve using Ohms Law. R=V/I.
V = 5.4 – Vfled – VCEsat
I = 18mA * numOfLEDs in string or 18mA * 5
R = (5.4 – 3.23 – 0.9) / (5 * 18mA ) = 14.11 ohms
Subtract off typical Fuse resistance of 0.9ohms and you arrive at ~13.3ohms. Repeat the calculation for a 4 led string and you arrive at ~16.9ohms.
The only “gotcha” with this theoretical calculation is that what happens to the LED if we are drawing near 18mA? There’s no real ground plane or PCB to draw heat away from the LED. The only “heat sink” is the hot glue used to encase the LED and the wire LEDs connecting the LEDs. I’m hoping the low duty cycle of the LEDs will help keep thermal runaway in check. This is something I’ll have to watch in the final assembly. If LEDs start dying… we’ll know it’s either getting too hot… or the parallel LED VF vs current is a problem.
Here’s the Fab A boards as committed to OSHPark. This is their render of the boards as I don’t have them back from fabrication yet:
I’ll post more on the circuit boards in Part 2 when I get them back from OSH Park and have them built. We’ll see if my little experiment bears fruit.
Now onto the actual chase lights themselves. I thought about using polyurethane blue tubing from Granger.com as I posted in Pat’s Pinside thread on post #19. But honestly, I don’t really feel like the blue rope light fits the ST:25th theme very well. To me; it looks like some attention getting feature to draw in the eyes of a would be quarter-dropper in an Arcade. Not that is a bad thing; I just figure since I’m not going to an have original rope light assembly with the proper Light spacing… I might as well try to make it fit the theme a little better. If not blue, then what? Well Clear is definitely an option… but it won’t really hide the led wiring very well. I had a lot of extra 1/2″OD, 1/4″ID rigid Acrylic tubing left over from the guide plastics from Star Trek: The Mirror Universe custom pinball project. I was thinking of doing that with some custom Acrylic etches on the tubing. The problem is that Techshop.ws failed due to piss-poor-Management so I can’t really go there to use their rotary attachment on the Tortec. My new-to-me Epilog Laser doesn’t have a rotary Attachment, yet. Not sure what I’m going to do about that yet… Even if I get a Rotary… what would I put on the Tubing? Then there’s the issue of bending it properly… which shouldn’t be too hard given the Youtube videos for Hard Tubing in water cooled case tutorials.
Then I stumbled across ENT Corp on Ebay who seems to have colored versions of this rigid tubing available in cut-to-order. I went ahead and ordered 6 pieces of the smoke acrylic tubing custom cut to 15.25inches. After I experiment with the etching and try my hand at the clear tube bending… I’ll proably finalize the chase lights using the smoke tubing on the final machine.
That’s it for Part 1 of this blog series… I’ll start Part 2 when I get some tubing experiments done and/or when I get the PCBs back from OSHPark. For now, Peace and Long life…