Receiving sometimes stop
Isao SEKI - JM1WBB
I am a newcomer and am setting WSE and Linrad on Debian.
One of my problem was fixed by kindness advice from Leif.
At this moment I have another problems. One is receiving
sometimes stops (noise level downs -5dB and receiving
internal birdie only) for 5-20 minutes and returning. I
guess my power supply cause a trigger and already ordered
new one. But anyone has same problem and fixed, please
let me know your detail.
Two is a internal birdie on center frequency of each pass
band. In problem one situation, also it is.
WSE 4 boxes for 144MHz
Delta 44 Sound card
Debian GNU/Linux 5.0.1, OSS-linux-4.1-1052, Linrad-03.05b
Thanks in advance.
73,
Isao / JM1WBB
Leif Asbrink
> At this moment I have another problems. One is receiving
> sometimes stops (noise level downs -5dB and receiving
> internal birdie only) for 5-20 minutes and returning. I
> guess my power supply cause a trigger and already ordered
> new one. But anyone has same problem and fixed, please
> let me know your detail.
In case your power supply becomes overheated so it switches
off or reduces the +15 or -15 DC supply (or both) you should
see it easily with a volt meter. I am not aware of anything
else that could produce a similar effect. Maybe better
positioning could improve coolign?
> Two is a internal birdie on center frequency of each pass
> band. In problem one situation, also it is.
The center frequency corresponds to DC voltages and they
are not amplified by the Delta 44.
Near the center frequency there is 1/f noise. The normal behaviour
of the Delta44 without anything connected to it is shown here:
http://www.sm5bsz.com/linuxdsp/hware/delta44a.htm
The center spur can be slightly reduced by modifications on the
Delta 44.
During normal operation on 144 MHz when your preamplifier should
lift the noise floor by about 15 dB the center spur should not
be visible or very weak.
73
Leif / SM5BSZ
Isao SEKI - JM1WBB
Thank you for your advice. I repaired (changing) DC power supply and
an electric fan is installed in side of the systems. So, the problem
is not reproduced. And, I am going to modify the first step of Delta
44 (GND cutting at Dsub connector) soon. I'm not old age but my eyes
are presbyopic...
73,
Isao / JM1WBB
Isao SEKI - JM1WBB
I have still have the problem without any error message.
I'm using two Kenwood PA18-3A Regulated DC Power Supply
now and their volt meters are OK (about +15.5 and -15.5).
Of course, I get many signals while normal operating.
Can you see a screenshot?
http://gongon.com/JM1WBB/linrad-1.png
In waterfall it was normal operating at orange ellipse
and other times were receiving stopping... Can you
anyone advise me?
tia 73,
Isao / JM1WBB
On 2009/05/15, at 4:33, Leif Asbrink wrote:
Isao SEKI - JM1WBB
It seems that my RX10700 has problem. While trouble condition,
Linrad get a signal on 2.5 MHz and doesn't get a signal on
10.7 MHz from signal generator. (HF Rig with dummy load)
And I saw two other statuses of the waterfall. One was 3 times
only:
http://gongon.com/JM1WBB/linrad-2.png
Two is continuing:
http://gongon.com/JM1WBB/linrad-3.png
last one is like self oscillation or using super preamp and I
get big noise from PC's speaker. It will go to normal level if
outputs from RX10700 is offline. I was aware that WSE boxes are
heating. Then, WSEs are cooled by electric fan and was reseted
by power OFF/ON, WSEs is working well now. But I got cold :)
http://gongon.com/JM1WBB/linrad-4.png
Must WSE boxes have forced air cooling? I suspect the gap is
needed more and someone said it is needed 2cm gap at least
before... Gap of current setup at test bench is about 1 cm.
73,
Isao / JM1WBB
Leif Asbrink
> It seems that my RX10700 has problem. While trouble condition,
> Linrad get a signal on 2.5 MHz and doesn't get a signal on
> 10.7 MHz from signal generator. (HF Rig with dummy load)
into the RX10700 to see if you have the same problem.
(Both units have 4 crystals.) Select 144.0, 144.5, 145.0 or
145.5 and test with the signal generator for signals at the
RX144 input which should be centered at 10.675, 10.7, 10.725
or 10.75. (Ignore the error messages for non-connected units)
In case there is a poor connection in the dsub(s) chances are good
you will detect it this way.
If the RX10700 does not show anything even with a very strong
input signal, the LO is most probably not running. Did you
try to change the frequency in the 10.7 band to see if another
frequency works.
> And I saw two other statuses of the waterfall. One was 3 times
> only:
> http://gongon.com/JM1WBB/linrad-2.png
> Two is continuing:
> http://gongon.com/JM1WBB/linrad-3.png
> last one is like self oscillation or using super preamp and I
> get big noise from PC's speaker.
When a WSE unit is powered up and before Linrad has sent a
command to it, the state of the latches holding the info about
which crystal to use is undefined. It could be no crystal at all
or it could be several crystals. When several crystals are
selected, several oscillator transistors are powered simultaneously
and the total gain in the LO becomes too high causing various
peculiar phenomena.
Every time a unit has been without power it has to be programmed
by Linrad to start working again. If you are in a receive screen,
type 'X', then 'B'. It is not enough to just select a frequency.
Linrad would only send data to the units that need re-programming
and trusts that other units still remember their settings.
> It will go to normal level if
> outputs from RX10700 is offline.
Yes, the RX2500 only has one crystal so it does not have any
memory.
> I was aware that WSE boxes are
> heating. Then, WSEs are cooled by electric fan and was reseted
> by power OFF/ON, WSEs is working well now. But I got cold :)
> http://gongon.com/JM1WBB/linrad-4.png
>
> Must WSE boxes have forced air cooling?
No, but they must not be allowed to heat each other.
Normal temperature is 45 to 50 degrees with boxes vertically
and a 3cm air gap between boxes with provision for air to get
in from beneth and out at the top.
> I suspect the gap is
> needed more and someone said it is needed 2cm gap at least
> before... Gap of current setup at test bench is about 1 cm.
Check the temperature. The units should work properly at much
higher temperature, but the lifetime of electrolytic capacitors
would suffer.
In case the RX10700 oscillator stops working at some specific
temperature I think you would find that the other three
crystals still work at that temperature. If they do not there could
be some other temperature related error. One of the serier regulators
might turn off for example.
73
Leif / SM5BSZ
Isao SEKI - JM1WBB
Thanks for your support. I tested the RX10700 again in this
evening. (it is mid-night in JA)
1) Gap is 3cm by my son's old Lego blocks.
2) When had a problem, result of your suggested test was NG.
When had no problem, it was OK.
3) Local oscillator testing with the spectrum analyzer
- When had a problem, LO didn't work or maybe worked with
incorrectly X'tal.
- When had a problem, voltage of R4, R5, R6 and R10 were
minus level. And all of X and Y at the U6/U7 were High.
But it seems to me that hardware test in Linrad was OK.
- When had a problem, the X'tal selection occasionally
changed if the coil was touched by my finger.
- Changing normal/problem by D-sub on/off.
- working normally at the start but will go problem and
return to normal suddenly.
I was aware after re-connecting 4 boxed that 143.975,
144.025-075, 125-175, 225-275, 325-375, 425-475 were NG and
others were OK between 143.975 and 144.500 MHz. But all of QRG
are NG now.
I'm very confusing... What is no good in my systems? PC?
tia 73,
Isao / JM1WBB
Leif Asbrink
> 3) Local oscillator testing with the spectrum analyzer
> - When had a problem, LO didn't work or maybe worked with
> incorrectly X'tal.
> - When had a problem, voltage of R4, R5, R6 and R10 were
> minus level. And all of X and Y at the U6/U7 were High.
????????????
When all X and Y on U6 and U7 are high, the voltage on R4, R5,
R6 and R10 must be high too. It is the same test point;-)
Anyway, the same voltage on all is an error. One, and one only
should be negative while the others have to be positive.
> But it seems to me that hardware test in Linrad was OK.
> - When had a problem, the X'tal selection occasionally
> changed if the coil was touched by my finger.
Touching the coil should not affect anything.
> - Changing normal/problem by D-sub on/off.
> - working normally at the start but will go problem and
> return to normal suddenly.
Do you really mean it will return to normal suddenly without
any activity on the computer from your side?
I could imagine a loose connection or a short in the control interface
between the computer and the WSE unit, but it seems unlikely that
the unit would go back to normal operation by itself.
If there is an intermittent error you might find that Linrad sometimes
can send a command sucessfully but that random pulses due to some
short or loose connection could store incorrect data in U3.
If something causes an error on the SELECT line to the RX10700 the data
in U2 will be transferred to U3 and that will likely cause an error.
U2 only has the correct data when Linrad uses the SELECT line to clock
in the new data.
Maybe you have some other software that writes to the parallel port
occasionally? Perhaps the parallel port is in some kind of unusual
mode that is not compatible with the way it is used in Linrad.
> I was aware after re-connecting 4 boxed that 143.975,
> 144.025-075, 125-175, 225-275, 325-375, 425-475 were NG and
> others were OK between 143.975 and 144.500 MHz. But all of QRG
> are NG now.
>
> I'm very confusing... What is no good in my systems? PC?
Connectors, cables or the PC. I think. Can you monitor the voltage
on the RX10700 SELECT line? Does it change by itself? It should
stay permanently low except when Linrad reprograms the RX10700
frequency.
(Just be careful to not send +15V or -15V into the computer
parallel port if you measure on the d-sub. More safe to do
measurements on U4 in the WSE box.)
73
Leif / SM5BSZ
Isao SEKI - JM1WBB
Yes, I must monitor SELECT line for RX10700. I will setup
later.
- LO of RX10700 sometimes has the hiccup without command
from Linrad. I've seen changing frequency or losing in
display on spectrum analyzer.
- Hardware test on Linrad is no problem. HC03 is maybe OK.
- When working stable, no problem with joggling the Dsub
and the cable.
- I'm running terminal window(s) and Linrad only but I do
not know any daemon of the fedora 10 touches parallel
port...
Today RX10700 box works fine so far!
73,
Isao / JM1WBB