STD 'chaser' lights


    Every year, not long after production of the new model started, it was time to start planning the next year's machine. Starting with the USC1 jukebox, Seeburg employed a consulting industrial designer named Bob O'Neil to develop new cabinet designs. Usually a cabinet design lasted between two and four model years, with facelifts every year to keep the design fresh and to discourage third parties from offering dress-up kits to make last year's machine look like the current model. The trick was to make the follow-on year designs look different while at the same time changing as little as possible to save tooling money. Bob was quite good at this, generally delivering three or four artist's concepts each year for management approval. One or two of these would be chosen for further development, and Bob would come back in a few weeks with a small scale three-dimensional model of it, made out of foam core material. The 1974 model year STD160 Vogue II introduced a brand new cabinet design, replacing the SPS2 Matador whose basic cabinet only ran two years; it was expensive to build due to the large number of castings used. The STD160 design featured simple shapes which were easy to form in sheet metal as were the rounded cabinet corners.  Back-lighted graphics panels were used for the front of the cabinet, held together with extrusions made from aluminum with a bright anodized finish. Extrusions are very cheap, made by forcing semi-molten aluminum through a flat die with the desired profile cut into it. The dies are very cheap to make. The only expensive tooling was that for the rounded corner 'extrusion' pieces, and that was only moderately expensive. These were not actually extrusions, since that process only works for straight sections. Rather, they were small cast aluminum pieces with the same grooves on the back as the extrusions, and the same anodized finish to match the other pieces. The grooves hold sheet metal screws, used to fasten the assembly together.

    For 1975, Bob came up with an infinity mirror effect for the cabinet front, using the exact same shape as had been used in the STD160. To do this, the graphics panels were replaced with semi-transparent glass. Behind this was a U-shaped channel made of mirror-finish sheet metal, with a row of small light bulbs positioned in one wall of the channel. The lamps were the same as those used in the FC1, a type 2182 grain-of-wheat bulb. The lights, semi-transparent glass, and mirror channel created the infinity effect, with multiple reflections of the lamps repeating off into the distance. He wanted it animated, and I was assigned the task of coming up with something that would control the animation. This became the 5BS1 or '5 Bit Sequencer'. At first glance, it appeared to require a lot of circuitry to drive 60 or so lamps (the STD2 used 60, 58 for the STD3, and 59 for the STD4), but then it dawned on me that a fewer number would work if the shortened string repeated. I settled on a 5-lamp string, since there was a digital logic chip available that had five outputs. This works fine if the string is organized like this: 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, etc. The machine's wiring was designed so that there was a single off lamp traveling through a field of five lamps, starting at the left bottom and working its way up and to the right across the lower front panel, then up and to the left across the upper panel.

    The design of the digital portion of the controller was also pretty simple, using a serial-in parallel-out shift register chip. The only even semi-tricky part was to insure that no matter what state the part came up in when power was applied, it would settle into the one-off/four-on sequence quickly. I put together a state diagram, and verified that all 32 possible states that the chip could come up in would result in the desired sequence, after five or six clock cycles. The clock was generated by a 555 timer chip, with a control that permitted speed adjustment. With hindsight, I think the adjustment was unnecessary. All machines I've ever seen run the display at the same rate.

Below is a scanned schematic of the controller.


    Here's how the logic works. Assume that timing capacitor C3 is discharged. When power is applied, it starts to charge through R24 and R1, the 'Flash Rate' control, and R13 to the +5 VDC power supply. When the voltage across C3 reaches about +3.3 VDC, the Q13's Out pin will switch to about 0.8 VDC (logical 0). At the same time, Q13's D pin will turn on, connecting the junction of R24 and R1 to ground. C3 now discharges through R24, until the voltage across C3 reaches about 1.6 VDC. Now, Q13's Out pin switches to about 2.4 VDC (logical 1), and the D pin turns off to allow C3 to charge again. This continues as long as power is applied. Q13's Out switching between logical 0 and logical 1 is a clock, and each low-to-high transition of this clock shifts the value on Q12's S input one position to the right through Q12 outputs Qa, Qb, Qc, Qd, and QR (note that QR is actually mislabeled in the schematic; it should be Qe). So if the S pin is high at the low-to-high transition, Qa will go high. Otherwise, it will go low. Qb will have whatever Qa was last time, etc. This is how the data (pin S) is shifted through each of Q12's five outputs. Note that the first four outputs of Q12 are connected through diodes CR5 through CR8 and R26 to the base of Q11. Whenever any of Qa, Qb, Qc, or Qd of Q12 is high, Q11 is on, so there is a low applied to Q12's S input. When all four outputs of Q12 are low, Q11 is off so a high is applied to Q12's S input. In other words, a logical 1 (high) is 'walked' through the outputs of Q12.

    Each Q12 output is connected through a 470 Ohm resistor to the base of a transistor. This transistor is used to invert the logical 1 output by Q12, converting it to a logical 0 for the following power transistor. These are medium power transistors, mounted to the chassis of the 5BS1, and are the same as those used in the SAS2 Auto-Speed Unit and the +27 VDC power supply of the machine's Digital Control Center. The chassis helps to dissipate the heat generated by the power transistors. Whenever a logical 0 is applied to the base of one of the power transistors, it turns off, turning off the lamps in the string connected to it. Recall that the other four power transistors are on. Since all five inverter and power transistor circuits are identical, I will describe the operation of only one, Q6/Q1. When Q6 is off (which is most of the time since there will be a logical 0 output on Q12 Qa most of the time), Q1 is on. This is due to the base current supplied by R14 (68 Ohms), connected to the 9.6 VDC power supply on J2 pin 3. Base current flows from the supply, through R14, Q1 base/emitter junction, and R3 to ground. When Q1 is on, current flows from the +9.6 VDC power supply, through J1 pin 2 to one side of all the bulbs in the string. Current flows through the bulbs in string #1 (12 type 2182 bulbs connected in parallel for the STD2), and enters the 5BS1 on J1 pin 6. From there, it flows through the Q1 collector/emitter junction, and finally through R3 to ground. R3 is used as a current sensing resistor. This resistor operates together with the voltage divider made up of R14 and R28 to act as a 'safety valve' for the circuit, protecting against shorted bulbs or wiring. The voltage divider provides a fixed DC voltage to the base of Q1 when Q6 is off and Q1 is on. An increase in current flow through Q1 above safe levels could destroy the transistor. When a shorted bulb occurs, the excessive current will cause an increased voltage drop across R3, .reducing the base/emitter voltage drop of Q1, so it starts to turn off. An equilibrium condition will be reached where the transistor is partially on and generating lots of heat which is dissipated through the chassis and jukebox cabinet, but the current through the transistor is limited to a safe value so the transistor will not self-destruct. While Q1 is off, R22 keeps the bulb filaments warm by passing a small amount of current around Q1. This helps to insure that Q1 will not immediately current-limit due to cold bulb filaments. Q6 is used since Q12 cannot provide enough current to turn on Q1.

    For the following year, Bob wanted something in addition to the animation. I suggested that we vary the lamp brightness at the same time as doing the animation, and he asked what it would look like. I came up with a variation on the 5BS1 using another 555 chip and additional digital logic to vary the on/off duty cycle of the lamps to alternately ramp up and down their brightness. Nobody liked the result since it looked as though the machine was 'breathing', so the idea was dropped and the unmodified 5BS1 continued in production for the STD3 and STD4.

    The only part of the 5BS1 left to explain is the power supply. The STD2 used an LPS3 power transformer; the STD3 and STD4 used the LPS4 instead. A scanned schematic of an LPS3 is below. Thanks to Ron Rich for providing the copy.



The schematic shows Line (Mains) connections for both 120 VAC 50/60 Hz and 240 VAC 50 Hz. Notice that the fuse rating and primary winding connections to T2030 are different between the versions.  Note from the 5BS1 schematic above that P1 pins 3 and 4 are connected together. When the 5BS1 is plugged into the LPS3, this connects the blue secondary wire to the Red/Yel wire, connecting the two T2030 secondary windings in series. Here, pins 2 (Grn) and 5 (Grn/Yel) are not used. As originally designed, the 5BS1 was to be an optional accessory, so most STD2s were to leave the factory without it. Jumpers on the lamp string connector (that plugs into 5BS1 J1) connect pins 3 and 6 together, and pins 2 and 5 together. When the lamp string is plugged directly into the LPS3, this puts the two secondary windings of T2030 in parallel, with pins 1 (Red) and 4 (Red/Yel) not used. This applies a lower AC voltage to the lamps, making them glow at about the same brightness as the DC switched voltage when using the 5BS1. In the event that the 5BS1 develops a problem, the lamp strings can be plugged directly into the LPS3, so that the 5BS1 can be taken back to the shop for repair. I believe all STD2s left the factory with the 5BS1, and few developed problems, making the extra secondary winding of the LPS3 unnecessary.

    For the STD3 and STD4, the 5BS1 was made a standard part of the machine, and the LPS3 was replaced by the LPS4, which was cheaper to manufacture. A scanned schematic and exterior view of the LPS4 is below.


   As you can see, the secondary wiring of the LPS4 is quite a bit simpler. Externally, the LPS3 and LPS4 are identical, except for the silkscreen legend.

    For the STD2, the cabinet engineers came up with a scheme whereby the lamps were installed in a plastic lamp holder, and held in place by an insulation displacement terminal which also served to connect the lamp to the wiring string. This was quite inexpensive in terms of parts, but required a fair amount of labor to assemble; and it was difficult to repair in the field. Replacing chaser lamps on an STD2 can severely try one's patience. For bulb replacement, see the drawing below, which is a scan of a Figure in the STD2 Installation & Operation Manual.


    What follows is how Seeburg suggested lamps be replaced: 'For replacement of these lamps, open the upper lid and/or the front decorative panel. Turn off phonograph power, or unplug lamp assembly to be worked on. Lift lamp holder sections and slide out from under retainer springs to gain access to defective lamps, see the Figure above. Remove the terminals on each side of the defective lamp using pliers; care must be taken to prevent breaking the insulated wire, see the Figure above. Remove the defective lamp and replace with the new lamp. Leads of the  new lamp must come out thru slots on the side of a tube that sits in and projects down into the terminal holes. Press terminals back into holes to make connection to the lamp leads. Be sure that lamp lead does not slip into slot in round portion of the terminal resulting in poor connection. Be sure that terminals are completely down into holes so that short circuits to lamp box will not occur. Re-assemble lamp holder sections by sliding angled edge under retainer springs and lining up lamps with holes in the reflector boxes.' Here's a couple of tips:  Use Chicago Miniature 2182D bulbs. The leads are thicker and will make better contact, and make SURE that none of the wires on the bulb or at the ends of the "tray" can short out.

    After many complaints by the operators and the folks on the assembly line, printed circuit boards replaced the lamp holders and insulation-displacement connectors for the STD3 and STD4 machines. The lamps are held in small plastic cups with a slot for each lamp lead. The cup fits into a hole in the printed circuit board, as do the lamp leads which are then soldered to the board. Each board has a post connector for each lamp string circuit plus the common. Push-on wires are used to interconnect boards, and to connect them to the 5BS1. To replace a lamp, unplug the LPS4 from the power junction, and remove the wires from the posts on the printed circuit board. Then unscrew the board from the mounting frame to get access to the bulbs, unsolder both bulb leads, remove the bulb and the plastic cup. Dress the leads of the new bulb into the slots in the cup and push the pin on the end of the cup into the hole in the board, along with the leads from the lamp. Solder both leads in place. Install the board back into the mounting frame and plug the wires back in. Make sure you match the wire color with the etched legend on the printed circuit board.