开云体育

Interesting findings Paul.

I have found that the oscillator frequency is very prone to temperature changes. I have two kits and this is more noticeable in one of them than in the other. I think the reason for the difference is that one has very short crystal leads allowing heat transfer from the board through the base of the crystal, while the other is plugged into header sockets and has longer leads so it is thermally separate from the board.

When the frequency moves from the one you set up the kit with (Freq setting and FSK.ADJ) the width of the wpsr modulation changes. This is part of the nature of the kit. The frequency, FSK.ADJ value and actual width of the modulated signal are all intimately connected. This is due to the fact that the characteristics of the red led being used as a varacter diode are not linear. The capacitance/voltage characteristics are not in a straight line. Because of this, when the set FREQUENY differs from the real frequency, the effect of the micro-controller trying to adjust the frequency alters the fsk width.

By keeping my kit in a stable temperature (no direct sunlight shining on it, or diurnal temperature changes) it stays very stable both in its frequency and in the width of the wspr signal as long as it was properly set up in the first place.

I don't worry about minor changes in signal width. If it is within about 5 htz to 8 htz you will get very many reports if your antenna is OK. Obviously it is ideal for it to be 6 htz, but the receiving stations seem to be tolerant of some deviation.

I have also seen the wrong LOCATOR on the LCD screen for a while, but the kit was transmitting the correct data. I am not certain how often the screen updates if your GPS loses lock and for a time a wrong LOCATOR is on the screen. The kit can correct its LOCATOR, transmit the right one and still show the wrong one on the LCD until the Menu button has been pressed.

In starting up from cold, I find it best if using the GPS NOT to go straight into WSPR transmit mode. Often there is a spurious LOCATOR at start up and the WSPR transmit process can not start until this is corrected (if the value of LOCATOR is an illegal one). Once in this condition, the kit will continue to refuse even if you go through the menu list and press the right button to start. All the while, it seems inhibited from updating the wrong LOCATOR value. Instead of getting into this mess, I usually start up my GPS connected kit in TEST mode for a few minutes. Two or three minutes in TEST mode and the kit gets all of the data correct, and is ready to operate properly. I then initiate WSPR mode and it goes on fine.

I have noticed on two UQRSS kits that I have built that the oscillator frequency when in non TX mode during wait periods is about twenty htz higher than when the TX ON period starts. This may be because in WSPR the processor tries to steer the oscillator to match the FREQUENCY you entered into the kit's FREQUENCY parameter. Whether the processor controls the oscillator or not at a particular time, depends on the mode in use. Read Hans's comments about how the GPS controls the frequency.

In WSPR the frequency is controlled in between transmissions only.

In QRSS, FSKCW and DFCW modes, frequency stabilisation only occurs when the RF output is ON, and the FSK is zero. Maybe this explains what you have observed about the changes of frequency between different modes and transmit states.

73s

Tony G0BZB


--- In QRPLabs@..., "paulwijnand" <paulwijnand@> wrote:

--- In QRPLabs@..., "paulwijnand" <paulwijnand@> wrote:

Hello all,

Build the UQRSS for 30 M a few days back and started without the use of GPS.
Nice kit to build! Thanks QRP-kit team!

Changed the capacitors C4 for 22 pF N750 and C5 for 15 pF N750.
Looks it stays better on QRG using N750 then N0.
Have to experiment with C3 with a N750, see what it brings in Temperature compensation.
But that is a long process, will take a few weeks...

Anway I started with WSPR, in the menu put in the Time (hours and minutes) and hit the button there when at a DFCC clock the seconds jumped from 59 seconds to 00.
It seems that in this way the DT is nor far off in WSPR, about 1 second.

Results were fabulous running 0.2 Watt on my Indoor magnetic loop of 1.5 by 1.5 meter.

2012-08-28 06:52 PA0OCD 10.140179 -23 -1 JO22li +23 0.200 KC9NBV EM69oe 6775 4210 .

So a 6775 km is not bad hi.

The time without use of GPS stayed very precise! during a day or so.
But have seen in the messages here, that it will stay longer precise.
I monitored WSPR on a second notebook and a precise homebuild transceiver locked on a double oven OCXO 10 Mhz standard.

After that I used GPS to see what that brings.


I have a old GPS Jupiter T Firmware V1.80 and used 10 K pull up resistors from the +5 volt from the Data and 1 PPS line. This one uses an extern antenna and has for both the module and antenna + 5 volt powersupply via the UQRSS. Without the use of these pull-up resistors the UQRSS goes wild in adjusting the QRG hi

The other GPS is an EM-406A also using 10 K pullpup resistors and has a internal antenna. Nice little module and with a very high sensitivity. Also +5 volt power supply.

Used them both.

For WSPR it is with firmware UQRSS 1.04 nice to see the right QTH locator and precise timing.

After that I put on FSKCW3 with Frame 10 and Start 1 at 10.140.080 with a shift of about 6 Hz. In the menu GPS ON.

Let it run for a few days and could see my signal on the grabber of Bill W4HBK (Tks Bill!), also a nice distance of about 7522 km...No stacking of the pictures needed yet.

The GPS adjustment was nice to see and was soon withing 1 Hz.
The timing was also good. The first Character was send after
a 3-5 seconds (about) after the startminute (Start 1).

BUT:

A: After a few hours, the display showed a not changing time of 12.28.36 or 05.28.36 or 08.28.36..... The character being send were not showed any more...
This was the case with both GPS modules.

The GPS adjustment an timing were still ok! and the characters were still being send. (hi)

After starting the FSKCW3 mode 2 times , not one time!, all was running well and the time showed ok in the pause when the message had been send (PA0OCD JO22) and the start of the minute 1.
(Frame 10 Start 1).
This time it stopped at 11.18.36 (put it on at 10.24 utc)...


B: Then I removed the powersupply and put it on after a while.
The FSKCW 3 started automatically BUT not at the right time and the display was also not showing the right time.
After starting (with the powersupply on) the FSKCW3 2 times all was
well again.
Perhaps I had to wait a bit longer then 10 minutes to see if the Time was right again by itself...

I still have to test the other modes.

That was my experience up to now.

Will build the other UQRSS modules, but perhaps with 1 display, not sure if I can switch the display between the other modules while they are running...


73

Paul, PA0OCD

Hello all,

In my former message I mentioned using WSPR without and with GPS.

I had first used WSPR without any connection to a GPS unit:
(later on I only connected GPS in mode WSPR to see if the QTH locator and time were right, but did not test WSPR GPS further then.....)

This use, without connection to a GPS unit worked well both for QRSS/FSKCW and WSPR, and the adjusting of the FSK shift (Fskadj) in Testmode was also good for running WSPR:

First C-8 had to be tuned to the QRG which was set in the Menu of UQRSS.
After that the FSKadjust had to be right using a shift FSK of 4 (or 5 Hz).
Then choose Mode Test or FSKCW with a low speed, say 009, and Frame 00 and Start 00.

After a lot of adjusting the FSKadj was found, let us say 00,800.

For WSPR the Frame had to be set then on 10 and Start on 0,2,4,6,or 8.
I choose Start 02.
After setting the Time using a pretty precise Clock, all was running well in WSPR and also the other modes QRSS, FSKCW.

BUT:

After I connected with these adjustments above, the GPS unit ( EM-406A with 10 K pull-up resistors from the +5 Volt to both the Data and 1 PSS line), and after powering up the UQRSS and the connected GPS unit, some strange things happened:

(BTW: Use the Start in the menu a few times after powering up the QRSS with the connected GPS !, so that the Time on the UQRSS display is in sync on the precise Second with a DCF77 Clock or precise Internet Time....)

The QRG dropped now on WSPR TX not only about 25 Hz ( in my case from 10.140190 to 10.140.165) but also the Shift had gone up !
This was now for WSPR about 7.1 Hz. ( had to be around 5.9 Hz).
In the Menu the QRG was still 10.140.190....and FSKadj still 00.800...
Of course no detection of my to wide WSPR signals on the WSPRnet.org.

So now I tried first to set QRG in the Menu to the new QRG 10.140.165 (no change of C8!) to see if the Shift was back again around 5.9 Hz.
But on TX now the "real" TX QRG dropped even further down hi.
This was not the way!

So I put the QRG in the Menu back to 10.140.190 and tuned the C-8 up to 10.140.190.
Now even more strange things happened.
On TX the QRG shifted regulary about 100 Hz up and down like a warning signal hi and of course no good WSPR was being send.

So I put with C8 the QRG back on 10140.165 ( in the Menu still 10.140.190) and adjusted
the FSKadj to a smaller value (from 00.800 to 00.730).
This adjusting took a lot of time but finally I got on WSPR the 5.9 Hz using Argo Speed 10.

Why not use Test for shift adjustment with the connected GPS and Menu GPS ON?

Aha: In Test mode with te connected and running GPS unit the Shift is a bit more then in mode WSPR … so best is to adjust FSKadj in mode WSPR.


Use Argo Speed 10 and aim at the middle of each 2 most outside WSPR tones to see if it is around 5.9 Hz.

Yes once you know the trick it is not difficult anymore. Hi.

BTW: I tried another GPS unit ( Jupiter T) but the result was about the same.
Changed both IC1 and IC2 the result was the same.
Had ordered a few more...for other bands.


Hope it helps others a bit.
It was nice doing it and with very good results.

73

Paul, pa0ocd


Nice result now with WSPR:

VK4TMH at 15690 KM with 0.2 Watt on the Indoor Magnetic Loop


?2012-09-05 16:12?
?PA0OCD?
?10.140169?
?-25?
?-1?
?JO22li?
?+23?
?0.200?
?VK4TMH?
?QG56fj?
?15690?
?9749?