Occam’s Razor and a USB-C to HDMI adapter

The simplest explanation is likely the best, even if its validity isn’t always intuitively obvious in advance.

Speaking of USB-C, and the cables (including splitters) that connect to it to other things…

My long-in-the-tooth Intel-based 2018 Apple Mac mini is still sitting on my desk, humming to my right, even though its M2 Pro Apple Silicon-based successor has been sitting downstairs in storage awaiting its turn in the spotlight for going on three years now. The legacy hardware is no longer a candidate for new Apple operating system releases, but it’s still receiving bug fixes and security patches, until sometime in the second half of next year, if past-history trends remain valid. And as the saying goes, “if it works, don’t touch it” (I’d also be tempted to haul out the “a penny saved is a penny earned” quote, but since I’ve already bought its replacement…).

In conjunction with my transition to it from its 2014-era Mac mini predecessor roughly 2.5 years ago, I upgraded the two-display suite above it to a set of Dell P2415Q 4K LCDs.

The Mac mini leverages the displays’ HDMI inputs; a “Project Volterra” Windows-on-Arm dev kit 2023 stacked above it connects to those same LCDs over their DisplayPort connections, and front panel buttons toggle the displays between the two systems when running concurrently.

Look back at that 2018 Mac mini rear panel “stock” photo a couple of paragraphs ago, however, and you’ll only see one HDMI output, into which I’ve plugged a “straight” HDMI cable running directly to one of the LCDs. How, then, did I connect the system to the other display’s HDMI input? That’s where today’s teardown victim enters the picture.

It’s an Anker A8730 6’ USB-C to HDMI adapter cable; I’d bought and pressed it into immediate service way back in mid-2021. I sourced from Anker’s outlet store on eBay, where it cost $25.90 as a refurb. So, you can say I’ve gotten my money’s worth out of it! Speaking of fiscal topics, there’s a 0.75′′ (19.1 mm) diameter U.S. penny alongside in the second (and non-stock) image that follows, along with subsequent others, for size comparison purposes.

A multimedia-centric repurpose

How it works is (at least to me) an interesting story in and of itself. It fundamentally leverages a DisplayPort technology called “Alt Mode”, which repurposes USB 3.x’s “SuperSpeed” data and Configuration Channel (CC) lines to transport dedicated audio and video streams sourced from the graphics and sound subsystems. The CC leverage explains why, among other implementation reasons, it’s only applicable to USB-C, not also to prior-generation USB physical connector standards.

Macs don’t additionally support DisplayPort’s Multi-stream Transport (MST) mode, either in the operating system (generally) or in hardware (for newer Apple Silicon-based systems). If MST support had existed, it would have conceivable enabled me to extended-mode tether both displays to a single USB-C (or Thunderbolt 3, aka TB3, in this case) system connector, in combination with a separate inter-display connection. But again, per the system back panel photo, I had plenty of spare connector candidates available, not even counting those on the separate expansion hubs (the latter added subsequent to the snap of the earlier “stack” picture).

But what’s with all this DisplayPort talk? I thought we were discussing connecting the computer to a display’s HDMI input! We are, and that’s where the bulge surrounding the adapter cable’s HDMI connector end comes in. Conceptually, and I hope to definitively confirm today, inside is likely a DisplayPort to HDMI protocol-converting bidirectional transceiver; a particularly robust one, it seems, since the Anker A8730 touts 4K 60 Hz specs. HDMI once also had an “Alt” mode for USB-C, at least on specification paper, although actual products never made it to market.

Erratic function = dissection rationalization

Since the adapter cable had been operational in my abode since mid-2021 (and, considering it was a refurb, maybe even earlier than that with its original owner), why’d I decide to take it apart? Curiosity was as always one motivation, although I generally hold that particular urge at bay for as long as the gear remains operational. But this one had eventually failed, in another story-in-and-of-itself, through whose telling I aspire to also rationalize the “Occam’s Razor” title reference.

As previously mentioned, the 2018 Mac mini is still receiving periodic patches for its last-supported MacOS 15 “Sequoia” operating system release. A few updates ago, the left-side display of the pair (the one leveraging the Anker adapter) was no longer recognized by the system after the upgrade and subsequent reboot. Online research revealed that mine wasn’t a unique post-update quirk, and swapping the adapter cable to a different TB3 system connector got the display going again.

So, I figured that an obscure “Alt Mode” bug (specifically, since the first TB3 port still worked fine for generic data) introduced along with the upgrade had made it through Apple’s pre-release QA checks. This purported scenario was unsurprising albeit still disappointing, given that Intel-based Macs in general, and this system in particular, were on their “last legs” and likely no longer receiving their fair share of QA attention. Why my surmised bug only affected one TB3 port but not another wasn’t clear, but…

The next time I did an update, however, the exact same thing happened, this time to the second TB3 connector. Swapping the cable adapter back to the first TB3 connector didn’t resurrect the display. And I no longer had any other spare TB3 ports on the system available to try. At this point, I began to fear I had a fundamental system hardware degradation issue on my hands.

But after unplugging the cable adapter from the Mac mini and trying it with the “Project Volterra” system instead, where it also didn’t work, I was reassured by the now-nonfunctional repetition across multiple systems (and O/Ss) that a dead cable adapter was the likely culprit. An “Amazon Basics” cable adapter replacement fired up fine, sealing the Anker A8730’s fate.

What was behind the cable adapter’s initial failure, temporary resurrection, and eventual permanent demise? I hoped the pending teardown might provide visual Achilles’ Heel evidence (hold that thought). But I suspect it has something to do with MacOS’s use of display-supplied Extended Display Identification Data (EDID) information for software interface purposes versus Windows’ leverage of (when available) display-specific drivers.

My guess is that EDID (re-) enumeration in MacOS happens both when a display is first-time plugged into a new system connector and after each sufficiently impactful operating system update. In this particular case, the aforementioned DisplayPortHDMI bidirectional protocol translator in-between the system and display initially began operating erratically and eventually failed completely. But that’s just my conjecture; reader theories in the comments are as-always also welcomed!

Eyes-on conceptual validation (or not)

Speaking of tearing down, let’s dispense with further abstract chit-chat and get to dissecting, shall we? Here are a few more real-life device photos as prep. Packaging is long gone at this point, along with any potential originally accompanying literature. The USB-C end, to start.

With a product-code sticker behind it.

Intermediary cable markings next.

And now the bulge-augmented HDMI other end, starting with the seam-less cable-intro side.

The connector side conversely does have a visible seam at its circumference.

But attempts to surmount it with hair dryers and heat guns, along with spungers and such, were for naught. So, I escalated my attack by breaking out the hacksaw with hopeful deft technique.

That’s what I’m talking about!

For anyone following in my footsteps who prefers a less Neanderthal-reminiscent dissection approach, here’s the HDMI connector-surrounding piece I was unsuccessfully trying to extract earlier. Note specifically the locations of the retention tabs.

Finally, what you’re all most interested in, the PCB. As I’ve mentioned before, “top”, “bottom” and other orientation terms are particularly nebulous where HDMI is concerned. So instead, here’s the side corresponding to the wider HDMI connector edge.

Now zooming in for a closer look.

I was admittedly surprised to encounter a preponderance of passives, given that the most common adapter-usage orientation would have put this side on top, with the remaining (and proportionally higher heat-generating) stuff you’ll see next on the other side and below it. Given that heat rises (don’cha know), and that there’s no passive ventilation venting available, that’s a seeming premature failure-inducing decision. But not in my case, since the HDMI connector points downward with this display, not horizontally straight out the back as with many others.

Components of particular note include a five-lead SOT23-packaged step-down (buck) DC-DC converter at lower left, marked as follows.

JWA5J
91D5T

There’s also a “2R2” (2.2 µH) inductor to its immediate left, and a 27 MHz oscillator to its right. And what does that clock chip likely drive? Let’s flip the PCB over to the other side.

Now that’s more like it (unless you’re into passives, that is). Again, we zoom in for a closer look.

At lower left is Via Labs’ VL100, a (surprise, surprise…not…) DisplayPort USB-C Alt-mode controller. To its right is (surprise, surprise…truly, this time, at least a bit…) another oscillator, 24 MHz this time. Continuing to the right is, I’m guessing (readers?), a serial interface nonvolatile memory for code and data storage purposes, marked as follows.

125S40
BG17K8
P19030

In the lower right corner is another JWA5J DC/DC converter. Above it is (once again, no surprise) Paradise Technologies’ PS176 DisplayPort to HDMI video interface converter. And again for the passives fans among you, there are plenty more examples to see on this side of the PCB, too!

That’s a “wrap” for today, folks. Share your thoughts with me and your fellow readers in the comments, please!

—Brian Dipert is the associate editor, as well as a contributing editor, at EDN.

Related Content

The post Occam’s Razor and a USB-C to HDMI adapter appeared first on EDN.

Source link