The LaserActive That Almost Wasn't

Not the most pristine of specimens but this LaserActive has seen battle and lived to tell the tale.

At any given moment I probably have three or four "projects" sitting on the back burner, either because I bought parts but by the time they arrived, I had lost my motivation, or because life happened and hobbies had to make way for more important priorities. 

One such back catalog project was restoring my LaserActive PAC-S10 module closer to its original function.

Just a quick primer on LaserActive in case this is the first place you're reading about it. Pioneer partnered with both Sega and NEC to integrate game consoles into a Laserdisc player. The result was a player with an expansion slot for various modules. To my knowledge only five modules were released for it. The PAC-S1 and PAC-S10 are the Japanese and North American versions of the Sega Genesis. The PAC-N1 and PAC-N10 are the Japanese and North American versions of the Turbografx-16. Finally the PAC-K1 is a module supporting karaoke Laserdiscs.  When equipped with a game module, the Laserdisc player was capable of playing standard game cartridges, standard game CDs, and special Laserdisc games.

The LaserActive platform had fascinated me for years before I decided to check it out. In the early 2010s The retro gaming community was just starting to realize many dead or dying consoles could be brought back simply by replacing the capacitors, and the LaserActive PACs were notorious for failing due to bad electrolytic capacitors. Since they used SMD capacitors, the rumor of the day was that you needed specialized equipment to remove them. 

In 2016 I finally found a reasonable deal on a still somewhat working CLD-A100 with the Sega PAC-S10 module. I want to say it was around $300. Before it even arrived, I also ordered my very first hot air rework station. The system worked okay with CDs, Laserdiscs and the copy of Hyperion I snagged to try out a Mega LD game, but it wouldn't load Genesis cartridges. It was about eighteen months later that I finally set aside time to tackle replacing the capacitors in my PAC-S10 module. 

In order to maximize longevity, I decided to replace all of the electrolytic capacitors with ceramic capacitors (with the exception of one where I had to use tantalum). Ceramic capacitors are known for much higher reliability and lifespans. However capacitors are used for a lot more than just storing electrical charges, and while two capacitors may have the exact same capaticance and voltage tolerances, they may behave very differently when used in AC circuits like audio stages. 

There's surprisingly little in the way of straightforward information on the real-world effects of swapping ceramic capacitors for electrolytic capacitors. There are plenty of forum know-it-alls out there who will sagely tell you that you just shouldn't do it, but precious few that can give a coherent explanation of why. My theoretical research took me down the rabbit hole of DC voltage bias, and I concluded that I could probably mitigate any negative effects by choosing ceramic capacitors with roughly double the voltage tolerance of the original electrolytics. The substitution seemed to function fine, and my PAC-S10 was restored to full function. 

Fast forward a few years and the obsession over proper audio output from Genesis sound mods, emulators and FPGA implementations led to the great conversation ender: MDFourier. This tool can tell you objectively whether your implementation is properly outputting sound that matches a genuine Sega Genesis. This software runs on the Genesis, or Sega CD and sends a series of audio instructions which you then capture with an audio capture device and analyze. The plot will show you the whole frequency range and can easily be compared with a plot of the same audio instructions run through real hardware. 

I came across a conversation Artemio and Zaxour were having about the effects of ceramic capacitors on the audio output of these systems. Since I already have a PAC-S10 with all ceramic capacitors I was able to easily provide some test results for comparison. The comparison validated their assumptions - the output from my PAC-S10 was clipped in places, distorted in others and generally out of whack. It worked and sounded "OK" to my ears so I wasn't in a hurry to do anything about it, but it nagged at me a little. 

Over the next couple of years Zaxour became something of a LaserActive guru, adding tons of new content on the Console Mods Wiki, including official service bulletins etc... He posted an image of a PAC-S10 where he had replaced the capacitors using half electrolytics (like the originals) and half ceramics - leaving the electrolytics for the audio circuits where it actually made a difference. This was exactly what I had been hoping for and someone else had done all the research. Between that, the audio service bulletin, and the SRAM to FRAM mod (to eliminate the need for a save battery) were enough to convince me to crack my PAC-S10 open again and get to work. 

The process was fairly easy, and it was kind of fun to revisit one of my earliest SMD soldering projects to note how much my skill has improved in the last 7-8 years. I neatened up some of my old work, swapped in the new electrolytic capacitors, swapped the SRAM for FRAM, swapped out the components for the service bulletin and fired it up. It was at that point that I noticed my LaserActive machine was struggling. It was able to boot Sonic CD, but it took a couple tries, and a very long time when it finally did work. The opening FMV sequence was skipping and halting all over the place. Once the game loaded, though, I could instantly hear the improvements made by putting the correct capacitors back in the audio circuits and by implementing the service bulletin. 

Just the same I was concerned about the LaserDisc player struggling to read the discs, so I took the cover off, cleaned the lens and re-lubricated the laser transport and tried again. This just seemed to make things worse. Part of me wondered if the modifications to the PAC-S10 were having some adverse effect on the player, but fortunately I have another CLD-A100 which is working perfectly and the PAC-S10 I just got done working on performed flawlessly in that unit. It looked like it might be time to replace the capacitors in the entire machine. 

The top layer of internals of the CLD-A100

Recapping a CLD-A100 is not a small task. Console5 refers to the capacitor kit (which doesn't even include the capacitors for the power supply) the "Final Boss" of capacitor kits owing to the fact that it has over 200 capacitors, including three varieties of specialty capacitor. There are numerous capacitors that have to be short enough to clear the space between boards, several bipolar capacitors, and several that are both. Once I learned that, I decided to take a more thoughtful approach and just refresh the capacitors on the power supply to see where that got me. So I ordered the PSU capacitor kit, but I wasn't content to just wait around for it to arrive. 

In the mean time, I noticed the problem was growing worse. There were now decoding errors when the system tried to play redbook audio - not skipping, but bursts of distorted music or static. There was definitely something seriously wrong. It was about that time that I made a gigantic mistake. One of the telltale signs of one or more bad capacitors is the tendency of a device to work better after it warms up. So I decided to power on my LaserActive player and leave it for a little while and come back to check performance. I ended up forgetting about it and leaving the house for a couple of hours. When I got back I could hear a loud clicking, rattling sound coming from the room where the LaserActive was. 

As I walked up to it, the loud knocking coming from inside the machine was sickening as it indicated something had gone very very wrong. A warm plastic smell like melted band-aids hung in the air. My brain could not conceive what could possibly have been making that sound - there was no disc in the drive, so nothing should have been moving. With the cover off on my work bench the CLD-A100 was powered up again and I watched in horror as the laser rapidly bounced in its housing like a focus cycle gone terribly wrong. 

The rattling sound was coming from the laser lens assembly rising to its maximum height and then violently smacking back down into the housing. In the years I've been working on optical drive systems I've never seen anything like it. The realization that this had potentially been going on for hours brought me to the soulcrushing revelation that the laser would probably never work again. (Thankfully I was wrong).

Bad capacitors can do really strange things to a device. By providing too much or too little power, digital systems gradually lose the ability to distinguish between a one and a zero, so instructions get garbled. It's a common problem with a fairly straightforward solution. Whatever happened with the laser, I decided to at least try to bring the thing back to life by replacing the capacitors. 

Recapping the power supply made things a little better and a little worse at the same time. If I left the PAC-S10 uninstalled, the CLD-A100 could play a CD for the first couple of minutes after powering on, but still with errors and would eventually go into that horrifying focus loop and slamming the laser against the bottom of its housing. But now there was a high-pitched whine coming through all of the audio circuits. 

This is a video showing the crazy laser behavior which persisted even after the power supply was recapped. As you can see it starts slowly and then increases in intensity over time. If I had not cut the power it would have sped up to a blur. (Please ignore the fact that I had installed the spindle clamp upside down >.<)

Facing a potentially long troubleshooting process, I decided to at least try asking around to see if anyone had already seen and diagnosed this behavior. Zaxour, who is in my opinion the current foremost expert on the LaserActive, was extremely generous with his time. When I first contacted him, he was convinced that there was some problem with the FTSB board (where the laser controls are located), and asked if I had recapped it. Since I had not, I ordered the full kit for the CLD-A100 and got to work when it arrived. Zaxour shared some disassembly instructions which helped immensely. 

The LaserActive CLD-A100 uses a type of ribbon connector that I've not seen before. Instead of a plug and socket, or a plastic edge with pins laminated to it, the ends of the ribbon are bare, tinned wires pressed into latch sockets. It's very easy to bend these bare wires out of shape when moving parts around, but the thing that really made me uneasy is that the wire has to be very stiff to hold its shape long enough to be inserted, and to my mind that makes it more susceptible to fatigue from being bent back and forth. 

The unusual ribbon connectors are just tinned wires at the end of the ribbon.

The player has about 9 PCBs with electrolytic capacitors on them. My approach was to replace the capacitors in the hardest to access first, leaving the easiest to access for last. In general it's smart to recap one board at a time and test after each, so if you screw something up you have some idea where to start looking. But with 9 boards and those potentially fatigue-prone ribbon cables, there was no way I was going to take that approach here. Instead I meticulously verified that the silkscreen had the correct polarity for each capacitor, and that the capacitor values matched the list I was working from as I removed each one. Then I manually checked every newly installed capacitor for the correct value and polarity. 

The "Video" board sits on the very top of the CLD-A100 and has something like 100 capacitors all by itself. It is the first to come out and the last to go back in. When I got to that stage, before starting to recap the Video board, I decided to go ahead and fully assemble the system and see what kind of progress (or damage) had been achieved by replacing the capacitors on the rest of the system. If anything, it had become even worse. The system could no longer play a CD for the first couple minutes after being powered on, and it was still going through that horrible focus loop where it wanted to start beating the laser against its own housing. 

When I reached out to Zaxour, I caught him at exactly the right time, and he very generously spent nearly the whole day (off and on) troubleshooting with me. He was so polite when referring me to the Console Mods Wiki that it took me a little while to realize that many (most) of the things he recommended I try were already detailed there.  (This as opposed to being frustrated that I didn't RTFM before asking for help). Although I had been looking there for some information, I apparently missed the troubleshooting area for LaserActive. 

The sneaky little resistors that like to go bad.

It turns out there are four ceramic resistors on the PSU that can die out, one of the four was measuring wrong on mine so I replaced it. The other rookie mistake I made was forgetting about the legendary unreliability of factory solder from the 90's to the mid 2000's. Non-leaded solder is notorious for forming weak bonds which time and vibration can shake loose creating high-resistance points all over the PCBs. The general solution to this is to "re-flow" which means to re-melt the solder to re-form the bond. I always introduce some leaded solder while I'm doing it. The best solution would be to remove the old solder and replace it with leaded solder, but I'll take my chances this way.

On Zaxour's advice I broke the system back down to the FTSB board and re-flowed every ribbon connector socket, and every through-hole IC on it. I also re-flowed every ribbon connector on every other board, and the socket connector that connects the vertical board to the one on the very bottom.

After that I treated every ribbon wire with a super thin dab of Deoxit D100, and reassembled everything to try it. The system turned on, the laser did its startup operation normally and after several minutes sitting there was no longer going through some hideous focus cycle and trying to slam the laser into its housing. The high-pitched audio whine was gone completely. Progress! 

However, my worst fear seemed to have come true. The laser was no longer recognizing that a disc was inserted. It would go through a normal focus cycle and just give up without lighting the disc indicator (the light on the front that tells you whether a CD or LD is inserted) or spinning the spindle. Cleaning the laser was one of the very first things I had done, and I had been very cautious not to contaminate it, but I had put it in and taken it out of the machine about three times since then, so I decided to assume nothing. I removed the laser assembly, cleaned the laser with a cotton swab and distilled water again, rechecked the lube on the rails, cleaned the ribbon connector with D5 then treated it with D100, and made sure every connector was fully seated and latched. 

This clip shows the drive finally working again. 

Then I put it back together and to my utter amazement, the system could now see and play discs. No skips, no distortion, no misbehavior. The system now works better than it has since I've owned it. No potentiometers needed to be adjusted. Sega CD and Mega LD games load instantly. There is no skipping, no distortion, no high pitched whine in the audio, it's in perfect working order. I played hours of Sonic CD, Popful Mail and Road Prosecutor to be sure there weren't any residual issues. I've always said that optical drive lasers are orders of magnitude more resilient than people give them credit for, but even I am amazed this thing was still working after that.

The opening animation in Sonic CD is a good practical test of the laser - even minor problems with the laser will result in stuttering, choppy video and audio. Here it's playing smoothly.