I’ve used and editorially covered TRENDnet’s indoor security cameras quite a bit over the years:
That last link marks my most recent coverage (I think)…eight years ago. Why nothing more recent? Part of the reason is my wife’s unswerving reticence to the idea of having security cameras installed inside our home, for privacy reasons, and even if they were only operating when we weren’t home. If this were a remotely located second home (an example of which we no longer own), I could probably evolve her perspective. But all in all I can’t say that I blame her, given all the ongoing hacking that I regularly read about.
And part of the reason for the dearth of recent coverage is that, in contrast to the Blink outdoor camera system that she was comfortable with me installing, and which has continued to work like a charm in the more than 2.5 years since I first did so, getting conventional security cameras like the TRENDnets initially set up, not to mention reliably accessing them on an ongoing basis, is just too much of a hassle. You need to have close-proximity access to AC power, first off, or run PoE-compatible cabling and networking equipment. If Ethernet cable isn’t your thing, you also need to make sure you have reliable Wi-Fi with sufficient coverage.
And then there’s the software side of the story. An embedded web server in each camera is conceptually cool, but accessing it via a web browser requires either Windows-only (and, as of Windows 11, Windows-no-more) ActiveX or cross-platform but user-installation-required Java support. Dedicated apps for computers and mobile platforms can get you around the browser-based access limitations but demand long-term support in responding to operating system evolutions, a concept that isn’t exactly in consumer electronics companies’ DNA. All of this perhaps explains why, when I visited TRENDnet’s product page this morning after a long time away and in doing preparatory research for the writeup, the suite of products I found still offered and supported by the company was a shadow of its former self.
That all said, cameras with embedded web servers are still conceptually cool. To wit, I thought I’d devote this teardown to the TRENDnet TV-IP551WI, which has some intriguing hardware capabilities: both 802.11n Wi-Fi (2.4 GHz-only, and reflective of the “W” in the product name) and 10/100 Mbps wired Ethernet connectivity, “night sight” support (reflected in the product name’s last-digit “I”, for “Infrared”) and an integrated microphone to give the camera both visual and audible remote monitoring capabilities.
I’ll start with a stock image:
Now here are some shots of my device, first with the stand still attached and-usual accompanied by a 0.75″ (19.1 mm) diameter U.S. penny for size comparison purposes. This camera model, unlike others I’ve used, doesn’t include a convenient-but-unreliable motor-driven swiveling base and pivoting lens for pan-and-tilt remote control purposes:
Note the WPS configuration button and recessed hardware switch on the other side:
Before continuing with the camera, here’s the companion “wall wart” power supply, both in overview and a close-up of the label:
The camera’s metal stand has a ~3” diameter base and adds ~0.5” of height to the dimensions listed on TRENDnet’s specs page, 75 x 115 x 30 mm (2.95 x 4.5 x 1.2 in.). Here’s what the camera looks like with it removed, beginning with the front. Note the four infrared LEDs around the lens, one in each corner, the ambient light sensor above the lens, and the microphone below the lens. Not obviously visible (but trust me, it’s there) is a two-status-LED cluster embedded in the silver ring around the lens, below the microphone aperture, to be precise:
And here’s the backside, also including a close-up of the product label with (among other things) the always-informative FCC ID XU8TVIP551:
Did you notice the four screws, one in each corner, in the previous backside overview shot? Remove them, carefully release a few tabs, and the rear half of the product’s plastic shell lifts right off:
The half-shell itself is largely unmemorable, save for the aforementioned WPS button and overall slavish use of passive ventilation vents. The insides are (unsurprisingly) more intriguing, beginning with an iPassion (believe it or not) iP2970-L camera control processor IC in the upper right corner, apparently also used in D-Link’s DCS-930L camera. Even more dominant, albeit less informative at least for the moment, is the ventilation-hole-ridden Faraday Cage below and to the left of it. I’d initially harbored aspirations of reassembling and then donating the camera to charity afterwards…oh, well:
The “guts” consist of a two-PCB “sandwich,” as these side views make obvious:
See the silver screw in the lower right corner of the top PCB in the “sandwich” in the earlier image? Remove it and the “sandwich” lifts right off, too:
Technically, the “sandwich” and front half-shell are still tethered together by that long rectangular green thing glued to the shell’s inside. Careful application of a flat head screwdriver acting as a lever arm freed it and confirmed my suspicion that it was the 1×1 Wi-Fi antenna:
This half of the shell is a bit more interesting than its back-half counterpart. Note the openings for the four infrared LED light pipes, the ambient light sensor and the microphone:
And underneath the latter, speaking of light pipes, is the pair associated with the status LEDs. See, I told you so!
The front view of the “sandwich” is still much more interesting:
First off, let’s unscrew that lens assembly, whose specs include (from the product page):
- Focal Length: 4 mm
- F/No: 1.5
- Minimum illumination: 0 lux
- Focus Depth: 20 cm ~ infinity
- View: Horizontal: 38.0 degrees
- Vertical: 28.7 degrees
- Diagonal: 46.5 degrees
It further separates into two distinct pieces. Note the explicitly labeled optics support for the infrared light spectrum:
And with it removed, we have our first glimpse at the image sensor underneath it:
TRENDnet is vague on the details, noting on the product spec sheet only that it is a “1/5” CMOS Sensor,” along with the following sensor-plus-processing capabilities:
- Encoding: MJPEG Resolution: 640 x 480, 320 x 240, 160 x 120
- Max frame rate: VGA @ 20fps, QVGA @ 30fps, QQVGA @ 30fps
- JPEG quality: Very Low, Low, Medium, High, Very High
- Digital Zoom: 4x
This particular photo also provides perspective on the four infrared LED light pipes (with operating range up to 7.5 m), the ambient light sensor, the status LEDs (labeled D7 and D8) and the microphone. The TV-IP551WI dates from mid-2012 so I’m guessing the microphone is an electret model versus the now-common MEMS alternative. Here are its relevant specs:
- Built-in omni-directional microphone
- Sensitivity: -38dB +/- 3dB (5 meters max)
- Frequency: 100~2 Hz
- Format: PCM
Onward. If you go back and look at the rear overview shot of the two-PCB sandwich, you’ll see two more screws, both black in color this time. Remove them, separate a 22-pin connector assembly normally linking them, and the “sandwich” peels apart:
Once again there’s a lingering two-wire tether between them, which was easily disconnected. Going forward, I’ll be referring to the PCB on the left as the “camera” PCB and the one on the right as the “system” PCB. Let’s look more closely at the former, first, beginning with its front side:
The significant IC here is a Spansion S29GL032N 32 Mbit MirrorBit (multi-level cell) NOR flash memory presumably used predominantly-to-completely for code storage purposes, since the camera doesn’t support native still or video image capture. Spansion, a joint venture formed out of AMD and Fujitsu’s flash memory programs, is now part of Cypress Semiconductor.
Next, let’s revisit the image sensor area, home to one endpoint of the two-wire harness. It’s glued in place; the flat head screwdriver took care of that:
The module comprises two main sub-assemblies: the portion that the lens screws into, and what I’m guessing is an electronically movable infrared filter, similar to the one we saw in the earlier Wyze camera teardown (during daylight hours you want it in front of the image sensor, to block infrared light from unnaturally affecting the captured video, but after dark infrared is the only illumination option, so the filter gets moved out of the way). Removing the two screws seen in the earlier image allows for sub-assembly separation:
Here are closer views of the filter sub-assembly’s ends:
One more close-up of the front side of the camera PCB, before flipping it over:
And here we go…
We’ve already mentioned the camera control processor:
Now, let’s get the top of that Faraday cage, once again relying on the flat head screwdriver instead of my fingernails. Zooming in…
…we find even more design commonality with the D-Link DCS-930L:
Also note the generous dallop of glue holding the antenna connector in place!
Last but not least, let’s take a look at the system PCB. Front side first:
Here’s another angle on that ambient light sensor (the camera can either automatically switch between normal and night mode based on this sensor’s discernment, or the user can manually control which mode is active):
And a few other closeups, before turning the PCB over:
Without further ado:
The only IC of note here is a TS8121A 10/100 Mbit Ethernet pulse transformer:
And with that, we’re done! See, I told you that a nearly 10-year-old security camera could still provide some interesting surprises! As always, I look forward to your thoughts in the comments.
This article was originally published on EDN.
—Brian Dipert is Editor-in-Chief of the Edge AI and Vision Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the company’s online newsletter.
I’ve used and editorially covered TRENDnet’s indoor security cameras quite a bit over the years:
That last link marks my most recent coverage (I think)…eight years ago. Why nothing more recent? Part of the reason is my wife’s unswerving reticence to the idea of having security cameras installed inside our home, for privacy reasons, and even if they were only operating when we weren’t home. If this were a remotely located second home (an example of which we no longer own), I could probably evolve her perspective. But all in all I can’t say that I blame her, given all the ongoing hacking that I regularly read about.
And part of the reason for the dearth of recent coverage is that, in contrast to the Blink outdoor camera system that she was comfortable with me installing, and which has continued to work like a charm in the more than 2.5 years since I first did so, getting conventional security cameras like the TRENDnets initially set up, not to mention reliably accessing them on an ongoing basis, is just too much of a hassle. You need to have close-proximity access to AC power, first off, or run PoE-compatible cabling and networking equipment. If Ethernet cable isn’t your thing, you also need to make sure you have reliable Wi-Fi with sufficient coverage.
【Download】Aprisa place-and-route tool for low-power SoCs
And then there’s the software side of the story. An embedded web server in each camera is conceptually cool, but accessing it via a web browser requires either Windows-only (and, as of Windows 11, Windows-no-more) ActiveX or cross-platform but user-installation-required Java support. Dedicated apps for computers and mobile platforms can get you around the browser-based access limitations but demand long-term support in responding to operating system evolutions, a concept that isn’t exactly in consumer electronics companies’ DNA. All of this perhaps explains why, when I visited TRENDnet’s product page this morning after a long time away and in doing preparatory research for the writeup, the suite of products I found still offered and supported by the company was a shadow of its former self.
That all said, cameras with embedded web servers are still conceptually cool. To wit, I thought I’d devote this teardown to the TRENDnet TV-IP551WI, which has some intriguing hardware capabilities: both 802.11n Wi-Fi (2.4 GHz-only, and reflective of the “W” in the product name) and 10/100 Mbps wired Ethernet connectivity, “night sight” support (reflected in the product name’s last-digit “I”, for “Infrared”) and an integrated microphone to give the camera both visual and audible remote monitoring capabilities.
I’ll start with a stock image:
Now here are some shots of my device, first with the stand still attached and-usual accompanied by a 0.75″ (19.1 mm) diameter U.S. penny for size comparison purposes. This camera model, unlike others I’ve used, doesn’t include a convenient-but-unreliable motor-driven swiveling base and pivoting lens for pan-and-tilt remote control purposes:
Note the WPS configuration button and recessed hardware switch on the other side:
Before continuing with the camera, here’s the companion “wall wart” power supply, both in overview and a close-up of the label:
The camera’s metal stand has a ~3” diameter base and adds ~0.5” of height to the dimensions listed on TRENDnet’s specs page, 75 x 115 x 30 mm (2.95 x 4.5 x 1.2 in.). Here’s what the camera looks like with it removed, beginning with the front. Note the four infrared LEDs around the lens, one in each corner, the ambient light sensor above the lens, and the microphone below the lens. Not obviously visible (but trust me, it’s there) is a two-status-LED cluster embedded in the silver ring around the lens, below the microphone aperture, to be precise:
And here’s the backside, also including a close-up of the product label with (among other things) the always-informative FCC ID XU8TVIP551:
Did you notice the four screws, one in each corner, in the previous backside overview shot? Remove them, carefully release a few tabs, and the rear half of the product’s plastic shell lifts right off:
The half-shell itself is largely unmemorable, save for the aforementioned WPS button and overall slavish use of passive ventilation vents. The insides are (unsurprisingly) more intriguing, beginning with an iPassion (believe it or not) iP2970-L camera control processor IC in the upper right corner, apparently also used in D-Link’s DCS-930L camera. Even more dominant, albeit less informative at least for the moment, is the ventilation-hole-ridden Faraday Cage below and to the left of it. I’d initially harbored aspirations of reassembling and then donating the camera to charity afterwards…oh, well:
The “guts” consist of a two-PCB “sandwich,” as these side views make obvious:
See the silver screw in the lower right corner of the top PCB in the “sandwich” in the earlier image? Remove it and the “sandwich” lifts right off, too:
Technically, the “sandwich” and front half-shell are still tethered together by that long rectangular green thing glued to the shell’s inside. Careful application of a flat head screwdriver acting as a lever arm freed it and confirmed my suspicion that it was the 1×1 Wi-Fi antenna:
This half of the shell is a bit more interesting than its back-half counterpart. Note the openings for the four infrared LED light pipes, the ambient light sensor and the microphone:
And underneath the latter, speaking of light pipes, is the pair associated with the status LEDs. See, I told you so!
The front view of the “sandwich” is still much more interesting:
First off, let’s unscrew that lens assembly, whose specs include (from the product page):
It further separates into two distinct pieces. Note the explicitly labeled optics support for the infrared light spectrum:
And with it removed, we have our first glimpse at the image sensor underneath it:
TRENDnet is vague on the details, noting on the product spec sheet only that it is a “1/5” CMOS Sensor,” along with the following sensor-plus-processing capabilities:
This particular photo also provides perspective on the four infrared LED light pipes (with operating range up to 7.5 m), the ambient light sensor, the status LEDs (labeled D7 and D8) and the microphone. The TV-IP551WI dates from mid-2012 so I’m guessing the microphone is an electret model versus the now-common MEMS alternative. Here are its relevant specs:
Onward. If you go back and look at the rear overview shot of the two-PCB sandwich, you’ll see two more screws, both black in color this time. Remove them, separate a 22-pin connector assembly normally linking them, and the “sandwich” peels apart:
Once again there’s a lingering two-wire tether between them, which was easily disconnected. Going forward, I’ll be referring to the PCB on the left as the “camera” PCB and the one on the right as the “system” PCB. Let’s look more closely at the former, first, beginning with its front side:
The significant IC here is a Spansion S29GL032N 32 Mbit MirrorBit (multi-level cell) NOR flash memory presumably used predominantly-to-completely for code storage purposes, since the camera doesn’t support native still or video image capture. Spansion, a joint venture formed out of AMD and Fujitsu’s flash memory programs, is now part of Cypress Semiconductor.
Next, let’s revisit the image sensor area, home to one endpoint of the two-wire harness. It’s glued in place; the flat head screwdriver took care of that:
The module comprises two main sub-assemblies: the portion that the lens screws into, and what I’m guessing is an electronically movable infrared filter, similar to the one we saw in the earlier Wyze camera teardown (during daylight hours you want it in front of the image sensor, to block infrared light from unnaturally affecting the captured video, but after dark infrared is the only illumination option, so the filter gets moved out of the way). Removing the two screws seen in the earlier image allows for sub-assembly separation:
Here are closer views of the filter sub-assembly’s ends:
One more close-up of the front side of the camera PCB, before flipping it over:
And here we go…
We’ve already mentioned the camera control processor:
Now, let’s get the top of that Faraday cage, once again relying on the flat head screwdriver instead of my fingernails. Zooming in…
…we find even more design commonality with the D-Link DCS-930L:
Also note the generous dallop of glue holding the antenna connector in place!
Last but not least, let’s take a look at the system PCB. Front side first:
Here’s another angle on that ambient light sensor (the camera can either automatically switch between normal and night mode based on this sensor’s discernment, or the user can manually control which mode is active):
And a few other closeups, before turning the PCB over:
Without further ado:
The only IC of note here is a TS8121A 10/100 Mbit Ethernet pulse transformer:
And with that, we’re done! See, I told you that a nearly 10-year-old security camera could still provide some interesting surprises! As always, I look forward to your thoughts in the comments.
This article was originally published on EDN.
—Brian Dipert is Editor-in-Chief of the Edge AI and Vision Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the company’s online newsletter.