Oct 21, 2020

Last Salute to the Samlex SEC 1223

Back in 2008, shortly before I got my Technician license, I bought some of the basic ham equipment in anticipation of getting on the air. One of the first things I got was a Samlex SEC 1223, a 23 amp power supply. Not too big, and not too small, with enough juice to power my Yaesu FT-857D without complaint. 

Twelve years later, the Samlex continues to work well, in conjunction with a RIGrunner 4004U attached to the topside. In fact, this is what keeps the lights on for the homebrew 20 watt repeater and the 25 watt APRS digipeater iGate. For the sake of modernizing and staying better informed, a digital volt and amp meter is ready for installation into the center of the face. Or should I say, "was ready".

So the other day, right after I got the duplexer installed on the homebrew repeater, I turned everything on in preparation for testing. My buddy Stu KD9MNK, was standing by at his place ready to help with the testing and signal reports. We talked back and forth for maybe a minute and suddenly, POP! And just as suddenly, the repeater and APRS went dark as the smell of ozone crept into the air. A quick text to Stu, "I think I blew a fuse!" just to let him know I'm still alive.

With fast thinking, some fine-motor dexterity, and a couple of powerpole positions switched to another RIGrunner, I was back on the air, this time on the Yaesu FP-1030 linear power supply. "Cool!", I told Stu, "That was really loud for an internal fuse!" as we continued on with testing, thinking nothing more of it, except for the ozone in my nose.

A blackened PCB at the fuse clip

Later in the day, I removed the Samlex cover with the RIGrunner attached to it and noticed the 6.3 amp glass tube fuse next to the input power had popped, I mean really popped. A virtual trip to Amazon Prime and two days later, I had a small pack of fuses in my fist. Minutes later, the Samlex was ready to rock, so with the cover back on, I plugged in the power and flipped the switch. 

Reminiscent of a California brown-out, the lights in the living room dimmed for a second. Then the switch light on the Samlex turned off without a sound and the living room lights returned. "Hmm", I thought, "That's weird." I pulled the power cord and re-removed the cover. Being a little annoyed over the whole thing, I just kind of tossed the cover upside down on my desk, having no regard for the RIGrunner on top. As I was about to go after the second burned fuse, for some reason I glanced over at the inside of the cover. "What is that?", I thought, as I noticed a small amount of white residue that, based on the printed pattern, was applied at pretty high-velocity. That's odd, a mark I never noticed before. 

Looking at the location and pattern, I quickly examined the matching area components on the PCB, and there it was, the problem. And what a problem it was! Three capacitors in a row, all with their tops popped open and looking rather burnt, way across the PCB from the blown fuse. Wow! I guess that explains the loud POP! No wonder the fuses blew. While wondering what other parts have been compromised, I thought, "This is a good time for an upgrade". 

After a semi-virtual, COVID-modified, trip to HRO, a new Diamond GZV4000, 40 amp, 100% continuous duty cycle power supply is powering the repeater and the APRS digipeater without even thinking of breathing hard. And the RIGrunner 4004U has no complaints either.

If we could have a moment of silence, please, as I offer one last salute to the Samlex SEC 1223 before it goes to the great recycle bin in the sky... or the recycle bin at the township. Thank you.

Ah, no.  That part is not OK, and neither are the other two.

Oct 12, 2020

HughesNet Satellite Internet Tests

I've never been a big fan of satellite ISPs, and the latency tests (last pics) show one of the reasons why. And the rest of this rant are some of the other reasons. 

According to their website FAQ's page, the HughesNet geostationary satellite is 22,000 miles away. The stated max speed is 25 Mbps down and 3 Mbps up. Right.

So, if it takes 10 seconds to get 25 Mb of data (2.5 Mbps), but then 60 minutes to get a second 25 Mb of data (69.4 Kbps), the average speed should not be 1.375 Mbps (data/time)+(data/time) / 2 data sets. The average speed should be 0.135 Mbps, or 135 Kbps (total data / total time fetching that data).

Remember this one from a couple weeks ago? Here's a screen shot while updating a Raspberry Pi. Note the speed. This was a 55 minute download of 84.8 MB, with about 40 minutes of this. Notice it's on the 4th attempt to get the data? Yes, 4th attempt.

Below is a HughesNet speed test. The actual time to complete the 2 MB download was 24 seconds, from the time the progress bar started, to finished. HughesNet states the speed is 1.41 Mbps. Hmm. My math says 2 Mb / 24s = 0.083 Mbps, or 83 Kbps. What kind of math do they use? Do they subtract all latency time, buffering time, break time, lunch time, etc?

Funny. Even so, by their standards, the speed on this run is way into the "crappy speed" zone, even for their crappy 12 Mbps, top-o-the-dial bliss zone. Oh, notice the Ping Time. It didn't even register. Welcome to every evening at my house on satellite.

These are some screen shots of the last 30 days using the Google Wifi app. Here again, calculations must be made only when the status is "active", as opposed to "waiting" or "buffering", or having a file restart 8 times because it thought the connection was completely lost (this happens frequently). These numbers? These are way off from reality. Again, even so, this is pathetic.

  

More good news. We live 2.0 miles east of a brand new fiber optic cable, and 1/2 mile north of a high-speed cable, but the ISP won't bring service this way because they don't want to, unless all the neighbors pay $3,500 each to get it installed (even though they equip subdivisions for free, knowing they will soon have customers).

AT&T has really old telephone lines in our area, and not enough nodes for us all to get a telephone line. They also offer DSL service, but unless one of the neighbors gives up their connection, no hard-line phone or DSL is possible. Who even thinks about being happy with DSL any more? Right?

Okay, back to latency. Is almost 4,000 milliseconds too slow? Who'd of thought? It's slow enough. That, combined with the constant buffering means, "no VPN for you". No VPN, no WIRES-X, nothing that needs any sort of constant, or low latency connection. Nope.

Even something as simple as Netflix must be set to the lowest quality video available, and still there's plenty of time to make the popcorn, while you "watch" the movie. Oh, and if it's Friday, Saturday, or Sunday night... Honey, where did you put the deck of cards? Wanna play 500 Rummy? We got time.

ELON MUSK!!!! Where are you?!?! I hope you're reading this! Where's my STARLINK?!?!

Okay, I'm done. Thanks for listening to my stupid rant. So, let's get on the radio! No LATENCY there!

Ha! ...just havin' fun!