A new, new amplifier: Advance Paris A12 Classic

The A12 front panel. Yes, those are tubes in there! 

I only recently started buying new gear for my stereo system as opposed to my habit of buying (and sometimes fixing) used gear. In July I bought an SVS 3000 Micro subwoofer to go with a pair of Canton Ergo 22DC bookshelf speakers my brother gave me. And more recently, I splurged on an $80 SMSL SU-1 DAC and an Advance Paris A12 Classic integrated amplifier. 

I got the SU-1 DAC to connect my TV PC to the stereo system before I decided to buy the A12 amp. More on that DAC, here. 

I've been using an LG sound bar with the TV, but have grown unsatisfied with its performance, especially with regard to dialog that I frequently can't understand. The sound quality from the stereo system is so much better than the sound bar, that it almost hurts to use the sound bar any more. It seems redundant to have the sound bar and the stereo system in the living room, so I thought it would be best to connect the TV to the stereo system and get rid of the sound bar. That means I need an amp that has HDMI ARC input. The Krell KAV-300i that I recently recapped, and has been powering the stereo system is a great integrated amp, but it only has analog inputs.

I am not a true-believer when it comes to audio gear. I trust measured specs more than subjective reviews full of purple verbiage. To me an amp is an amp, as long as it isn't poorly built and has adequate power to drive the speakers. Any subtle differences that may exist in sound character will be minor (especially compared to differences in speakers and their room interactions) and I probably won't be able to hear them (see my recent listening test of the SU-1 DAC vs 30 year old VRDS-20 CD player), or will quickly get used to them. 

I look for adequate power output, ability to drive low impedance loads (because who knows what speakers I may end up with), adequate input/output capability, quality of construction which includes components and layout, and near the bottom of the list, the way it looks. With a lot of high end audio amplifiers, you pay a lot for artistically machining the amp's face plate, or maybe even the whole chassis. Some look really spectacular, but that's just not how I want to spend my money.

I looked at a lot of different integrated 2 channel amplifiers to see if any had HDMI ARC inputs. There aren't many out there, unless you want to go full surround-sound with 5-17 amplifier channels and speakers. I saw the Advance Paris A10 and A12 amplifiers and really liked their I/O capability. I found a deal on an open-box A12 that was only about $200 more than a new A10, so I ordered the A12. Either one would more than meet my requirements.

Rear view of the A12. There are eight analog inputs, two of which are balanced, and including phono (good for MC or MM cartridges). There are also eleven digital inputs, including three optical, three coaxial, two USB, and two HDMI jacks. There's a balanced preamp output, two outputs for subwoofers with selectable LPFs, and two headphone jacks with selectable gains and output impedances. Note: this is not my amp. I found this photo online. Look closely at the labeling under the speaker connectors. It says "A or B Speakers Impedance: 4-16 Ohms" and A+B Speakers Impedance: 6-16 Ohms". My amp says "A or B Speakers Impedance 8-16 Ohms" and "A+B Speakers Impedance: 16 Ohms"

Since I have a subwoofer, one feature of this amp that I am especially interested in is bass management. There are two subwoofer outputs on the back of the amplifier with switches to control the LPF cutoff frequency at either 75Hz (for larger speakers) or 150 Hz (for smaller speakers, such as my Canton Ergos, though they really should be OK down to 45 Hz). The spec sheet and manual don't say much about bass management. Do the subwoofer outputs switch on and off with the speakers? What is the slope of the roll off? Does it also apply high pass to the A12 amplifier stage  I sent an email to Advance Paris asking these questions and got no response. Hmmm.

I also liked the preamp-out/power-amp-in I/Os. That can be useful for doing stuff like bi/tri-amping a system or adding an equalizer.

Advance Who?

Advance Paris is a French company, with manufacturing in China, like many many brands of audio gear, including some uber high-end stuff. They've been making audio gear since 1995 and selling mostly in the EU. They recently started marketing world-wide, including the US. 

The look of the A12 has been compared to McIntosh, and they are similar, but I think the A12 has updated that look. 

McIntosh MA12000 hybrid integrated amplifier. Notice any similarities to the A12? Two big meters, tubes visible between them (backlit in green here), knobs and a digital dot matrix display. In this model there are handles, but in other products there are always a couple vertical metal bars along the left and right edges of the front panel.

I'm not entirely sure I like the way the A12 looks - it reminds me of futuristic kitchen appliance that you might see in a 50 or 60s era sci-fi movie. Not that that's bad... Hey, maybe it will go well with my 3D printed Sci-Fi Lamp! Maybe I need to print another with a blue shade to match the VU meters in the amp. We'll see if the design grows on me when I've had the amp for a while.

Sci-Fi Lamp, printed on UMMD using 1mm nozzle and PETG filament. The shade is two identical prints, one green and one transparent edge-glow glass. The red parts are actually 15 identical pieces, stacked inside the shade pieces.

The A12 has a dual-mono (meaning that each channel has its own power supply) class AB power amp that delivers 190 W/ch into 8 Ohms and 280 W/ch into 4 Ohms, so it should be able to drive just about any speakers ever made (more on that, below). A lot of newer amplifiers are class-D circuits, which tend to be small and very efficient, but I'm not so sure about their reliability. I found the full digital amplifier in the Technics SU-G700 interesting, but the Stereophile measurements turned me off - they looked great except for lots of ultrasonic noise in the output. It also doesn't have HDMI input. Maybe the next generation...

The DAC that's built into the A12 is a PCM1796 made by Texas Instruments (originally Burr-Brown, bought by TI in 2000). That chip has been around for over 20 years, and it may not measure as well as some of the newer chips from AKM or ESS, but are the differences audible? Maybe to some, but probably not to me. When you run the incredible SINAD signal from a modern DAC through relatively noisy tubes, does the DAC's high performance really matter?

Yes, the A12 has tubes. For many people that's a selling point, but I'm not sure if it's good or bad. Some people think tubes have a euphonic sound,  In the A12 the tubes are biased in class A in the preamp section of the amplifier. If tubes have a "sound". I'd guess it's from noise and distortion- they tend to produce even order harmonics that for some reason sound nicer than the odd order harmonics produced by bipolar transistors. But in a preamp you're probably not going to drive the tubes hard enough to distort audibly, so is there really going to be a difference in the sound other than a poorer S/N? Another different "sound" you might get with tubes is microphonics. That happens when the tubes vibrate. Small changes in the physical position of the elements in the tubes cause changes in capacitance and that can modulate the audio signal. If it gets severe enough (maybe by turning the volume up to 11 so the subwoofer vibrates everything in the room) it can actually cause oscillation.

The tubes are 12AT7 or ECC81 type made by JJ Electronic. That type is still made by more than one company and not too expensive, and should last many years before they need to be replaced, so they're not too much of a problem in that regard. Some tube audio fanatics are into "tube rolling" which means trying tubes by different manufacturers, or old stock tubes, to see if one sounds better than another. I don't suffer from that affliction.

Construction

You know, I hate audio equipment reviews that don't show me the insides of the unit being reviewed. To me, quality of construction is a much more important factor in deciding whether to buy than the looks of the front panel. Why don't more reviewers open up the gear they review? Maybe their agreements with the equipment maker prohibit them from showing off the guts. 

Well, I have no such agreement! You're going to see pictures of the guts here.

There are good ways to build electronics and some not so good ways. Here's an example of not-so-good- this is an 80's vintage Soundcraftsmen DX4000 preamp that I recently recapped and sold on ebay. All the I/O switching is done with switches mounted on the front panel, with long wires from the connectors on the back panel. Does it work? Sure. Do I want to buy something like that, at any price, even if it specs well? Nope.

No! This is not the A12! It's a 1988 Soundcraftsmen DX4000 preamp. Can you imagine wiring this? or repairing it? Honestly, it looks like something I would have built when I was in high school and couldn't afford proper tools or parts.

Here's an example of more preferable build quality- the Krell KAV-300i amplifier:

This is more like what I want. This amp wasn't without its problems, but this type of construction is a big improvement over the Soundcraftsmen preamp in the previous photo. The I/O switching is done by relays at the back of the amp, very close to the connectors. There aren't a lot of wires running all over the place. 

As of this writing, there aren't many photos of the internals of the A12 online, but what I can find looks like good quality construction. Everything looks modular with PCBs used to connect them together, and minimal cabling.

One of the few images showing internals that I was able to find. 

In this video from Advance Paris, at about 4 minutes in, you can see some of the insides of what might be an A12. It looks well laid out, and I see many small relays used to switch signals, similar to the Krell KAV-300i amplifier. I also see some Bennic XPP series film caps used instead of nonpolar electrolytics. While there are many surface mount parts on the PCBs, it looks like the electrolytic caps are through-hole type that will be relatively easy to replace in 20-30 years when they start failing (assuming they last that long, and that radial lead electrolytic caps are still available in 20-30 years). Nice!

 

More reviews (you may need to translate):

https://www.hifitest.de/test/vollverstaerker/advance-paris-a12-classic-22057

https://www.on-mag.fr/index.php/topaudio/tests-auditions/24521-test-ampli-hifi-advance-paris-a12-classic-une-oeuvre-musicale-complete-et-imposante

https://www.whathifi.com/reviews/advance-paris-a12-classic

Video reviews in French:  

https://youtu.be/-pFpBgrVjsM?si=2d2rLTvVqJLoVSOG

https://www.youtube.com/watch?v=0v3YV-tX0MQ

It's Here!

When the amp arrived, I connected B&W 703 S3 speakers and a CD player to both the balanced analog and a coax digital inputs on the amp, and a Squeezebox Touch to another coax input. Everything sounded great, and worked perfectly. Then I tried connecting my TV to the HDMI ARC input. That's when things turned not so great.

I managed to get the HDMI ARC input working a couple times, but have no idea why it started to work and why it stopped. I emailed AP about it, and got a reply within an hour (great!), that the amp is stereo only and I should disable all surround sound multichannel audio capabilities in the source. It would be nice if they included that sort of info in the manual!

I have an Nvidia Shield TV Pro (STVP) connected to my TV via HDMI because the STVP has better video upscaling than the upscaling built into the TV.  I have been using the HDMI ARC connection on the TV to connect to an LG Dolby Atmos sound bar, and it was all working fine. After the email from Advance Paris, I went through both the STVP and the TVs setup menus and disabled all multichannel audio. 

I messed with it and messed with it, and changed settings, powered down and back up, etc. and could not get it to work. I tried connecting the TV audio to the amp via optical fiber and it worked without and problems, so that's probably how I'm going to use it. Maybe they'll update the manual or there will be a forum with info online somewhere. For now, optical is the way to go.

Photos and some details

I wanted to take pictures from the top and bottom sides, but after removing the top cover, realized that the bottom cover is part of the chassis structure and I'd have to take the whole thing apart to get the bottom off. It seems it is intended to be serviced from the top side only, which requires removing a bunch of stuff I'm not willing to remove right now, so the photos are mostly taken from the top side.

The binding posts on the output are very heavy, and seem to be very good quality. The posts are 7 mm in diameter and the vertical holes that you might insert raw cable or pin plugs into are 3.7 mm in diameter. You can also insert banana plugs into the ends of the posts, and they seem to grip well (with my bananas, anyway).

The power transformers are 120 mm D x 80 mm H. There are multiple windings all called out on the label. The power amp is powered by a dual 46.5V @ 2.5A windings. There are four, big amplifier filter caps (2 for each channel), each is 12,000 uF @ 80V. It looks like all the electrolytic caps are made by Decon and all (everywhere in the amp) appear to be rated for 105C operation, which is good because this amplifier runs warm. It looks like all the electrolytic caps in this amp are through-hole type, easy to replace when the time comes.

For those who aren't aware, electrolytic caps have finite lifetime, usually specified as something like "2000 hours at 105C, max voltage, and max ripple current". That doesn't sound like long, but operating at less than that temperature, voltage, and ripple current extends the operating life. For every 10C below the maximum temp, lifetime essentially doubles, so operating at 45C (about where this amp seems to sit most of the time) multiplies the operating life by 2^6= 64x, which means the caps should last 64 x 2000 hours= 128,000 hours. That's equivalent to 8 hours per day for 16,000 days or 43.8 years.

The tube preamp is on its own PCB and it looks like you have to remove 4 screws and a couple cables to get it out of the amp. Access to the tubes would then be from the front side, I think. There's a metal cover over the tubes that is soldered to the PCB that holds the LEDs that light up the tubes (you didn't think that light came from the tubes, did you?), so you won't be removing that cover when it is time to change the tubes (hopefully many years from now). There are a couple nuts on top of the metal cover that have some shellac on top. If you look closely into the tube window you can see what looks like conical springs that appear to keep the tubes pushed down in their sockets. I think those shellacked nuts are on the ends of the springs.

There are 6 output transistors bolted to the heatsinks in each channel. I can't see the numbers on them.

Inconsistencies and information needed

The manual is lacking in some critical details. There's nothing about how to use the HDMI ports, it doesn't specify exactly what the hi-bias switch does, it doesn't explain whether the preamp is still functional if you turn on the power amp input (could be useful for biamping). It doesn't say if the tubes are on the preamp or power amp side of the power amp input. 

The specs on the web site say the amp can deliver 190W/ch into 8 Ohms and 280 W/ch into 4 Ohms, but two different versions of the amp seem to exist, one in which the speaker terminal labels say 4 Ohms is OK and one that says the minimum speaker impedance is 8 Ohms.

A12 with the top cover off. There are 11 screws that hold the top cover in place.

The back panel (gold) is a structural component, as is the bottom pan (black). There's no taking the bottom off without taking most of the amp apart.

One of the two power transformers in the A12. I would guess that the 210V windings are used to bias the vacuum tubes. The 46.5V 2.5A windings are powering the power amplifier.

The preamp board. The tubes are located inside the metal box at the top of the image. It looks like there are 4 cables and 4 screws that have to be removed to take the preamp board out and to access the tubes. The metal cover over the tubes is soldered to the circuit board.

The preamp board appears to be held down by 4 screws, one of which is seen near the center of this image. The other boxed area is one of the two shellacked nuts that hold the tube hold-downs in place.

The small HDMI board sits on top of the larger digital input/DAC board.

This is the HDMI board.

The Digital input/DAC board. You can see the PCM1796 DAC chip and some NE5532A opamps among the parts.

The left side (viewed from front) meter board.

The right side meter board.

Speaker binding posts. Good quality, can handle pin plugs, banana plugs, wire or spade lugs.

If anyone has any questions, ask away. I'll try to answer.

Recapping a 1982 Luxman RX103 Stereo Receiver

Another recapping project: Luxman RX103 stereo receiver from 1982.

In 1982, no one at Luxman ever would have thought their product would still be operational after 42 years. This receiver has 157 electrolytic caps that will all eventually fail, though it only takes a couple to stop the receiver from working. The only way to continue using this receiver (which is working and sounding great!) is to replace all those electrolytic caps.

This job looks like a major project. There are a lot of caps and a lot of PCBs, and accessing them is going to require studying the thing and figuring out what needs to move to get to each of the 20+ circuit boards to be recapped. This reminds me of the first day of Calculus class I took in high school. You get the book and start flipping through it and your first thought is "I'll never figure all this out". But have patience, do a little at a time, and before you know it, you'll be through the whole thing.

I could disassemble the whole receiver right from the start and start recapping boards. However, there are too many screws, cables, and circuit boards to keep track of and I'd probably never get the thing back together again. My approach is to recap one board at a time, put it back including all the screws and cables that have to be removed to get at the board, and test it to make sure it's working. That way if it stops working I'll know where to look for the problem. It takes a lot more time to do it this way, but probably not as much time as it would take to troubleshoot after recapping everything and running into a problem, or maybe being unable to put it back together again.

When you do this sort of project, it's always best to have at least a schematic diagram that you can compare to the parts that are actually on the circuit boards. In this case, I actually had the service manual that someone was kind enough to copy and place on the internet for all to download. The manual includes parts lists broken out for each of the 20+ circuit boards in the receiver.

Luxman did several thing to make this job easier. First, the service manual contains lists of parts that actually include the capacitance and voltages instead of just cryptic Luxman part numbers. They also labeled all the components on the circuit boards and marked the polarities of all the caps. The manual has separate parts lists for each PCB, and caps on each PCB have numbers that are grouped for that board, making it easy to keep track of all 157 caps and where they go.

10/10/24:  I'm writing this post and publishing as I progress, so on any given day, if you get to the bottom and the recapping isn't complete, come back in a day or two and check again. The power amp board will be one of the last because I ordered the wrong caps for it and I'm waiting for delivery of a new order.

VERY IMPORTANT NOTE:  There is a button labeled "Remote Control" on the front panel of the receiver. If you push that button, you will not be able to hear any output from the amplifier. When you replace caps and start testing the receiver, if you don't hear anything coming from the amp, check that switch!

First step

Select and order replacement caps. I made a spreadsheet and copied the part numbers, cap values, and voltages of the original parts from the manual. Then I sorted it by part number, cap value, and cap voltage. That grouped all the same value caps next to each other. There were some caps with same value but different voltages. For those I looked for replacements at the highest (sometimes higher) rated voltages. For example, there were 47 uF caps at 16, 25, 35, and 50 volts used in the amp. I simply ordered some 35V and a few 50V caps for all of them. By doing that, the caps will last longer, and I will be ordering a larger quantity of the same part which drives down the cost, and I'll have fewer parts to sort through when I am replacing the caps. 

I was able to find Nichicon caps for some of the parts, but not all, so the new parts list is a mix of Nichicon, Rubycon, Panasonic, and Kemet parts, all selected for low ESR and long life. A few low capacitance electrolytics are going to be replaced with film caps. Here is the spreadsheet of the parts I ordered. Total cost was about $70 from Mouser Electronics.

Note- I didn't receive the 22 uF caps I ordered, and I ordered the wrong physical size replacements for the 10,000 uF main filter caps. I also accidently order nonpolar caps for all the 0.47 uF 50V parts. I put together another order to replace all those, this time from Digikey. There's a separate page in the cap spreadsheet showing the parts I ordered from digikey.

Step 2: Power Supply (PS) Board Recap

I took the top and bottom covers off the receiver, set the screws aside in a tray, and looked at it for a while and saw that the PS board looked relatively easy to access and had lots of caps to replace, so that's where I started. 

The PS board is accessed on the bottom of the receiver. In order to get it out to where you can turn it over for soldering, you have to clip a couple zip ties on a large cable bundle on the top side of the board, clip a few zip ties on the cables on the power supply board, and take out a bunch of screws- you have to remove the audio input connector screws on the back of the receiver, and the screws that hold down the PS board- see photos. You also have to remove the switch pushers that extend to buttons on the front panel.

Once you do that, you can wiggle the board out of the chassis and turn it over to access the solder side. Note: there are a lot of connectorized cables used throughout the receiver. The white ones will unplug, but the brown ones are soldered down and you can't unplug them.

Brown connectors used in some places in the receiver. These are soldered down, so leave them alone.

Clip these zip ties on the top side of the receiver to loosen up the cable to the PS board so you can get the PS board out of the chassis.

Power supply board on the bottom of the chassis, before recapping.

The switch pushers are glued to the switches. Push them down to release them, then slide them to the left to remove them for PS board access.

Remove screws at circled locations to release the PS board from the chassis. Once the screws are out you can wiggle the board out and flip it over (sorry, no photos of that).

When I recapped the board, kept a copy of the spreadsheet with caps sorted by part number handy, and a copy of the PCB layout from the service manual. I removed one or two caps at a time, checking that the polarity marked on the cap matched the polarity marked on the PCB, and verifying the part numbers marked on the PCB. Once a cap was off the board I placed a check mark on the spreadsheet, checked the replacement value to insert in its place (voltages may be higher than the original parts), inserted the new part and soldered it and clipped the leads. Then I marked the layout diagram so I would know that that part was replaced. It's slow going, but helps ensure there are no errors.

Spreadsheet and board layout marked as each cap was replaced. 30 down, only 127 to go!

Old caps from the power supply board. The brown stuff isn't leaked electrolyte, it's glue. 

Recapped power supply board back in place. Don't forget to put a drop of hot melt glue on the switch pushers after the whole receiver has been recapped (you will have to take them out again to access other PCBs, so don't glue them right after the PS board is recapped).

After putting the PS board back where it belongs, I powered up the receiver and checked to make sure it was still working. Success! Now on to the next PCB...

Step 3: Remote Control (RC) PCB

The remote control PCB is located right next to the PS PCB, and is easy to access, so that one comes next. Now it could be argued that I'll probably never use the functions of the RC PCB, so I could just leave it alone, but if one of the caps on the board shorts out, it might take out a power supply that's needed for another part of the receiver, so replace the the caps!

The RC PCB circled in green. The only thing that holds it in place are the studs that mount the connectors on the back wall of the receiver. 

4 little plastic studs hold the RC PCB into the receiver. Release them by pushing them out with a screw driver.

After removing the studs, the RC board comes out easily.

Underside of the RC board. No problem!

There are just 6 caps on this board- all 1uF 50V parts. They are strangely labeled R080-R085 in the parts list, and some are labeled with R on the PCB, too. This is the only place these weird designations showed up in the service manual and on the PCB (as far as I know).

I replaced all 6 caps, put the board back into position, inserted the studs, and it was good to go. Only 121 caps left to go.

Next up: The "Decorder" Board (DB)

I think they meant "decoder", but I will continue to use "decorder" to refer to this board because that's what it's called in the service manual.

This one requires a bit of work to get loose. Start by clipping zip ties holding cables together over the board. Then disconnect the cables.

The "decorder" board sits right under the headphone socket.

Zip ties cut and cable unplugged.

Remove the screws, and the headphone jack- see next photo.

Remove two bezel screws (one in the center of the pic, the other not shown farther to the left) and flex the bezel outward exposing the headphone socket clip (circled on the left).

Slide the headphone socket clip out to release the headphone jack. Set it aside.

You won't be able to get the decorder board out until you move the amplifier support rail out of the way a little. There are two screws at each end of the rail (circled, toward the right in this photo, only one end of the rail shown). remove the screws from both ends of the rail. The circled hole on the left is the bezel screw location that was removed to release the headphone socket.

Remove these amplifier board screws from the support rail and the rail will come loose. You don't have to remove it, just loosen it so you can take out the decorder board.

With the amplifier support rail loose, you can get the decorder board out of the chassis.

Underside of the decorder board, ready for cap replacement.

Uh oh! Both these caps labeled C962 on the PCB, and the service manual says C962 isn't supposed to be used, but C693 is, but is nowhere to be found. Hmmm.  2.2 uF 50V. I believe one of them is actually C963. I replaced both, hopefully I ordered extra caps and won't run out when I need that value on a different PCB. I have updated the spreadsheet to include C962.

I removed and replaced the old electrolytic caps including C962 in two places (the manual says not to use).  Hmmm. I think one of them is supposed to be C963. It was working with the caps on the board, so I replaced the caps. I may have to order more of that value if I run out when replacing caps on another PCB.

Reverse the process and put it all back together. Note- when you put the bezel screws back you have to make sure that all the radio station preset buttons are free to move. It takes a bit of wiggling to get them all into their holes in the bezel, so be careful.

Once it was back together, I hooked up speakers and tested it and it worked fine. 7 down, 116 left to go (with C962, the total number to replace went up to 158).

Next: the Tone Control Board (TCB)

Now we're going to flip the receiver over and work on the top side. The tone control board is low hanging fruit. All you have to do is remove two screws and the board comes out. You don't even have to remove the knobs.

The tone control board is mounted bottom side up. Remove the two circled screws to lift it out for cap replacement.

Top side of the tone control board. Spray some cleaner in the pots while you have easy access. I removed the knobs, but it wasn't necessary to get the board out or back in.

I ran into a small problem here. It seems the 22uF caps I ordered didn't arrive with the rest of the parts. I'll add them to an order with the others I missed.

TCB recapped, except for the two circled 22 uF caps. I'll have to come back to those after I get them ordered.

Once again, I hooked up speakers, powered it up and tested it to find it working fine. 10 more down, 106 left to go.

Next up:  Power Relay (PR) PCB

More low hanging fruit- the PR PCB has only one cap to replace. Take out 3 screws and carefully slide the board off the voltage regulator IC leads and it's ready to be recapped.

The board is mounted right next to the power transformer, accessible from the top side of the receiver.

There are three screws holding the board down, circled in green. The cap that needs to be replaced is circled in blue.

Screws removed, the board is loose. Carefully slide it off the 12V regulator chip (blue circle) leads so you don't bend them.

Flip the board over and unsolder the old cap and solder in the new one. 2.2 uF 50V.

Be careful when reassembling to slide the socket over the voltage regulator leads, then screw the board down. That one was easy!  1 more down, 105 left to go...

Next up: Full CAT PCB

More top-side low hanging fruit. The Full CAT board attaches to the chassis with just 3 screws which also help hold the receiver board down. Disconnect the cables, then remove the three screws and the board will be fully accessible.

The Full CAT PCB, mounted along side the radio PCB. Disconnect the cables then remove the screws that go through angle brackets that hold it down.

Three screws hold the Full CAT board to the chassis. They're a little easier to access if you disconnect the cables first.

Full CAT board with cables disconnected and screws removed.

Underside of the Full CAT board, ready for recapping.

Top side of Full CAT board, recapped. Instead of reinstalling it, might as well work on the radio board next...

I ran into one problem on the Full CAT board. C305 was listed in the service manual as 4.7 uF @ 50V, but the part that was installed was 47 uF @ 16V. Hmmm.  I installed a 47 uF @ 35V part.

That's 7 more down, only 98 left.

Next comes the Radio PCB

With the Full CAT board out, the radio PCB is half way out. There are a few more screws to remove, including the screws that hold the antenna terminal block against the back wall of the receiver, and one sneaky ground connection screw.

Screw locations to release the Radio PCB. The two at the top of the photo are accessed from the back of the receiver. The middle one on the right is a ground screw that is accessed from the right side of the receiver.

Radio PCB ground screw. The board won't come out until you remove this screw and bend the ground strap down.

Radio and Full CAT PCBs out. The radio board is ready for recapping.

Bottom of the radio board, ready for recapping.

I ran into several of the 22 uF caps that I didn't have, and realized that my 0.47 uF 50V electrolytics are all non-polar, and I don't want to install those where there was a polar cap, so I wasn't able to install all the caps in this board. I did install all that I could and will come back to it after I place another cap order. 

There wasn't anything tricky about the recap and reassembly. Just get the radio board back into position, reconnect the ground terminal on the side of the chassis, plug the cables into the Full CAT board then screw both boards down. I hooked it up to speakers and an antenna and it's working fine.

28 more caps installed, leaving 70 more to go...

Next board: Speaker SW (SSW) PCB

This one is a little harder to get to and requires removal of the front bezel and power and speaker select buttons. It's also a good idea to remove all the selector buttons along the bottom front edge of the receiver. You can leave the radio station preset buttons alone.

Remove the bezel (3 pan head screws on top and 3 flat head screws on the bottom. Then loosen the screws on the both sides of the chassis so that you can slide the front panel assembly forward. Once it is forward you can remove the steel switch pushers from the SSW PCB. Disconnect T4 connector on the underside of the receiver on the decorder PCB. Finally remove three screws that hold the PCB into the chassis, loosen up the cables that connect to it, and work it out of the chassis. 

This board has the speaker select switches, power switch, and speaker protection relay. 

I wanted to do a little additional work on this board- the speaker A switch was intermittent and I wanted to fix that. I ran into similar intermittent switch when I recapped a Soundcraftsmen DX4000 preamp. The problem was that the solder joints had cracked. So when I had the SSW board out of the chassis, I resoldered all the switch/PCB joints. It worked. Both speaker switches are now reliable. Maybe I should go through all the other boards and resolder all the switch joints....

There are only 3 caps to change. C351, 352, and 353. Swap them out and put the pushers back on the switches, then put a drop or two of hot melt glue on each to ensure they don't come off the switches. Then carefully work the board back into position with the switch pushers protruding through the front panel. Push the front panel back and tighten the screws that hold it. Then install the three screws that hold the SSW PCB, and put the button caps back on the pushers.

Speaker switch (SSW) PCB viewed from the top of the chassis.

SSW PCB from a better angle. You can see two of the three caps that should be replaced.

Switch pusher- one of three. They are glued to the switches, but come loose easily if you pull them upward, close to the switches.

Front bezel removed, the speaker switch and power buttons can be removed.

Front panel screws released so the front of the chassis will slide forward.  Also T3 pulled from the Decorder PCB on the underside of the chassis.

These 3 screws will release the SSW PCB from the chassis.

SSW PCB loose, pushers and buttons on the table. I got a little out of order with the photos, sorry. You should be able to wiggle the board out and flip it over to access the solder side and replace the caps.

Hot melt glue on the switch pushers before reinstalling the PCB in the chassis. Don't forget to plug in T3 and push the front part of the chassis back into position- you'll have to wiggle the switch pushers a bit to get them to fit into the slots in the front of the chassis. Once you get it slid back, tighten the screws that hold it there.

That's 3 more caps replaced, 67 to go.

Update 10/12/24

The new cap order has arrived, so I went back through the boards I already had out and replaced the 22 uF caps, and some of the 0.47uF caps with the new ones. That was 5 or 6 caps on two or three PCBs.

While I was doing that I ran into a couple 0.22 uF caps on the radio board that were listed as 0.47 uF on the parts list. Hmmm. I'll take a look at the schematic and see how they are used. I'll either order some 0.22 uF caps or install 0.47uF like the parts list says.

I also checked the new main filter caps for fit - perfect! I unsoldered the old ones from the power amp board without having to remove it from the chassis, then just soldered the new parts in. The pinout matches the old caps, and the new ones are 8 mm shorter than the originals. I ordered 12000 uF 80V caps ($14.10 each) vs the originals that were 10000 uF at 75V. The new parts are rated for 3000 hours at 85C, so should last many years at their normal operating temperature (life doubles for every 10C below rated temperature, so the new 85C parts should last 96000 hours at 35C).

Old and new caps. Identical pinout and spacing, easy replacement even without removing the amp board.

After replacing those caps I tested the receiver again. I momentarily panicked when I couldn't get any sound out of it. But I checked the front panel switches and realized I had inadvertently pushed the "remote control" button. That killed the output, so I pushed that button again, shutting off remote control mode, and the receiver sprang back to life. 

Next: Power amp (PA) board

Removing the PA board from the bottom side of the receiver is relatively easy- remove the switch pushers, take out all the visible screws holding down the output transistors, driver transistors, temperature sensors, etc., from the heatsink, and then a few more screws holding the PCB down on the heatsink and support rail toward the back of the chassis. 

Removing these screws lets the PA board move enough that you can replace the electrolytic caps. While you're in there, clean the old silicone grease off the transistors and thermal washers so you can replace it with some fresh stuff.

The PA board can tilt up when all the screws are removed from the transistors and the ones holding the PCB down on the heatsink and the support rail. It's a little too hard to access the caps that need to be replaced without freeing the PCB like this. Those are the new main filter caps already installed.

When I was unscrewing the transistors from the heatsink, one of the driver transistor screws refused to unscrew and snapped off. I tried grabbing the end of it with vice grips and it still refused to come out. Hmmm.

This is the screw that decided not to come out. Notice how long the transistor leads are- I can drill a new mounting hole closer to the PCB to remount that transistor.

I went through the board and replaced all the electrolytics, then used paper towels soaked in IPA to remove the silicone grease from the heatsink, transistors, and thermal washers.

One of the screws holding the driver transistors down on the heatsink decided not to come out. I grabbed the nub with a vice grips and couldn't get it to move. The solution is to drill another hole, closer to the PCB, and tap it, moving the transistor closer to the PCB.

Now what? Looking at the way the transistors are mounted provided the simple answer. The transistors are mounted far from the PCB meaning their leads are quite long. All I have to do is to move that transistor closer to the PCB, drill a new hole in the heatsink and tap it so I can screw it back down.

The drill and tap will arrive tomorrow...

If that fails, I can always make a metal clamp that screws down at the adjacent transistor and will hold the problem part against the heatsink.

Still another disaster averted! The output transistors have a large metal pad on the back of the package to improve thermal transfer between the die and the heatsink. That metal pad is connected to the collector of the transistor so it has to be electrically insulated from the grounded heatsink. Thin mica washers are used to insulate electrically, with some silicone grease to improve thermal transfer while still providing electrical isolation. I was putting things back together and applied some "arctic silver" heatsink compound to the output transistors and mica washers. As I tightened down the screws, excess material was squeezed out. It occurred to me that Arctic Silver heatsink compound is probably just silicone grease loaded with powdered silver, which means it is probably electrically conductive. If the excess that was squeezed out connected the back side contact to the screws, it could short the transistors to the heatsink. I checked the output transistor lead resistance to ground (i.e. the heatsink) and found that the collectors of both 2SA1215 transistors were shorted to ground! I checked the schematic and no, the collectors of those transistors are not supposed to be grounded. BADBADNOTGOOD! I cleaned off all the Arctic Silver from all four of the output transistors, the heatsink, and the screws. What a mess! I'll be using this silicone pad material to replace the mica washers- no grease needed.

update 10/15/24

The silicone thermal gasket material arrived, along with a tool to punch neat holes in it. I marked it up for cutting and punching and got it done. I reinstalled the board, including the driver transistor that I had to move, checked to make sure there were no shorts between the output transistors and ground, powered it up- no smoke, sounded great!

Silicone gasket material and spinning punch used to make new gaskets for the output transistors.

Power transistors reinstalled with new silicone gaskets. Driver transistor reinstalled in new location because its screw broke off in the heatsink. Rectangle at the bottom shows some minor damage to the PCB occurred when I removed the old main filter caps. These boards are brittle!

That's a total of 14 more caps replaced, leaving 53 (I think) more to go...

Next up:  a nightmare!

The rest of the boards that need recapping are all located in the front of the chassis. 7 or 8 of them are all in a sub chassis that is screwed to the main chassis. Getting those board out requires removing dozens of cables, and it's getting too complicated for me to document it because I'm having to try different things to get at the sub chassis, some of which work and others don't.

The front of the chassis can be folded down to get access to the back of the subchassis where most of the cables connect. You have to loosen the screws on the sides, then pull the front forward and tilt it down. The cables are working against doing that. So I think the best approach is take a bunch of photos, then clip zip ties holing the cables together, then tilt the chassis down and disconnect the cables. Then unscrew the sub chassis and take it out of the main chassis, then start attacking the individual PCBs in the sub chassis. 

The Speaker Switch board has to come out again to get the sub chassis out.

There are screws on the sides of the subchassis that have to be removed to get some of the boards out and I don't think you can get at them without removing the subchassis, so out it goes!

RX103 receiver chassis with the subchassis removed. I clipped a lot of zip ties to loosen up the cables and make tilting the front of the chassis down easy. The subchassis comes out from the back side of the front of the chassis, and has to go back in that way, too.

This is the bottom/front of the subchassis. There are 8 PCBs attached/packed into this little assembly. Some of the PCB slide out toward the front, some slide out toward the back, and a few are screwed down.

This is the back of the subchassis. You can see many of the connectors and some of the labels on the plugs. There were a total of about 50 capacitors that had to be replaced scattered among these PCBs.

Getting it back together shouldn't be impossible- each connector is marked with a unique identifier and the matching connectors on the PCBs are also marked. I marked a few of the connectors on the back side and marked the PCB where they plug in. The real trick will be not missing any of the cable connections, and getting the cables back into place and zip tied when it's all recapped. 

Forget about recapping one PCB and then testing at this point. It will be far too much effort to put it all back together. I'll just do the recapping, reassemble, hope for a miracle, then test.

Wish me luck!

Update 10/17/24

I have finished replacing every electrolytic cap in the receiver! Now I have to put it back together so I can test it.

Update 10/18/24

I spent a few hours putting the subchassis back into the front of the receiver and reconnecting all the cables. There was one cable with the label smudged beyond reading, and a couple that had duplicate labels that required tracing the wires through the schematic diagram to figure out where they go. Once I got them all reconnected, and triple checked to make sure I wasn't missing any, I crossed my fingers, plugged in the receiver, and hit the power button. The standby light flashed for about 5 seconds, then the speaker relay clicked and out came very nice sounding music from the FM tuner. All the front panel lights worked and all the buttons worked. The vacuum fluorescent displays worked fine, too. Tone, volume, balance, and loudness controls all worked fine.

I put all the button caps except the preset memory buttons back, reinstalled the three long switch pushers, remounted the front panel, put in the preset memory button caps, and put in a couple zip ties on some of the cables- not nearly as many as there were originally, but I still have to fix the suckface when the belts I ordered arrive, and I'll need to tilt the front part of the chassis out of the way to get at the mechanism. That will be easier with the cables hanging loose. Finally I put the bottom and top covers back on the chassis and powered it up and let it run for a few hours to see if anything would quit working. No problems!

I marked the bottom cover with a permanent marker to indicate that all electrolytic caps were replaced in October of 2024. 

Woohoo!

Looking back on the whole process, here's what I recommend to anyone thinking of recapping one of these Luxman receivers. From the start, remove the top and bottom covers, front panel and buttons, all knobs and extenders except for the volume control knob (the extender for that one is screwed to the potentiometer shaft), and the three long switch pushers. Clip most of the zip ties off the cables- it will be a lot easier to move the PCBs around. Be sure to have new thermal washers ready in advance for the output transistors. Have fun!