Spectrum Analyzer v1.0

This is version 1 of a spectrum analyzer I built based on Wes Haywards W7ZOI and Terry White's KZTAU architecture. I have made design changes mainly because some of the parts are way to pricey for me to purchase. Further refinements and features are are added as the problems or need presents itself. Most of the simulations and initial calculations are done on EMRFD Ladpac tools.

Calibration oscillator is connected to the SA

Architecture

I chose a 2nd IF of 25Mhz there are disadvantages in using this frequency.1. The chosen log amplifier AD8037 has better linearity at 10Mhz
2. Narrow band 10.7Mhz IF filter are readily available.


Other things that are highly recommended when building an SA.
1. Signal Generator with frequency range that can cover the BPF filters
2. Highly sensitive RF Power meter.
3. Lots of BNC connectors
4. A copy of EMRFD


People that helped me a lot.
1. Wes Hayward- http://w7zoi.net
2. Todd Gale -http://http://www.qrp.pops.net

Here are a few more pictures.

I am using a 100Mhz clock IC for my 2nd LO. To remove the harmonics a low pass is used. The latest version is using a 9th order filter.

5th Order low pass filter.

The 2nd LO is being tested for proper output levels.

Sending Morsecode via an Atmega

A few weeks ago I wrote an openBSD client to connect to CWCOM servers and called it irmc. 

A new version of the client was implemented using an Atmel microcontroller. I decided to use an arduino board that was gathering dust on my work bench. I was initially planning to use the arduino as a testbed only and build an actual board later, But I usually only have time to do weekend projects and the coding already took me the entire weekend. So I decided to put building a standalone hardware for later or maybe never.

The source is available here http://fernan.bitbucket.org/irmc-avr.tgz. The code is written in C compiled using the avr-gcc tools. Orginally it was a port of the original irmc, but the limited memory of the atmega required a considerable rewrite. In addition to the limitation of the original irmc, here are few things to note.
1. Only 12 presets are programmed, instead being able to connect to an arbitrary frequency/wire.
2. PD6 is a GPIO that generates a ~650Hz square wave tone.
3. Some clock jitter results in 4ms variation in timming was observed.
4. I have it configured to interface to a separate sidetone generator for keying so the key input is expecting a 5V pulse.
5. I did not like the way I implemented the MorseKOB latch function in the original irmc, so I did not add that function into the avr version. (maybe someday)

BUGS?: Probably a few.

The final version was placed into an enclosure. The green LED blinks to indicate that it is connected to the cwcom servers.

Internet Relay Morsecode

I have been using cwcom to practice sending morsecode, unfortunately my main computer at home is running openbsd and it has gotten harder to gain access to a reliable MS Windows machine. So I wrote my own client that works on openbsd.
This is written in C and although not tested it should compile under other OS. If any one wants to try it and send me feedback. you can download version 0.01 here http://fernan.bitbucket.org/irmc.tgz
Les Kerr of https://home.comcast.net/~morsekob/ has been very patient in answering my questions regarding the cwcom protocol and setting up a test server during debugging.

There are few difference between this and the official cwcom client
1. It is a command line tool
2. It does not send the characters of the message to the receivers screen.
3. It does not translate CW for you. You can try fldigi if you just want see the transmission.
4. Tone pitch is currently hard coded to 650Hz
5. It does not have a way to show you who is listening. 

Bugs
There are a few that I am working on right now, but feel free to send me feedback if you get a chance to try it out.

File(s)
http://fernan.bitbucket.org/irmc.tgz


Resources
The following people have helped me a lot to learn the protocols and setting up test servers.
Les Kerr -  https://home.comcast.net/~morsekob/
Bob Denny - http://morse-rss-news.sourceforge.net/
John Samin - http://www.mrx.com.au/

Updated DC Receiver

Some changes from the original receiver:
1. Uses the latest multiband vfo with +15 to 19dbm
2. replaced the LM386 amplifier with an NE5532
3. a single copper clad was used with separate pcb for the 100db audio amplification

Some notes:
This comes out as 4rth rebuild since the SA612 to diode bridge. The amplification seems a bit low I can barely hear weak stations. There is still a few spurious whistle, but very low volume, mostly are external interference and compared to the LM386 implementation the hum is barely noticeable. I can connect the receiver to an external amplifier and not much hum is heard. There is currently no filter installed and the hope is by adding a 300Hz high pass filter will further remove the hum. This will also be test bed for learning about CW and SSB audio filters.

Multiband VFO ver 1.2

New Multiband VFO

Some changes on the VFO
1. Replaced the SA612 with a hombrewed diode ring mixer
2. Replaced the Dual tuned 1.0 Mhz bandpass filters with a Triple tuned 2.0 Mhz band pass filters
3. Diode switched crystal oscillators to change band.
4. Change from 2 stage to 3 Stage RF amplification to provide 13 to 19dbm output.

Some notes
The construction on a single Copper clad board instead of multiple modules. a purely manhattan construction was change to a mix of manhattan and ugly construction.  The original plan was for the crystals to provide 7dbm to 9dbm output this was adjusted to 10 to 13dbm. The intention was to allow adding a 3db attenuator and still have enough power to drive the mixer.

Frequency Coverage Plan

VFO		HFO	IF	Coverage		Band	Target
3.500	4.005	11	0	7.4997	6.995	40m	7.300	7.000
3.500	4.005	18	0	14.4997	13.995	20m	14.350	14.000
3.500	4.005	25	0	21.4997	20.995	15m	21.450	21.000

Power levels between stages (roughly measured during construction)
1. VFO = -8dbm this is further reduced by a 3db pad.
2. Crystal oscillators/buffer = +9 to +10dbm originally a 3db pad was planned
3. Mixer output = -26dbm This is after a 6db pad.
4. First 2 stages RF amplifier = +6db to +7dbm
5. 3rd stage RF amplification = +13 to +19dbm - I did not have any other transistor available so I maybe loading the 2n3904 more than it can handle.

Diode switched crystal oscillators

VFO, Mixer and crystal oscillators. Some leveling
of +/- 1db was sacrificed to get 9 to 10dbm output

Nice textbook waveform from the new mixer.

Unofficial measurement of RF leak of -46dbm

Upper band has a lot more noise

The output of the new mixer is a lot nicer than the old SA612, but whether it is better I will not know until I do some measurements. Most of the measurements right now are quick check as I build the mixer.