High-End Active Monolith Prototype: Help Isolating Internal Streamer PCB from Hypex FA123

[Adrian from London]

Hi everyone,

I am building a high-end active floorstander prototype with the long-term ambition of creating a small, designer-maker series. The design relies on a strict AirPlay/wireless workflow housed entirely within a minimalist, all-in-one enclosure. I am looking for some peer review on my technical layout, internal bracing, and component shielding before I begin machining the timber.

The Architecture:

• Cabinet: 110cm (H)×28cm (W)×34cm (D) built from Baltic Birch Plywood (24mm front baffle, 18mmpanels). The cabinet will feature extensive internal window bracing to ensure it is structurally inert.
• Acoustics & Drivers: Each tower houses two 8" woofers and a single tweeter, utilizing a front-facing horizontal slot port at the very base for bass reflex.
• Amplification: A Hypex FusionAmp FA123 plate amp recessed into an 8cm deep, airtight pool on the upper rear panel. This pool is covered by a removable powder-coated steel mesh grille.
• The Technical Vault: At the absolute bottom of the cabinet, there is an isolated, independent plywood box/vault that slots completely out of the cabinet for servicing or future hardware upgrades. It is secured by four screws from the outside.

The Layout Concept & The Tech Challenge: A core feature of this speaker is a single, clean mains power cable input at the bottom. No external boxes hanging off the back, and no visible jumper cables.

Inside the removable lower plywood vault, I am integrating:

1. Main AC power inlet with an illuminated rocker switch.
2. An internal step-down power supply (e.g., Mean Well) to deliver 5V DC.
3. An internal streaming transport board running Toslink straight up into the Hypex deep pool.

My current plan is to use a naked, stripped-down PCB from a WiiM Pro (casing removed) mounted on standoffs. While I am not worried about the static magnetic fields from the 8" woofers, I am highly concerned about the physical realities of combining high-output drivers and wireless components in one cabinet.

I would love your engineering feedback on three specific bottlenecks, and I am entirely open to alternative streaming boards/modules if you think the WiiM isn't the best choice for this application.

My Core Concerns:

1. Mechanical Vibration vs. Digital Clocks: With two heavy 8" woofers moving massive amounts of air, physical cabinet vibration is a serious threat to the clock oscillators on a bare streaming PCB (microphonic jitter, long-term solder fatigue). Is isolating the PCB inside an independent, modular plywood "cassette" at the base of the speaker structurally sufficient to decouple it from these low-frequency energy transfers, or should the PCB standoffs themselves feature elastomeric damping?
2. The Antenna vs. RFI Shielding Paradox: Since the lower vault is made of wood and sits completely outside the rear steel mesh zone, the Wi-Fi signal can easily escape to connect to the network. However, those same radio waves can travel upward through the vault's wooden ceiling straight into the Hypex input stage. How do I achieve directional RF shielding? Is lining just the inner ceiling of the vault with grounded copper foil enough to block RF "chirping" without trapping the Wi-Fi signal and killing network connectivity?
3. Internal Power Routing & Noise: I am running AC mains power into the vault, stepping it down to 5V DC for the streamer, and then routing AC power up out of the vault to power the Hypex. Given the tight space inside a sliding plywood enclosure, what are the best practices for shielding the AC lines so they don’t induce hum into the audio board or the optical lines?
4. Wireless Stereo Sync: To avoid a physical cable running between the Left and Right independent speakers, I need a rock-solid wireless stereo sync. If network-grouping two independent streamer boards introduces too much latency or drift for a premium product, what alternative internal modules (like WiSA) should I look at?

I’ve attached my scaled graph paper schematic showing the front/back configurations and the spatial constraints of the sliding vault vs. the Hypex pool. Any advice on grounding topology, component layout inside the box, or RF/vibration isolation would be incredibly helpful.

Thanks,

Adrian

 

[Vincent Kars]

AirPlay

Airplay 2 has the habit of converting to 256 AAC when communicating with other Airplay 2 devices. Have a look at UPnP.

 

[Adrian from London]

Airplay 2 has the habit of converting to 256 AAC when communicating with other Airplay 2 devices. Have a look at UPnP.

Thanks Vincent, that's a brilliant catch. I had no idea AirPlay compressed everything down like that. Since the WiiM board handles both, I guess I can let the user choose: UPnP for serious lossless listening, and AirPlay for casual convenience.

Since you know this architecture well, how does switching to UPnP affect the wireless stereo sync between two independent cabinets?

If I link two separate WiiM boards over Wi-Fi as a Left/Right pair, does streaming those heavy, uncompressed files via UPnP cause latency or timing drift between the channels? If network sync is too risky for a high-end speaker, what setup would you use to keep the two boxes completely wireless?

 

[Vincent Kars]

Since the WiiM board handles both, I guess I can let the user choose: UPnP for serious lossless listening, and AirPlay for casual convenience.

I expect it to work different.
Basically Airplay/UPnP is about choosing a source.
I expect the WiiM uses the Linkplay Music Streaming Platform to communicate between WiiM devices.

 

[Adrian from London]

I expect it to work different.
Basically Airplay/UPnP is about choosing a source.
I expect the WiiM uses the Linkplay Music Streaming Platform to communicate between WiiM devices.

That makes total sense about Linkplay. Knowing the Left/Right sync is handled natively under the hood is a massive relief, thank you!

Regarding the physical vault at the bottom, I am trying to figure out if cabinet vibration transferring through wood-on-wood contact is something I should actively design around to protect the bare PCB clocks, or if I am overthinking the microphonics issue entirely.

If it is worth addressing, my thought was to either line the sliding tracks with thin rubber/neoprene foam tape to damp the whole vault, or simply use silicone anti-vibration standoffs directly under the bare PCB.

I would love to hear any thoughts or experiences the community has on this, or if it's a non-issue in practice.

 

[Redacted]

I still stuck on seeing Airplay2 and high-end in the same paragraph.

But: (and sorry if this has already been put to bed)

1) I doubt acoustic vibration will affect Xtal controlled clocks; there is slight frequency mismatch there! Solder: Look at the millions of applications in harsh environments where solder holds up. Sure, there is theory v. praxis.
2) RFI - if worried, a simple grounded metal box or mesh works wonders, but nothing in your system will affect networking. Sure, there is theory v. praxis.
3) You can twist mains wires together. And work out what is common mode and what is not!
4) Just use wire. It works with no perceivable quality loss, delay/sync, or interference susceptibility

• Mechanical Vibration vs. Digital Clocks: With two heavy 8" woofers moving massive amounts of air, physical cabinet vibration is a serious threat to the clock oscillators on a bare streaming PCB (microphonic jitter, long-term solder fatigue). Is isolating the PCB inside an independent, modular plywood "cassette" at the base of the speaker structurally sufficient to decouple it from these low-frequency energy transfers, or should the PCB standoffs themselves feature elastomeric damping?
• The Antenna vs. RFI Shielding Paradox: Since the lower vault is made of wood and sits completely outside the rear steel mesh zone, the Wi-Fi signal can easily escape to connect to the network. However, those same radio waves can travel upward through the vault's wooden ceiling straight into the Hypex input stage. How do I achieve directional RF shielding? Is lining just the inner ceiling of the vault with grounded copper foil enough to block RF "chirping" without trapping the Wi-Fi signal and killing network connectivity?
• Internal Power Routing & Noise: I am running AC mains power into the vault, stepping it down to 5V DC for the streamer, and then routing AC power up out of the vault to power the Hypex. Given the tight space inside a sliding plywood enclosure, what are the best practices for shielding the AC lines so they don’t induce hum into the audio board or the optical lines?
• Wireless Stereo Sync: To avoid a physical cable running between the Left and Right independent speakers, I need a rock-solid wireless stereo sync. If network-grouping two independent streamer boards introduces too much latency or drift for a premium product, what alternative internal modules (like WiSA) should I look at?

 

[Redacted]

While anticipating every conceivable problem in advance feels the right thing to do, there is something to be said about building kit and then seeing what the real problems are.

 

[Vincent Kars]

I would love to hear any thoughts or experiences the community has on this, or if it's a non-issue in practice.

Have a look at products like Dutch&Dutch 8C or Kii Three or your car. All run using processors and function correctly....

 

[Adrian from London]

I still stuck on seeing Airplay2 and high-end in the same paragraph.

But: (and sorry if this has already been put to bed)

1) I doubt acoustic vibration will affect Xtal controlled clocks; there is slight frequency mismatch there! Solder: Look at the millions of applications in harsh environments where solder holds up. Sure, there is theory v. praxis.
2) RFI - if worried, a simple grounded metal box or mesh works wonders, but nothing in your system will affect networking. Sure, there is theory v. praxis.
3) You can twist mains wires together. And work out what is common mode and what is not!
4) Just use wire. It works with no perceivable quality loss, delay/sync, or interference susceptibility

• Mechanical Vibration vs. Digital Clocks: With two heavy 8" woofers moving massive amounts of air, physical cabinet vibration is a serious threat to the clock oscillators on a bare streaming PCB (microphonic jitter, long-term solder fatigue). Is isolating the PCB inside an independent, modular plywood "cassette" at the base of the speaker structurally sufficient to decouple it from these low-frequency energy transfers, or should the PCB standoffs themselves feature elastomeric damping?
• The Antenna vs. RFI Shielding Paradox: Since the lower vault is made of wood and sits completely outside the rear steel mesh zone, the Wi-Fi signal can easily escape to connect to the network. However, those same radio waves can travel upward through the vault's wooden ceiling straight into the Hypex input stage. How do I achieve directional RF shielding? Is lining just the inner ceiling of the vault with grounded copper foil enough to block RF "chirping" without trapping the Wi-Fi signal and killing network connectivity?
• Internal Power Routing & Noise: I am running AC mains power into the vault, stepping it down to 5V DC for the streamer, and then routing AC power up out of the vault to power the Hypex. Given the tight space inside a sliding plywood enclosure, what are the best practices for shielding the AC lines so they don’t induce hum into the audio board or the optical lines?
• Wireless Stereo Sync: To avoid a physical cable running between the Left and Right independent speakers, I need a rock-solid wireless stereo sync. If network-grouping two independent streamer boards introduces too much latency or drift for a premium product, what alternative internal modules (like WiSA) should I look at?

Thanks for sharing your thoughts on this! It's a great reality check. It's really reassuring to hear that the digital clocks and solder points will easily handle the cabinet energy, and that a simple grounded copper shield will clear up the RFI.

I completely agree with what you're saying about just getting on with the build instead of over-analyzing everything on paper. My main goal was just to avoid some complex isolation fixes if they could be easily skipped, so this saves a lot of unnecessary effort.

Since you mentioned twisting the mains wires together to cancel out the hum, I’ll definitely do that for the internal routing. To make sure the vault is completely safe, what is the best practice for tying that copper ceiling shield into the main AC earth ground pin on the power inlet within a wooden enclosure? Is it as simple as a single ground wire running to the incoming earth?

 

[Redacted]

The main reason to ground any part of your kit at the mains, is down to electrical safety, in the event of a failure that means someone could touch something that might be live.

I'm a bit loathe to get too prescriptive about anything to with mains, as while I have a relevant degree, I don't know your local regulations - I could be in a difficult position if you burn down a street. You may also want to look up common mode interference with a view to basic ideas, as it's hard to just say "do this and not that" when I can't really visualise what you've got.

I keep getting a picture of a guitar player trying to add a Faraday cage around his pickups really ineffectually! Try to explain grounding, and how no RFI should get in or out of a cage, and how all that might affect pickups, when a guy is asking about how copper tape affects tone! That guy is not you.