Pioneer Elite Laserdisc Player CLD-97 Repair, AC-3 RF Out and DNR Mod

The CLD-97 is one of the best performing and best looking Laserdisc players I've ever used.  I've been an enthusiast of the format for many years now.  When I first got into it, players were cheap and plentiful, but even with people dumping the format for DVD, the CLD-97 and its follow-up the CLD-99 maintained a high resale price.  It was for that reason that I never planned to own one.  Well, plans change, I guess.

Sections

• Why I Bought A CLD-97
• Fixing "No Disc Spin"
• Fixing The "Button Door"
• Replacing The Remote
• AC-3 RF Output Mod
• Digital NR Mod
• Answered Questions

Why I Bought A CLD-97

My wife is a photographer so visuals are really important to her; she notices everything.  It took a lot of convincing just to get her to watch Laserdiscs with me because, while she doesn't really keep up with the technology, she knew we had Braveheart in 4K so why in the world were we watching this blurry mess instead?  

She begrudgingly acclimated to the shortcomings of the older technology when I explained the nostalgia factor and my fascination with the sheer amount of engineering that went into making it work at all. Then she started to notice flaws that aren't necessarily inherent to the technology. The most noticeable of these is the "ghosting" where a luminous outline appears just to the right of certain picture elements. It's particularly noticeable where there are hard transitions between light and dark like in on-screen text. Not being an expert on composite video, I assume this has something to do with the chroma and luma separation getting out of sync.

My main Laserdisc player for the last few years has been a Pioneer CLD-D605.  It was originally connected directly to the composite input of an Onkyo AVR which handled analog-to-digital conversion and scaling. The Onkyo did such an amazing job I took it for granted until it broke and I had to replace it with an AVR which did not have composite input at all.  After witnessing the horror of letting my Sony KDL-70R550A LCD TV handle the scaling through it's composite input, and trying a few cheap  composite-to-HDMI adapters from Amazon, I eventually landed on the professional Extron DSC-301 HD to handle scaling of composite video from the Laserdiscs. The output was on-par with the Onkyo AVR with the added benefit of having quite a lot of control over scaling, cropping and image correction. 

Later on I added a Sony VPL-VW295ES 4K projector to the mix.  Blown up to a 120 inch screen, Laserdisc can be...an acquired taste.  Still, the ghosting with this setup was incredibly bad.  After trying a few things - isolating power supplies, adding ferrite chokes, the best improvement came from placing  Extron scaler directly on top of the Laserdisc player and connecting it with the shortest possible (12") double-shielded RCA cable I could find.  This reduced the ghosting to a minimum. It was still noticeable but easy enough to tune out.  Well, it was easy for me to tune out.  The next time I watched a Laserdisc with my wife, she asked me about the ghosting as though I had not spent 30 minutes explaining how I had tried half a dozen things and rearranged my equipment rack to get it as good as it was. 

Then she laid this on me: is this the best it gets?

So immediately in the back of my mind I think about those insanely expensive MUSE players which supposedly have the best picture quality of any player. Then I came back down to earth a little and started wondering if the Elite series of players really were that much better.  Never having seen one in person, I wasn't sure, but I did a little reading and figured that my best bet was either a CLD-97 or CLD-99.  The only thing I read about the 99 in terms of being better than the 97 was that the 99 had a 3D comb filter so the S-Video output was better.  Since I had no intention of using the S-Video output (it's kind of pointless for Laserdisc), I wasn't too hung up on getting the 99.  That turned out to be a good thing as there weren't any CLD-99's for sale when I went looking.  

There were a stack of CLD-97's but the sellers were all expecting in the neighborhood of $500 for units with serious problems. Eventually a listing popped up for one that wouldn't spin up discs - something I thought *might* just be a failing loading belt. (I had no idea at the time that the loading mechanism for the Elite players was completely different from every other Pioneer LD player I had ever worked on). It was missing the cover across the bottom for the buttons, but the seller told me he still had the cover and would include it.  We negotiated a pretty reasonable price for a non-functioning player.

Assuming I could get it working again, I would still have to figure out how to fix the flip down panel on the front, and find the correct remote control.

Fixing The Player

Fixing No Disc Spin

As I waited for the player to arrive, I downloaded a service manual (the LDDB is a great resource for this kind of thing), looked up belt sizes and searched for forum threads that might be related to the problem the seller described.  It was then I came across this thread on AVS Forum.

When the player arrived I was very happy to notice the dust on the screws and in the plastic cups that hold the wood sides on. This meant the player had not been disassembled recently, so I wasn't inheriting someone else's repair project.  

Laserdisc players (and many other devices) have sequences they follow when you power them on, and when you load a disc. In software we call these "state machines". I'm not sure what the correct term is for it in this context. 

The power on procedure might be something like:

1. Check the pickup (laser) position, if the center limit switch is not closed, move the laser to center.
2. Check the tray position, if the tray limit switch is not closed, close the tray.

The load procedure might be something like:

1. When the tray limit switch is closed...
2. Raise the disc up to clamping position
3. Move the pickup to the 10" LD outer position
4. Activate the laser. 
1. If there is a reflection, activate LD mode
2. If no reflection seek to the CD outer position, activate the laser
1. If there is a reflection, activate CD mode
2. If no reflection indicate "no disc"
5. If LD/CD mode
1. Light or show "LD" or "CD" as appropriate
2. Seek to center
3. Spin the disc
4. Read the table of contents
5. Standby

If any part in the process fails or provides a bad result, the player will not move to the next step.

On this unit, the tray was operating normally, the clamp was fully engaging but the disc would not spin.  The most likely explanations were the sensor was not "seeing" a reflection from the disc, or the spindle motor was shot and would not turn.

Spindle motors commonly fail for a couple of reasons. Lubrication failure can cause them to seize up. If you catch them quickly enough to fix the lubrication problem, they'll often go back to operating normally, but if you don't catch them in time, all that energy being dumped into the commutator to move the motor will become heat and eventually melt part of the coil winding. Other common reasons are oxidation, which can do anything from creating dead spots where the brushes are no longer making good contact with the commutator, or in extreme cases they can corrode the commutator arms with "copper rust" and they fall apart. The former can be fixed by carefully disassembling the motor and cleaning the contacts with a wire brush, the latter is a death sentence.  

The spindle moved smoothly when I turned it by hand, so I knew it wasn't seized up.  The pickup lens was trying to focus, but it was covered in dust and visibly hazy.  I used a power duster to blow the dust out of the pickup, then cleaned the lens with a q-tip and some distilled water. It took several passes to get it clear. But discs still were not being detected, and the spindle never turned.

Following the advice in the AVS Forum thread linked above, I put the player in maintenance mode by shorting the "TP" to ground. 

To access the TP pin, the "digital memory" PCB needs to be unscrewed and folded up towards the outer edge of the player, and the disc tray needs to be fully ejected. The pin is located almost all the way to the front of the player.

To put the player into maintenance mode, with the tray ejected and the player powered off, I grounded TP by attaching the female end of a DuPont wire over the TP pin, and jammed the male end under the head of the large copper screw circled above, allowing the rubber damper to hold it in place, then powered the player on. Be careful because as soon as the TP is ungrounded the player will pull the disc tray in - just be prepared to either pull the wire quickly, or move it so it doesn't catch on the tray.

My objective in putting the player into maintenance mode was to determine whether the spindle motor was not spinning because the player wasn't trying to spin it, or because there was a fault with the motor itself. Once in maintenance mode, if you load a disc and press the play button twice, it will send the instruction to spin whether it detects a disc or not. 

I set the player into maintenance mode, inserted my Pioneer GGV test disc and hit play twice. The disc started spinning right away, so I knew the spindle motor was still good. 

Although it was technically "playing" the video output was complete garbage - mostly static with occasional flashes of color that revealed it was reading something from the disc.  The test disc is encoded in CAV and the front panel was trying to display a frame count, but it was somewhat garbled. I ejected the test disc to see what it might do with a CLV disc, and that too provided distorted output.  I tried a CD, and it appeared to play (I only had video connected to a monitor, not audio).

There was definitely something wrong with either the control circuitry or the pickup itself.  I took a couple more passes at the lens with a swab and distilled water, and noticed that the CCV calibration disc was considerably less distorted afterwards. Then I tried the CLV disc, which was still heavily distorted, but images were becoming recognizable. It occurred to me that there may be an alignment issue - possibly something was knocked loose in shipping.

To test this theory, I switched the player to Side B play for the first time in this process and to my utter amazement, the CLV disc played perfectly.  I did a little more experimentation, switching to Side B immediately upon inserting the CCV calibration disc - again, it played perfectly.

I expected that it would still be unable to detect discs when I took it out of maintenance mode, but the player now behaves completely normally.  If you go back to the AVSForums post I linked above, you'll note the same exact thing happened there.  I can only theorize that some mechanical part was stuck and forcing it to play was enough to re-lubricate it, or some passive component like a capacitor was dormant too long and somewhat corrected by being energized again. 

The CLD-97 has some of the highest quality capacitors I've ever seen installed at the factory for a consumer device.  Not only were none of them showing any signs of leakage, or failure, I'm not sure I'd be able to find comparable replacements from modern sources. I made the decision not to mess with that until I have to.  Likewise on the lubrication. A spot check of the lubrication showed it was holding up fine, so I also decided not to mess with that for the time being.

Foam Bits

There are PCBs on the left and right sides of the chassis which were originally covered with a thin black foam, however time had not been kind to them, and they crumbled when touched.  Concerned bits of foam might eventually work their way into the mechanism, I went ahead and removed the foam from both sides of the player.  I'm not sure what function it served, but I assume it was noise dampening and electrical insulation from the metal sides. If this had been made in the 1970's I expect I might have found paper in those spots instead, so that's what I ended up replacing it with - just some thick textured cardstock cut to size.  I'll continue to look for a direct replacement, but I'm sure this will be fine for now.

Fixing The Button Door

The lower part of the front of the player has a hinged door that hides the buttons and generally keeps the front of the unit looking cleaner.  The service manual refers to this as "the pocket". 

Parts diagram from the service manual

The service manual lists these parts as VNK1714 and VNK1715.  As far as I've been able to tell they're not available anywhere.

Many, if not most, of the units on Ebay have this door broken off. It's not surprising once you try to handle one of them.  It's super easy to forget that it's not load bearing when you reach down to pick it up. 

The row of buttons at the bottom is supposed to be concealed behind a cover when the buttons are not in use, but the cover was broken off.

The one I received was no exception, but as I mentioned, the seller was able to include the detached door with the player so I could make some attempt to re-attach it.  Unlike a most similar devices I've worked with, this door is not a flimsy piece of plastic, it's a strip of aluminum, and it's rather heavy. It pivots on pins that stick out from either side of the door, but it has a tension clasp at one side and a damper gear to ensure it moves smoothly.  

The "pocket hinge" parts that broke off of the door are long since out of production, so my only option for getting new ones was to find some intact hinges and have them 3D modeled and 3D printed.  Fortunately for me I have a buddy who is really good at this sort of thing and he agreed to create some replacements for me when I found a pair.

To acquire hinges to duplicate, I had to resort to buying a whole front panel from Ebay which had the door still attached and intact.  Even though this was nearly $100, I was still considerably under that $500 threshold for a working player so I bit the bullet. When the replacement face plate arrived, I realized the seller had kind of screwed me a little.  The hinges were gold and one of them was bent and had stress cracks. My best guess was that the original, good condition hinges had been swapped out with a CLD-98 (Japanese model which is gold colored instead of black), and the half broken ones slapped slapped onto a parted-out face plate to sell on Ebay.

This is the "right" bracket. Notice it's visibly bent, and the little copper tongue (the "Pocket earth" according to the service manual) that's supposed to touch the pin to ground the door has been mangled. It looks like someone tried to "reinstall" the door by forcing the hinge over the peg instead of properly disassembling it. Clearly this was swapped in from a CLD-98 before selling the face plate it to me.

With careful application of heat and pressure I was able to straighten the bent hinge, but before attempting to reinforce the cracks, I brought the hinges to my friend to work on 3D printed replacements.

The brackets removed. The one on the right was straigtened with careful application of heat and pressure, but as you can see it's still cracked.  Thankfully this was enough for my friend to 3D model and print replacements.

Unfortunately 3D printing is the only option for replacing the buggered brackets for the button door. This is an in-progress shot from my friend who's amazing at doing this stuff.

Here's the finished product, printed in ABS-like resin.  My unscientific tests show that they are stronger than the original parts.

With the new hinges printed and installed, those pedestrian buttons are safely hidden behind a sleek metal cover until they're needed.

Replacing The Remote Control

The unit came without its remote control. While pretty much any Laserdisc player remote will control the basic functions, there are a couple of features of this player that can't be accessed without the matching remote which is the CU-CLD090. Unfortunately these are very rare and there were none for sale at any price when I went searching for them.  When they are available they can cost upwards of $90.  Fortunately the Japanese CLD-98 player is functionally identical, and it's remote the CU-CLD062 has an identical set of buttons/features. What's more these are a lot less rare on the Japanese market so it was Yahoo Auctions Japan to the rescue.

The matching remote for the CLD-97 is the somewhat rare CU-CLD090.

The Japanese CU-CLD062 is functionally identical to the CU-CLD090 and is a bit cheaper if you don't mind the gold color instead of black.

Buying the Japanese version of the remote was considerably cheaper than if I had bought one in the US, but that's not to say that it was cheap.  

The original remote is NOT essential to the experience. I would strongly recommend not doing what I did. Instead and just grab any Pioneer LD remote that looks vaguely like these.

The special features unique to the original remote are kind of useless.  The HLITE/INTRO is a bizarre function that I can't really conceive of ever wanting to use outside of maybe a showroom where you're trying to sell LD players- it either plays the first 8 seconds of every track, or it skips to two minutes in to each track and plays 8 seconds.

The remaining features all seem to be engineers straining to find an excuse to include a rather sophisticated (for the time) video memory system in the player.  The "Strobe" function freezes an image on the screen while the disc continues to play in the background, and also allows for refreshing the freeze frame every couple of seconds. Again, I can't really envision what you would use this for.  Finally the One-Shot Memory saves a freeze frame to the player's memory and holds it on screen, even after the disc is ejected and another is inserted. The manual suggests you might want to do this so you can play music to a single frame of video.  I can kind of see that being a desirable thing, except that it's basically a recipe for CRT image burn-in so again, I would never use it.  

I have a high-end Betamax VCR with these same "digital" features, so it was clearly a desirable feature in the 80's and 90's. In the case of the VCR I kind of understand how this would be useful for creating interesting effects on home movies, but it still seems less useful to me here.

Modding The Player

AC-3 RF Output

First, a bit about AC-3. In 1995 the Laserdisc standard was updated to include a home cinema version of the 5.1 surround sound that was being used in theaters. Developed by Dolby, the signal was originally called AC-3 and then later Dolby Digital, though they both refer to the same thing. Although this is the same Dolby Digital used by the DVD format later on, it had a somewhat more complicated implementation on Laserdisc.  By 1995, the analog audio tracks on Laserdiscs were barely in use (because most Laserdiscs made since the early 80's had far superior digital audio tracks) so the engineers decided to repurpose the right analog audio track to contain an RF modulated digital bitstream. If you listen to it it just sounds like static.  To demodulate it into a digital bitstream requires an external demodulator which uses proprietary logic, so these things are unfortunately rare, expensive and no modern replacement currently exists. 

Chalk it up to impatience, but I was kind of caught unawares when I realized that the CLD-97 did not come with AC3-RF output for playing titles encoded in Dolby Digital Surround Sound.  I managed to get all the way through getting the player back into working condition before I noticed this was missing.  

Many years ago a German company sold mod boards that could be installed into older players to enable AC-3 RF output.  About 5 years ago I tried to contact them but received no response.  Fortunately, the mod itself isn't the rocket surgery it may seem like at first. 

As of the time of this writing there are two sellers on Ebay who make and sell these mod boards.  I decided to get the board from swilber because it includes logic for the muting circuit.  While the circuit itself has been floating around on the Internet for some time, the design files necessary to order them from a fab house like OSHPark or JCLPCB aren't readily available.  The seller assures me he plans to open-source the files on Github in the near future.  I'll try to update this post when/if he does.

After looking at what the mod circuit does, it's apparent that the AC-3 RF output is considerably less specialized than I had originally assumed.  The mod is a transistor in an emitter-follower configuration that, in essence, just copies the whole RF signal at the test point and makes it available to a port on the back of the machine.  I had always assumed that some special filtering was happening inside the player to identify and isolate AC-3 signals, and output JUST the RF-encoded audio, but apparently all of that filtering happens on the demodulator side.

Many players, including the CLD-97 have an internal signal that changes from high (5v) to low (~0v/ground) depending on whether the player is in playback mode or not. This is known as the "mute circuit".  The CLD-97 holds its mute signal high (5v) whenever the player is NOT playing, including when it is paused.  While the disc is playing, the mute signal is pulled low.  The mod board uses an additional transistor to switch the RF output of the board off when the mute signal is high, this way.  Apparently there can be transient voltage or other garbage in the RF signal when no disc is playing or during pause that can be interpreted as sound and cause noise, so shutting the output off in these situations makes a lot of sense.

Installing the mod board was extremely simple.  A guide already exists for this exact player that illustrates all of the installation points, but it leaves a little bit to be desired so I'll give a clearer illustration below. 

This seemed to be a good place to attach the mod board. It's thin enough that it won't contact the player lid when reassembled. (But I did insulate the top of it with Kapton tape before reassembly just to be safe).

Here is the same image with the connection points labeled.  The RF "in" wire has only the core of the coax connected - this allows the shielding to shield the wire while not being directly connected to ground.  The blue wire is the "mute" signal, the red wire is the "+5v" line, the green wire is the "-5v" line, and the black wire is the "ground". The ground wire is isolated by a .1uF capacitor that helps prevent noise on the ground plane from interfering with the RF output signal. "J+" and "J-" correspond with the coax line connected to the RCA jack on the back of the player.

The coaxial wire is RG178 purchased here.

 

Probably the hardest part of the mod was cleanly drilling out the hole for the RCA jack.  I picked this location because it was in line with the other digital audio jacks, but I would not recommend using it because it left barely any clearance for the bracket and one of the electrical connectors on the other side (see below). Note: this RCA jack does not directly contact the chassis - this is to prevent noise from the ground plane from interfering with the signal.

I got very lucky that this actually fit on the other side. About 1/4" closer to the digital out coax jack would have been a better fit internally. 

With everything connected back up, my receiver is detecting and decoding the AC-3 tracks just fine.

I use a Pioneer receiver with a built-in AC-3 demodulator. It's very convenient because it will automatically switch between Digital Stereo, AC-3, and DTS depending on what the player is outputting.

Reversing the DNR Logic

The CLD-97 has what the engineers obviously considered a premium feature where it uses its digital memory to analyze and try to remove "noise" from the image before output. This is probably a desirable feature for some older discs which do have a lot of video noise, and it probably did subjectively improve picture clarity when viewing on a CRT. Unfortunately in our modern world of fixed-pixel displays and video scalers, it creates irritating picture artifacts. The difference is subtle but noticeable, especially when the image is blown up to 4K and projected onto a 120" screen. Without actually knowing anything about how this "DNR" circuit works, a result of it seems to be turning the video stream into pixels (analog video doesn't actually have pixels, it has an AC waveform that transitions up and down across the horizontal). Scalers "see" these pixel boundaries and exaggerate them causing noticeable blocking - at least that's my guess about what's causing the image artifacts I'm seeing.

By default the player's "DNR" function is switched on. The ability to disable it seems to have been an afterthought, as it cannot be done from the remote control. There's a button on the front panel which can toggle this feature off, however it can only be switched off while a disc is playing, and it gets switched back on automatically seemingly every time a disc is fully stopped, or the tray is ejected.  I frequently found myself forgetting to shut this off until I got a few minutes into a disc and noticed the pixelation noise.

The LDDB documentation archive for the CLD-97 contains a guide to modifying the player to reverse the logic that turns this feature on and off, so it's off by default, and shut off automatically when the disc is fully stopped or a new disc is inserted. 

The DNR mod guide is a little outdated now as the 2SC1740 transistor is an obsolete part, but it's just being used as a "switch" so any NPN transistor with similar specs will work just fine.  The transistor I chose was the very common BC337 because I happened to have some of them on hand.

The guide also advises to cut a trace between IC502 and J4, but I'm not a big fan of cutting traces if I can help it, so instead, I just lifted pin5 of the IC from the board using a dental pick to gently pry it up while I heated the pad beneath it, then used some de-soldering wick to clean the pad off beneath it.

Pin 5 of the IC lifted rather than cutting the trace that connects it to J4,3.

The DNR Mod installed. Pin 5 of IC502 is de-soldered from its pad to disconnect it from J4,3 and gently bent back to be soldered to the base of the BC337 transistor. The emitter of the BC337 transistor is soldered to the nearby ground point. The collector of the BC337 transistor is wired to the test pad for J4,3. A 3.3k Ohm resistor is soldered in-line between +5 and J4,3.

Although the solder on the ground point holds the transistor very securely, some hot glue is added to the transistor for a little more protection against movement. The wire with the resistor is held flat with Kapton tape to prevent it from getting snagged on anything during reassembly.

Answered Questions

Before installing these mods for myself I had a lot of questions about how they would function. 

• Question: Since this player didn't originally have logic to output AC-3 RF, do I have to manually switch to the right analog audio channel to activate it?
• Answer: No.  The AC-3 RF output mechanism (whether a factory feature or enabled by this mod) is not selective at all - it seems to send the whole RF signal (or at least a good deal more than just the right analog audio signal) from the laser to the demodulator and leaves it up to the demodulator to discriminate and identify whether a Dolby Digital signal is present.  Furthermore, the audio selection logic in the player happens downstream of the RF signal. Simply put, the AC-3 RF output is not affected by the audio settings of the player at all.

• Question: Is there a risk that it will mis-identify the contents of the right analog audio channel?
• Answer: Not unless the external demodulator is malfunctioning. My best guess is that normal audio looks nothing like the digital bitstream to the filter process.

• Question: Some available mods and circuit diagrams feature a mute circuit, some do not. Is this feature necessary?
• Answer: The "mute" part of the mod board is not essential to it's function.  Just the same, it is a good idea to have this feature when possible, and it is the way that players that output AC-3 RF from the factory work.  If the mod has a mute input, that means it is designed to "watch" what the player is doing, and shut off the output unless a disc is playing. The mute circuit on the player itself is used to silence the standard audio outputs under the same circumstances. I'm not sure what kind of RF signals would even exist when the disc is not playing, but presumably there's a possibility these could exist (otherwise why would players have a mute circuit?) and could be annoying or even harmful (if they send bad signals to your amplifier).  From what I understand different players implement the mute signal differently. The CLD-97 pulls the signal low when playing, and leaves it high (5v) in every other scenario, so the transistor on the mod is configured to turn on when the voltage on that line goes low.  Although I don't have direct knowledge, I have read that other players handle this the opposite way, where the signal is pulled high (5v) when playing, and held low in every other scenario. In that case the transistor logic would have to be reversed to have the same effect.  If you installed a board that was designed for the opposite logic as your player, it would not work because the signal would be shut off while the disc was playing.

• Question: Is this in any way inferior to the AC-3 RF output of a player that had this feature from the factory?
• Answer: Functionally, no. This seems to work identically to the AC-3 RF output on my other players. About the only drawback is that there's no easy way to make the output jack look "factory".

• Question: Does the DNR really degrade the video output from the CLD-97?
• Answer: On a 15kHz CRT to my eyes, the DNR looks fine - it even even marginally improves the image by removing video noise on some scenes.  But when using a scaler to output to a modern display it's a different story.  Even when I forgot all about the DNR feature I would find myself watching a disc, and wondering why I was suddenly seeing pixelation on fast moving scenes - something you would normally only see with a digital format like DVD.  My best guess is that the feature is digitizing the picture in order to identify and filter out noise. Digital video uses pixels instead of a smooth waveform to transition horizontally across the picture. These pixels create more distinct boundaries which the scaler picks up on and exaggerates.