This initial entry in a planned dissection series analyzes the insides of a long distance-capable fiber optic cable networking adapter containing a focused-beam laser.
Last week’s tutorial, the conceptual kickoff to the teardown coverage cadence that begins today, provided what I hoped was a concise but comprehensive tutorial into the diversity of implementation options available with small form-factor pluggable (SFP) modules. Today’s premier patient, as mentioned last week, is a 10 Gbit LX SFP+ module. Specifically, it’s TP-Link’s TL-SM311LS, a member of the company’s business-tailored Omada product line.
Does X mark the multi-cable spot, or not?
Two things particularly interest me in perusing the TL-SM311LS product page. First off, the company refers to it there not as a SFP module but as a “Mini GBIC Module”, using the less common alternative naming convention that I’d noted upfront in last week’s coverage.
Secondly, it’s an “LX” module, with the “L” referencing long range transmission capabilities, specifically in this case specified as up to 20 km (~12.5 miles). The “X” typically references a module that’s usable with both single-mode and multimode cable, albeit at differing max transmission distances and, in both mode cable cases, at a common 1310 nm wavelength. However, TP-Link only documents the SM311LS as usable with 9/125 μm single-mode fiber cable, a seeming characteristic of the alternative “LR” module designator.
I sourced the SM311LS from the Resale (formerly Warehouse) section of Amazon’s website, where it set me back $16.99, versus a $19.99 MSRP. Here’s how it arrived, within a generic cardboard shipping box, and as usual accompanied in the following photos by a 0.75′′/19.1 mm diameter U.S. penny for size comparison purposes.
And here’s our patient, with dimensions of 2.1×0.5×0.5 (55.4×13.7×12.9 mm), after freeing it from its Styrofoam and antistatic bag sarcophagus. Top.
Right side (with “right” referencing its orientation when inserted in a network switch or other gear’s SFP port).
Bottom.
And left side.
The double-sided 20-contact (10 per side) connector is on one end.
The other end begs for a bit more explanation.
Removing the dust cap exposes to view the fiber optic connector mate sites, from two perspectives: first right-side-up and then upside down.
How do you tell which is the transmit one, and which is for data reception? One way is to look at the directions of the two arrows at far left in the earlier top side shot. And what’s with the blue-color lever beyond them? For that, we’ll need to revisit the earlier bottom side shot.
Tab-restrained removal, and internal access
See that tiny tab sticking out of the bottom, to the left of the product label? It’s spring-loaded, temporarily retracting flush with the bottom edge during module insertion and returning to its original position once the module is in place (thereby also locking the module in place). The other means of retracting it, thus enabling module removal, is to move the lever midway through its arc, meaning halfway between the two positions shown in the dust-cap-installed and -removed photos.
And speaking of tabs, did you already notice the tiny ones halfway down both sides in the earlier shots? They’re our pathway inside. Use a small-but-solid “poker” of some sort to pull them back out, and you can then slide the outer chassis away from the interior assembly.
Turning the PCB right-side up again caused the black plastic piece normally covering the fiber optic cable female connectors (and visible in the previous photos) to fall away.
Here’s the underside of the PCB again, standalone and with that plastic piece temporarily back in place.
Now with all supplemental pieces removed, revealing the details of their bottom-pin retraction function operation.
The right side again, this time absent the outer enclosure.
And the left side, upside-down compared to its normal-operation orientation.
One more screw to go.
And the PCB is finally free.
I’ll start with the simple stuff: the SFP and fiber optics connector ends.
Frickin’ laser beams
Let’s next revisit the PCB underside, now free of its metal surroundings.
The orange-color assemblage at far left is the fiber optic receiver. Below it is the chrome-color laser transmitter. And the ICs to their right? The larger eight-lead one, in a TSOP-1 package, is the GT24C08A-2ZLI, an 8K-bit serial I²C EEPROM from Chinese supplier Giantec Semiconductor. Third-party SFP modules typically use memory such as this to, among other things, “spoof” networking equipment that would normally accept only same-name-brand (translation: expensive) SFP modules.
Remember my last-week comments about how single-mode technology has dramatically dropped in price in recent years, thanks in no small part to the “availability of non-proprietary, widely compatible modules”? Here’s a case study example. And the smaller six-lead IC below and to its right? If I’m right and the PCB mark next to it is “Q1”, then it’s likely a dual-transistor device.
But the “371H” mark on top, which I’m guessing is a date code or lot trace code versus an actual product code, isn’t helpful from an identity-sleuthing standpoint, at least to me. Reader assistance here is as-always appreciated.
I’d wondered, after first seeing all these ICs (and particularly before realizing that the larger one was “only” a nonvolatile memory), why they were on the PCB underside, as I’d assumed they were heat sources. Then I saw the PCB topside for the first time and instantly realized why.
The far more sizeable, albeit sole, device on this side, packaged in a 28-lead QFN and marked “3320C” on the top line, is the Uxfastic UX3320, a G/EPON ONU fiber transceiver developed by another Chinese manufacturer, Xiamen UX High-Speed IC. Presumably, given the lack of other logic anywhere on the PCB, it has integrated SFP interface capabilities.
To that point, and in closing, keep in mind that what we’re looking at today is a conventional (and legacy?) 1 Gbps SFP module. Recall from the “laundry list” in last week’s post that I also have a number of both higher-bitrate SFP and SFP+ modules in my teardown queue, and even a QSFP+ version, along with multimode modules to contrast against this and other single-mode offerings. I don’t know about you, but I’m intrigued to learn not only about each of them in an absolute sense but also relatively speaking, in comparing (including contrasting) them against each other.
With that “teaser” now out there, I’ll close for today. Please let me know your thoughts in the comments!
—Brian Dipert is the associate editor, as well as a contributing editor, at EDN.
Related Content
- SFP: Add-in module delivers diminutive performance, flexibility
- TP-Link MC220L: Media conversion keeps the network well
- The whole-house LAN: Achilles-heel alternatives, tradeoffs, and plans
- Design considerations in high-speed fiber networks
The post SFP: The single-mode module autopsy appeared first on EDN.