Od: Vladimir.Petrzilka [ok1vpz@seznam.cz] Odesláno: 10. ríjna 2001 21:13 Komu: jan.urbig Predmet: DF0MTL interference
Dear Jan
During UHF contest more czech stations had on 1296 MHz heavy
interference from your contest station. It was not any splutters,
but just "broadband noise". In our QTH it was appear as
unmodulated noise across whole 23 cm band in level up to S5, when
our dish was directed anywhere outside to you.
This noise was occured immediately, when your transmitter was
switch, even if your sinal was not modulated. So under my
suspection it look as badly designed transmitter way, where is
too high gain, probably limited only by ALC, or by similar
regulation. We already have similar experience with new TS 2000
from Kenwood, where transmitting route is not designed very well
from this point of view.
Please check this problem, test your equipment again in real
conditions and fix it, to prevent any future disturbance between
contesters.
Thank you for understanding
Vladimir
OK1VPZ
(OK2KKW Contest club)
---------------------------
Od: Jan Urbig
Odesláno: 11. ríjna 2001
Komu: ok1vpz@seznam.cz
Predmet: Re: DF0MTL interference
dear Vladimir, at first congrats to your excellent results in october contest. I apologize for any problems you have had with our signal on 23cm's. I forwarded your mail to our 23cm-OP but have no answer at the moment. He said that no one complained about our signals during the contest so he didn't know about signal problems. To me it seems that the problem was that our power supply line is very lossy (700mtrs up the hill) and so there was low voltage in our QTH (we had several problems with power supplies that didn't work properly).Our Equipment is a FT1000MP with DB6NT tvtr directly form 28 to 1296MHz and then a PA with 8 times the motorola modules. I know these are not the best choice in case of intermodulation but we checked the spectrum at home an it was within the limits. So actually I can't say more than sorry and wait for the results of discussions between us und results of equipment test. Thank you for your patience, 73's de Jan, dl3jan
PS:we used your vusc-program for the first time on both 70,23 and microwave. I think the network function is very useful, but from our point there are 2 weak points: no 9cm included and you have to log a qso if you want to have the direction in degrees for a specific locator. Jan ---------------------
From: tilo.arnhold Sent: Friday, October 12, 2001 To: ok1vpz@seznam.cz Subject: VY SRI fer qrm
Dear Vladimir,
we are very sri that you had problems with our signal on 23cm. Jan has send your email to me because I have built up the 23cm-equipment. During the contest (when I was on 23cm) only one station told us about problems with our signal. It was OK1KZE in JN79FX. We switched the antenna (from the yagi-group east) to the quad-group (west)) and he says, it´s better. So we believed, the problem is, that his RX cannot good handle with strong singnal. After your email I know, the problem must be in my TRX-configuration. We used FT1000MP (IF 10m) > DB6NT-TVTR > Filter > 1x M57762-module > 8x M57762 modules by DL2AM > ANT. One of the 8 modules didn´t gave power-output. That was the result of some measuring - sessions 2 days before the contest. We had not the time to change this module, so we decide to isolate it by 50 Ohm and to work with 80W and not with 100W. I have spend some days and a lot of hours to find the reason for the "low" output. And of course we had the PA-output on the analyser for checking the signal (2 days before the contest). On the analyser the signal seemed to be ok and we decide to use the pa in the contest. (Maybe the signal-quality of the M57762-modules is not good enough for contests in areas with many stations on small distances ?) I don´t know the exactly reason for this noisefloor-problem at the moment, but I will change 1 of the 8 module during the wintertime. And after that I will check the signal by the analyser again to find the reson. Ok ? So, pse don´t worry. In the next year this problem will not be anymore. Again: we are vy sri for the problems in this contest. All the best ! GL in contest es best DX ! vy 73 de Tilo, DH1FM (23cm-Op at DF0MTL / JO60LK)
PS. Just 2 questions: Were there differences between cw and ssb or was the noisefloor in ssb and cw ? You wrote more czech station had problems. Thats not vy exact. Who are the stations, I should say sri ? ---------------------------------------
From: "Steffen Koehler" To: <ok1dfc@tesmail.cz>; Sent: Monday, October 15, 2001 Subject: TS870 Noise
Hi Zdenek, tnx for your TS870 noise report. We know about the problem of wide band noise, but we think it is not that easy to overcome with a simple solution. Problem with noise is evidently thanks to TS870 because this TRX has wide band oscillator and only DSP selectivity. Thanks to switching LO (each 100kHz) in FT1000 we do not have problem between station in OK and stations witch are very closer. But problem is, when TRX TS870 produce wide noise in LO, this effect must be on RX and TX some. Allthough you might think that the TS870 TX/RX concept does not perform well compared to the FT1000MP, it is somehow different to your description. The Problem is in fact the noise from the VFO for converting the 8.83 MHz last IF to the final frequency, which is - as in the FT1000MP - a hybrid solution of a switched PLL combined with a DDS circuit. The LO's are also implemented variable, but there is additional selectivity by crystal/ceramic filters on 8.83 MHz and 455 KHz IF respectively, not only DSP selectivity. Since you can hardly (and would also probably not plan to) implement an additional tunable narrow filter in the 28 MHz range, reducing phase noise of the VFO is the only way with significant influence on wide band interference. Producing an excellent phase noise of -148 dBc is simply not reachable for us, we are unfortunately unable to measure it too. The TS870 has a carrier subpression of approx. 80 dB, but at -148 dBc noise is again more than 60 dB below, which can't be measured with our best spectrum analyzer. We can potentially reach about -120 dBc, that seems to be generally acceptable. Also, the noise problem of small signal bipolar transistor stages seems to appear significantly in the TX path of the transceiver and the transverter. For as low noise levels as -148 dBc, we would have to set up a pure MOS design, which is far beyond our budget. If you have found any good solution for VFO/DDS improvements that might be adaptable to the TS870 - please let me know. Regards Steve, DH1DM - contest crew DF0MTL - ------------------------------------
Od: Vladimir Petrzilka [ok1vpz@seznam.cz] Odesláno: 22. ríjna 2001 3:41 Komu: tilo.arnhold Predmet: RE: VY SRI fer qrm
Dear all
Just I got copy of e-mail (in annex) from Zdenek OK1DFC, which created some suspection about observed noise floor discused few weeks ago. Really on your side FT 1000 Mp was used for 23 cm and not TS870 ? TS 870 is very well known as bad designed TRX, produce too much noise - it is not matter of sideband noise of local oscillator (but as well), but rather as bad designed TX chain, where very low signal after pass through filter and mixer filter is amplificated. And of course first stage has quite high broadband noise, which probably generate a noise floor, indicated by us on the band. This nose floor is of course irrelevant on HF(because of dynamic), but very substantial on VHF/UHF. From that perspective, just superior HF transceivers are usable on VHF/UHF bands. Just very short calculation:
You were 9 km from us - let's say 10 km Your had aprox.100 W You had let's say 20 dB antenna gain
How strong signal was at our receiver?
Let us use simple (but quite precise) attenuation formula:
Au = 20 × ( - 4.622 - log f - log L)
f is frequency in GHz L distance in km
for our example it is: Au = 20. ( -4.622 – 0.11 – 1)) = -114 dB
Radiated power: 100 x 100 = 10 000 W ERP = 40 dBW = + 70 dBm
In case, that we were not turned to you our antenna had let's say 0 dB gain.
From that we can calculated approximate PWR on our receiver input:
70 - 114 = - 44 dBm
For VHF / UHF transceivers S9 level is defined as - 93 dBm. So in that case, your sigs were 50 dB above S9. (In case of OK1KZE observation, when they were 100 km away, it was almost exactly the same level, but, of course, if both of your both 20 db antennas were directed to each other).
If your broaband noise was (generated by TS 870 ?) let's say on the same level as sideband noise of L.O. at 20 kHz distance (TS 870 has as measured approx. -120 dBc/Hz), we received this noise through 2.7 kHz wide SSB filter as 2.7 x 1000 more intensive so 34 dB stronger => -120 + 34 = 86 dB bellow carrer, so bellow S9 + 50 dB => it is approximately at 36 dB bellow S9 => S3 !!! It is very close to our observation.
Conclusion: to prevent interference, your spectral purity should be minimaly 20 dB better, then it was. It mean FT 1000 as low level sideband noise generator is principal condition, but as low as possible gain after FT 1000 transvertor output is necessary as well.. Transvertor output of FT 1000 MP has approx. - 6dBm, and so up to 100 W PWR you don't need more, than 60 dB overall gain from - 6 to + 50 dBm) with some 4 dB reserve for interconnection and regulation. I believe, that if you will use that configuration, no problems will appear in practical operation any more .
So my suspection is that : either not FT 1000 MP, but TS 870 was used for 23 cm on your side, or much higher gain was adjusted in the 23 cm TX chain (in case of disconnected ALC loop, if used).
No differency in noise floor were observed in case of CW or SSB operation - just noise, which had not change in relation of your modulation - so it take us to evidence, that not phase sideband noise, but noise of bad designed TX chain was observed. But of course, even if it is unprobable, maybe DDS synthesizer of TS 870 could produce broadband noise as well, who knows...
Measurment of broadband noise is not problem, but it is not task for spectrum analyzer, because analyzer has ever much worse sideband noise compare to any transceiver. For test kindly use dummy load, from which sample of RF signal is going through very sharp rejection filter (help to attenuate carrier) into measurment receiver without AGC and with excelent local oscillator spectral purity. We use for those tests another transceiver with very good stopband (two crystal filters on different IF's), crystal as receiver local oscillator and blocked AGC. On the AF output of receiver is for observation connected AF milivoltmeter with dB scale. For calibration of this measurment system simple noise generator with known output level is used.
I hope, that those ideas could help you to prevent similar problems within next contest.
Best regards
Vladimir OK2KKW team ------------------------
Od: Steffen Koehler Odesláno: 22. ríjna 2001 Komu: ok1vpz@seznam.cz; Predmet: TX noise
Hi Vladimir,
thank you for your interesting e-mail. Yes, we know about RF wave distance attentuation difference between SW and VHF/UHF/SHF and yes, it is true that 3...6 dB performance improvements are not enough. Of cause, we can distinguish between a TS870S and a FT1000MP, they even look different :-). The reason for our FT1000MP driven "bad signal" on 23cm can be easely seen below. As you will point out, the FT1000MP has far not 20dB better TX noise than other transceivers. But ok, the FT1000MP's RX dynamic range is significantly better.
vy 73's de Steve, DH1DM
------------------------------------------------------------- The March 2000 issue of RadCom has an in depth technical review of the new ICOM IC756 PRO. I have summarized three key technical measurements on a variety of transceivers, all made by the same reviewer, Peter Hart, using the same methodology. The three measurements are the two-tone dynamic range (in a SSB bandwidth), a reciprocal mixing measure, and the amount of transmitter broadband noise (also in a SSB bandwidth). The rigs are listed with the best at the top in terms of close-in performance. The IC756 PRO ranks near the bottom.
Spacing 3k 5k 10k 15k 20k 30k
Corsair I - November 1984 - no synthesizers DR - 90 90 90 90 90 Recip - 98 98 98 98 98 TX noise 95 95 95 95 95 95
OMNI-VI - January 1994 DR 88 88 88 88 88 88 Recip 86 86 - 96 96 96 TX noise 83 87 96 97 98 98
FT1000MP - January 1996 DR 78 82 92 103 99 98 Recip 80 86 94 99 103 108 TX noise 76 81 89 94 96 99
IC737 - September 1993 DR 74 80 88 92 94 98 Recip 90 96 104 105 107 110 TX noise 79 87 93 95 96 97
FT990 - April 1992 DR 70 78 88 92 94 96 Recip 83 87 93 97 100 105 TX noise 74 79 84 86 87 88
TS690S - November 1992 DR 67 77 87 91 93 94 Recip - - - 98 101 105 TX noise 70 77 82 85 88 88
TS930 - May 1986 DR - 77 87 - 91 92 Recip 77 83 89 - 97 101 TX noise - 79 83 - 85 88
TS940 (with Lowe modification to improve reciprocal mixing by 12db)
- May 1986
DR - 76 81 - 88 91
Recip 83 89 98 - 105 110
TX noise - 72 72 - 72 74
TS850 - October 1991 DR - 75 89 95 95 96 Recip 85 89 100 105 108 112 TX noise 75 83 90 95 96 99
IC756 - May 1997 DR 74 74 75 79 82 86 Recip 79 84 91 95 98 102 TX noise 73 80 87 91 94 97
IC756 PRO - March 2000 DR 72 73 76 82 86 90 Recip 83 86 92 96 98 101 TX noise - - - - - -
IC775DSP - January 1996 DR 72 72 78 87 91 93 Recip 90 96 100 103 106 109 TX noise 78 80 87 92 97 100
FT920 - August 1997 DR 70 72 83 95 97 96 Recip 85 90 97 101 104 108 TX noise - - - - - -
TS50S - May 1993 DR 67 69 74 80 84 92 Recip 78 84 95 - 103 104 TX noise 74 79 85 90 93 97
FT847 - August 1998 DR 65 67 78 89 96 94 Recip 76 80 90 94 97 101 TX noise 70 76 86 90 93 97
FT1000 - June 1991 DR 65 65 72 82 88 93 Recip 81 85 91 95 98 102 TX noise - 76 81 - 84 -
TS870 - April 1996 DR 58 61 73 84 94 95 Recip - - - 100 103 107 TX noise 75 80 88 91 92 93
IC781 - July 1990 DR 52 57 76 92 97 - Recip 90 96 103 106 110 113 TX noise - 85 92 - 98 -
Notes: DR - two-tone dynamic range on 7 MHz in SSB bandwidth Recip - reciprocal mixing for 3dB increase in receiver noise TX noise- dBC in 2.5kHz bandwidth
spacings refers to spacing of the two tones for dynamic range and the offset from the carrier frequency for reciprocal mixing and TX noise
Dates refer to issues of RadCom containing the Peter Hart review.
Rigs listed in order of decreasing dynamic range with 5 kHz spacing. -- CQ-Contest on WWW: http://www.contesting.com/_cq-contest/ ---------------------------------
Od: Fred Siegmund Odesláno: 22. ríjna 2001 Komu: ok1vpz Predmet: Re: VY SRI fer qrm
Gentleman,
just to clear up the things a bit, a few words: The noise problem of 23cm and 2m has got nothing to do with each other. We are using a TS870 + transverter(10m IF) on 2m and FT1000MP with transverter (10m IF) on 23cm. The problem of 2m was discussed by DH1DM sufficently. I dont think that a -148dBc LO in the transverter solves the problem(the following stages produce a higher noise for the total), as well as we can not afford to buy a FT1000MP for every band. Additionally i can say that also the phase noise of same SW Transceivers differs especially under TS850 TRX's(we used one before on 2) there are significant differences. We have to think about a solution.
On 23cm we had a fault in the TX chain, which has to be found yet. But we saw the noise floor already on the analyzer, i hope we can solve this problem easier.
73 Fred DH5FS
-----------------------------------