Gauntlet 320 wifi connection issues - possibly bandwidth/speed related?

[Ygolonac]

I bought a Gauntlet 320 back in February 2013, for use (mainly) with my rooted Nook Color running Cyanogenmod when away from my home network. From the very start, it had what I considered a minor issue - wifi connections would die about 6-8 hours in, when streaming MP3s to the Nook. (Using the power adapters for both devices, obviously - no magic "overtime" batteries here.) "Die" in this case means the wifi connection would still be visible, but no actual data transfer would occur; as well, new connection attempts (or reconnects from the previously-connected device) would fail with authentication errors - as if the connection had "locked". The device did get somewhat hot (not too hot to touch) in the area directly behind the printed "battery charge" icon on the top cover, and I assumed that charging it while running streaming media over wifi was overheating a component, and so I limited my overall usage time - this seemed to work, and I've had no significant issues until recently.

Within the past six weeks or so, the Gauntlet's time-to-lockup-when-streaming has decreased dramatically - randomly occurring anywhere from the original 6-8-hours timeframe, down to two hours or less. (Using the Nook Color/Cyanogenmod 10.1 or higher, or using my laptop running Windows Vista Home/Win7 Ultimate.) As well, upon powering up, the disk-activity light began to turn on and start steadily lit - as opposed to the normal "flickering" behavior previously observed. Direct USB connection and error-checking/drive analysis showed no significant errors, and no change was noted after corrupt files were removed. The drive shows as 4% fragmented, with 93% in use.

Last week, I purchased a Samsung Galaxy Tab 3 7.0 tablet (stock Android 4.1.2), and attempting to stream MP3s to this device results in audio dropouts (ranging from about half a second to 5-6 seconds or more, followed by the resumption of playback), beginning within the first few minutes of streaming and continuing indefinitely until lockup occurs approximately 2 hours after streaming begins. As well, straight data-copying, to or from the Gauntlet, will randomly either run steadily, or suffer very frequent "recalculations" as to file size/completion times, with a wide variety of speeds ranging from 300 kb/sec or higher, all the way down to 9 kb/sec at the slowest observed. (Audio dropouts were previously only observed when nearing the "lockup" time when streaming to the Nook or laptop, after a much longer period of time. Erratic data transfer was not observed until recently, and neither issue is occurring when connected via USB. The Nook and Tab 3 do not exhibit this behavior when connecting to other wifi sources.)

The Gauntlet firmware was updated to V21_1.2.4.5 after purchase, and this past weekend was updated to the latest version (V_21_1.2.4.6). No change was observed. This morning, I did a factory reset via the pinhole, and manually reconfigured the device. No change was observed. The same kind of audio drops occur when directly under high-voltage power lines, but stop when I move not less than ten feet from directly below. (Data transfer issues have not been tested in that location.) I utilize the Patriot app strictly for settings changes on both the Nook Color and Tab 3 - data transfer on both devices use ES File Explorer, media streaming uses BSPlayer or Neutron. (Laptop transfer is via Windows Explorer, media streaming is Winamp and Media Player Classic.) Direct wifi connect to the devices only - no passthrough has been used.

The feeling I'm getting from this is that the wifi signal or actual hardware is getting overwhelmed - even now, the Nook Color (which is much less powerful than the Tab 3) handles the dropouts/transfer glitchiness better; I'll try a wifi-throttling app if I can find one, just to see what happens. Media quality is set the same for both Nook and Tab, so it's not as though I'm suddenly trying to pull HD video when I'd previously been pulling low-res.

Is it possible that the Gauntlet's firmware is trying to write to some godawful-huge-by-now logfile hidden away somewhere? That might explain the solid drive activity on boot...

[mat]

Have you tried another HD? How did you test, chkdsk? There might be a corrupted fat table. Maybe try a full reformat while usb tethered. Interference from power lines is normal, Marconi effect, wide frequency noise.
If a component is overheating, try blowing compressed air to clear out dust.

[Ygolonac]

Have you tried another HD? How did you test, chkdsk? There might be a corrupted fat table. Maybe try a full reformat while usb tethered. Interference from power lines is normal, Marconi effect, wide frequency noise.
If a component is overheating, try blowing compressed air to clear out dust.

I don't have any spare 2.5" drives to test with; no appreciable dust was found when i went ahead and voided my warranty by opening the thing up. :V Speaking of which...

With the casing opened, the drive hanging off the data/power cabling, and the battery pack removed, I could expose the copper foil attached directly beneath the battery. This is connected to a small PCB? that connects to the main circuit board via a single wire. Doing so produced a *significantly* higher power output, as reported by the Wifi Analyzer app. Replacing the battery and drive resulted in a drop to the normal signal levels. (While exposed, I noticed no audio dropouts whatsoever, while the usual dropouts resumed on closing everything back up.)

Is that wire and small PCB bit the actual antenna? If so, would it be possible to simply attach some form of external antenna (whether a proper one or just a short length of insulated wire hanging out of the case) and perhaps improve the signal, which may be the issue I'm having?

[mat]

I don't think opening would void the warranty as a variant is sold without a drive. And an RMA with support may be in order, this tech is old enough I assumed the warranty lapsed!

It seems there may be rf interference from either the drive or the battery. Marconi effect is that any spark creates a radio signal on all frequencies. Rf interference can be from a motor, rheostat, fluorescent light, CRT monitor etc.

If you have some extender cables to move the HD only away from the antenna, or know someone with a field strength meter to pinpoint the rf.
An electrician may have an inductive amplifier which is what I use. (Progressive electronics model 200ep)

[Ygolonac]

I don't think opening would void the warranty as a variant is sold without a drive. And an RMA with support may be in order, this tech is old enough I assumed the warranty lapsed!

It seems there may be rf interference from either the drive or the battery. Marconi effect is that any spark creates a radio signal on all frequencies. Rf interference can be from a motor, rheostat, fluorescent light, CRT monitor etc.

If you have some extender cables to move the HD only away from the antenna, or know someone with a field strength meter to pinpoint the rf.
An electrician may have an inductive amplifier which is what I use. (Progressive electronics model 200ep)

2 year limited warranty, I'm within the timeframe but the warranty terms state it voids for "unauthorized modification" and/or "physical alterations of product." Generally, once you crack the casing you're void. <shrug> Oh well. Popping the top shows that it does seem to be possible to get the proper performance out of this thing, I just have to tweak around and find out how...

I'll take some pictures when I open it here today - can't upload to anywhere until later, though. First thing I'm going to try is reorienting the battery and rerouting the power wiring - that should help with the localised overheating, as where the wiring crosses components is where I find the hotspot.

No available FSMs or such - best I can do is a 1980's Radio Shack pocket multimeter.

[mat]

An old transistor radio tuned between stations can help localize rf noise. Or an old guitar pickup (single coil not humbucking) wired to an amplifier. You might be able to hook up a small coil of wire to the multimeter and see a change in ac millivolts.