Elecraft Transverters

I decided to stop my tweaking and take some measurements in a slightly more scientific format.  --aka... I was stumped and needed to noodle through my data and step away for a little bit.

• I went back to each transverter and reset the PA bias to 20mV.
  
• I did not adjust any of the filters...  good or bad... measured As-Is
  
• I measured each transverter with three different R22 values Full Counter Clockwise, Centered, and Full Clockwise. (R22 is a variable attenuator for the IF input.)
  

The fields below that are yellow indicate that I don't trust the results of my oscilloscope. (It is an old slow 60 MHz scope so I am pushing my luck measuring the LO and Mixed values on the 222 and 432 transverters.)

All oscope values are with the probe @ x10  -  K2 drive from the K2 display



Good news: The 2m transverter (XV144) is working good. I am seeing what I expect for power levels and it should be near it's final form.

Bad news: The 222 and 432 transverters need some help. I am not sure of the exact problem but I have two ideas to investigate:

1. Go back thru the filter adjustments. I think that I need more signal going into U6.
2. U6 - is it bad or not getting enough signal to do much real work?

73 de NG0R


PS... I need to go back to remeasure the power out of the XV144. I thought that I had a 50w slug in the bird 43... but I think that I actually had the 10w slug in there. --That will mean that my power is still pretty low but at least it is measurable.  :-(  

PPS... C53 in the bandpass filter was acting weird on the XV222. I found it while poking around with the oscope. I top soldered the joint and resolved that little issue.

PPPS... I tweaked up the power level on the XV222. It is still running less than QrpP levels. I am going to take a break for the evening and ponder this some more.

PPPPS... After some pondering and discussion with N0FP I think that I have a plan of next attack to diagnosis this issue. Notes to follow once I have time to revisit the work bench.

APRS RF vs Cell

I was out running some errands so I decided to compare APRS on an RF path vs on my Crackberry.

Red: D700 2m @ 50w
Pink/Purple: Crackberry on AT&T EDGE network


Clearly the traditional RF path worked better today. To be fair the Crackberry client that I am running is an early Beta so it is possible that influenced the results.

Since I was looking at the data I decided to see what the longest path that my signal worked during my short trip. 100 to 200 miles on 2m FM is a pretty impressive path and not a normal APRS path.



Links:
APRS FI (for the Google Map + Data)

Crackberry APRS Client:
http://aprsbb.bluearray.net/

APRS IS Password Generator:
http://www.rimboy.com/rimdistro/rimigate/calculate/

73 de NG0R

Elecraft Transverters

My setup/troubleshooting quest continues.....

Showing the bandpass filters inside of the the transverters.



I was have a problem getting the full 1mW (0 dBm) of drive from the K2+K60XV into the transverters. I was never seeing more than about .50mw (-3dBm).  After some searching on the internet I saw a note that pointed me back to the Errata for the K60XV. Don the Elecraft mail list confirmed the detail.

R6 on the KAT6 board needs to be changed from a 47 ohm to 470 ohm resistor or the op-amp detection circuit may not give you the full 1mW of drive do to some interactions between the K60XV and KAT2 boards.

So I cracked open the K2...


I removed the top lid, the KIO2, K60XV, and KAT2 interfaces.


The KAT2 is actually made up of two boards so that need to be split apart to get good access to R6.

I had forgotten how much labor was invested inside the K2 case. It always makes me a little nervous popping the cover off as there are a lot of add-on modules in there some of which are attached to the top lid. The R6 change was very easy and the radio went back to together fine.

It appears that I am now able to get 1mW from the K2. So I am back to my original issue of figuring out why I can't get the proper power out of the transverters.  --arghhhhh.

73 de NG0R

Elecraft Transverters

A little bit of progress tonight.....


My transverters were not talking to my K2 as I trying to work through the initial alignment at the workbench. After some testing and a couple of emails with the Elecraft reflector it was pretty simple.

Dooh.... the plastic shell on the DB9 connection was not allowing the pins to go far enough into the connector on the radio.

It turns out that I needed to move the connector to the forward slightly in the DB9 case to a different notch. That slight tweak in the alignment solved the connection issue.

When time permits I will need to work through the alignment steps. With some luck maybe I can get them tested and on the air this weekend.

73 de NG0R

APRS: RF vs Blackberry

I had to run an errand over my lunch break so I decided to run an APRS experiment.

NG0R-9 : Kenwood D700 @ 50w transmitting every 3 minutes
NG0R-12 : Blackberry Bold running APRSBB sending every 45 seconds (the default)

There is an APRS receive gateway here in Kingston, MN so coverage should not be a problem. The trip today was a maximum of 12 miles in a straight path to the receive station.

Red is the RF beacon
Pink/Purple is the Blackberry


I was surprised at how much of my trip was missed or not received on the RF path. I expected to see more cell phone/data drop outs because there are numerous black holes in my area on the AT&T network.


The weird part is that I was heard by three stations ranging in distance from 12 miles away to 129 miles away. I would have expected the two closest sites to be able to hear my whole trip.

The table below show that with the D700 in my truck I have seen APRS paths as good as 200 miles over the past couple of weeks. Those are some pretty extreme distances as 20-40 miles is more common.


Links:
Blackberry APRS client:
http://aprsbb.bluearray.net/

APRS Maps & Data:
http://aprs.fi/info/NG0R-9

This data really surprised me as I figured that I would have better APRS coverage near my home since there is are two nearby receive sitez. It makes me want to do some more experimenting.

• Map out a regional trip and compare the APRS coverage of the RF path vs. cell path. (AT&T cell coverage gets pretty spotty as you move further west across rural Minnesota)
• Put up two receive stations at my QTH (attic and tower) and compare what I can hear direct at my QTH vs the nearby APRS receive nodes.
  

73 de NG0R

Kingston QTH in August

I took the point & shoot camera with on my afternoon laps around the block. It is a nice summer afternoon in Minnesota with the temperature around 80 degrees (F) and 54% humidity.

We got 2 inches of rain Thursday night & Friday morning. That should make the plants perk up a bit given how warm it has been recently.


The plants in this corner of the yard are going a little crazy.


We should have some grapes that are ready to be picked in the next couple of weeks. We made a lot of jelly last year and it appears that we will have a larger crop this season.


This shot gives you an idea how big a hill we live on. This is only about 1/2 the elevation of our yard... the rest of the drop is to the right of the image.


For the folks that are interesting amateur radio you can see the HF tower in the foreground and if you look carefully you can see the VHF tower in the background. (I am active on 160m-->23cm or 1.8MHz-->1.2GHz with a couple more bands likely in the future.)

73 de NG0R

More... Reverse Biased Diodes

I am still looking for some other options for a better method of handling my tuning for my next QRSS prototype. I decided to revisit the bread board and take some additional measurements.


I was wondering if a series red LED + 1N5817 (both are reversed biased) might give me the range that I need within a 0 to 5vdc window. (I will be using a PIC 16F628A for pulse width modulation so 5vdc is my upper range.)



The results are promising based upon the graph below. My design goal is about 12-20pF. I think that I need about 18pF based upon some earlier prototypes.



This picture shows you what a pain it is try to measure these parts and then swap out new combinations. At some point I will design a little prototyping adapter for my LC meter  so that I can do this right on the front of the meter without any jumpers.


I tried one other combination that provided a very granular range.
Reverse biased in series 1N5817 + Red Led + 1N5817
0-5vdc = 6-18pF


I have four different variations that I have tested over the past couple of days that will get me into the ball park to differing degrees. (granular tuning vs. tuning range vs. parts count)

1. Reverse Biased 1N5817... a bit too much capacitance by it's self
2. Parallel 10-30pF variable cap + rev biased Red LED
3. Series rev biased 1N5817 + rev biased Red LED
4. Series rev biased 1N5817 + rev biased Red LED + rev bias 1N5817
   

Option 2 provides the best flexibility
Options 3 & 4 provides more granular tuning

I will need to ponder this a bit to determine which path I like the best.

73 de NG0R

Skywarn on a Friday

K0MPX asked us to activate the Meeker Co. Skywarn group twice this afternoon.

Here are some images of the second trip near Pigeon Lake south of Dassel, MN

Overshooting tops + rainbow


Crackberry photo of the radios in the truck


Crackberry photo of me in the mirror sitting in the rain


Crackberry photo thru the rain of the rainbow


Overshooting tops near Dassel, Mn


Trip1: Kingston, Kimball, Watkins, Eden Valley and back
Trip2: Kingston, Dassel, Pigeon Lake


73 de NG0R

Reverse Biased LED for QRSS

I managed to sneak to the workbench after everyone went to bed  :-)

I wanted to see if I could find a good diode or LED + capacitor combination that might work for my QRSS project. So I placed some likely parts on the breadboard and attached my AADE LC meter.


Since I plan to use a PIC 16F628A to pulse width modulate the LED I figured that I should keep the DC voltage on the breadboard between 0-->5vdc since the PIC will use a 5 volt regulated source. It turns out that I have about 200 red LEDs from an eBay purchase that work nicely in parallel with a 10-30pF variable capacitor.



With some real quick testing it looks like I can dial this in between about 8-50pF pretty easily. I think that I need about 18pF for the current QRSS design that I am working with. This looks like it should allow me to vary the frequency of the transmitter fairly easily.

The QRSS sub band is 100Hz wide. If you are doing any sort of FSK activity the general rule is to only run about 5Hz of shift since the sub band is so small. The last time that I tested this concept I think that 5Hz of shift at 10MHz is about .5pF.  That is a tough value to try to achieve without something like varactor diode. A reverse biased LED can fill this role pretty nicely.

The variable capacitor on the left should allow this to be rescaled for other bands fairly easily.

The LED will allow for the 5Hz frequency shift.

At some point in the near future I need to write a little test harness for the PIC and put on a breadboard. That would allow me to measure this as a complete circuit with the PWM running.

73 de NG0R

Reverse Biased Diodes

Tonight I was playing with the idea of using reversed biased LEDs and Diodes as varactor diodes or variable capacitors. Hans Summers has done a lot of testing around this concept for use in CW and QRSS transmitters.

Here is an example of what a particular diode exhibited during a test:



I mocked up a simple test on a bread board and used my AADE LC meter to take some basic measurements. This is largely based upon a picture that I saw on Han's site.


I used a variable bench power supply to run the tests. I might build a simple printed circuit board fixture for future tests so that I have solid tool with reproducible results.

Here is the raw the data for this test:



Overall it was a fun experiment that shows some practical promise for some future projects.

73 de NG0R