Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
Category: Science
If you measure something often enough, it becomes science
APRS tracks for my rides around Poughkeepsie in early November 2015:
APRS Coverage – Highland to Hopewell – 2015-11
Turning on the topo data and squinting at the Red Oaks Mill area:
APRS Coverage – Red Oaks Mill area topo – 2015-11
The topography isn’t in my favor, with two ridgelines between Red Oaks Mill and the two APRS nodes near Poughkeepsie. APRS coverage southwest of Red Oaks Mill along the Mighty Wappingers Creek (basically, Vassar Road) ranges from spotty to nonexistent, because that route has even worse topography.
Seems to me an APRSiGate in Red Oaks Mill, running Xastir (perhaps headless) on an RPi, conjured from my heap (perhaps with a shiny new TNC-Pi atop the RPi, rather than an ancient Kantronics KPC-9612), and using a vertical VHF antenna in the attic (because lightning), might improve the situation.
That whole project continues to slip into the future, but at least I have more motivation and linkies…
This sheaf of tests shows three of the four STK NP-BX1 batteries deliver about 4 W·h during a constant 500 mA discharge, with battery B trailing behind:
After the three most recent bike rides, I popped the partially discharged battery into the tester and used the same test current:
Sony NP-BX1 – STK ABD – charged vs used – Wh scale – 2015-11-22
The longer curves come from the top chart (with different colors), the shorter ones from the partially discharged batteries. In an ideal world, the shorter curves should give the energy left in the battery after the ride, so subtracting that from the before-ride capacity gives the energy used during the ride.
The results for battery A may not be typical, as the camera turned off before I rolled into the garage. The camera may run with a battery voltage below the 2.8 V cutoff in those tests, so it can extract more energy than the tests. The slope of the curve toward the end suggests it won’t get much, but that will still bias the results.
In round numbers, the bike rides required:
A: 3.8 – 0.1 = 3.7 W·h
B: 3.6 – 1.4 = 2.2 W·h
D: 4.2 – 1.0 = 3.2 W·h
I generally turn the camera off during the mid-ride pause (Protip: never wear a helmet camera into a Port-a-Loo), so at least two of the rides have discontinuous usage. I figured the total run time from the video file sizes at the rate of 22.75 min/4.0 GB, blithely ignoring issues like the battery recovering during the pauses, the effect of ambient temperature vs. camera heating on battery temperature, and so forth and so on.
In an ideal world, dividing the total energy by the run time (converted from minutes to hours and not venturing into pirate·ninja territory) should produce a nearly constant value equal to the camera’s power dissipation:
A: 3.7 W·h / 1.25 h = 2.96 W
B: 2.2 W·h / 1.0 h = 2.1 W
D: 3.2 W·h / 1.4 h = 2.25
Ignoring the suspiciously high result for battery A, it looks like the HDR-AS30V really does dissipate a bit over 2 W while recording 1920×1080@60fps video. That’s with GPS, WiFi, and NFC turned off, of course.
Which turns out to be pretty close to the test conditions: 3.7 V x 500 mA = 1.85 W. I could goose the test current to 600 mA = 2.2 W/3.7 V for the next tests, but maybe long-term consistency is a virtue.
One might reasonably conclude all six came from the same factory; the STK B battery looks like a dud. The two replacement batteries from STK performed slightly better than the first pair.
The Wasabi and SterlingTEK batteries all carry a 1600 mA·h rating that’s far in excess of their actual 1000-ish mA·h performance. If they were advertised as 1.0 A·h batteries, they’d meet their specifications (for small values of “meet”), but nobody would buy a second-tier battery with less capacity than the Sony OEM battery’s 1.24 A·h.
If you rummage around in previous posts, I did verify that battery capacity does increase with decreasing test current, but definitely not by the 60% needed to reach 1600 mA·h.
Because most devices these days operate at constant power from a boost supply, presenting the results against a watt·hour scale would make sense:
They should survive for hour-long rides with the GPS tracker turned off, which is about as much as I want to ride at once. I’ll eventually autopsy the STK B battery, which won’t last all that long.
Credit where credit is due: after I sent the first test results to STK, they sent a pair of replacement batteries and, based on the second test results, refunded the entire purchase price. I’m reluctant to give a five-star rating for customer service, because shipping mis-advertised products should carry a zero-star rating.
Because the first pair of Wasabi NP-BX1 batteries for the Sony HDR-AS30V camera faded to the point where they weren’t useful for a typical bike ride, I bought a pair of SterlingTEK (a.k.a. STK) NP-BX1 batteries that, like the Wasabi batteries, claimed to have a 1600 mA·h capacity. These are “second tier” batteries, not the cheap eBay crap I’ve already dismissed, and run a bit under $10 apiece.
Here’s the picture from their product description:
SterlingTEK – STK NP-BX1 battery – as advertised
Here’s what arrived:
STK NP-BX1 batteries
Huh.
That’s a red flag, right there. It’s remarkably tempting to ship a good product for a while, then swap in much cheaper junk that can ride on the good reviews. Not saying that’s what happened, but it’s a possibility.
Here’s how they performed:
Sony NP-BX1 – Sony Wasabi STK – as received
The red and blue curves show that the STK batteries produced less than 1000 mA·h in their first two charges, with the blue battery (I labeled it B) showing considerable variation that suggests it’ll suffer early failure. The green curve shows one of those Wasabi batteries and the purple curve is the OEM Sony battery, both in as-received condition.
SterlingTEK will send two more batteries, in the belief that I received two sub-standard samples. We shall see…
All the tests are at 500 mA, approximately half the camera’s load. Oddly, the numeric values along the mA·h axis work out pretty close to the actual runtime in hours:
Sony – 1:30
Wasabi D – 1:15
Wasabi B – 0:40
Given that a typical bike ride takes an hour, the two year old Wasabi B battery’s 40 minute runtime isn’t useful. The Wasabi D battery is a bit over a year old and looks very much like the B battery did last year.
The Wasabi batteries march through the camera and charger in order, so each one gets used about once a week. The Sony battery gets used once every half-dozen complete cycles, just so I have a standard “good” battery.
The Sony and Wasabi B cells over the course of two years:
Sony NP-BX1 – OEM Wasabi – 2015-10 2014-10 2014-01
Much to my surprise, the Wasabi batteries started out slightly better than the Sony OEM battery, at least as measured by the available voltage and energy. The camera runs from an internal switching power supply, so the area under the curve (basically equal to energy in W·h) above the cutoff voltage is all that matters.
In round numbers, I can expect 100 cycles out of each battery before the run time drops below the ride time; at $10/battery, that’s a dime a ride. Any claims that the batteries can be recharged “1000 times!” may be true, but they’ll have a useless fraction of their original capacity by then.
As part of a discussion on the M2 forums about using rice to dehumidify 3D printer filament, I replaced the 500 g bag of silica gel in the basement safe with a bowl containing 200 g of long-grain brown rice from our rice supply and let it sit for a while:
Basement Safe Humidity – Rice vs. Silica Gel – 2015-10-31
The abrupt drop in humidity from 52% to the logger’s minimum 15% marks the point where I replaced the rice with a fresh bag of silica gel, with a door opening shortly thereafter. The basement air outside the safe varied between 52% and 54% during that time, so the air inside the safe trended upward toward that goal.
The rice still weighed exactly 200 g after its stay in the safe, so we can conclude it hadn’t absorbed or released any water.
Conclusion: nope, rice doesn’t work as a dehumidifier…
The discussion about scrubbing hard drives suggested I really should be using larger block sizes to wring better performance from the hardware.
So I ran variations on this theme:
time sudo dd if=/dev/urandom of=/dev/sdc bs=4K count=32K
For the BS (“block size”) parameter, 1K = 1024 and 1KB = 1000. Similarly for 1M vs. 1MB.
The results, viewed as a picture because WordPress seems unable to import a formatted spreadsheet from LibreOffice like it used to:
USB-SATA Adapter – Barracuda 7200.10 drive
Each operation transfers 128 MB (128 x 220 = 131 x 106) bytes. The variations probably come from other stuff going on, most notably the USB-to-serial adapter driving the plotter while I’m testing a tweak to the Superformula demo code.
Reads ever so much faster than writes, so the USB adapter definitely isn’t getting in the way; I assume the drive accepts the commands & data as fast as its little heads can carry them away. The data, being relentlessly pseudo-random, won’t get compressed along the way.
So, in round numbers, the block size just absolutely does not make any difference.
dd if=/dev/urandom of=/dev/zero bs=4K count=32K
32768+0 records in
32768+0 records out
134217728 bytes (134 MB) copied, 9.63064 s, 13.9 MB/s
dd if=/dev/urandom of=test.bin bs=4K count=32K
32768+0 records in
32768+0 records out
134217728 bytes (134 MB) copied, 10.018 s, 13.4 MB/s
dd if=test.bin of=/dev/zero bs=4K count=32K
32768+0 records in
32768+0 records out
134217728 bytes (134 MB) copied, 0.0385358 s, 3.5 GB/s
dd if=test.bin of=test2.bin bs=4K count=32K
32768+0 records in
32768+0 records out
134217728 bytes (134 MB) copied, 0.45044 s, 298 MB/s
I installed an SSD on this box a while ago, so the 3.5 GB/s disk-to-discard speed represents the SSD’s read rate. The 298 MB/s disk-to-disk speed would be its write speed, probably with some clever buffering going on.
So the real bandwidth limitation in wiping a disk comes from the pseudo-random generator behind /dev/urandom, not the disk or USB interface. It would probably be faster to fill a 1 GB (or more) file with noise at 14 MB/s, then copy it enough times to fill the drive at whatever speed the drive can handle it.
The Smell of Molten Projects in the Morning 