USB video adapter, or almost a graphics card.

Since my work involves computers quite a bit, I'm interested in all sorts of unusual computer accessories. Not long ago, I learned about an interesting device: a video adapter that plugs into a computer's USB port. As soon as I got the chance to review one, I jumped at it.
So, the usual routine: review, teardown, testing.

There's nothing particularly unusual about this kind of device. Video is sent over the USB bus, converted to an analog signal, and output to a monitor. But USB 2.0 doesn't have enough bandwidth for these devices to work properly at high resolutions, so they're made for USB 3.0, which offers a noticeably higher data transfer rate.
This adapter can also work when connected to USB 2.0, but the maximum resolution is limited to 800x600.
What are these devices for? Lots of things. For example, setting up a multi-monitor configuration where you can't install an additional graphics card, or where installing a full-size graphics card simply isn't worth it.

But let's get to the review. I'll try to keep it brief and to the point.
As usual, I'll start with the packaging. The device arrived in a small bag.


The package includes the USB video adapter and a driver disc. That's it.


This device is about the size of a matchbox, and its case is the same as the HDMI-to-VGA converter I reviewed some time ago.


One side of the case has a standard 15-pin VGA connector, with screw posts to secure the cable.


A short cable comes out of the opposite side, ending in a USB connector covered by a protective cap.
The whole device, including the cable, is blue—presumably to match the blue color of USB 3.0 connectors.


As everyone here knows, I almost never skip a teardown. Besides, I was curious to see what was inside. I even decided to open it in style, so I grabbed a knife that matched the case color :)
The little box opens very easily: just pry up the front along the seam, then push it outward a little, and that's it. It still holds together well, though. I saw the same design in my review of the HDMI-to-VGA converter.


The board is very well made—or, well, almost very well made. I didn't spot any obvious flaws or sloppy work. The wires are secured with hot glue.
The soldering is neat, and the board has been cleaned of flux.


As always, here are a couple of photos of the board, in case it needs repair or you just want a closer look.
The USB-to-VGA controller is visible in the center. It's powered by two small DC-DC converters (the controller needs two voltages, 3.3 and 1.2 V), and there's a footprint for an EEPROM chip.


The back of the board is mostly populated with passive components, apart from four transistors. The blob of solder under the controller presumably acts as a tiny heatsink :)


The USB-to-VGA controller is the FL2000DX chip from Fresco Logic, a company we already know from a previous review. This is the second of the four chips they make :)
As last time, I didn't use the included CD and downloaded the driver from the manufacturer's website. Just in case, I've also mirrored it on my Yandex Disk.


The controller's block diagram suggests that a version with a digital output should also be available, which could be useful. I didn't look into it, since I specifically needed an analog output.
Calling it a video card is a bit of a stretch. A digital stream is sent over USB, converted into a format the monitor can understand, and then we see the image. There's no actual video processing; everything is handled in software on the computer.


A brief digression. Sometimes when using devices like these (including VGA converters), you run into a situation where a monitor connected to the converter won't go to sleep. In other words, you turn off the device, but the screen keeps showing a message like “Check cable connection,” and the monitor never turns off.
This happens because some manufacturers don't connect pin 10 of the VGA connector to ground.
It's actually the common pin for the sync line, but whether the monitor is connected to the computer is detected by checking whether this pin is grounded.
If you have this problem, simply connect pins 10 and 5 of the VGA connector. I had the same issue with an HDMI-to-VGA converter; connecting these pins fixed it completely.
This device already has that connection, so no modification is needed.


As is often the case, it all starts with installing the driver.
I was pleasantly surprised to find that the installer is in Russian. It's not essential, but it's nice.


During installation, the program asks twice whether we agree to install the drivers.
First it's the adapter driver itself, then the display mirroring driver.
Since I didn't have many options, I agreed. Before trying anything, though, I made a system image just in case—better safe than sorry.


The ddmgr.exe background process also runs in the system.
The screenshot shows CPU usage while video is being displayed on the additional screen.
While video was playing, it varied between 9-11%.


First I connected my main monitor to the adapter and checked all the available resolutions (listed below), then I connected it to the test Samsung monitor.
After rebooting, the driver identified the test monitor's model, but didn't recognize the main monitor, even after a reboot.
This doesn't affect anything, of course, but it showed that in at least half of cases the adapter can retrieve the monitor's model information.


Everything was going smoothly: the board looked good, there were no driver issues, and it even managed to identify the monitor model. But that was where the positives ended.
So, first I'll list the resolutions I tested and that are available with this adapter.
800x600 1360x768 1024x768 1366x768 1152x864 1400x1050 1280x600 1440x900 1280x720 1600x900 1280x768 1600x1200 1280x800 1680x1050 1280x960 1920x1080 1280x1024
That's a pretty good list, covering most of the monitors and projectors out there.
But testing showed that every resolution worked perfectly except the ones I needed—or rather, the ones my test monitors are designed for.
At 1680x1050 and 1920x1080, both monitors displayed the image as shown in the photo.


When I tried to set the resolution to 1600x1200, both monitors cursed me out and refused to work, since that resolution was outside their supported range.
At 1600x900, my main monitor (native resolution 1920x1080) displayed the image without any problems. The test monitor, with a resolution of 1680x1050, showed a small black bar at the top.
At all other resolutions, both monitors worked perfectly.


As an experiment, I connected the monitor to a tablet. I wanted to check how it worked in USB 2.0 mode and whether it was compatible with Windows 8.1.
Both tests were successful, but the monitor ran at 800x600; it simply wouldn't let me select a higher resolution. All the display options were still available: I could either clone the main monitor (but then the main monitor's resolution would match the secondary one) or use the secondary monitor as an extended desktop (in this mode, the resolutions can be selected independently).
The photo shows a document open, with part of it on the main screen and part on the secondary screen.


Next, I'll cover a few things that weren't included in the description above.
I originally planned to assess how blurry the image was, and perhaps even take some comparison photos, but the failure to work at high resolutions forced me to change my plans. Blurriness usually isn't noticeable at low resolutions, and I couldn't get this little box to work at high resolutions. The highest resolution I managed was 1600x900. But since none of the resolutions I got are native to my monitors, it's hard to draw any conclusions. At one resolution (1400x1050), I was able to check horizontal sharpness, since 1050 is one of the native resolutions of my test monitor. The lines were sharp and undistorted.

Naturally, I also checked whether it could display video on the secondary monitor.
At first, I couldn't get it to work. I even tried disabling DualView, which I use to output video to my TV. When I moved the player to the secondary monitor, the video played for a few seconds, then froze, although the player's timer kept running. The problem turned out to be Media Player Classic itself. When I tried the native Windows XP player, everything worked without a hitch: the video played perfectly, with no stuttering or freezing. It felt just like using a regular graphics card, both in windowed mode and full screen—there was no difference.

I didn't measure power consumption, but the adapter barely gets warm during use, even at 1600x900. Based on its temperature, I'd estimate its power consumption (and therefore heat dissipation) at around 0.3–0.4 W at most, or about 60–80 mA.

Summary.
Pros
Good build quality
Easy to install and connect
Works well at all resolutions up to and including 1600x900.
Low power consumption and, consequently, little heat.
A wide range of available resolutions.
With some monitors, the device can automatically identify the monitor model.

Cons
Just one: I couldn't get video at 1680x1050 or 1920x1080. The 1600x1200 resolution is a question mark, since the monitors couldn't display it, but I suspect the result would be the same as with the resolutions mentioned above.

I can't say whether this is a defect in this particular unit or a problem they all have, since I tested only one adapter. But the problem was there, and I did my best to show it. It's most likely a hardware issue: the screenshot looked normal, so perhaps the monitor can't work properly with this adapter. Still, I suspect the adapter itself isn't working correctly.
Overall, though, I was left with a positive impression. If you use it with a lower-resolution monitor (the idea of outputting a high-resolution image over an analog connection isn't really a good one anyway), it should work well.
The CPU load was of course higher than with a hardware graphics card, but I wouldn't call it particularly high (around 9–11% with an E7500 processor). Keep in mind that everything is done in software; the adapter is just a DAC connected via USB. Not so long ago, people were measuring CPU load with a hardware sound card versus an AC97 codec and seeing a difference. These days, that seems funny.

The adapter was provided for review and testing by the gearbest store.

I'm thinking of reviewing another adapter from a different manufacturer for comparison. Not sure if that would be interesting—let me know, and I'll try some other models.


I hope you found the review interesting and maybe useful :)

The product was provided by the store for this review.