Black Knight Pinball Restoration - From BK to BK-Extreme


Index


Initial Restoration and Prototyping

Confident that I could fix the issues with the machine (CPU and driver board related) I started searching for parts to refurbish it and considering that the machine was free, I had a bit of extra headroom with my budget. I first picked up a CPR (Classic Playfield Reproductions) playfield set from Bay Area Amusements (CPR only had seconds left) and a new reproduction backglass from Mayfair Amusements. I managed to fix the initial problem which was actually just related to a bad battery holder (no circuit board damage though). I had to tweak a few other things here and there but was able to get it to a fully functional, fully playable machine with the original parts. It was at this time that I began replacing most of the larger original electrolytic capacitors because they are known to leak/dry out over time which reduces their effectiveness. I continued to gather parts for the restoration. It wasn't long before the machine was out-of-order again so to speak and thankfully my brother, Jeff, lent a hand in replacing sockets/connectors, installing a new set of ROMs and diagnosing a bad crystal. So now the boards were all up and running and stable.


Prototyping - Audio Enhancements

It was at this time that I started thinking of ways to improve certain aspects of the machine, especially the sound. The original machine had a single 6" woofer firing towards the floor. The voice of the Black Knight sounded very muddy and lacked clarity. The game sounds were also band limited as well. Even prior to the major modifications which I will discuss next, I added a 2" full-range speaker above the top of the playfield to enhance the clarity of the sound. It was clear that this still wasn't optimal but added much more clarity to the game and voice sounds.

It took me about 2 months to design and implement the multichannel audio system that will be inatalled in the finished machine. I had first figured out how to separate the game audio from the voice audio into two individual line-level signals. These two signals feed 8 channels of a mixer which then distrubutes and mixes those 8 channels down to 5 channels which are powered by the HT receiver (5x100 Watt ) and sent to the various speakers. I removed the existing 6" woofer and replaced it with a 12" DVC (Dual Voice Coil) subwoofer as well as a pair of 6.5" coaxial car speakers which are mounted facing down as well, located between the coin box area and the subwoofer. These two speakers carry "surround" information which is a blend of voice and game sounds processed with an onboard mixer effect (I usually select chorus/delay). The subwoofer carries the game sounds (the galloping horse really shakes the room) as well as the voice sounds processed with a Zoom 505 guitar effects pedal which produces a sub-octave tone and echo which makes the Knight's voice sound very ominous. The rest of the game sounds come through a small 2" fullrange speaker located between the bottom of the backbox and the top of the playfield and they provide a focused sound that is very crisp, enhancing all of the laser-like sound effects. Finally, a custom topper will house two 3" fullrange speakers for the unprocessed voice audio which focuses the image of the voice much higher, closer to the center of the backglass relative to the playfield sounds. Below are a few pictures of the prototype audio system, the final configuration will be arranged in a much neater fashion and I have plans to create a separate ported enclosure which will encompass the subwoofer and receiver (the current configuration tends to rattle the playfield glass and backglass if the bass is turned up too much due to the sound pressure levels created by the 12" woofer cone).





With all of that said, I always get the same reaction when I try to explain what I did and how good it sounds... they all think that I went way overboard and the only people that disagree are the ones who have played it. Personally, considering that I am an audiophile I was very surprised at how good the onboard audio data is. It covers a full spectrum from below 20 Hz (the horse gallups make the speaker act like a kick-drum which indicates a near 0 Hz, DC impulse). Also, the high frequencies go very high, probably beyond the range of a CD. I just added better speakers, better amps and a way to mix things and process certain speakers for an enhanced audio experience. It didn't hurt that I stumbled upon a used Sony 500W receiver for $100 as well as a used Behringer UB1622FX-Pro Mixer for $100 which work very well for this application (prior to these purchases I had been experimenting with many different combinations of amps, mixers, equalizers, etc. but the mixer/receiver combo is great in terms of both performance and flexibility). I can't imagine going back to the original 6" speaker now.


Prototyping - Lighting Effects

So, during my quest to improve the audio as well as add a topper to my Black Knight machine I eventually decided that I was going to add a few other modifications in an attempt to somewhat modernize the "look" of the machine without altering the "play" of the game. The next area that I would focus on would be the lighting effects. Certain effects like flasher lamps and controlled GI (general illumination) lighting didn't show up or weren't extensively employed until a few years later. Other than animated playfield lighting effects and a GI relay that could turn off all of the playfield and backglass general illumination during multiball and a few other modes, the Black Knight lacked some of the "dynamics" that new machines tend to have as the lighting interacts with the ballplay a bit more. As well as adding interactive lighting effects, I also wanted to improve both the playfield plastic lighting as well as the backglass lighting. I had realized that the 6.3VAC power supply to the GI section couldn't handle the current from the original 60 or so #44 bulbs and the voltage was dropping below 5VAC, dimming all of the bulbs. I had considered adding an extra power supply but I realized that I wasn't too happy with how the original #44 bulbs lit the backglass because it looked very "spotty" with dark sections, especially in the corners. Long story short, I picked up a 16.4' roll of warm-white LED strip (300 LEDs, three every 2", highest intensity available) and managed to use just about every inch/LED in order to light the backglass much more evenly. I also purchase a three channel dimmer and sectioned off the LED strips so that I can light them independently during multiball. This removed 39 lamps from the power supply and allowed me to add extra lamps to the playfield to improve the existing lighting as well as cover the areas in the upper right which were never lit at all (I'm pretty sure I found out why they left those lights out based on what I found above regarding the excess current draw causing the voltage to drop). There was a related issue with voltage drop along the wires between the transformer and the GI sections due to the resistance of those wires under a heavy current draw so segmenting the GI according to my design will solve that problem because each voltage feed will only have to power a few lamps. Around this time I just happened to stumble upon a great deal on NOS 7 digit displays so I replaced mine which had signs of burn-in. At the same time I extended the displays further from the circuit board in order to try to minimize the shadows created by the larger board. This completed the backbox/backglass modification.

Below is a picture which shows the comparison between an original (not mine) Black Knight backglass lit using the original thirty-nine #44 bulbs at the top revealing the "spottiness" that I spoke of. The lower picture is of my reproduction backglass lit by 291 warm white LEDs.



As you can see in the picture, the LED strips do a much better job of lighting the entire backglass more evenly although they do introduce a slight flourescent tint which I may try to adjust using tinted tape and boosting the brightness (the optimal backglass brightness shown in the picture is set below the maximum possible brightness). However, some of the differences in appearance are due to the fact that the reproduction backglass uses different paints in certain areas which allow it to illuminate much better (especially the horse/knight armor as well as the yellow/gold accents). Overall I'm very happy with the look of the reproduction backglass lit via the LED strips because it really enhances the artwork when lit (and makes for some great multi--ball strobing effects!).

Around the same time I purchased a few "Neon" 12V Cold-Cathode tubes which were designed to "strobe" to audio via a microphone. I bought a 20" red one which reacts to the voice sounds and will be located on the topper (see pic below). I bought a 20" blue one which reacts to the game sounds (low sensitivity, rear of cabinet) and a 15" ultra bright game sound CCT (high sensitivity, under cabinet). I removed the microphones and attenuated the inputs which are sent from the mixer via the headphone output (left channel is voice, right channel is game). All of these lights strobe in sync with the audio and the two different game sound CCT's react differently as well due to the sensitivity difference (rear one takes a louder sound to reach full illumination).


I might as well add at this time that I prefer to play my pinball machines in the dark so that all of the lighting effects are not drowned out by the ambient light. Instead, I have created ambient light which reacts to the machine and enhances the experience (often without realizing it because it isn't distracting at all). I have the same goal regarding the playfield and the visibility of the ball when the room is dark. I added about a dozen extra GI lights to the upper/lower playfields but these primarily serve to light the playfield plastics and not the playfield itself. For the playfield lighting I chose to use three spot-lamps (#555 bulbs) placed strategically and controlled independantly. The first one was added soon after getting the machine to play for the first time and is located above the upper right flipper and really does a great job of illuminating most of the upper playfield without obstructing your view of the ball. The other two are mounted to the left/right outlanes, just below each Magna-Save (which looks really cool when the ball is grabbed under the spotlight). These two spotlights not only light up the playfield well, but they also light up the spinner and targets. See the picture below for a rough example of the lighting provided by the extra GI (upper right) and three spot lamps. Note that the added GI lights will be better spaced when installed on the new playfield.

At this point I've established an 8 speaker sound system (12" Subwoofer, pair of Vifa 6.5" coaxial w/0.5" tweeter, Tymphany 2" fullrange, pair of HiVi B3N fullrange) mixed/processed via the Behringer UB1622FX-Pro mixer and Zoom 505 guitar effects pedal and amplified by the Sony STR-D505SE 100 watt per channel (5 channels) receiver. This sound system not only reproduces the full range of frequencies from the original game and voice sound but enhances it through equalization and processing as well as distrubute the sounds throughout various parts of the machine providing a multichannel surround sound experience from two signals (voice/game). The cold cathode neon tube lighting effects which react to the game and voice sounds enhanced the experience a bit more and while it wasn't very evident or distracting while playing, you would notice it immediately if it was turned off as if something was missing.

As for the lighting effects, the backglass was now fully lit using around 300 warm white LEDs (16'+ strip) and the playfield was well lit with additional GI bulbs as well as three spot lamps. It was time to add effects that would react to the machine as well as gameplay. One of the first effects that I added, and probably one of the coolest, was to add two strips of cool white LEDs to each side of the apron, facing towards the playfield. These LEDs are only activated when the GI relay is triggered to turn the GI lighing off which offers a huge contrast from the warm white of the incandescent bulbs. This effect looks very nice during multiball where the GI and LEDs strobe in an alternating fasion. Tilt warnings are also much more evident with a bright flash of light as all other bulbs go dark. The picture below is a bit poor but still provides a good example of this effect during attract mode.

I managed to find a way to modify the ringer such that I could choose between the normal continuously ringing mode and a mode where the ringer would only ring once (night mode). The ringer solenoid is sent a constant voltage but as the solenoid is triggered to ring the bell, the lever opens a switch which in turn opens the circuit feeding the solenoid causing it to release, close the switch again and repeat. In order to get the ringer to ring only once, I added a 50 ohm, 50 watt resistor in parallel with the solenoid switch so that once the solenoid fires, hits the bell and open the switch, there is a current loop through the 50 ohm resistor which holds the solenoid enough for it to stay energized without opening the switch again. This idea is similar to flipper solenoid circuits which use lower current coils/circuits after the flipper is held. So now that I set up that circuit and provided a switch between normal, night (single ring) and off, I figured that it would be nice if there was a visual indicator along with the bell (especially if the bell was turned off or in night mode). I found a set of 4 LED flashers with a controller unit at a local auto parts store and plan to mount it in the topper when I'm done. I used a relay to trigger this light show whenever the bell would ring for a nice dramatic effect.

Next I would begin designing and prototyping circuits for the playfield. I first determined which solenoids and playfield lamps I would utilize as triggers for playfield lighting effects. Then I determined how each trigger would do in terms of GI and additional lighting via LEDs and/or LED strips. I was able to segment the GI lighting into several sub-sections that could be turned off independently. I treated the added playfield spot lamps the same way. I then added LEDs or LED strips which would light momentarily as the associated GI lighting section would turn off, enhancing the "flash" effect.

The trickiest part of this process was figuring out how I would trigger these effects as it relates to the voltages provided by the various different triggers (ex: solenoid voltage = 28VDC, lamp voltage = 18VDC w/25% duty cycle, etc...). I was able to prototype each circuit and through the use of relays accomplished the goals mentioned in the previous paragraph. However, I did ponder the idea of using TTL (Transistor-Transistor-Logic) chips fed by the lamp/switch/solenoid matrix pre-drive circuits along with Triac driven GI sections which would be a bit more eloquent but would also open up a lot of possibilities. I finally concluded that, along with a fairly elaborate matrix of relays, I could achieve all of my intended goals with much fewer parts and less wiring in general.

For each relay trigger, I tapped off of a solenoid or playfield lamp signal and first used a series diode to isolate the driver circuit from the relay circuit. Then I placed a capacitor to ground in order to extend the length of activation (and in the case of the playfield lamps, to remove the 25% duty cycle that would normally cause the relay to trigger on/off/on/off very fast). Finally a resistor is placed before the input of the relay to ensure that the input voltage doesn't exceed the maximum voltage of the relay (12VDC/24VDC). I also place a diode across the terminals of the relay coil in reverse bias as is common practice to improve switching. While I didn't get a chance to prototype and evaluate all of these lighting effects, I did get far enough to know that all of my prototyped circuits will work.

I'll provide a rundown of the lighting effects that will be included on the finished BK-Max pinball machine. When either slingshot is triggered, four GI lamps from the respective side of the lower playfield are turned off along with the associated spotlamp while two strips of cool white LEDs light up (one under the slingshot plastic and one under the nearest side plastic). When the lower eject launches the ball, most of the lower GI lights go out while a strip of Green LEDs flash near the exit. When the upper eject launches the ball, most of the upper GI lamps go out while a strip of Green LEDs flash near the entrance to and within the ball lock. When the pop-bumper is triggered, the upper section of upper playfield lights go out, the upper flashlamp goes out and two strips of Green LEDs flash (one around the pop-bumper cap and one along the top edge of the playfield glass, the full width). Each of the ramps are highlighted when the associated "Mystery" or "2500 spinner" arrows are lit with localized LED lamps while the nearby GI lights turn off.

Next Page - Disassembly and New Parts