Saturday, August 17, 2019

Home Station Build - The Stability War

This is an old blog I never quite finished, but I think it's important more than ever with folks going to automatic doppler control, and wanting to use remote stations in Satellites. Therefore I've decided to put some polish on it and get it published in the hopes it may help someone else down the line.

--------

Moving from manual doppler control to automatic doppler control on Satellites was supposed to make things easier. Once the migration was done, sure, but getting there was a battle. In retrospect I suppose this all came down to me doing things unconventionally, and in the end it was going the more conventional route that won the day. Here's where I went with the battle, and how I finally overcame the challenges of building an automatic doppler control station.

SatPC32

There's a lot of satellite tracking and frequency adjustment programs out there, but I think SatPC32 is probably the best. When you're first getting it going you will need to calibrate it a bit to work with your radio. To do this, find a quiet spot in the passband on the bird and go into the Accuracy menu and mess with your uplink until you get it tuned just right. Once it's on, you shouldn't have to mess with it anymore assuming some other factor doesn't change. Other factors might include the things I'm going to talk about later, but this is where you'll keep coming back to to get your adjustments right. Learn this. 



Accurate Keps and Clocks

I already talked about the importance of up to date Keps in my blog HERE which should definitely read if you haven't. But knowing where the bird is in relation to you is how programs like SatPC32 to adjust your frequency for you. In addition to accurate Keps, you need to keep a very accurate clock on your PC. There are a ton of different ways to do this, which you can read about in some of my other blogs, but automatic doppler and digital modes both relay on your computer clock being accurate. Also consider when your keps get updated, you might need to go tweak with your SatPC32 calibration. It's usually not much, but any change in the factors might require a recheck. I wouldn't update your keps right before you're going to chase someone with a 30 second window at your AOS. Update and Calibrate on nice long passes where you've got plenty of time.

Extremes!

When using a rig in the field, you're exposing it to all sorts of different kinds of environments. Hot, Cold, Wind, moisture, you name it. This is a big reason why I stress manual doppler correction in the field, each of these factors messes with things like the cpu clock in your computer, or the stability of the oscillator in your radio. Adjusting by ear is using the CPU in your noggin to adapt. However at home when having a computer do it, you need to keep these factors absolutely minimized. Keep your computer and rig in a normally controlled temperature (65F-78F is usually recommended) and keep it away from direct sun, sources of moisture or abnormal air currents. Now that's not to say cooling isn't a good idea, cause it is. Keep good ventilation to your rig so it can cool itself off. You don't need a fan blowing right on it (unless you're doing jt65 on your 857 for hours like I used too) but having the air around it moving is a good idea. 



LO Crystal Replacement

This one is optional, but useful for Sats, and *really* useful for things like Meteor Scatter and EME. In my 821h I upgraded to the CR-293 high stability reference crystal for the Local Oscillator. I'm not positive my original crystal was causing the problems, but the upgrade was fairly inexpensive (~$90 on ebay) and I had heard from other Icom users that it made a world of difference. 

Here's the new crystal and the old one (after removal)


Where it goes inside the guts of the radio. Big Dashed Square is for the CR-293, small square is the original factory crystal. 


After using it for 2 more years with the upgraded crystal I never noticed a difference, but other users told me they did. I considered this to be a good, although maybe not critical investment.  


Voltage

The final change that got me over the hump to solving all the stability issues came down to one simple, but crazy important thing I neglected to consider. If you'll remember at this point in my ham experimentation career I still relied on batteries to do everything. I had several 70ah banks of 12v slabs that powered things in my shop. I would use some of the higher performance lifepo4 Zippy's for portable work, but everything was still charged and discharged thru chemical means before being turned into RF. One day I was testing output on my rigs and noticed I never could get full power out on the higher watt settings on the higher frequencies. The Zippy was better, but still lagged a bit. This came down to voltage under load - which is a simple concept. The more power you're pulling from a battery, the ability it has to push that power drops, ie voltage. Logic follows then, for electrically powered clocking type things (like a cpu or an oscillator) voltage variations will change the rate at which they clock. In the perennial words of Gru - "LIGHT BULB".




I finally gave in and got a power supply. I didn't opt for anything huge, just a 30A 13.8V constant supply from Powerwerx. Based on the specs on my radios I figured this would be sufficient to get full voltage even under load, and it was. I went from being 500hz low every time FO29 came over, to be spot on every pass. And while it did put out a little bit of noise, it was nothing that a few ferrites couldn't fix for me.  I had finally figured out the key to having good stability on a home automated satellite station and it was as simple as good power in, good power out.

The Final Recipe

If you're just considering an automatic station for the first time, I would recommend in this order things you need to do to keep your doppler correction on point. 

1. Stable Power
2. Accurate Clock
3. Frequently Up to date keps
4. Stable Temperature
5. Upgraded Oscillator Crystal. 

Do this, and you should be good to go. Good luck in your own battles with stability, once you're won, the spoils of victory will be yours. 

Monday, August 5, 2019

QRP EME - An incomplete Guide, written by a blind guy leading the blind.

So while I have about 40 drafts of blogs I've attempted to start over the past year and just haven't gotten done, somehow I've convinced myself to try and write something about QRP EME. Since finishing my 488 on Sats I've really been debating what to try next, EME is the natural evolution for a weak signal VHF/UHF guy who enjoys pointing his Yagi's at the sky.


OK, where to begin: Probably first, a disclaimer. I have made a total of 13 EME contacts in my entire ham career using my own equipment. And among those, only 7 were unique stations (I've worked a few stations twice, and one thrice). I did sort of make about 130 more EME QSOs on the W7D rove with Wyatt, AC0RA, but that was his gear and his setup, I was just the guy clicking the buttons and moving the beam. All this is to say, I'm hardly an EME pro. I know just enough to squeak out a handful of Q's a couple times a year with modest equipment. If that's why you're here, great - read on. I'm going to write this from the assumption you're a moderately skilled to above average AMSAT operator like me too. I'm going to assume you've got similar gear as me, and point out places where gear absolutely makes a difference. I'm also going to assume you're semi computer literate as doing this requires some mouse button mashing.

Computer Stuff

You're not going to make an EME QSO with a QRP station without digital modes. It just won't happen. If you want to make your first QSO off the moon with CW or SSB, then go read something else. If you're still here, and haven't downloaded WSJT-X yet, go do that now. Click this: https://physics.princeton.edu/pulsar/k1jt/wsjtx.html find the distro for you computer, download and install. If you don't know how to configure your radio and computer to make them play nice together, google your setup and you'll find hundreds of step thru guides on how. Once you've got it all working, you're going to want to play with mode JT65, Submode B. You're going to want to send 'short' messages too, so click that. JT65 works a lot like FT8, only slower. The TX/RX cycles are 1 minute long. The station transmitting will TX for the first 50 seconds of a sequence, and have a 10 second decoding period at the end of each minute. Depending on your Distro, JT65 will likely not decode everything in your waterfall automatically. You'll need to find a trace, click on it after the 50 seconds is up, and then it will decode. You'll also need to punch in their call manually, generate messages, and move the message sequence down the chain. No Auto-sequencing here. 

The messages themselves are a bit different from FT8 too, you still exchange the grid, but instead of signal reports you'll send essentially just confirmation of reception messages that look like this:


Since practically every EME qso at QRP level will be somewhat scheduled, you should know who you're calling, and they should know who you are. Both stations will start with the 1st sequence, until someone decodes the other, the first to decode will go to the 2nd sequence, the "OOO" message (I think this means OVER OVER OVER). Once the OOO is received, that station sends back RO (Roger Over?) and then finally the original OOO station will send RRR (Roger Roger Roger). 73s are optional as the QSO is complete, but if you've got a path go ahead and send it to be cool. OK, so now that the QSO methodology is out of the way, how to do you find when, who, and where is a good time to make an EME QSO, great question!

Situational Awareness Forever

The chances of making a random EME QSO at QRP levels are basically zero. You need every little bit of help you can get to pull it off. First question - When. Obviously the moon needs to be visible to you, and the person you're trying to work. It doesn't necessarily need to be visible to the eye (a New Moon is still good to bounce RF off of) but both stations need to be able to 'see' it with their RF.. even the big guns can't send signal thru the dirt. Next, the moon, like basically any orbital body we're aware of in the universe, wobbles just a bit when it's up there going in circles. At it's closest point in it's orbit around Earth the moon is about 359,000km away from me, and it's furthest it's about 406,000km away. Unlike Satellites where we wish we had something higher, 46,000km is a ridiculously far distance and can introduce a whole lot of extra path loss to our RF. As every great Nobel laureate will tell you, this is bad. You want to find a day when the moon is as close to smashing into you as it's going to get. The typical Apogee/Perigee cycle takes about 28 days. You've normally got a couple of days each month at or close enough to Perigee where the path loss is the lowest. So pick those days to play around. Also to take into consideration when answering the 'when' question are things like "Sky Noise" and "Declination". Sky Noise is just like it sounds, RF Noise coming from the Sky. The sun, for example, puts out a crazy amount of RF. If the moon is really close to the sun during your window, it's highly likely the Sun will drown out reflections from the moon. Also, the Milky Way Galaxy disc also puts out a noticeable amount of RF, if the moon is traversing along the disc, it too will drown out the echo signals. Pick days neither of these things are happening. Declination is where the moon is in relation to the equator. Obviously if the moon is at perigee, but you're at a highly northern latitude and the moon is over the southern hemisphere, You're not going to see it very long... or if you're at a *really* high latitude, you might not see it at all. 


What this all means is there's a ton of factors that go into finding a good time to work EME, that are beyond your control. If you live in the US or Europe, probably the best way to find a good time is to look at an EME Calendar by one of the many operators smarter than me. I like the calendar built by Bernd, DL7APV, Here:  http://dl7apv.darc.de/moon2010/moon2010.htm But any of them will do. He takes what I just talked about, and more stuff to build a calendar of good weekends to work EME. Weekends are important because folks that are good at EME typically have spent a lot of money on their station. Those same people that have lots of money to spend on their station often have to work and sleep during the week. Therefore, EME activity is typically up on the weekends.

After figuring in mutual windows, perigee timing, declination, noise and the effects of capitalism you're now down to maybe a dozen good weekends a year to try this. Now consider things like weather, family obligations, and your own need for sleep, and you might have 15-20 hours a year when everything lines up as a QRP EME operator. Scared yet? Good, you should be... you should also be excited.. most folks will never make it past the 'when' question. 

Non-RF Communications

Since you're going to need to know who to look for, and where people are, here's a few links to get you going. First - https://www.livecq.eu/default.asp - This is basically an online 'spotter' for stations being seen off the moon calling CQ. It's a good place to watch to see who's active, and where and on what frequency they're talking on. There are also numerous chat pages where EME ops congregate and setup skeds: https://logger.hb9q.ch/ is one very popular site, and primarily used by 432 and 1296 folks. https://www.chris.org/cgi-bin/jt65emeA Is another mainly used by the 144 bouncers. My personal favorite is the VHF-Chat Workspace on Slack which you can join here: http://chat.n5tm.com/. You might have to look around a bit, but between those 3 usually you can find a big gun online looking for skeds. The kind of station you're going to want to find is probably going to have 8, 16, 32 or more Yagis, and probably running legal limit (or more!) on that particular band. Find the biggest station you can to try and work. They're going to have to do most of the heavy lifting here. I won't name particular stations to try, but if you look and ask around in the chat pages it won't take long to find some real serious players on EME. The good news is so far every op I've run into that runs a monster station, is super cool, and happy to show you what they got. When you've sunk as much cash into something as these guys have, it's for a reason.. and that reason is to work you!

It's also worth mentioning that each website / chat system has their own rules and procedures for doing stuff, as well as accepted etiquette. Spend a few minutes watching and listening before busting in and going 'OH HAI EVERYONE I R QRPer AND WANT TO MAKE QSO!'. There's a certain language and flow for skedding an EME contact, so make sure you get it before piping up.  One rule that is absolutely ubiquitous though is you're not allowed to exchange any QSO or Report information until after the contact has been completed. Things like one way propagation are very common in EME, and if you're exchanging reports via the internet, you're cheating. Do. Not. Do. This. It is a cardinal sin that will not be forgiven more than once. Also pay attention to the weekends of the EME contests. Contests have some of their own rules about 'confirming' contacts once completed and 'self spotting' your frequency and sequence. EME contests are much more lax than HF contests due to being about 60db more difficult, but pay attention and make sure you're not saying anything that will get frowned upon.


So, you've got JT65 working, You've found a few hours on a good condition day that you don't need to do anything else more important, and you've even talked with a big gun who's willing to try and work you.  What now.. Oh yea, equipment!

Broken Record VHF/UHF Operation

If you've ever heard me talk about satellite operation, get ready for a lot of the same. The rules aren't that different.

1. Prepare ahead of your window: This should be a no brainer, but it's worth mentioning anyway. Get to where you're going to operate from way ahead of time and test your setup. Make sure everything is working the way it's supposed too. Make sure there's no computer glitches, your SWR is good on your antenna, and your batteries or generators are full up. You want as few things to manage as possible when it's crunch time. 

2. Location, Location, Location: In QRP EME, as in Satellite DXing, Location is absolutely key. This is also the single biggest advantage you have over a multi-yagi high power EME station as a QRP operator. Find a spot with no external RFI, far away from power lines and fluorescent lights. Find a spot up on a hill with a clear view of the horizon and no trees to get in the way. Find a spot where you can setup a tripod or small mast, and run the shortest run of coax you can to your rig. Find a spot where you can sit comfortable for an hour or two and not be disturbed by others. 

3. Low Loss Coax: Every single fraction of a db matters in EME. You need the shortest run, of the lowest loss coax you can afford. LMR240 is probably an absolute minimum, with LMR400 being better. And while there's obviously a balance between how close you're going to sit to your antenna and power, if you're running more than a couple hundred of watts at VHF/UHF you're probably not reading this anyway. 

4. The Biggest Beam You got: For a lot of us Satellite operators, this is going to be an Arrow. If you've got the Alaskan Arrow, break that sucker out instead. If you've got something bigger, great, use that. You can normally find 10'-12' 432 and 144 Yagis for around 100 bucks used at hamfests or online. You can always build something too, I put together my 17 element 432 yagi for about $50 in McMaster parts. You can beg the local VHF/UHF guy that has every kind of antenna in the world in his barn, but spends his days on 146.940 for an a proper beam too. Make sure to show him your awesome QSL card with Germany on 70cm a few weeks later and he'll probably let you keep the beam he'll be so dumbstruck. No matter what path you take, More gain = more success on EME.

5. Mount up:  This is optional, but I do recommend putting your beam on a tripod, or some sort of a mast. Wait, is CCI actually suggesting the use of a Tripod? For Sats, I say no.. for EME, I say yes. A Tripod allows a stable platform for your beam pointing at something that's moving really slow. Remember, this ain't a tumbling LEO cruising across the sky in 14 minutes, the moon takes it's sweet ass time going east to west, so you you can too. A tripod also allows you to use a beam bigger than you can hold in your hand, and it's a bit safer if you're running more than 50w on VHF/UHF, which is personally where I draw the line. Additionally, while it's nice, you don't *have* to have elevation control on your mount. Since Luna is slow walking her roll, you've got an hour or so when she just hangs out on the horizon low enough that she'll still be in the beamwidth of most normal yagis without elevation adjustments. There's also this thing called 'ground gain' that sort of happens, sometimes, on some bands, over certain ground.. maybe... so it's worth pointing straight at moonrise or moonset, and just leaving your beam fixed in that spot for a bit. Finally, you don't have to, but if you can adjust polarity on your beam, it might also help from time to time. Since these signals are so weak and take a few seconds to detect, I would be very slow and deliberate with your polarity changes. Go 45 degrees at a time, and then watch for 10-20 seconds to see if a signal shows up... you'll know it if it's working or you're making it worse. In summary: Your mount is probably the easiest place to get creative. Just whatever you do, keep short elements closer than the long elements to the big cheese in the sky. 

6. All the Wattz: Unlike satellites, there's no electronics you have to worry about frying on the lunar surface. You're literally bouncing RF off of rocks. Therefore, bounce all the RF you can. At an absolute minimum the power required on your end seems to be around 30w. I've made a few 30w QSOs with my 17 element 432 yagi, and others have made similar. I've never heard of any made < 30w. Obviously 50w is better, 75w is even better, and 100w or more are better yet. The 910 and new 9700 will both do 75w on 432 and 100w on 144 and should be enough with a decent beam to make at least a token qso or two. Give your rig a good voltage feed, keep them cool, and they'll run JT65 at full smoke with no issues. If you've got a small brick style amp, something in the 130-170w range, bust it out and hook it up. I know this is a different mindset than what we're used to as Satellite operators, but there is no such thing as too many ERPs when you're talking about EME. More is better, in every single case. 


Good Luck, we're all counting on you

If you've made it this far, congrats. You're obviously nuts, and I commend you. The final piece worth mentioning is that this all still requires luck. There are a ton of natural factors that can still degrade the signal that I haven't even discussed, or really even understand. Sometimes, despite all your preparation, a contact just isn't going to happen. You need to be ready for failure when attempting this, and be ok with trial and error that might take you the better part of a few years to figure out. You can always buy bigger better stuff and increase your odds, but that's some serious investment and dedication to pull it off. However, the rewarding feeling of bouncing a signal off the freaking moon is unlike anything else I've done in Ham Radio. If you're an AMSAT operator thinking of trying your hand at QRP EME - give it a shot. You've probably already got what it takes equipment wise, now just apply some knowledge and luck, and see what can be done.