• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

NUBUSPAS16-DB :: A daughterboard hack to restore some functionality of a standalone MediaVision PAS16


Well-known member
Introduction pitch
A member of the forum gave me this card (without the matching breakout box) for R&D purposes. Sometimes R&D gives birth to something so here we are.
The initial design for a breakout box was plagued with a few issues, like missing information about some custom chips and lack of time and interest to build a new external box without having the original hardware (you can guess so much from pictures)

Another member showed interest for an internal solution as well, so here we are: NUBUSPAS16-DB. It's a daughter board designed to be soldered on the original board, in lieu and place of the HD26 connector.

For people who are scared of such modifications, I've determined that the PAS line output and some input can be somewhat operational with a simple breakout cable. Such a cable wiring will be posted on the patch panel thread with a few other notes. Others feature do need active components

Some sound measurement were also made during this project, they will be posted in this thread
The design is open source and will be posted on GitHub. It only works for the NuBus


The board restore the following features:
  • Line output (buffered)
  • Headphone amplifier (including support for high impedance headphones)
  • Line in (switched between internal and external, more on this below)
  • Hardware mixing of inputs and sound enhancements (bass, treble, loudness, stereo enhancer, external mute)
The OPL is probably working as well in this configuration, but no idea how to double check this

The following features are not restored:
  • Microphone preamp
  • Hardware mixing of multiple line in
  • PC Joystick
  • MIDI
The following features are added:
  • debug LEDs (power, signal, peak for each channel)

Design choices
The board is designed to be as close as possible as MediaVision intended it to be and from what I could understand from the breakout box photos. It's using the same op amps (LF347s) with a balanced power design (± 12V). This ensures that the sound coloration is close to what MediaVision designed and is keeping the noise level as close as low as possible (to some extend…)

Please note that the breakout box is likely using a single supply with biased opamp, which would add noise, but I can't double check that
Jacks are more or less following the AC'97 standard (green = line out, blue = line in), with the headphone output beeing black

The original headphone amplifier used discrete component with a design that is known to not be the best. It was probably done as a cost saving measure.
In contrary, the headphone amplifier of NUBUSPAS16-DB is a modern twist: it is build around a modern op-amp especially designed to drive headphones (OPA1688).

It will provides up to around 3V RMS for high impedance headphones (600Ω), which the voltage delivered been proportional to the headphone impedance.
If your downstream equipment has sufficient headroom (at least +10 DBu) , the headphone output can also be used as a high level line output with a better SNR than the standard line output.

Input sources
There is actually only one real input on this board. Line inputs 1-4 were unusable on my board, either because i'm too dumb and missed something, my card was damaged or the proprietary chips on the breakout board are doing something that I don't know. Any sound input on any the left channels would result in a -20dB crosstalk on the right channels, which make that unusable

I so defaulted to be used the microphone input, which on the NuBus board side is a simple stereo line in input. You can then mix the mic input to the output using the PAS16 Control Panel. It allow you, other than plugging your old tape deck, to daisy chain your internal sound card. The PAS will then mix the sound coming from both sound card. It's useful if you have software that isn't compatible with the PAS and or is accessing the internal sound chip directly.

It's also useful for daisy chain internal CD Audio. Speaking of which, an additional CD Audio input is available at the back of the card which allow for either connecting a SCSI drive here, or @rabbitholecomputing zuluscsi with a @saybur DAC Attack! module

Plugin anything in the blue jack will disable the internal audio in

Overall this is a niche project, but I hope it will be useful for some people !


Don't mind the bad paint, my dremel died during the operation so I had to finish the bracket by hand. A 3D printed bracket likely be be better than dremeling the old one
Last edited:


Well-known member
Build log
First remove bracket. Then desolder C101, C110 and the HD26 connector
In my case C101 was dead, the card needed a recap anyway


Assembling the daughter board
This requires soldering all three jacks, both Molex connectors and both capacitors. Everything CMS can be assembled by your favorite PCB service. Otherwise it's 0805 and SOIC stuff, like the NuBus card



Stacking headers
This hack is using stacking headers to keep both cards together. You will need to cut the middle of some stacking headers like this, otherwise it doesn't match the NuBus card holes


Preparing the NuBus board
Solder the 4 stacking headers, and three dupont wires for where it's not possible to use stacking headers


Attach daughter board

First connect all three dupont cables to the internal connector (J1)

Then assemble both boards together. You may want to use some PCB as a shim here, it helps levelling the daughterboard a lot. Also it's ok if there is some room left on the stacking headers (as long as the pin goes through, it's ok)


That's the worst part. Dremel the bracket so the jacks can go out. Paint if needed.


It's done :) Enjoy your reborn PAS16


Well-known member
Amazing work! I have the breakout box but am tempted to do this!

Thanks a lot ! I would keep the box, the card is more valuable this way, but you do you :)

The gerbers files and design has been released on GitHub here
Will try to release the measurements this week-end as well


Well-known member
Some measurements and subjective remarks. This has been benchmarked against a Quadra 650 board (Wombat). I did it to double check something wasn't broken with the daughter board

Output levels
The last notch on the PAS16 Control Panel volume is an overdrive, 100% volume is last position minus one, with both bass and treble controls at 50%. Volume control at max level results in high output levels and likely to include clipping (including the PAS itself)
This is important to avoid clipping / distortion if your other gear doesn't have enough headroom.

Measured values

Volume positionBassTrebleVoltage out (RMS)dbV
max -150%50%0.810-1.8
vintage gear (reference)--0.316-10

Stereo enhancer and Loudness don't add much level
Headphone output is about 200% of the line output at high impedances

Sound quality

The sound quality of the card is pleasant, but it's not high fidelity. It's typical of late 80 hardware with FET amps (think TL072 sound), with a noticeable level of hiss and noise. You don't have the muffled sound present from most 68k Macs.

At 22k sample rate, the card would output frequencies up to 16-17 kHz, which should be impossible, that's what triggered me to do all theses measurements

The DAC is a CS4216-KL which is closely related to the one used in ... Beige G3s (The whisper personality card used a CS4212-KL). The DAC in itself itself, but the card on a whole is noisy

The noise floor is about -52dBFS from my measurements. Also note that the Wombat isn't much better, because the damn audio click bleeds into the stereo output


Frequency response
So enter the first round of tests. Here is the frequency response of the card vs a few others ones


  • Green is the PAS16 (with DB)
  • Red is a modern Pro sound card (RME HDSPe AIO)
  • Blue is a Quadra 650
Notice how the PAS is increasing both the top end and low end to have a more dynamic sound
The wombat start giving up at 4kHz, which explain the moffuled sounds. Phase reversal is also really bad

Harmonic Distortion
Had a lot of trouble here because I couldn't believe the measurements. I rebuild a few daughterboards and probed around the board to double check. This happened:

The number was consistent after a few days of measurements so that's where the ballpark is. The card is so noisy and the THD so high that the measurement software couldn't even measure at a moderate level.


  • Yellow is a Wombat board​
  • Blue is the headphone output (OPA1688)​
  • Orange is the line output (LF347)​
That's where the trick is. The PAS is generating a ton of odd/even harmonics, which explain the frequencies which shouldn't be there at low sample rates (11kHz and 22kHz). This technique was used on harmonic exciters and was commonly used in the '80s and '90s in recording studios

It distords the sound a lot, so quite the opposite of HiFi, but gives a better sound especially when the high frequencies and some of the mid ones are missing like in our 68k Macs. However, with "modern" samples rates (44.1 kHz) it's just more noisy


So this is really a gaming sound card before gaming sound cards where a thing 😝. It's very nice for games, and of course Player Pro, but as a production card that's not it. This is an Apple G3 era quality sound card in a 68k with a few bell and whistles


Amazing stuff, love it! Might I suggest drilling three holes in a blank bracket?

Had the bare NuBus card in the backwhen. Luckily Shreve Systems did a blowout sale, so I got the boxed LC version for a song and was up and running for stereo output from the DuoDock.

Thanks, this reminds me that I owe someone here a closeup pic of the resistors on the breakout box PCB.


Well-known member
Amazing stuff, love it! Might I suggest drilling three holes in a blank bracket?

Thanks, this reminds me that I owe someone here a closeup pic of the resistors on the breakout box PCB.

Thanks ! I didn't have a blank bracket around but that's a good idea !
The "someone" is me. Don't bother, I found the resistors values. I did this project using your previous pictures, so thanks a lot !


Well-known member
I will switch to a few others projects for now, but maybe someone can take it from here !


Well-known member
Just wanted to say, interesting project! I like the idea. May be useful for "lonely" cards that lost their breakout box.

Amazing stuff, love it! Might I suggest drilling three holes in a blank bracket?
Do you know sources for blank NuBus brackets? I have some ideas and that might come in handy.


Well-known member
I guess that works in a pinch, but it's not the same thing. Having the actual NuBus bracket is nice as it helps secure the card in the slot.


Well-known member
@Arisotura They used to sell them on Mouser, Digikey, etc. Don't know if it's still true.

Edit: Never mind. I don't see any.
Last edited:


Well-known member
This is awesome. For anyone a fan of Audio Science Review these charts are a fun comparison. How far we've come!

Now I just want one even more after missing one go for $65 recently :(.