Working on oscillators
Jan 07,2007 Been wrassling with oscillator drift in LC and xtal oscillators. Man, there must be some tough guys out there to put up with the pain of designing a stable low drift oscillator. I have a feeling this is going to take me some time. I made a calibration chart for my germanium diode RF probe because the probe becomes nonlinear below about 2.5 volts or about 100mW. With a calibration chart the RF produces reliable readings to much lower readings. RF Probe Calibration Measurements Also made a spice model of a 1N34A based RF probe that you can use to see that the probe is nonlinear but predictable below the germanium diode cut-in voltage.
Cleaning up website a bit.more 2.57 rtnloss/impedance
Dec 10,2006 I found out that my K1 no longer worked on 20 meters. After some sniffing around it looks like the 22MHz lo for 20 meters is kaput. I ordered a new one from Elecraft. It appears that this is a common problem with K1s. Elecraft got in a bad shipment of 22MHz xtals. Cleaning up the web site a bit. Trying to have less buttons on the left by placing items in larger categories. I am working on EMRFD Fig2.57 still - trying to show techniques for measuring in and out impedance using a return loss bridge both experimentally and in LTspice.
EMRFD Fig2.57
Dec 02,2006
Been spending a lot of time with Spice and the books and have made some
headway. I built the amplifier circuit in Experimental Methods p2.20 Fig 2.57.
This is a great little amp. It has 50 ohm input and output impedance,
has 20 dB gain flat across the HF spectrum and with the values specified
abd puts out about 100mW. From what I saw in LTSpice the amp seems to perform well
with variations in component parts.
I used slightly different resistors and a different transistor from what was in the
book and still got performance that closely matched what was seen in LTSpice and EMRFD.
I put its spice file with my other spice models.
If you click on the spice button to the left you can see the other models.
The circuit worked as designed. I didnt use exact values of resistors
as I have more of some flavors that I want to use up instead.
I got +18 dB gain flat through HF with +21dBm (125mW) out.
I have been interested in trying to experimentally determine input
and output impedance of amps to see if what I see in spice really
happens in the real world.
I was shown that input and output impedance can be measured by
using a return loss bridge to determine return loss which is
related to VSWR and impedance.
The input return loss can be measured by the following recipe:
Attach a 50 ohm dummy load to the amplifier output
Attach a siggen to the RLB "RF in" port
Attach a power meter/oscope to the "Det" detector port
Turn on the amplifier, siggen and power meter.
Note the power reading on the power meter ("Unknown open").
Attach the RLB "Unknown" port to the amplifier input and note the power.
The return loss is the difference between the two power readings.
The output return loss can be measured by the following recipe:
Attach a 50 ohm resistor to the amplifier input -important-
Turn on the amp. With a scope, insure that it is not oscillating.
Attach a siggen to the RLB "RF in" port
Attach a power meter/oscope to the "Det" detector port
Turn on the amplifier, siggen and power meter
Note the power reading on the power meter ("Unknown open").
Attach the RLB "Unknown" port to the amplifier output and note the power.
The return loss is the difference between the two power readings.
That is it! I couldnt believe it was that easy
I understood that you could place a return loss bridge at the input of
an amplifier to determine input return loss/VSWR. What I didnt understand
is that you could use the same return loss bridge on the output of an amp
to determine output return loss. In addition to not understanding that
this was possible at all, I was concerned that I might blow up my siggen
by attaching it to the amp output. That turned out not to be an issue
at the 100 mW power level. At higher power levels it may be necessary
to put a 20 dB pad betwen the siggen and the RLB to protect the siggen.
Same goes for the power meter. It may need to have a 20 dB pad between
it and the detector output.
I think that as long as the amp is quiet and not oscillating and putting
out more power than the siggen/powermeter/RLB can handle then you should
be ok directly measuring output return loss/VSWR.
It is kind of wierd to me to be applying low level power to the output
of an amplifier in order to determine "reverse SWR" but it works. I think
there must be some sort of "dynamic biasing" that is going on. The
collector wiggles in voltage due to the siggen voltage and that causes
the base bias voltage to change which causes the transistor current to change
which has an effect on impedance/return loss. Just a guess on my part.
If someone really understands how this works, I would love to hear about it.
I saw 18 dB input return loss which is something like a 1.3:1 VSWR
across HF. I saw 20 dB output return loss - about a 1.2:1 VSWR.
This is a big deal for me. To actually be able to measure output
impedance gives me confidence that what Im doing is actually working.
Im going to do some work at this power level to get some more
experience then am going to try to graduate to higher power levels
and see what happens.
Built a KD1JV Variable Power Supply
Nov 11,2006 In case you are not aware, KD1JV is an amazing QRP designer. On his website he has a cool schmatic of a current limited power supply. I built one with a few modifications: I added a meter to measure the current through R5 and via a switch to measure the output voltage. I used a throw away Dell laptop power supply as my source of +20V. Other changes I did: Q1 changed to a 2N6594 junkbox PNP power transistor. This transistor required a lot of base current so D1 was dropped and R1 was changed to 500 ohm 2 watt Q1 was changed to an IRF510. U1 was changed from a single quad LM324 opamp to two dual LM358 opamps. R3 changed to 30K. The spst switch for the 50 microamp meter has a 5K going to the top of R5 for current measurement and a 400K going to Vout for voltage measurement. I havent yet gotten the current limit circuit to work properly It maxes out at 2.5 A and 20 volts. I got the current limit to work. I had inverting and non inverting pins swapped on the current sense amplifier. After burning up a bunch of gear here, I found out that my KD1JV was not putting out clean DC. I added a 100uF and a .01 to the output of the power supply and stopped it fom acting like a 0-20V pulse width modulated signal rather than a variable DC power supply.
Not much building going on
Nov 7, 2006 Been busy but haven't built much lately. Doing some studying. We'll see what comes of the study. Got some feedback on the Manual Spectrum Analyzer - need to spend some time adding a pad to the input of the mixer and on the output of the mixer before it hits the xtal filter. Also need to route mixer inputs/outputs better using rg174. Cleaned up the web site some. Working on a KD1JV power supply. Need to work some on the DDS30 SigGen. I added a 12v input through diodes and the thing generates wideband noise when running off 12V now works fine from 9v internal battery - may just add a switch.
Homebrew Manual Spectrum Analyzer
Oct 25, 2006
I have been pretty busy.
I got some 10MHz xtals in so went through chapter 3.4 of Experimental Methods
and determined the crystal parameters for the xtals. I used the AADE Filter
design program to make a 10 MHz 4 pole butterworth 3kHz 50 ohm in 50 ohm out
xtal filter.
The schematic for the filter is in the manual spectrum analyzer docs.
Once I built the filter, I thought I would try to make a spectrum analyzer
using it, a mixer, a DDS30 LO, and the W7ZOI power meter:
IT WORKS!
Here is the documentation for the Manual Spectrum Analyzer
LTSpice SM0VPO
Oct 21, 2006 Been cleaning up the website. Moved some things from the left menu to Various. Created a "Spice Files" left button. I plan to add models and circuits there. This week I have been experimenting with SM0VPO's half watt amp. I have been re-organizing my bookmarks. I created a section there called "Biasing". The W7ZOI amp notes there.
Bitx20 / W7ZOI
Oct 18, 2006 Looking at the dates it is clear that I need to spend more time keeping up with the journal. Winter is coming here in Juneau time to think about winter building projects. I put in an order to Digikey and to KitsAndParts for some balun core ferrites and some interesting transistors. Been working weekly on SolderSmoke. It takes a fair amount of time each week for Bill and I to do the podcast. We recently had an Echolink interview with Ashhar Farhan the designer of the Bitx20 SSB transceiver. Check out Farhans Page at http://www.phonestack.com/farhan. In preparation for the podcast we had an opportunity to have an email dialog with Wes W7ZIO who sent some long and inspiring notes. Thanks Wes. One item that Wes sent was I think of general interest to the QRP community. It concerned the problems that Bill and I were having with getting sufficient gain out of amplifier stages for sideband gear. Here is a link to what Wes had to say on the subject of amplifier design using the tools in Experimental Methods in RF Design. W7ZOI Amplifier Design Notes Spent some time with the layout of the left window on the home page.
Catching up with Journal
Aug 26, 2006
I took a photo of the new shack layout, it also wraps around to the left
Had a bunch of trouble with SolderSmoke.
Moved our mp3 files to www.soldersmoke.com
Been experimenting with QRSS. I put up a QRSS beacon on 10.140.030 Mhz.
I saw a couple of signals last night which got me interested in QRSS again.
Resurrected an old Heathkit VTVM.
Finished the schematic for the Motley1
The Motley Special
After Re-Flooring my house and demolishing/rebuilding my shack, I am finally back at soldering.
I made a simple - Very Rockbound - transciever in the basic design of the PSK Warbler. This
rig is for our "Motley Group" frequency here in Alaska.
More info is available Here
AFU Knoppix under VMWARE
I experimented with runnning AFUKnoppix under VMWARE and posted the results to QRP-L
Two Tone Test mp3
I used audacity to create a Two Tone Test audio file to be used for transmitter testing.
HF/6M Diplexer
I Made an HF and 6Meter Diplexer using ELSIE from the 2006 Handbook to design the HF lowpass and 6M hipass butterworth filters.
Soldersmoke etc
SolderSmoke has turned into a weekly thing. Been working with AA0ZZ on a PICEL frequency counter for the past few weeks Im at the limit of my abilities on this one. Craig has taken over and is making progress... Im thinking about taking a rest from that project for a while and working on ones that I can finish in a weekend..
KX1 EPS-1 DC-DC Converter Mod - 12v internal
[Elecraft] KX1 Mod / EPS1 12V DC-DC converter - WORKS! I got an EPS-1 DC-DC converter in a horse trade about a year ago and thought it would be a cool addition to my KX1 to jump the 9Volt internal batteries up to 12Volts. The EPS1 was either dead when I received it or I killed it because when I plugged it in to a battery source as the switching FET surface mount transistor caught on fire! It looked like one of those 4th of July "snakes" burning only with sparklies. I let it set on the shelf for a while and then tried a couple of times to resurrect it with no luck. I tried replacing the blown FET with a 2N7000 but it didnt work. I then ordered a new MAX1771 DC-DC converter controller chip and when replaced the one on the EPS-1 board with the new one it started putting out 12 volts with about 4 volts in. I experimented with load resistors and it was clear the 2N7000 switching FET would not handle the current so I swapped it out for a Radio Shack IRF510. The IRF510 didnt even get warm with an amp load so I tried running the KX1 off of the modified EPS-1. It worked great. I was getting 12 volts from a little 9 volt battery and was getting full power out of the KX1 rather than the lower output I get when running off the internal 9Volt batteries. I thought I could just disconnect one of the 4.5V AA packs in the KX1 and then put the EPS-1 inside of the KX1 but it turned out that 4.5 volts was too low to get reliable switching from the EPS-1. I then went to Radio-Shack and picked up a 4-AAA battery holder (RS270-411) which fits nicely into the space of one of the AA original battery holders. I removed both original battery holders and installed the EPS-1 where the other original holder was. It worked FB. I had to cut a small hole in the back of the case so that I could access the RS770-411 on/off switch and used double sided foam tape to secure the EPS-1 and battery pack. I have no idea how long the batteries will last. I do know that I have had the KX1 for a couple of years and have only changed the batteries once in that time so I imagine that they will last long enough for my use. Ill see. EMBRES no longer makes the EPS-1 but the schematic is pretty simple and comes pretty much right off the datasheet. Maxim-IC makes a litle MAX1771 evaluation kit that might be a direct replacement for the EPS-1 although I dont know how much it costs. Useful links: http://pdfserv.maxim-ic.com/en/ds/MAX1771EVKIT.pdf http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1030 http://pdfserv.maxim-ic.com/en/ds/MAX1771.pdf Have Fun Mike KL7R
Been busy...
I got back a bunch of the orders. I built a PONG game using a 16F84 for my son for Christmas Made a page: http://kl7r.ham-radio.ch/scrimshaw. Describes a novel technique for making prototype circuit boards. Got back my FT817 - need to mount it in here somewhere. Been working on Atmel Butterfly code for DDS VFO. - Hacked on Steve Webers code to remove his CW menu entries. Now it is just a siggen.
Monster Mail Order
I sent in orders to Digikey, Mouser, OceanState, www.sparkfun.com, Expand Spectrum System, ON Semiconductor, Analog Devices, AMQRP, and Elecraft. I sent in my ATS3 for the latest firmware. I should have lots of stuff to keep me busy this winter. I ordered a few different types of pics and a programmer so plan to experiment with them a bit. Should have lots to do this winter.
BCD LED Clock
I built an LED Clock I started with the information on this page but had almost none of the parts that he used in my junk box so wound up designing one myself I am going to give it to my son for Christmas so Keep it a Secret :) I cant believe that Juneau Radio Shack had wirewrap tools and wire!!
Fixed up some old test equipment
Had a dead counter - replaced front end FET - works great now. Had a dead logic probe rebuilt it. Sort of works - looking for a better design. Built a return loss bridge like page 7.23 of Experimental methods. Did some looking into the soundcard based impedance bridge and curve tracer that was mentioned in QEX/QST and CircuitCellar magazine. Look for "QEX Steber LMS bridge". That QEX article pointed me to his curve tracer program on CircuitCellar magazine. Downloaded the two programs I need to get a better soundcard.
OM3CPH Counter
I built an OM3CPH Counter based on comments by PIC Elmer WB8RCR on Echolink, It works FB! Thanks for the hints and the great work with the PIC Elmer course John!
SolderSmoke Podcast
M0HBR and I have been working on a podcast stream of the conversations we have on Echolink re building projects Look for more information at http://www.geocites.com/soldersmoke
DC to Daylight Binaural Receiver
Update: Aug 27 05
I built the second mixer/audio chain and put the quatrature phasing choke
to generate I/Q like on page 9.20 of "Experimental Methods in RF Design"
IT WORKS! Ill have to put up an MP3 of the resulting audio...
I may draw up my schematic also.
----
So far this is not binaural but does receive and sounds great.
It takes the signal right from the antenna and mixes it with the
output of the PIC/EL DDS VFO via an SBL1 mixer.
The output of the mixer is audio which is amplified with a lm386
in a special configuration so that the LM386 gives 70+ dB of gain.
The output of the lm386 goes to a set of Sound Blaster Speakers.
I tried to make it into a transciever but had not too good luck.
If I can get the transmitter to work, I will need to hack the VFO
code some more to get a 700 Hz Transmit/Receive offset.
If you want to experiment with doing this, Check out Ham Radio Deluxe
It is an amazing program and has a BandScope feature which does exactly what Im trying to do.
Also Check out This doc on setting up HRD to act as a poor mans spectrum analyzer
