A teardown story of two not-so-different switches

June 17, 2025

2

A teardown story of two not-so-different switches

Eleven years in the past, my spouse and I skilled the aftereffects of our first close-proximity lightning blast right here within the Rocky Mountain foothills, clobbering (amongst different issues) each five-port and eight-port Gigabit Ethernet (GbE) switches, each of which ended up going beneath the teardown knife. The failure mechanism for the primary swap ended up being non-obvious, in sharp distinction to the second, whose controller chip ended up with a number of holes blown in its bundle prime:

One yr (and a decade in the past) later, lightning struck once more. No Gigabit Ethernet switches expired this second time, though we nonetheless misplaced another gadgets.

Quick ahead to 2024, and…yep. This time 4 GbE switches ended up zapped failed, two of them eight-port and two extra five-port (destiny was apparently taking part in catch-up after beforehand taking a cross on switches). The previous two will likely be showcased at present, with the others following quickly. Then there’s the three-bay NAS; you’ll have have already got seen that teardown by the point this piece is printed. And one other CableCard receiver (we’re three for 3 on these), together with one other MoCA transceiver…you’ll get teardowns of these within the close to future, too.

At the moment’s dissection sufferers are from the identical provider—TRENDnet. They hail from the identical product household technology. And, as you’ll quickly see, though their outsides are (considerably) dissimilar, their insides are primarily equivalent (given the naming and launch date similarities, that’s not precisely a shock). Behold the metal-case TEG-S82g ({hardware} v2.0r, to be exact):

which, per Amazon’s itemizing, dates from September 2004, and the plastic-case TEG-S81g (once more, {hardware} v2.0r), additionally initially out there that very same month and yr:

Let’s begin with the metal-case TEG-S82g. Following up on the inventory pictures proven earlier, listed here are some views of my particular system, as traditional accompanied by a 0.75″ (19.1 mm) diameter U.S. penny for dimension comparability functions (the TEG-S82g has dimensions of 150 x 97 x 28 mm/5.9 x 3.8 x 1.1 in. and weighs 364 g/12.8 oz.). Entrance:

Left aspect:

This subsequent one, of the system’s bottom, begs for a bit extra clarification. Port 8, the one initially related to one of many two spans of shielded Ethernet cable operating across the outdoors of the home, is unsurprisingly the one which failed (due to this fact the electrical tape I utilized to establish it).

The opposite ports really nonetheless work, at the very least for the primary minute or few after I energy on the swap, however finally all of the entrance panel LEDs start blinking and additional performance ceases:

Onward. Proper aspect:

Prime:

and backside:

Right here’s its “wall wart”:

These screw heads you may need seen on each system sides? They’re our pathway inside:

Right here’s our first view of the PCB inside:

4 screws maintain it in place. Let’s eliminate these subsequent:

Let’s see what we’ve received right here. On the backside are the eight Ethernet ports, subsequent to (on the backside proper) the DC enter energy connector. Thick PCB traces operating from there to the circuitry cluster within the higher proper quadrant recommend that the latter handles energy technology for the rest of the board. And above, every two-port combo is a Bi-TEK FM-3178LLF twin port magnetic transformer. Right here’s the precise one (at far proper) related to failed port 8:

On the prime edge are (at far proper) the ability LED, subsequent to eight exercise LEDs, one for every of the ports. And under them is the system’s “brains”, a Realtek RTL8370N 8-port 10/100/1000 swap controller. It might very properly be the identical because the IC within the 8-port swap teardown from 11 years in the past, though I can’t say for positive, as that one had chunks of its packaging (due to this fact topside markings) blown away! That mentioned, this design does use the identical transformers as final time.

Right here’s a close-up of the RTL8370N and the aforementioned circuitry to its proper:

Now let’s flip the PCB over and take a look at its bottom:

No apparent proof of injury right here, both. Right here’s one other port 8 space closeup (as I used to be penning this, I paused to revisit the {hardware} and ensure that these white globs are simply mud):

Now for its plastic-case TEG-S81g sibling, with listed dimensions once more 150 x 97 x 28 mm/5.9 x 3.8 x 1.1 in. (albeit this time tapered within the entrance), though the load is (unsurprisingly, given the shift in case materials building) decreased this time round: 186 g/6.6 oz.:

This time, port 5 failed. The opposite seven ports stay totally practical to this very day, though for the way for much longer I can’t say; due to this fact, I’ve determined to retire it from energetic service, as properly, within the curiosity of future-hassle avoidance:

The “wall wart” appears to be like totally different this time, however the specs are the identical:

No screws on the case sides this time, as you might have already seen, however take away the 4 rubber “ft” on the underside:

and beneath the entrance two are seen screw heads.

You recognize what comes subsequent:

And we’re in (with the tape nonetheless caught to the highest):

Let’s put that tape again in place so I can preserve observe of which port (5) is the failing one:

The sooner-shown two screws did double-duty, not solely holding the 2 halves of the chassis collectively but additionally serving to preserve the PCB inside in place. Two extra, towards the again, additionally should be handled earlier than the PCB could be free of its plastic-case captivity:

That’s higher:

One other set of closeups, first of the affected-port area:

and the majority of the topside circuitry:

And now, flipping the PCB over, one other set as earlier than:

I hope you’ll agree with me on the next two factors:

The 2 PCBs look equivalent, and
There’s no visually apparent cause why both one failed.

So then, what occurred? Let’s start with the plastic-case TEG-S81g. Reality be advised, the tape on prime of port 5 initially existed in order that I may keep in mind which port was unhealthy down the street, after I pressed it again into service, and in the identical “use it till it utterly dies” spirit that prompted my current UPS restore. That mentioned, long-term sanity aspirations finally overrode my traditional thriftiness. My guess is that, given the rest of the ports (and due to this fact the frequent controller chip that manages them) stay operational, port 8’s related transformer received zapped.

And the metal-case TEG-S82g? Right here, I believe, the lightning-strike spike results made it by means of the port 5 transformer, all the way in which to the Realtek RTL8370N controller nexus, albeit curiously solely with derogatory results seemingly after the chip had been operational for a bit and had “warmed up” (observe, as beforehand talked about within the earlier eight-port GbE swap teardown, the dearth of a heatsink on this design). Because the block diagram in this RTL8370N datasheet makes clear, the chip is very built-in, together with all of the ports’ MAC and PHY circuits (amongst different issues).

~1,300 phrases in, that’s “all” I’ve received for you at present. Please share your ideas within the feedback!

—Brian Dipert is the Editor-in-Chief of the Edge AI and Imaginative and prescient Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the corporate’s on-line publication.

Associated Content material

The put up A teardown story of two not-so-different switches appeared first on EDN.