K4KPN-10 Robust Packet I-Gate / Digipeater Update and APRS-IS Gotchas

May 26, 2018

K4KPN-10 has clearly benefited from the combination of antenna system repairs and noise floor (NF) reduction work at the site last Fall. Performance is back to what one would expect from it.

For those that ask why Robust Packet? One word….performance. Below is a screencap of what the 30m RP APRS world looks like from the Kentucky gate’s perspective over a few days.

RPR_K4KPN-10_20180519

K4KPN-10’s view of the 30m Robust Packet APRS world. Why RP? One word….Performance.

Around 25 to 30 of those stations on there have been heard direct on RF at some point. Single hop stateside coverage is easy. Europe is in to some degree nearly every evening and EU mobiles are often heard. Even the challenging polar path over to RT9K-15 is in there at times.

I’m hesitant to use APRS-IS for serious coverage analysis since it doesn’t begin to reflect everything heard on RF, especially on HF. Regardless K4KPN-10’s “heard” and “heard by” data is interesting:

RPR_K4KPN-10_heard_201805

APRS-IS perspective on RP HF APRS stations heard by K4KPN-10 so far this month.

RPR_K4KPN-10_heard_by_201805

APRS-IS perspective on RP HF APRS stations that have heard K4KPN-10 direct so far this month.

Yes some of the soundcard modems will do good work on 300 baud AFSK for single hop HF APRS and some occasional DX can get through. Some can do even better with bit fixing. The problem with bit fixing (guessing) is it breaks spec and can easily cause corruption if done too aggressively even on APRS. Thus why we so often trace corruption on the VHF APRS feeds back to soundcard modem equipped gates. Even with bit fixing the limitations of 300 baud AFSK become apparent over long haul DX paths.

Where Robust Packet shines is it uses Pactor-III like modulation that allows error correction without breaking spec. RP offers good noise immunity and deals with the multi-path on long haul DX packets better than a typical AFSK modem can.  Since RP uses space and power efficient dedicated hardware it is also well suited to stand-alone portable/mobile tracker uses without the need for a PC/laptop.

Robust packet is a mode and hardware built from the ground up for improved performance on HF and well suited to HF APRS. DX reception is not occasional, it is commonplace. I admit to sort of taking 4,000+ mile APRS DX as NBD till one of the HF AFSK guys sent me a “holy $#@$ what are you using that hears across the pond nightly” email LOL.

For those asking “where is K4KPN-10 hearing RT9K-15? Oh it is in there on RF, but due to his gate configuration it’s going to be difficult for stations other than a few lucky EU gates to ever get credit (APRS-IS perspective) for hearing him.

RT9K-15_at_K4KPN-10

RT9K-15 being heard direct at K4KPN-10.

Well here is an opportunity for a good lesson on the flaws of assuming APRS-IS gives a complete picture of what is going on at the RF level.

It is important to realize that APRS-IS does dupe checking. If I-Gate A and I-Gate B hear the same packet only one of them can get credit for it from the APRS-IS perspective. Which one gets credit? The first one that gets that packet to APRS-IS. The one with the lowest latency to the APRS-IS system.

If you run an I-Gate or home APRS system on both IP and RF you can easily screw yourself with bad timing values. The RT9K-15 I-gate is doing something that will make it very hard for coverage analysis via APRS-IS. It appears to be aggressively beaconing to APRS-IS every few minutes? I’m told that it is heard often in EU, but rarely does an EU gate get credit for it in the eyes of APRS-IS. This tells me it is probably making the 2nd mistake of beaconing to APRS-IS and RF at the same time. This means even if another gate hears his packet on RF then it’s unlikely outside of internet congestion on his link that another gate can decode that packet off RF and inject it into APRS-IS before his internet beacon gets into APRS-IS.

Another problem this creates for APRS-IS analysis is it throws the packet counts off on the above pages. A good example is it shows a monthly total of 5 packets for DF1CHB /AM.  He was in direct on the eastern USA I-gates for hours earlier this morning, but also hitting EU gates so NOAM will only get occasional credit for gating him even after dozens of packets gated in. Tonight he is airborne over Serbia and K4KPN-10 is hearing him direct like clockwork on 30 meters:

DF1CHB_30m_rpr_aprs_20180526

DF1CHB /AM over Serbia and coming in on 30m robust packet almost as if he’s a local on VHF LOL.

Don’t get me wrong, this is not a complaint towards APRS-IS. It’s just how the system works and how it has to work. It’s just important for APRS operators to understand that APRS-IS doesn’t give you the full RF picture due to the necessary dupe checking going on. This is especially true on HF where the same packet can easily be heard by multiple I-Gates hundreds or thousands of miles apart.

Also noteworthy that some stations run beacons marked “RFONLY” or “NOGATE” paths. A properly configured I-Gate will not gate those beacons into APRS-IS. This is another gotcha if you’re trying to get a full picture of RF level activity via APRS-IS based systems and apps.

Recommendations based upon lessons learned here if you are interested in getting decent coverage analysis via APRS-IS?

  1. There is normally no need for an APRS station to position beacon on their internet port (APRS-IS) more than once or twice an hour unless it is moving. Serves no useful purpose for fixed stations and it only adds to the cumulative load on the APRS-IS system.
  2.  I-gates should consider beaconing on their internet port at a rate of just under 60 minutes.
  3.  Avoid beaconing via RF and internet ports at the same time.
  4.  If you are not moving, then you don’t need to be pounding the heck out of the RF side either. 2-3 RF beacons an hour from a fixed station is more than enough to keep you on most maps and fresh in APRS-IS for messaging/gating purposes if you are in range of an I-Gate.

WA4ZKO


Packet Update, Winlink, Is Ham Radio Dying? Puerto Rico

May 26, 2018

Since the inbox shows some got worried when we went west early with only the APRS gate/digi project done I guess an update would be a good idea.

We went west for a friend’s wedding.  Pro Tip: Few things can compare to a wedding set against the Teton Mountains. Since we spent most of the post-holiday Winter season in 4-land I also had several 7-land biz/personal items needing attention. I flew back and the XYL is driving back as she wanted to stop and visit with her Iowa family.

For the worries about the packet stuff getting done this Spring/Summer? Well I had been waiting on two Liebert UPS systems to come in so I can finish up the power and remote control side of the “packet/HF” rack. They are in and this looks like a somewhat slow weekend/week ahead so we’ll see what gets done this week. completion time = good question. Between work and Dad’s health issues I’m not going to commit to any timelines on hobby stuff as I can barely nail down a work schedule beyond a week out. It is what it is, family and work come first.

What’s going to be put on the air?  Well I can say for sure that a 10m RP port and a UHF 9.6k port are coming. Our core occasional EMCOMM needs can be met by a local packet BBS with a 10m RP port plus a UHF 9.6k port. I’m still back-n-forth on spinning up the 2m/220 1.2k stuff. Granted we have lost several local ops in the last decade or so due to SK and job related moves, but I’m frankly more than a little shocked at how dead ham radio seems around here and across Kentucky.

A lot of area packet and voice infrastructure has came and gone within the last few years. “Lack of usage” is usually the reason given. Even worse I often hear sysops of existing gear commenting that they really can’t justify keeping unused stuff on the air. Several have said they’ll run it till they take storm damage then it will not be repaired. Use it or loose it folks.

APRS activity for “Dayton” seemed to be down somewhat. On that note, one of the western Kentucky ops emailed me that he came up for Dayton and could not hit a single Winlink packet gateway with his D710. I looked at the Winlink gateway map below and could only ask where are all the gateways? Lexington/Frankfort is a dead zone??? Several of the WKY gateways seem MIA. The SE KY gateway doesn’t exist, it is actually just a misconfigured Georgia gateway.

ky_rms_pkt_map_dayton_wknd_20180520

The other day I was at a tower site and tapped into a VHF-Hi antenna port to a nice true 6 dBd omni up at 210′ AHAG that I can hit repeaters from Lexington/Richmond, Louisville, and up into Dayton/Columbus from. Fired up Winlink Express, pulled the freshly updated channel list below and could not hit anything. Granted I was just using 4-5w from a FT-817 (very possible 25/45w could change things) to a 150-160 MHz optimized antenna. Still that’s not much of a gateway selection list compared to a few years ago. Lot of previously active systems around here are gone, misconfigured, or off the air for some reason. Yes I checked for the EMCOMM group.

RMS_Packet_few_20180526

Past chats with some of the Winlink sysops I know all reveal a common thread. Folks talk of wanting this or that infrastructure and it sees a brief burst of interest then activity just fades away. One sysop that shut his gateway down told me it was seven months before someone asked about his gateway’s status LOL.

In fairness this is not just a packet problem, we also see similar with the current state of D-Star, DMR, and the analog repeater scenes. I ran our group’s portable analog repeater from a couple sites from October 2017 till early May 2018. I kept it patched into a couple systems so I could both record and monitor any activity on it.  Over all that time the repeater’s total transmit time was just over 52 minutes. 99% of that was just the IDer running from lots of kerchunking, only four unique callsigns heard beyond mine, and two brief QSOs. As I told the guys, no need of risking that to storm damage for that amount of activity so it is back in storage till this Fall.

Several have commented that the hobby is dying a slow death. While I’m not going to be that dramatic the hobby is definitely facing some serious activity related challenges in most areas of the country. This even includes areas which used to be hotbeds of activity. We now have a boatload of paper hams and hams that have gone or remain inactive for a wide variety of reasons.

Kentucky has nearly 10,000 licensed hams so “where the heck are they?” is a fair and obvious question. I found it revealing awhile back when the primary ham radio mailing list for the state revealed only about 450 subscribers when they began a move to another mailing list system. I just looked and only around 190 have migrated to the new system after several months. One does not need to be a Harvard MBA to be floored by those numbers. Yes it’s easy to be a Negative Nate on this stuff, but Step 1 to fixing a problem is recognizing that you have one.

When asked about the latest ARRL licensing proposal I can only respond with “I  don’t think lowering our standards once again is going to change much. The downsides are many for a hobby of technical pursuits. I do thank the ARRL for the free lunch.”  For those wondering about the “free lunch” portion of that? Well I made a lunch bet back in 2016 that if ARRL membership continues to decline then expect another push to lower licensing standards.

To be blunt, the further “dumbing down” of the hobby is not the fix unless the problem revolves around fixing the flow of subscription/advertising money. You know it’s not like passing the current General class test is so difficult. Common to hear stories of folks that take the Tech test, pass it, then be handed the General for giggles and either pass it or barely fail it. It’s not like during the last solar cycle we had Techs piled up deep on the sweet chunk of 10m spectrum they already have access to. Since chasing “quantity” (ahem money) hasn’t worked out so well maybe it’s time to try focusing on quality?  I could go on and on but this post will likely be long enough as is LOL.

So back to VHF/UHF packet…..

Considering all the above I’m obviously going to have to ponder how much 1200 baud VHF gear I want to put on the air. Would it see enough BBS and Winlink usage to justify it? If it wasn’t for already having the gear and putting in the extra rack space for the remote HF-UHF station then it wouldn’t even be on the table for consideration.

 

My stance on Winlink….

I’ve never been overly warm to the VHF/UHF side of Winlink. It doesn’t make as much sense as the HF side of Winlink does. One of the locals did make a reasonable argument that since the Dry Ridge site serves several counties across a handful of power grids and ISPs then there are a few EMCOMM scenarios when VHF/UHF packet Winlink access could be useful, especially at 9600 baud.

I’ve never been a huge fan of Winlink for a variety of reasons. Some are security related, the abuse that goes on, software stability, software roadmap stability, and ability to audit what content is flowing over systems under my license. Let’s examine each one of these..

  1. Security remains a big concern. Moving things onto a certain cloud provider doesn’t help make one feel all warm and fuzzy but I get why they did it.
  2. The abusive use of Winlink on HF remains a concern. Sounds like it’s gotten worse now that the solar cycle has forced more and more of the HF Winlink folks down onto already crowded lower HF bands.
  3. Software stability issues remain. Client software glitches are one thing, but gateway software needs to be rock solid. We are now how many years into the development of the RMS software?
  4. The software roadmap appears as spastic as ever. There is really no reason why the version silliness needs to go on to the degree that it does. I mean come on after all this time surely we can have long term production versions that do the basics and do them well. Then have a beta fork for those that wish to play with whatever new shiny soundcard mode is the current rage. If I put RMS services on my gateway and it’s a never ending PITA of having to upgrade and revalidate new software versions then RMS will quickly go bye bye. I can always use the SMTP functions in BPQ32 for my own limited “Plan D” backup email needs.
  5. The ability of the sysop to audit content flowing over his/her RMS gateway is a mandatory requirement in my book. If you can’t inspect the content flowing over your system then how do you detect abuse and ensure rule compliance. While it’s not as good as it could be I will  credit the WDT for giving RMS gateway sysops some ability to review messages flowing over their gateways. Does it need improvements? Yes, but it is a start in the right direction.

So as you can hopefully see there are many things being considered in regards to what will be reinstalled and what new features will or will not be available.

 

Packet TNC options…..

Some have commented they would consider saving up for some packet gear if they knew with some certainty about what is going to be available long term. 10m RP (Robust Packet), 9.6k UHF, and maybe the 6m/220 1.2k baud AFSK ports should be around for the long haul. Due to how noisy the VHF band can be in some of our needed coverage locations I may convert the 6m 1.2k port to RP also. If I put the 2m 1.2k and 220 1.2k ports back online their future will be dependent upon their usage levels which I will examine in a couple years to see if they are worth maintaining.

Two decades of experience showed the 220 band was more useful than the 2m port. This was due to being cleaner spectrum and greatly reduced desense/bleed-over issues when we worked along side our public safety systems which are predominately 150 MHz. 145 MHz also tends to be full of various noise/QRM sources and that problem is only getting worse.

 

The ARRL / ARES Puerto Rico Adventures?

My KP4 trip was months afterwards and unrelated to any of that hot mess. I’ve heard both some first hand accounts and a lot of second hand stuff including the thought provoking interview the HRN crew did with a couple ops that went into post-hurricane KP4 for disaster relief. In short I could write a mini-novel on that topic. I’ll just hit on a few key points here that will be lengthy enough.

First – The ham radio failures in more than a view major drills and disasters of the last couple years should serve as major wake-the-heck-up calls for the EMCOMM folks. I’m around a lot of LMR and EMA folks for my day job. Trust me today many of these leaders are not impressed by the local ARES groups and are going to call the MARS folks first. Lot of this is due to past bad experiences and the often correct observation that their MARS teams are both better equipped, better trained, and thus more likely to meet a particular EMCOMM tasking.

If ARES wants to sell itself as the ones to call for when all else fails then ARES better be equipped, trained, and well practiced at operating under such conditions.  Hint – it takes more than fancy radios and go kits.

Second – Please DO NOT send folks into disaster zones with only soundcard modes for Winlink needs. Yes I concede that it sounded like they had to put that HF gear together quickly (didn’t already have it????) with portability and accountability in mind. Still for the love of Pete equip them with quality Pactor modems that are going to work better under the often marginal conditions of such deployments. Keep the soundcard modems as backups if you wish. That “you get what you pay for” saying has some merit to it.  Hint – there are reasons why SCS doesn’t sell soundcard Pactor software when they could make a fortune if they did.

PS – SCS would probably deeply discount Pactor modems to such official staged HF disaster go kits if asked via the proper channels. They have been generous towards good causes in the past. Hint hint hint hint!

Third – Even at this point in the solar cycle, down there, that time of the year, the higher bands will often be your best friends so take appropriate antenna system(s) with you. This “we’ll do it all on 80/40m” mentality continues to be a recipe for failure. 60m NVIS will be handy and having a good antenna for the WARC bands is a really good idea. Be aware that 15-10 meters may not only be open some during the day but may provide some nice SNR on the link. The propagation predictions within the Winlink software are barely useful for the lower bands and can’t predict all the propagation modes on the higher bands. A good op learns HF propagation, understands both solar cycle and seasonal changes, is aware of single/multi hop short skip possibilities on the higher bands, checks the beacon subbands, and tries to use the highest band available for a particular path.

Forth – It seems time for AMSAT to start dreaming big again. Imagine what could be done with a 9600 baud store-n-forward digital bird or two up there. Yes that is easier said than done, but we also need to think beyond HF.

All the above aside, at least some aspects of the KP4 deployments in 2017 were moments ham radio could be proud of even if it was an ugly year for ARES stateside. We also got a real world lesson on the importance of knowing how to handle formal traffic and the NTS folks have things to be proud of.

I see some latched onto the lack of good ICS over the deployment cycle. Not thoroughly knowing both sides of the problems revealed afterwards I would refrain from bashing that aspect of things. As someone that was a first responder for nearly two decades I can tell you that regardless of what you are told beforehand you need to go in prepared to walk right into chaos. You’re not there because things are going well. You must be flexible, able to prioritize, work as a team, and able to adapt on the fly to various surprises. These are situations poorly suited to those that can only function off checklists in textbook scenarios. Ultimately they got a limited crew in, did some good, and then got everyone out safely. Could it of gone better? Of course it could of. There will always be room for improvement.

Non-ham politics of the Puerto Rico disaster? Well in many ways KP4 was a long neglected mess before recent events put it front and center. It did not get that way overnight and it will not be fixed overnight. I think the bigger question is will the repaired infrastructure be well maintained long term? Can their previously struggling economy recover from all that has happened?

WA4ZKO


K4KPN-15 I-Gate & WIDEn-n Digipeater now Operational

May 8, 2018

KYPN is happy to announce a new 2 meter APRS I-Gate & Digipeater callsign K4KPN-15 is now fully operational from EM78RP in Williamstown.

K4KPN-15 is a full 2-way I-Gate and WIDEn-n digipeater optimized for 1200 baud infrastructure use. System is transmitting pre-emphasized audio and the modem is fed flat receiver audio. This configuration helps provide improved compatibility and performance with the wide range of equipment and signal quality (wink, wink LOL) found on today’s VHF APRS network.

System is running 40w to 3db at 990′ AMSL. Performance testing with just a 5 watt tracker (Kenwood TH-D72) has shown good coverage to the west, north, and east. Coverage to the south is compromised due to both the rougher terrain south of Williamstown, terrain ridges, and some obstructions blocking a clear view to the S to SSW direction.

Low power tracker coverage around the Dry Ridge and Williamstown areas is nice although on 2m APRS I suggest 25w is going to stand a better chance against the congestion heard at your local digi’s receiver. Admittedly I had to find the “sweet spot” standing outside, but I managed to tag it from New Liberty KY (17 miles) with a TH-D72 HT on stock antenna which isn’t much of a VHF antenna to begin with.

Drive tests with the D72 on a quarter wave antenna has shown expected coverage. It hears better to the north than to the south and exhibits pretty much the same coverage of our central local public safety site when the remote receivers are not helping with the various coverage challenges our terrain creates. An interesting base-to-base coverage observation is it hears the WC8EMA gate/digi 60 miles away in Lebanon Ohio as good or better than the Edgewood digi at 24 miles.

I-75 coverage for a decent 25w tracker to an efficient antenna system should begin around the 168 mile marker southward till you get down around the 144mm. It is frequently hearing hilltop mobiles up to the 182mm. This has been verified with a 5 watt tracker to 1/4 wave roof mounted antenna. If you’re struggling to get in then either adjacent area QRM is really bad that day or you need a better antenna system.

Reference the above “better setup” note. One of the common observations of many voice/data infrastructure sysops over the last 20 years has been how often you will hear/see a ham mobile station running 25-40 watt mobiles yet have the talk out range of a HT. No joking, it’s increasingly common even on VHF where things tend to be more forgiving of poor installation practices. The graphic below is a textbook example. Here we have an unnamed 2m APRS mobile running a D710 passing through on I-75. He passes three different digipeaters and looks to only have maybe 1-2 miles of range. I’m thinking someone needs to examine his system as a 5w HT on the dash would do better than this LOL. Drastic example? Yes, but it is common to see mobiles with performance not explained by being on low power (bad idea on 2m APRS).

poor_aprs_range

Pro Tip:  Forget the fancy looking high gain multi-band VHF/UHF mobile antennas and stick to more simpler designs. Most of those overpriced things are lossy as all get out, look horrible on a vehicle, have long term durability issues, and “might” perform to 50% of their hype only if mounted in the most ideal location.  For 2 meters try a simple quality 1/4 wave antenna in a good location, fed with good coax/mount, and carefully installed connectors. You don’t need to spend big $$ to get good performance on VHF/UHF, you just have to do it right.

Okay back on topic……

Current plans are to put a higher gain antenna on it late this Fall, but for now this is working out well enough and let’s see if enough stations benefit from it. The noise floor at the location is still higher (S1-S3) than I’d prefer, but much better than the nasty S5-S7 powerline noise that used to plague the site. Some of this seems to come and go with dry/wet weather so I suspect we have another bad spot in the utility lines nearby. Being located in a commercial area with numerous offices (lots of switching PSUs, LCD monitors/TVs, cheap cell chargers, etc etc) nearby I doubt this site will ever hear as well as it could for its location.

The K4KPN-15 site has limited backup power and thus should not be counted upon for EMCOMM use. It is on a unique local city managed power grid that sometimes survives when adjacent areas are down, but generally suffers far more outages than adjacent area providers. The local joke is if a cow farts too hard in an adjacent county the power goes out in town LOL.

The Dayton Hamvention is coming up soon and that will give us a good look into K4KPN-15’s I-75 coverage. Gate/digi is not really intended to cover I-71 but I suspect it will help out some.

73
WA4ZKO