Happy New Year everyone!
No, it’s not deja vu, we’re in 2019 … despite the blunder that occurred in Sydney’s NYE celebrations. Now that 2019 has arrived, it’s time to reflect on the year that passed and look at the random stuff I’ve been doing in-between …
The year of 2018 went much faster than I expected. It was also a very important year for me, as I was free of my university “shackles”, but also, I had recently shaken up my life by moving to my Dad’s house. In doing so, I cleaned a lot of needless material goods from my possession and decided to more carefully evaluate what is valuable and what is not. I also decided that maybe I should begin to sell some of my excess belongings in the hopes that others may benefit … but I still haven’t gotten around to that.
Having returned back from my lengthy holiday in 2017, I started my job search in earnest. Part of this involved continually applying for jobs, to the point where I made this mundane task into something akin to a systematic process. Of course, I was not immune to making a few embarrassing mistakes, but I started off the year with an optimism that I might find employment within industry with a major company that I respected, that innovated and improved lives for all.
Unfortunately, the optimism wore down quickly. Aside from a string of rejections, the emotional and energy cost of having to attend numerous virtual interviews, assessment centers (sometimes multi-day) added up. Ultimately, the companies I felt were most interesting to me ended up turning me down, whereas others that had accepted ended up offering a lot less than I expected. It was rather draining to face the disillusion of reality – especially once I got to know just how industry is like. Perhaps the term perpetual probationary period probably describes this best, along with the new trends of unpaid internships.
Out of all of this, I ended up returning to academia, but not in a way that I expected. My appointment to the position of Postdoctoral Research Fellow (Biomedical Engineer) at the MARCS Institute for Brain Behaviour and Development at Western Sydney University was a completely unexpected turn to broaden my interdisciplinary engineering philosophy to a whole new field. In my role, I am “resident” at Liverpool Hospital, having the opportunity to better connect engineers with clinicians and real problems in the hospital. In some ways, I do feel a little overwhelmed with the change, but I’ve also been involved in doing what I do best – troubleshooting and building embedded systems. While I am being adequately compensated, the workload is sometimes high and the contract only lasts one year. Unfortunately, certainty is not part of research – but it was still better than the competing offers. I can only hope for positive results, but the future is unpredictable. At least, I can say I’ve made some positive efforts in this aspects of my life.
As a result, I didn’t have much time nor energy to blog. Instead, I spent most of the time and energy delivering four massive RoadTest reviews at element14 – being the lucky awardee of a Rohde & Schwarz RTM3004 Oscilloscope, B&K Precision BA6010 Battery Analyzer, Molex 2.4/5Ghz Antenna Kit and Rohde & Schwarz HMP4040.04 Programmable Power Supply. Big ticket items deserve thorough attention-to-detail and comprehensive reporting, but I suspect maybe my efforts are not quite as well appreciated as I would have hoped. Then again, your average Joe isn’t going out to buy one of these from their local Woolies.
That’s not to say I didn’t leave some time to blog, but often that entailed blogging about more recent/current items or events. This unfortunately, doesn’t cut down on the backlog of blogs that I have intended to make, so it always seems like I’m getting nowhere. I feel perpetually guilty to have some images prepared for a blog over a year ago … but it just hasn’t seen the light of day yet. That’s the case with my 2017 holiday – very little of it has seen the light of day. Unfortunately, this also means that some parts of the site have been missing updates … oh well.
One thing I’ve changed is that I’ve been watching more video content than before. I’ve been watching videos from a number of different YouTubers, most of them very nerdy and specific in some ways:
- Technology Connections – great channel for looking at vintage technology and how it works – especially Laserdisc, VHS, Betamax, etc.
- Geoff Marshall – knows literally everything about the London Tube system, which makes for very interesting experiments with public transport in the UK, such as the “all the stations” project and world-record attempts.
- HVACR Videos – think ElevatorAdjuster, but for air conditioners and refrigerators. It’s great to watch experienced people do their work, document their decision-making and troubleshooting process, explain some rather complex concepts and share their industry experience. Big respect for the effort in producing watchable videos and answering questions all the time.
- N.W. Ohio HVAC Videos – just as above, but from an equally humble and experienced engineer. I’m not in the HVAC business in any way, but it’s fascinating to watch and learn.
In something I don’t think I’ve admitted to openly (much) – I’ve actually spent the last two-and-a half years watching (subtitled) Korean TV shows. I find Korean TV to be quite entertaining compared to the local stuff. In this time, I’ve watched a very long list of shows – but the ones I actively follow are:
- 1 Night 2 Days
- Running Man
- Idol Room (formerly Weekly Idol)
- Problem Child in House
- Twilight Delight (Your Rosy Cheeks)
- My Daughters’ Men
- I Live Alone
- Thrifter’s Guide to Luxurious Travels (Salty Tour)
- Real Men 300
- Chart Go
- Michuri 8 1000 (Village Survival, The Eight)
- Law of the Jungle
- 4-Wheeled Restaurant China
- Guess My Next Move
The list of shows I have watched is a lot longer than that – I still miss some of those that have reached the end of their run, such as We Got Married, but we take what we can get. Some shows got bumped since they weren’t interesting, or the cast changed, but it’s not an issue since I’ve got plenty to watch. Lucky that I spend a lot of time on public transport, so I can enjoy it on the way to and from work – although nowadays, I do watch it at two-times speed to get more out of my day. I still don’t have the patience for a movie and frequently interrupt my viewing to attend to “stuff”.
Of course, I’m also partial to some K-Pop groups too – getting to see behind the scenes in the “systematic” process that is their music industry is rather interesting and makes for some rather endearing situations.
But despite this, my biggest failing of this year was that I never really had much time nor opportunity to develop the more “social” side of my life. I did occasionally go out to meet with a friend or two … but not being a social butterfly without so much as a social network of like-minded or even “compatible” friends, nor having much free time, I’ve basically “flown solo” most of the year. I’ve learned to become perfectly comfortable on my own, in my own room, with my own gear, pursuing my own interests – but there’s more to life than that.
Assimilating to a more social culture is difficult. When you find the latest trends rather trivial, when you can’t understand the narcissistic obsessions with selfies or the “you-only-live-once” extravagant nature of many people in my age category, it’s hard to find any common ground. Of course, to them, I’m equally as strange and alien.
But it’s always been my hope to meet someone I’m compatible with … sadly, the superficial nature of internet dating doesn’t do me any good either. Not being too photogenic and not taking selfies at exciting locales makes for a boring profile laden with text – a hard sell in a world where people can’t stop to spend more than a minute reading, as they’d rather watch a video.
New Years Eve – on Optus D2
Every year since I got into watching feeds on Ku band satellites, NYE is always an event I go looking for feeds to watch. It’s a higher-quality way to enjoy the event by receiving the signal sent by the satellite uplink van from the event – a stream with fewer watermarks, higher bit-rate, uncensored and with no ad-breaks.
This year was no exception, but the need to feed was compounded by the fact that my location is not great for getting DVB-T terrestrial broadcasts and my internet is quota-based LTE, so streaming the event wouldn’t be a good outcome – lower quality and high cost. As a result, I got my dish pointed at Optus D2, certain that the feeds would be exactly where I’d expect them.
But what I didn’t expect was problems receiving anything from the satellite. Rather unfortunately, my “beeper” style satellite finders all failed just last week during my attempt to scan the Ku band arc … but signals were all lower than expected. DTH carriers formerly with 10-13dB SNR were peaking at just 8-9dB, meaning a very small margin to failure and almost no chance of catching weaker feeds which normally have lower SNRs.
The cause was not entirely unexpected. My dishes have always been propped on caravan-style portable stands, weighed down in some way. The freak storms that have hit Sydney this summer resulted in rains and strong winds – sufficient to topple the dish twice onto the concrete. As a result, the dish is probably misshaped and the LNB holder has been destroyed. A little electrical tape does hold the LNB now, but not as firmly in alignment as it should be.
Nice rack, ABC!
No surprises, the feeds were exactly where I expected in a very common mode – 7200kS/s using DVB-S2/8PSK. Two carriers were used – one at 12673V and another at 12682V. The lower carrier is the “in-house” ABC feed for local broadcast, whereas the upper one came online later in the evening for “international” outside broadcaster use.
Both streams carry about 16Mbit/s of data in just under 9Mhz of spectrum, allowing three to fit within a 27Mhz transponder. It’s no surprise that another identical feed sat below the lower carrier – this one was used by a New Zealand broadcaster (it seems), but I wasn’t interested in that one. The local feed has no pilots, which makes it a little more difficult to lock – so I had to use my Prof 8000 card to decode it.
The broadcast is sent up by Telstra Broadcast Services (as usual) with a 1080i H.264 video stream of about 14Mbit/s and four stereo audio streams in MPEG-1 Layer II 256kbit/s. Nothing too special about the set-up.
The international feed has practically the same set-up, the exception is that pilots were turned on. With an SNR of 9dB+, my problematic TBS6925 was able to still lock the service, occasionally “bursting” in errors when it loses lock.
Unfortunately, this screenshot was taken at good conditions. This year’s NYE celebrations were dampened with some rain – this was a doubly-bad outcome for me as the rain-fade reduces the SNR locally (down to 5dB at some points) as well as the SNR of the uplink. As a result, there were quite a few times where I just lost the carrier entirely on the 85cm dish – below 8dB, the BER is non-zero.
The rain did clear at midnight so, no problems there … solid reception throughout the midnight fireworks!
Feed Content & Structure
I’ll try not to share more than necessary for fulfilling curiosities and documenting what I received.
The best part of watching a feed is tuning in before the event itself and watching them prepare. A lot of effort goes into making sure a program goes off without a hitch – including rehearsals and last-minute script modifications as above. Occasionally, you get a glimpse of their equipment as well.
The video switcher they used seems to emit blue background when no valid signal is input. But as a valid signal begins to be input, there are “stripes” of signal showing through. I wonder if this is a consequence of trying to switch between sources that are not genlocked?
As with every year, getting wireless video links to perform stably during an event like this is a problem. This year, there seemed to be fewer glitches – but they were still there. If you saw these, it wasn’t your reception – it was the drone-to-truck link suffering uncorrectable errors! I know it because I saw no TEI marked packets around this time, nor any CRC errors on the transport stream. No ABC logo on the feed – but the #SydNYE seems to be applied at the truck … so no escaping that.
The event coverage was pretty good – but when I glanced at the local broadcast, I did see the images were a little soft and the black levels seemed to “clip” resulting in blocky artifacts. But since I didn’t get a stable reception of my local terrestrial signal, I decided to check it out via iView today and what I saw was even worse in quality. In this “split” image, we can see just how much definition was lost in Daryl Braithwaite as he was singing Horses.
Unfortunately, I do think that ~14Mbit/s from a “live” H.264 encoder is still a bit paltry compared to the “heaviest” feeds that I’ve seen push close to 50Mbit/s (possibly because they’re looking to make a Blu-Ray from it) – it’s better than the 4-5Mbit/s you might be getting from a terrestrial transmission. But I’ll take what I can get – satellite spectrum is expensive …
Once the event is over and the terrestrial has cut back to the hosts, we are treated to aerial drone views and a view of the pylon projections. It’s not over until it’s really over.
On the other hand, the international feed starts up later in the evening, with a relatively modern test-pattern with audio sync indicator. The international feed does persist for a little longer as well, not going down until an hour after the end of the event.
In that time, the “90 second” short-clip is transmitted, followed by a replay of the full fireworks. The best part is that these replays are often more carefully prepared – in the case of the international feed, it starts ~60s before the new year to fully capture the countdown. The feeds have a different audio track layout as well – the primary track for international contribution seems to have different music, possibly due to rights issues or to avoid any music with vocals (avoid clashing with a news anchor narrating over the event possibly).
Stereo Audio Track 4
In the past, I didn’t spend any time analysing the audio tracks, but this year was a little different. The tracks seemed to be quite well used:
- Track 1 – Local Audio
- Track 2 – International Audio
- Track 3 – Ambient Microphone
- Track 4 – Timecode & Studio Comms
I wasn’t quite aware of this at the time, but I checked Track 4 to find mostly silence on one side and a continuous “blipping” on the other. At first, I thought it was a crashed decoder … but alas, it wasn’t.
This was the audio right at the point of the count-down – the studio comms is on the right track with a New Zealand accent.
Analysing the left track shows a repeating pattern of sorts – we can hear the pattern changes right at midnight as well. That’s SMPTE Time Code, as a Linear Time Code track. It’s the first time I’ve encountered it on an MPEG-2 encoded transport – surprisingly it fares perfectly well despite its limited amplitude. I wonder how useful it is, as theoretically you can carry some of this type of data in the transport stream (e.g. in the presentation time stamps (PTS) clocks, in the TOT/TDT tables). Then again, I guess it doesn’t hurt to have redundancy.
I decided to sit down with the specifications and attempt to decode the two packets either side of midnight. Unless I got something really wrong (I doubt it, as the sync words came out fine), the time code seems to be a very arbitrary time (not wall-clock) and seemed to have reset at midnight. At least I tried – my first time decoding time-code from an audio source manually.
Time to Let Go of Satellite?
Unfortunately for me, it seems I might be giving up the satellite TV hobby entirely. My dishes aren’t as healthy as they should be, my TBS 6925 cards have all become problematic, the costs of equipment has gone up now that satellite TV is not as popular as before, but the Ku band satellites are almost all encrypted with very few FTA services. In fact, the DTH market is really copping a big hit because of IPTV being a more cost-effective means of distribution and VoD services seem to be a big hit with consumers. As a result, many of the satellites are dedicating more bandwidth to VSAT data services using ACM/VCM, sometimes generic-stream/multi-stream features, none of which are particularly interesting.
Random: Radio Lost an Ear?
Earlier, this Chinese ethnic radio band receiver came in for a quick-fix since the antenna was snapped off at the base. This is basically a narrow-band FM receiver in the VHF band.
As I had no spare “rabbit ears” telescopic elements I could just swap in and the base was made of a metal that couldn’t easily be soldered to, I decided to repurpose a crimp lug and snap off the base to fill the inside of the base of the tube with solder. The crimp lug is soft and had to be doubled to make the necessary thickness, but if gentle, it worked just fine.
Despite this, there was a complaint that the reception was a bit weak. I provided some wire on a crocodile clip which helped things – but not before that, I wanted to take it apart just to see if there’s any obvious reasons. The radio comes from 2AC and seems to have their own custom PCB, dating back from 1994 as far as I know.
The receiver is a fairly simple one based on a Motorola MC3335P IC.
All three frequencies it was capable of tuning into are crystal locked – so no drift potential there. There may have been slight drift in the IF stages – but any tinkering could make things worse, so I put it back together and left it at that.
Random: DisplayPort Sleep Woes – Dummy Solutions
My recent upgrade to 4K monitors was not without some downsides. One thing I discovered was that DisplayPort behaves strangely when monitors are powered down – in the analog/DVI days, I would power down monitors manually to save energy and prevent nuisance screen “wakes” at night. Unfortunately, doing the same with DisplayPort monitors results in them disconnecting from the system and the GPU, which is reported to the OS.
This has the downside that if I am connecting via VNC, the framebuffer doesn’t exist anymore and I have no ability to control the computer or even wake it up! Losing remote access is a pretty big issue, but another one is that depending on how you power up your monitors, your windows can get rearranged because there is a time the computer thinks there’s just one monitor and not two. The icing on this cake was that the monitors also presented audio interfaces, so powering them down/up resulted in audio stuttering which wasn’t nice.
Searching for solutions seems to show that this is a widespread issue, especially with nVidia GPUs, with no real workaround and a variety of symptoms. As a result, I suspected the solution would lie in having an analog display always connected.
As miners and other GPGPU compute users would know, most modern graphic cards can’t be told to “assume” a monitor is connected without a “dummy” plug. It’s as if there is something sensing load on the R-G-B lines to determine if an old dumb monitor is connected before allowing a manual display configuration – so I built this dummy plug using some header pins, left over 56 ohm resistors (although 75 ohm is more appropriate, the difference shouldn’t cause any damage – but do it at your own risk), a DVI to VGA plug (to fit in my GPU’s unused port) and some hot glue for insulation.
Of course, I had to manually configure a “third” analog display, cloning my primary display to ensure that despite the two DisplayPort monitors being turned off, that there was still a monitor connected. This seems to work just fine – the downside is now, the graphics card is working harder to output a 4K 30Hz signal over VGA that’s not being used at all.
But now that this is solved, I’ve now found that my computer doesn’t reliably sleep and wake anymore. It fails about half the time, leading me to hibernate the system instead, but the cause is not known and a BIOS update didn’t help either. It happened after the Windows October 2018 update, so I wonder if it’s another driver incompatibility … or whether it’s also DisplayPort related.
Random: TP-Link Switch Contains Pork?
While sitting at my computer, I started to notice some squealing. You know, that kind of high pitched whine that comes and goes, but gets more persistent over time? The kind that eventually distracts you and drives you nuts?
At first, I thought my computer power supply might have been responsible. But after shutting down the PC, the sound was still there. Sniffing around (with my ears of course), I narrowed down the source to a switch-mode plug-pack for my main 8-port TP-Link “dumb” GbE switch.
In previous reviews of switches, I never tore apart the supply as it wasn’t designed to come apart for service and I didn’t want to damage it. But this squealing … this was something else!
So, into the vice of knowledge it goes and a teardown is in order. I was convinced there would be some bad capacitors inside … that normally leads to squealing …
With the seams broken all the way along the edge, it was interesting to see the metal prongs are connected to the PCB by forming sheet metal into a connection spring that “bites” onto the edge of the PCB.
A look inside the top cover seems to reveal some thick, yellowy, electronics-smelling deposit which … seemed conclusive … until …
I looked at the PCB and there were absolutely no visual issues with any of the capacitors! Yep, all the capacitors look good. In fact, I got my multimeter on them and all of them measured appropriately in-circuit.
The supply looked relatively simple on a single layer paper-type PCB. The number of electrolytic capacitors used was a little worrying, especially because they seemed to vary all-over the pace.
The output has two parallel capacitors, but despite having the same values, one is a Taicon while the other is Samxon. That just strikes me as odd. The remainders are Aishi branded … so generally not really problematic brands but also not premium.
The underside shows good isolation between primary and secondary, with a PCB spark-gap which seems quite long. I had to use the thermal-paste-on-finger trick to get the markings of the IC to show up – it’s an AP3706 Primary Side Control IC for Off-Line Battery Chargers.
Random: Kingston Datatraveller 128GB is Exhausted!
With my recent trip back to my other house, I found one my the TVs had malfunctioned but I didn’t have the equipment nor time to diagnose it fully. After partially pulling it apart, I decided to take out the 128Gb USB drive (previously reviewed) that had been used for PVR storage, in case I could reuse it.
Unfortunately, once I formatted it and decided to try testing it, it showed very slow writes before dropping out of the bus entirely. That’s basically game over for this drive – quite poor considering the TV was rarely used, so the unit would not have done much recording or timeshifting. I guess that’s the quality of modern TLC-based solutions.
I was going to recover the drive using the manufacturers’ tools, assuming it would be Phison (as most Kingstons usually are). Unfortunately, none of the MPTool/MPall/etc software would even list the information on the controller. Out of desperation, I tried the Alcor Micro software with no results either. So what’s inside the drive? This gives me an ideal reason to tear it apart. It comes apart very easily if pried at the edges near the ring end …
Inside, a single unbranded BGA flash package is used to store the data. The lettering along one side suggests this is probably IMFT flash. On the rear, there is a space for a leaded chip and a Solid State Systems 6130 controller – one which has no publicly available MP tools to my knowledge. Oh well, into the bin with this one.
The world is a fascinating place and I hope to be able to maintain the blog in some way and make postings, even though time does get short from time to time. I’ll continue to take apart, test, play with, modify, repair and torture electronics as I normally do, as that’s what I love to do. But I hope to be able to share that with you all in some way.
More than that, I hope to be able to advance my life in other domains as well, although sometimes it’s hard to maintain optimism in a world where economic conditions make it difficult for all to envisage a future of stability, prosperity or even security. I still hope that 2019 would be a better year for myself, my family, my friends and everyone (in general).
Regardless, I hope that you were all able to enjoy some of my postings made in 2018, even if they were more fewer and more randomly uploaded than usual. There are still a lot of intended postings I want to make, but it might well be years before they see the light of day – my holiday in 2017 is still conspicuously absent.
But no matter what happens on the blog, whether it seems quiet or not, I’ll always remain busy doing something. Not everything makes it on the blog – not everything can. But what does make it on the blog … sometimes doesn’t get much attention, which might not be so surprising given the niche nature of the content. But I’m still hoping that it would be of use to others and that it might be shared. Maybe some things take time to be appreciated, but I’ll do my best to carry on as I always have … into 2019!
Bonus: The Future Calls …
Before I go, I just have to post a picture of the relic that is the public payphone. Around Christmas, Telstra offered free calling from payphones … so I couldn’t help myself but give it a go.
Surprisingly, the phone worked just fine, the audio was clear but a little low in amplitude. The quaintness of a dialtone, DTMF weren’t lost on me, but … the future is calling … although not on landlines … and maybe not even in voice. Surprisingly, I thought there might be queues for free calls … but there was not a single person at the payphone. Besides me, that is.
Just this experience gave me pause to think just how far we have progressed. I’ve now reached the point of my life where nostalgia begins to take over … and it just seems strange to think that the younger generation may have never experienced using a smart-card phone card, fumbling for coins, using a payphone that doesn’t give change, or waiting to use a landline phone. To think we are now all practically wireless when it comes to telephony is … rather astonishing.
Will you answer the call of the future? Interestingly, I’d have to say that I’m becoming increasingly wary of new technologies … after all, progress is not always for the best. VoIP technologies over packet-switched networks don’t achieve the same reliability or quality as circuit-switched technologies – but the costs of maintaining circuit-switched networks have led to their phase-out. Likewise, wireless technologies are often contention-limited, so in cases of disaster, maybe you can’t get a call through at all. Maybe we need to critically evaluate what we have now and whether we’re even making good, appropriate use of it. Maybe it’s worth thinking not only about what you gain, but also what you lose.
I think the same kind of issues persist throughout a society which is seeing technological disruption at a breakneck pace – so quickly that regulation can’t catch up and when it does, damning revelations seem to be the expectation. Maybe we need to slow down, think carefully, tread lightly and avoid waste … just my two cents for 2019.