Wednesday, December 30, 2009

One 8 step sequencer running on the bench

I spent a bit too much time googling for synth music and sequencers over the last few days. I had a lot of fun just reading about folks projects and how they built up various kits or designed from scratch different arrangements. I have not really spent much time working with 4000 series or 7400 series ICs in the last ten years. Well not since first year Uni to be honest. That was a long time ago and my memory for truth tables is long forgotten.

So, I almost cheated but finding the friendliest schematic I could find and trying to build that up. You know how it goes, adapting part of this circuit as a stage one of your project. Well I had a few problems. My 4017 wasn't behaving at all! I spent several hours on the bench trying to figure out why I couldn't get the thing to behave. The first problem was that I missed a ground connection on pin 12, eventually running a bypass cap to the reset pin before it settled down and started to do the right thing. But the first LED wasn't behaving itself at all ...

Some further testing revealed that the LED in the first position was a flashing LED! So its now in its own draw, in the component cabinet. The 555 was a whole lot easier to work with and was pretty much doing what it was supposed to from assembly. The potentiometer has a dead spot right at the fast end of oscillator, so may need to swap the wires on the pads, or get a new 1megaohm 'pot'.

I will say that after having the 'friendly' schematic, I did change most the resistor values, add a few extra capacitors and lay it up differently on the plug-in breadboard. Its not a new circuit, but I learned a lot, mainly how to trouble shoot this little thing.

So what have I achieved here? More motivation to tinker in the shed/shack on the plug-in breadboard and play around with the circuit. Getting things working build up the motivation to do more. I have a tall ladder before me. I'm aiming to make the rungs nice a close together. Easier to climb that way ;)

Anyhow, happy homebrewing to all!



Monday, December 28, 2009

one dual rail 12V PSU on the bench and working

Spent a couple of hours researching and then building, testing/destroying, then researching and re-engineering a dual rail power supply for +12 volts & -12 volts.

Not sure what kicked it off, but I am kind of over my 10 meter QRSS receiver not working. I was looking at a ham radio blog, found a video of NE555 and 4017 based synth sequencer. Down the rabbit hole I went reading dozens of web pages, until finally finding the coolest sequencer I've ever seen!

On the Natural Rythm Music site I found a 16 step rotary analog sequencer. It is a visually amazing device, certainly feels more intuitive than you regular 4 blocks of 4 linear sequencer. The builder has a great presentation style and can put together some cool tracks.

This discovery then in turn inspired more and more reading until I found some basic building blocks. Everything seemed to need a dual rail power supply, so I though I should start there. Some web searching, followed by checking a couple of different editions of the ARRL handbook led to a simple dual rail regulated supply. I carefully noted down the schematic, made a parts list and marched off to the shed/shack to build.

Stepping back a ways, I purchased for two dollars, a two amp twelve volt power supply at the radio club meeting. I was a bit disappointed that it didn't work, but put it aside for a rainy days hacking I suppose its the risk you take when you buy gear from a deceased estate, no real history on the item. So yesterday was that day, except it was 32c, no rain and no wind. Makes the shack a little warm.

I amassed all the parts, dissected the old power supply, noted the two 10,000uf 24 volt electrolytic capacitors and carefully discharged then with a long screw driver... the spark was impressive for a power supply that had been unplugged and idle for six months. Note to self, be careful around these caps!

Spent a while working out the arrangement on tag strip, then manhattan style, but didn't find a nice way to mount everthing in the space available. Got a coffee then came back with the lost thoughts of some veroboard of yore...

Found the vero board and worked out a layout on the board from the schematic. Assembled it, checked it twice. Carefully hooked everything up and turned on the power. Everything looked Ok until the LM7812 regulator launched a small part of itself into orbit!

Ok, this wasn't part of the plan! Turned everything off, disconnected everything and started the autopsy. Everything was correct according the schematic. So I checked the schematic against the original. Sure enough the pin out of the LM7812 was wrong and had exchanged the ground and incoming unregulated power line, pins 1 and 2. No wonder it went FIZZ POP!

I downloaded the data sheets, just to be sure. Who would have thought that the LM7812 and the LM7912 would have different pin outs? I mean really?!?!?

Anyhow, I spent about an hour looking for a replacement regulator. Almost despondent that I'd never found one, I happened to find a little plastic draw that had been missing out of the component cabinet. Interestingly I must have put it aside when I first looked at the power supply and thought about reworking it as a regulated unit for 12 volts for my 2 meter radio.

Anyhow, it all went together more easily the second time and it passed the smoke test the second time, keeping the smoke on the inside! Along with providing 11.99 volts and -12.01 volts. The simple test load this time was a pair of 12 volt 20 milli amp four inch fans from a peecee power supply.

So, now to find all the parts for an audio VCO for tone generation, an LFO for effects, mixer for multiple channels, various high and low pass filters and an amplifier... The aim is that all of these will be building blocks for test equipment for RF down the track. Cunning plan huh?

Monday, December 07, 2009

MTB on the road

I spent some time over the weekend working on the mountain bike I picked up. The complete bike is less than half the weight of my old MTB! Its all aluminium, with the inch wide double walled with eyelets. Currently setup as a single speed, about 70 gear inches. Seems a good balance for pootling around the small hills here. Will fit road tyres at some point in time. Looking at fitting my rack this coming weekend. Then lay up all the gear for a short tour.

VK band plan

Was doing a bit of reading to confirm the frequency range for 15 meters to think through the IF article in this months AR magazine. So 15 meters covers; 21.000 - 21.450 MHz. I have a hand full of 4MHz and 25MHz crystals.

While I was reading through the band plan, I found something new, and entry for 2200 meters. But what grabbed my attention even more was that there was a mention of QRSS;

137.6 - 137.8 kHz Slow CW modes, e.g. QRSS

Sunday, November 22, 2009

bicycle hacking imminent

grabbed a few bikes from a neighbour that was having a tidy. two BMXs, mainly for the wheels, for the long wheel base(LWB) recumbent that I've been wanting to make for a while. I made a LWB bent over at Lyns place but it has issues and needs a fair bit of work, mainly brakes and gearing. The steering is now sorted.

Also grabbed a alloy MTB that fits well enough and has a disk up front. Put a straight 26" on the back and it feels ok. Need to tweak the chain line, or drop the 1/8" chain for some thing a little more modern.

Been putting more thought into some touring. Still don't have a destination in mind. I think I want to explore the top end of Tassie and some of the northern NSW coast. Those are both big trips, so I think I better start with something smaller, closer to home and flatter =)

So anyone with suggestions of interesting places to cycle tour around Adelaide, Hills and surrounds please let me know.

Adelaide launches a Perl Mongers

Well my last post about the map and what its missing, is now not missing the entry for Adelaide. Although there still isn't a pin on the map.

Once we have had a few meetings under our belt, Adelaide Perl Mongers will get an official mail list, currently hosted locally. Perhaps even a pin on the Map!

All the kudos to Justin Hawkins for setting up the web site and the mail list =)

So please join the mail list over at Adelaide Perl Mongers, so we can get a meeting rolling for some time early December.



Thursday, November 19, 2009

Interesting rant on where programming languages should go

Has some interesting ideas. I suppose it all depends on what kind of programming you do.

Perhaps an alternative would be a tiny thin language that hides the data type to library massaging.
Having to "import os" in python, is an example of why these "modern" languages are stuck in the 70's or 80's etc.

None the less, make me think about the pre-compile vs soft-dynamic languages.

Tuesday, November 17, 2009

There is something missing from this map

There is something missing from this map.

Can you tell what it is?

No, well we're working on it =)

20091117, quote of the day

"Is there and app for that? Yeah, on both. Wait, no, it looks like it was rejected from the iPhone store. Droid it is, then."

Foundation License limitations and restrictions

I have received a number of emails from concerned VK HAMs about the contents of my posts and my 'use' of homebrew transmitters.

It is well known in VK land that Foundation calls are restricted to using only commercial built transmitters with a maximum of 10 watts PEP. However, no other useful or relevant details of the Foundation license seems to be well known.

So let me make two things clear, firstly, I have explored the LCD on all the issues of transmitting and beacons as this is a topic that interests me greatly. Secondly I have not built any homebrew transmission equipment. Only two homebrew receivers and used my Kenwood HF rig and oscilloscope to check that they are not putting out noise on the bands, fundamental or harmonics.

The LCD is particularly vague about the issues around the use of beacons by Foundation license holder. Except that it precludes the use of "computer controller beacons". This does leave plenty of scope for experimentation in analog generation methods. To be honest, in this day and age I believe those most likely to advance this aspect of the hobby are those that grew up in a digital world and have been programming since they were knee high to a grasshopper. However the folks writing the LCD were no doubt born early last century and these new fangled computers will take some time to catch on... I'm not being rude, just a bit disappointed.

That said, I have the basis for a microprocessor-keyer sitting on the shelf, waiting. Waiting for when I upgrade my license.

Which brings me to upgrading from a Foundation license to a Standard or Advanced license. There is a hand book for studying for a Foundation from the WIA, but not the Standard or Advanced. Why is that?

Perhaps this is the reason that there are so many F-Calls out there now? And so many HAMs complaining about there being so many F-Calls. Apparently one more letter is too hard to deal with.

I've tried ordering materials, as advertised on the internet, but unresponsive is the nice way to summarise. Its too hard to get the materials to study to upgrade, unless you are lucky to be part of a club with materials. If you're not part of one of those few clubs, I've noticed they don't share. Why is that? There doesn't seem to be much collaboration in that space. Or I'm just not in a position to see it, because I'm only an F-Call. Its a bit of a catch-22 if you ask me. I;m sure they work hard to produce the material in the first place, but you will get more recognition and more amateurs if that material is shared around the clubs. Free or at cost or more, it doesn't matter. Get those materials out there!

I've been flamed on mail lists for asking about topics that only Advanced calls are allowed to use. So my response was how am I supposed to get my Advanced license if I don't know about these topics, their response is, you don't until you are a Advanced call. There is definitely a weird head space around learning for your upgrade. You are supposed to learn it all yourself and asking for help is cheating.

That said I want to upgrade to Advanced and to tinker with QRSS and build my own kit.
Its just a shame that there is so much attitude around being a F-Call with aspirations.
There are some local HAMs that are very helpful. They have been inspiring, but their time is quite limited. Which is why I try to make the most of my time at club meetings. There are other clubs north and south that would be beneficial, but getting to and from the meetings is a serious hassle. Perhaps I just need to get a car or motorbike of my own as public transport isn't an option, thanks to Adelaide's central hub transport model.

I know that AR wants articles on what F-Call are up to and think, but I doubt they want to hear about the plight of what F-Calls can't do. "Become a WIA member!" and "Upgrade! Upgrade!" I hear folks say...

Anyhow, I've dropped a lot of other interests and other clubs' activities so I can focus on work and amateur radio. Hopeful over the Christmas break I can finish my upgrade study and sit the exams...

Monday, November 16, 2009

building an oscillator

I'm built an oscillator, for 28.188MHz based on the Genesis Q5 from VK1AA. Well, I had smoke. I'm still working on the plug-in-bread-board. I think the lesson is not to use 12 volt gell cells for testing out new ideas. Can supply far too much current too quickly. Now I understand the little 9 volt battery role in life...

The little 74hc04 squealed then popped! Took only a second. I'd put the 7805 in the reverse position. Its a 50-50 chance ... Lesson there is to not turn the pluging board around all the time and loose track of the orientation of things, I think.

So with the regulator in the right way now, it gets pretty warm, so I've added a heat sink, but not tested it again. Will have another go when the 2 meter AHARS net is on this evening, in the shed.

At the end of this, I might order a G5 or two. I expect that I'll need a few for various prototypes anyway. Just want to get a couple of modular oscillators up and running for receivers =)



Sunday, November 15, 2009

new 10m dipole

Replaced the dipole this morning. Cut to length some pvc coated speaker cable that had once been an end fed long wire. Its light and grey so not so easy to see as the red 30amp power cable I was using before. Also ran coax straight up to the center of the dipole and will run a similar arangement to the QRSS Rx of Hans Summers, G0UPL. Now onto modifying the receiver =).



Saturday, November 14, 2009

Saturday works

It was pretty hot today, was forecast to be 39c. Didn't feel like it, but did slow me down running about the place getting the antenna up. None the less I have a dipole for 28.188MHz up. Its currently mounted as the third guy rope of the mast that I ran the old WiFi gear on. The router board that was running the WiFi stopped for no particular reason, so I'm suspecting the power supply section of the board, needs further investigation. I removed all the WiFi gear from the mast and reset two of the stainless steel guy ropes. Made the 10 meter antenna the third guy...

It works, probably not nearly as well as I'd like, perhaps I should build a Z-Match tuner for it instead. So, I setup the little Kenwood and tuned all around 10m. There was some fast CW but little else was heard on the band.

So, the 'ladder line' is a nasty hack. I grabbed a couple of plastic tent pegs and some cable ties. Its not so good, needs frequent adjusting. There was plenty of activity on 40 meters. This little dipole is about six meters off the ground, double the height of the previous dipole for 40m.

Actually it is the 40m dipole with the rest as the ladder line. And lots of string and a couple of eletric fence insulators. The old balun is currently in, but I'm not sure that the 4:1 is appropriate. Is a 1:1 better suited to a ladder line? Sorted out some coax and connectors too.

Dug through some boxes that I picked up at the last AHARS meet, grabbed some bits for to setup the audio into the peecee. Also found a 0C70 and a 0C71 ... I think they are germanium audio transistors.



Thursday, November 12, 2009

QRSS receiver crystal for 10 meters

The 28.188MHz crystal arrived today from Nick, VK1AA. Have been hanging out for it to arrive and would really like to get a receiver up and running.

This weekend looks like being a scorcher, about 38c, so a bit of indoors building will be in order. Which also leads one to think about oscillator stability. With night time temperatures of 15c to day time temperatures of 35+c degrees there is plent of room for drift.

Thoughts on work to do;

  • An antenna for 10 meters; vertical or dipole?

  • If heading down the di-pole route, balanced or OCFD?

  • Maybe a balun?

  • A transmission line should be straight forward, I have some spare clean RG-213.

  • How about a tuner, then how do I measure if its really working?

  • If a vertical, what about a copper cactus or a just a 5/8 vertical?

  • Should I make the local oscillator in a separate box?

  • If so I could drive the Kenwood to TX test ... wonder what specs are required to drive its VFO port?

  • Then modify the 'sudden tunner' style DC receiver. Based on an NE602 and LM386.

  • Hook up to peecee with 96KHz sound card.

  • Grab with glfer or baudline.

  • Announce its ready...

  • Later develop something more focused around web based delivery.

  • Find a design for a non-computer 'controlled beacon'. A condition of the Foundation Licence.

  • Perhaps a beacon controller as simple as a pair of NE555's and a 4017. The first NE555 at a 30 second on/30 second off interval, driving the clock line of the 4017. The first output pin of the 4017 could then drive the second NE555 with a 3 second on/3second off interval. Effectively sending a '5' every ten minutes in QRSS 3 second dots. The '5' being from the call sign of VK5. My call sign being VK5FNET. But thats a job for another time, but would be nice to know what folks think...

So, thoughts on work to be this weekend;

  • Get any antenna up

  • Get a length of RG-213 terminated with PL259 connectors

  • Blow the dust of my 4:1 balun I made some time back and check the connectors out

  • Modify the 80m DC RX to run on 10m

  • Patch up some audio cable for the DCRx to peecee sound card

  • Work out how to capture images or do screen dumps from glfer

  • Announce it...

Doesn't sound like much on 'paper', but on top of the domestic chores should be a busy weekend.



Tuesday, November 10, 2009

10 meters

On the Knights QRSS mail list theres been talk of activating more HF bands. Specifically 10 and 15 meters. There issues with activating 40 meters as there are access issues for non advance licences in the US and band plan issues across Europe. Not 100% sure how 40 meters is carved up elsewhere, but in Australia, Foundation, Standard and Advance all have access from 7.000-7.300MHz.

There have been a couple of contests on 10 meters with some success. This has lead to conversations on list to start putting SDR receivers and QRSS transmitters on 10 meters. The 100Hz QRSS band is 28.188,000 - 28.188,100 MHz. Folks have used other portions of the 10 meter band, around 28.322MHz if I remember rightly. Which seems to be around the QRP and QRO un-manned beacons.

Nick, VK1AA who is shipping the G80, G40, G2030, and G5 Genesis radios, has put up an offer to send out a batch of crystals on 28.188 MHz to folks free. So long as they get a MEPT or receiver up on the air for QRSS. I shot off an email straight away and am still patiently waiting for it in the mail.

I have started mods to the W1AW NE602/LM386 direct conversion receiver I build up for 80 meters. Needs rework on the band pass filter and local oscillator. Should be fun!



Wednesday, October 07, 2009

Wordpress-MU deployment at work

I'm rather happy that this project is finally coming to an end. I've moved jobs within the University of Adelaide, but I'm still working on our blog servers. It's funny how projects follow you around. Before the underlying infrastructure, now the actual blogs.

So it has an actual blog on it and we've put up a bunch of video from the Open Day we had a month back. The AV guys did a great job. So in time we'll move across other news sites from the main web site to the blog. While I've been hacking away in perl for the last few months and really enjoying that, Wordpress-MU or WPMU is in PHP. I haven't really done any serious work in PHP since I've been back in Australia, but its not hard to get updates in there =)

Kudos goes to Phil for content and plugins, and Scott for themes. Also the Open Day crew for great work organising all the speakers and video.

Tuesday, September 22, 2009

use Getopt::Long;

This mornings very short moment of enlightenment was related to the use of Getopt::Long.

I had been messing with passing command line arguments to my server deployment script and had completely failed to get the --verbose flag to work the way I expected.

So I read the perldoc on it and found that I had been going about this the wrong way altogether.

I had this;

my $VERBOSE = 0;
GetOptions( "config=s" => \$config_file,
"verbose=i" => \$VERBOSE,
"modules" => \$INSTALL_MODULES );

print "DEBUG: message" if $VERBOSE;

Then on the command line I had to specify --verbose 1 or --verbose=1, which really wasn't what I had in mind. So this morning the mist on the man page cleared and it all clicked.

If you use "verbose!" => \$VERBOSE, then you can use --noverbose to unset the flag if you have previously set it in your code above. I'm not sure why I moved to using "verbose=i" => \$VERBOSE, but "verbose" => \$VERBOSE was not working as expected.

Anyhow, I was talking to Andy about putting "!" character after the command line flag name and now --verbose sets the variable $VERBOSE and it behaves as I previously expected.

Tuesday, September 15, 2009

HF Lunchbox 1 lives again

This morning I dropped in to see Barry VK5ZBQ. We had talked about catching up to sort out one of the two Codan 6924 HF radios that I purchased at an AHARS members meeting auction.

So this morning was "the day". I had a minor ordeal to get there, including losing my bus ticket, missing the bus, then catching a lift after the car would not start. I arrived about 9am and was undoing the case screws to get down to business shortly afterwards.

Barry has a manual for the Codan 6924 Mark 1, so we looked up various sections that I was curious about. Barry set the rig up on the test bench, powered it up, connected it to the signal generator, selected the frequency for channel 1 and we heard the dual tone from the speaker on the front panel. Win number 1. Then we set about transmitting from the rig. The power meter on the signal generator shows out going power when hitting the antenna test position on the mode selector. So out with the multimeter, leading to the detection of a broken microphone lead.

The chassis can be separated by removing some screws and nuts, so you need tiny fingers or a little tool to hold the nut so you don't loose it. So Barry opted to replace the microphone lead with something a little more robust than the original lintz wire type. Apparently you can't solder it very well. So the new lead went in. Took a little while as its quite a fiddley job. We reassembled the chassis and microphone and connected up the output to the oscilliscope. Fired up the HF rig, tuned to channel 1 and tested it. Win 2. We have a nice upper side band trace on the scope and after a bit of fiddling with the HF radio antenna and dummy loads, we hear the transmitted voice from the HF rig.

At this point I'm very happy, as the radio works properly and we work out that it puts out 8-11 watts PEP. The antenna tuner varies the output power a lot. Need to sit down with it and work out how exactly the tuner works, it has a loverly wound toroid with wiper for continuously variable inductor. Now I'm ready to attack the idea of a VCO in stead of a fixed crystal oscillator.

We break for a cup of tea and Barry explains to me how the upper side band (USB) and lower side band (LSB) filters work in the 6924 and what I'll need if I change the oscillator to a 80 or 40 meter band. Apparently there were kits for this, but I've found no documentation. So I will draw everything up and go back to Barry and run it by him. I think I'll work out what crystals I need and populate the remaining empty crystal oscillator board first to get the hang of it first =)

Now that the weather is warming up, I can spend more time in the radio shed/shack. I've put together a dipole for 40 meters along with a 4:1 balun and suitable cable. Should be fun.


Sunday, September 06, 2009

Thoughts on servers and network services

This is a post that I sent to the Melbourne Wireless mail list. Thought it might be useful for others so I am reproducing it here;

Hi Victor,

Victor wrote:
> > Thanks everyone for the replies so far. At this stage will be sticking with a Intel Atom 330.

I run a dual core atom here, low on power, not particularly fast,but thats the balance you are after if you are going atom I suppose. Always spend as much as you can on RAM! 2GB+

> > When it comes to an OS for the server, I know a little bit (read NOT a
> > lot) about linux, enough to get me through things... Should I look at
> > setting up CentOS on this or is that possibly a bit too "much" for a
> > relative "youngin" into the world of Linux per se.

The only way to learn is to do. Don't let various folks opinions put you off =) There is plenty to learn and lots of different paths to take ;) If you're new to Linux/UNIX then maybe you can head along to your local LUG, Linux User Group. I'm assuming that you are in/around Melbourne so have alook at;

> > I read somewhere that CentOS isn't for the light-hearted?

Centos is geared to be a server OS. It has a few quirks and its configuration like RHEL, is a little different to other Linux distros. I have used RedHat RHEL3,4&5 a lot and Centos mirrors them in functionality, etc My only complaints revolve around; packages in the Perl and web areas are well behind the times and they still use RPM with broken meta package dependencies! If you stick to the default repositories you'll be fine, but the packages are limited. Have a look at EPEL if you need more packages/software...

Generally speaking RHEL is solid enough and stable enough to do most server tasks well. So if Centos lives up to its name of being RHEL with out the irritating license bits then its all good experience on the Resume too...

> > Other possible options is FreeBSD, Debian 4 or Ubuntu 8.x

I have run all the above at work and at home for various uses.

If you really don't have much UNIX/Linux experience then perhaps Ubuntu will give you the least troubles installing and getting around.

Ubuntu is primarily focused at the desktop. The server version has a number of buggy things that are tied to the design decisions around init/startup script and network interface management. Fine desktop. I would recommend it to anyone starting out in Linux. Then you can grow from there...

I run a number of Debian server at home. Solid, stable, but some packages are old. This may not matter in the beginning. Networking is rock solid, more packages that you can poke a stick at. Seriously lots.

FreeBSD I have not been keeping up to date with over the last four years... There was a lot of development in the desktop arena that wasn't interesting, as I primarily used it for servers. Its solid, networking was solid and fast. It was death by compile your own packages. Hopefully this has changed. They have been moving fast, hopefully the ports tree is uptodate. Was the main reason I dropped FreeBSD. Still have NetBSD boxes here for old school non-intel hardware, Alpha, Sun, VAX, etc.

Also, not on your list, but worth a mention is OpenSolaris. I installed it on my laptop. Was pretty happy with it, except no support for my WiFi.

Another thought, to get you going; virtualise! I know it was mentioned else where in this mail thread, though I'd add to it.

I have recommended to folks in our local LUG, put Ubuntu on your machine, install VirtualBox and install the other OSes in virtual machines. More doco;

Yes VirtualBox is targeted at Desktop installs with the guest OS in a window, but it gives you a feel for the guest OS with out much investment into building another real machine. You can shut it down and boot it up, tinker with it later. Some videos;

If you get real serious about virtualisation, perhaps the simplest remote managed VM server is the Oracle Xen based stack. I saw a demo of it at LCA2009 in Hobart. I was quite impressed. Managed through a web browser. Not through some heavily licensed Windows sever like VMWare =P You can play with all the iSCSI stuff and file system mirroring etc, lotsa fun =) Doco on Oravle VM;

VMs are only good if you *really* need separation of different server apps from each other, IMHO. There is a lot of overhead in running VMs, on machines that don't have the virtualisation instructions. I don't run VMs on servers at home any more. The only reason I would is if I have to tinker with different OSes for debugging. Another reason would be snapshot file system backups. I do have a few VMs on my laptop to demo, test code, or compile code...

I'd recommend figuring out what "network services" you want to run and run them on their own interface on the server. Have a look at the 'ip' command. You can run more than on IP address on each physical ethernet interface (NIC). Then you can make the application/service listen on that IP and then tighten up your firewall rules with 'iptables'.

Doco on iptables etc;
More info here;

Then once you've played with all that, theres the amazing arena of routing =)

Old school protocols;
- BGP;

New protocols;
- B.A.T.M.A.N.;

- quagga OSPF&BGP routing software;
- olsrd routing software;
- batman routing software;

Thought this talk might be interesting; David Rowe presented at Linux.Conf.AU in Hobart this year on the MeshPotato about the village telco model, using B.A.T.M.A.N. as the routing protocol;
talk notes;
talk slides;
talk video;

Anyhow, have fun with it =)


Melbwireless mailing list

Monday, June 22, 2009

Repairs and receiver experimentation

Its been a while since there was some progress on the workbench. This last weekend kicked off by reassembling the frequency counter after re-soldering every joint on the board. It took about two hours to do, but it has all paid off, its now working comparing against a known working source.

I started tweaking the direct conversion receiver, with less success, but got to test out a number of other crystals along the way. 2MHz oscillates stably at 1.999999MHz, while the 25MHz crystal does not. I changed a number of different inductors and NPO capacitors to similar values to other circuits I've seen, but to no avail for 25MHz... So much for the idea of 25MHz +/- 3-4MHz VFO. There are plenty more options to come though. Found what looks like a pair of FT37-43's. Will have to check the datasheet but they look about the right size and came from a mystery box from a junk sale a while back. Will tinker with it some more.

The DC receiver picks up a lot of terrible noise around 1.8MHz which I assume is AM broadcast with the 2MHz crystal. Need to tinker more with that, and see if I can find a better way to link up the front end, not using a band pass filter to give me more tuning range. Not really sure how this works yet, but theres a lot of draw design found from the net and build. Probably hitting impedance mismatch issues. I'm sure I'll figure it out eventually.

I think its time to migrate the DC receiver from the breadboar to a manhattan board. I'm finding that the more I tinker with it the less it works. Probably connections of the components being dragged around by CRO probes, etc... Well its too colder to hang around in the radio shack any longer.

Monday, June 08, 2009

mashup ideas: solar powered radio

From the maker blog a while back they ran a series of posts on joule theif.

Bill on the SolderSmoke blog ran a couple of articles on mechanically keyed transmitters, like the ET phone home rig, also a rew ideas around clock powered keying.

So my brain was buzzing away recently and it wasn't until I spotted the solar powered garden lights, that ceased to work a couple of months ago, that I thought about re-using them. The little solar cells only put out 0.8-1.1 volts. Three in series put out around 3.2 volts in full sun light. I need to go back and see what kind of current they can sink though...

To power something like an arduino keyed tx or a DC receiver with NE612, I would need a solid and regulated 5 volts.

It wasn't until this morning, on the irc channel of our local homebrewers, that I thought about mashing up the joule theif and the 80 meter DC receiver to make a solar powered radio =) It would make it limited to day light hours, but its certainly workable. Still have some other things to work on first, but certainly the next project on the list =)

Sunday, May 24, 2009

2 Meter Portable Ops

Yesterday Karl VK5FOSS and I VK5FNET, worked through the check list to do a test set up for the 2 meter portable rig. The aim was to get all the kit together for operating in the field. Why would we do that you ask? Well last year I wasn't really quite ready for either of the rallies that we worked at. The ROSA and the Classic Rally of Adelaide. Both time Karl and I worked as radio operators. The main function of the role is to record information on the stage and pass this information as 'messages' to Rally Control.

HAM radio operators work under the banner of WICEN. WICEN is an organisation for emergency radio communications to help in message passing for the emergency organisations like the Country Fire Services, Police, Ambulances, etc. The aim here is for radio operators to work well under pressure to get messages through under tricky conditions. I've been told that despite the technical problems I had at the rallies, I did a good job and should continue working in this field. So once I am confident I have useful portable kit I will join WICEN and do more formal activities with them. Here is a summary from the de-brief of the last rallies.

Anyway, so yesterday we made a list of things we are missing and working to narrow that down. Along with places to find said missing items. Things like; good rope, large non-black tent pegs, tubing for another mast, extra RF patch cables, another dummy load, a good box on wheels that can be pulled or pushed like a sack truck and fix my SWR meter.

So later in the day we put together a couple more patch leads and a 7.5watt dummy load for Karl. I have also spent a fair bit of time working on the SWR meter. I looks good to me, but I think I'll follow up with some of the local HAMs to confirm that its accurate as I had to repair the input connector. Which meant disassembling the whole thing and rebuilding it again, which means it will need re-calibration.

Was planning some more study for the next exam today, but a long list of around the house jobs are yet to be done. You know the ones, mowing, digging, weeding, tidying... Maybe tonight =)
The main reason I need to push to get these exams out of the way is that I really want to work packet at the next rally. This is where my radio/linux/networking/computing interests all meet. There is so much stuff to explore in this space, thankfully it doesn't seem to change very fast, so I will have a chance to catch up =) Then I can start work into the areas of HF and 2 meter packet radio with some of the local guys. Not sure where that will end yet =)


Progress on NE602 based direct conversion receiver

So today I stayed home due to brain pain, it had been building up for the last few days. Also I had difficulty with my balance. To the point that walking involved bumping into everything. Lots of water and pain killers helped a little, breakfast and coffee later helped a little too. However the head ache didn't wear off until late into the afternoon with more sleep. So I spent some of the day sitting down, tidying the shed.

The cool thing about that was I found a couple of little paper bags with parts that I had bought months back and promptly filed for safe keeping in boxes! Anyhow, that list of parts contained an SA602, 3.579MHz crystals, a variable poly-cap, a large breadboard and some zeners.

This was all the missing parts for the NE602 based direct conversion receiver I've been wanting to build. So late this afternoon I sorted out all the other parts, put them in a box and setup on the kitchen table to assemble. My soldering iron and other tools are currently at a friends, so I pretty much had to assemble everything up on the breadboard.

Made dinner, a tasty curry with shallot pancakes. Then assembled it as per the circuit from W1AW. I am not sure where I first found it, google has a few references to it about. There are also a few variations on the NE602 direct conversion receiver, like the MRX-40_Mini_Receiver and another that I like using LC tank circuits for both the RF input filter and the oscillator side.

I fired it up but heard absolutely nothing. Much comparing with the diagram led to adding the vcc and ground rail for the LM386, after which I could hear familiar hash with the speaker held very close to my ear. I dug out the Pixie2 hoping to create some kind of signal. However the oscillator from the Pixie2 did not create the expected signal or inteference on the DC receiver. So this needs looking at another day when I can focus for more than ten minutes at a time. Hopefully this coming weekend wont involve any more headaches.


Fun with radio - part 2

Now that I have built a receiver, as the first stage in a QRSS grabber, how do I calibrate it?

I commented in Bills blog, SolderSmoke News, about the how the two pixie 2's David Rowe and I built were not on the same frequency despite being the same circuit. They are about 2KHz apart, which we put down to the tolerances in the components.

My HF rig has a 'digial readout' that measures down to 1KHz increments. QRSS uses a 'band' of 100Hz wide. Thats right, a tenth of what my smallest increment on the HF rig. So how to I measure that?

I'm looking to build a frequency counter or meter. Are there homebrew designs out there?

kim vk5fnet

Fun with radio - part 1

Yesterday I spent an afternoon working with David Rowe building a Pixie 2;

The Pixie 2 a tiny direct conversion radio intended for training new HAMs in under standing radio and building electronics. The initial build was a little bit troublesome as the oscillator wasn't. This makes it quite hard to build the rest of the radio, as it can be built in stages; the oscillator, the PA/detector, the audio amplifier.

Well the afternoon turned out to be quite productive and with a working radio picking up Davids Pixie transmitting from a foot away on the work bench. It puts out a whopping 200milliWatts and the output waveform on the oscilloscope was very sinusoidal.

David has a full amateur radio license and is a very cluey electronics engineer. I have my foundation license, so while I'm not worried about building a transmitter yet, I am quite interested in software defined radio(SDR) and QRSS. Which is what this project/journey is all about.

Whats next? Well as I now have a DC radio with audio output, SDR being the end goal, the next thing to build is the audio interface to the laptop so I can get the Linux QRSS client running...

kim vk5fnet

Band Pass Filters

Last night I built up an input band pass filter. The point of this little beastie is filter out the RF from above and below the band or radio frequency that you want to receive on.

I found a page on the band pass filters.

So the original Pixie 2 circuit - a direct conversion receiver I'd like to use for QRSS - has a low pass filter. Firstly, the reason the BPF project was started, is to remove all the spurious noise from other bands. Mainly the huge signal from the commercial AM broadcasters around 1.8MHz. Secondly, I want to make pretty sure that I'm not spraying harmonics across higher bands.

Some testing with the original LPF from the crystal oscillator shows a nice sine wave like signal with some noise on the peak of the upper cycle. However the Chebyshev BFP has lots of harmonics and looks like its attenuating a lot. I need to get some advice on reading the output of my CRO. Also a spectrum analyzer would be a very hand tool. But I'm not real happy with the output of the Chebyshev, it actually looks worse than the input side from the oscillator.

It seems, on advice from VK5TR / VK5JST, that many of the generic toroids from Altronics, DSE & Jaycar, may not have a suitable Q factor. Jim suggested the FT68-41 toroid cores. The suggested core is the T94-6. Will have to find a source of toroids specifically for the purpose. Not found any local suppliers, but will keep looking.

BPF == band pass filter
LPF == low pass filter
HPF == high pass filter
CRO == cathode ray oscilloscope
harmonics == multiples of the frequency that you are generating


More on toroids

Been searching for info about the toroids from Altronics. Setting up a band pass filter for the QRSS receiver.

The L4534 seems to be the same dimensions as the T130-2.

OD = 1.30 / 33.0 mm +/- 0.02 in
ID = .780 in / 19.8 mm +/- 0.02 in
Ht = .437 in / 11.1 mm +/- 0.025 in

Also I've found a useful site on the standard sizes.

The T130-2 has a AL of 11 +/-5%. So then to calculate the inductance like so; uH=(AL*Turns2)/1000

The L4517 does not match up to any of the standard sizes. Its just a bit bigger than the T50-2. So I'll use that as a model and obviously subtract some turns ...

I have three of the L4517's would with 36 turns of 0.5mm enabled wire for an inductance of 12uH. Now to pick out the capacitors for the BPF. Then find the smallest way to build it all up on copper clad board for the Pixie2 in the Haighs tin.

kim vk5fnet

Fun with toroids

The aim here is to find an inexpensive source of toroids for kits for QRSS receivers. Jaycar and Altronics have a reasonable number of shops around the place and also do Internet and mail order, so folks should be able to get the parts easily enough.

Anyhow, some time back I picked up all the 15.2X8.53X5.94 toroid cores, L4517 the local Aztronics. They sell from the Altronics catalog. So these cores are not standard sizes or models that are used in HAM radio, so far that I can make out. They are pretty close to a T50-2, as they are supposed to be iron core, but are a bit larger diameter. Using 0.5mm/24bs enameled copper wire, WW4016, from Jaycar, I've spread the turns out over at least 80% of the toroid, values as follows;
  • 10 turns gives 0.001mH

  • 20 turns gives 0.003mH

  • 25 turns gives 0.005mH

  • 36 turns gives 0.012mH - for 3.5MHz BPF

I could probably squeeze on another 5 turns, but 10-36 turns covers the inductance that I need right now. Perhaps other gauges of wire will yield other usable values. These toroids as they are setup are intended for low pass filters or band pass filters. The Pixie 2 needs just one in its current configuration. I intend to build up a better filter as there is lots of AM broadcast signal there when down on the flat in Adelaide.

kim vk5fnet

dreams of arduinos

Its one of those local public holidays today. Cup day or something. The kind that make you sleep in, warm enough to organise firewood and mow the back lawn. Well whats left of the lawn from the water bans and hot summer, but then theres the go inside time. Cool down with a chilled beverage and surf the 'net time.

I've been following the SolderSmoke blog and podcast, Bill mentioned that he put in a submission to Hack A Day about QRSS. Hack A Day is one of those sites that just gets the juices in the brain a pumping and thinking about building various things.

The first thing I found this morning was pluggable modules on an arduino shield, like lego blocks to enable fast prototyping of hardware and software.

The second was a two servo, four legged robot powered by four AA batteries and an Arduino.

The third was a tiny parallel supercomputer; the non-von1-supercomputer. There was another project a while back implementing a basic stamp supercomputer.

From a link somewhere I found one of the things that I keep crossing paths with, a home made Arduino board project. I found one today that set my mind racing off down the path of pluggable modules; the cheapduino.

So while not specifically parallel, but pluggable architectures of embedded systems. It could simplify the design of the individual board if a backplane had all the power, clock and a MPI interface for each CPU.

Perhaps a single USB connected backplane with a serial interface for each pluggable daughter board, in a simple frame that lets each daughter board have an edge connected I/O space. Something like the cards in a QBUS VAX. Smaller obviously, about five centimeters or two inches a side.

On that you could build a service per CPU, say, temperature logging, writing to a SD card, inrfa red comms., two wheeled robot platform built with servos, xigbee Tx/Rx, bluetooth...

I'm sure that there are loads of other ideas in this space. Its all about context and filling a need, its just strange some days where ones mind wanders... parallel pluggable embedded systems, which I'm sure one call call a mini-frame. Hmmmm ... theres an LCA talk in that I'm sure.


Saturday, May 23, 2009

Podcasts I've been listening to recently

Recently I've been listening to a couple of podcasts about HAM radio, homebrew, computers and Linux.

I have been hanging out for the next podcast from Bill N2CQR. Bill posts usually once a fortnight and keeps an interesting blog called SolderSmoke. Homebrew amateur radio, some travel log, some amateur astronomy and other related geeky things.

A couple of weeks back I found Linux in the HAM Shack, which is self explanatory really. Russ K5TUX and Richard KB5JBV, aim to walk you through all the popular things you's likely do with computers in HAM radio. Assume radio experience, but not Linux experience.

I have at least a hour bus trip to work and then another hour home again, so these podcasts help fill the voids in my head with useful radio thoughts =)


Friday, May 22, 2009

SolderSmoke the book has arrived

My copy of the SolderSmoke book has arrived. Much reading ahead.