Some facts, tips and tricks for UHF Meteor Scatter [2008 & revised in 2022]

Written by OK1TEH


The first tests of 70cm MS were hold by G3HBW during early 60's. During 1967 K4QIF has heard first reflections from W4GJO on 70cm. On August 9 1968, after two months of daily schedules, Mel Baer, W6WSQ, and Don Hilliard, W0EYE, completed the first 220 MHz contact. Further tests on UHF continued and the first ever 70cm MS QSO was finished by W2AZL, Holmdel, New Jersey, and W0LER, Minneapolis on 12th August 1972 and it required 5 hours and 4 minutes to complete. First 70cm MS QSO in Europe was completed exactly 5 years later on 12th August 1977 between SK6AB and SM2AID - more about these  tests can be seen at  ok2kkw web.

The biggest positive step forward in 70's was the new LNA generation based on gallium arsenide field-effect transistors (GaAsFET) which allowed suppress the Noise Figure from 2 - 3dB well below 0,5dB which was quite remarkable especially on UHF, where the sky background temperature is very low.

During end of 70's further MS tests were rolled by Chris GW4DGU, Karl, SM3AKW, Claus, DL7QY and ON5FF. Probably the second European 70cm QSO was established between DL7QY and ON5FF/EI in late 70's and from UK in 1982 between G4VVZ/G4ZAP and EI2VPS (members: G8ROU, G8XVJ, G4VVZ, G8VHI, G6ABU and EI5FK). In OK the first 70cm MS QSO was established already in year 1986 on 11th August by Josef, OK1MDK & OK1KKH club with Carl, SM3AKW during Perseids shower. The contact took to finish about 3 hours + 15 minutes and OK1MDK heard 4 bursts (the best was 6s long) and about 16 pings. The over crossed distance was about 1400km and it was the new World record. SM3AKW used at that time his 16x21el F9FT antenna for EME and OK1KKH PA with RE025XA tube and 2x21el F9FT.

In 90's more contacts had been established (check out pa3dzl_1990_ms70cm.htm). In 1992 some (unsuccessful) Meteor Scatter tests were held even on 902 MHz and in year 1999 SM3AKW JP92 made the HSCW contact with well known UA9FAD LO88 on distance of 2141km, which is still the HSCW World record - Even 70cm MS SSB contacts were reported in 1999 thank to great Leonids outbursts with up to 1500 ZHR (compare to 70 ZHR in Quadrantids 2022). The next Leonids outburst we can expect on 17th November 2031 - 2033.

The 23cm Meteor scatter contact wasn't completed yet but on 4th January 2011 (Quadrantids) OK1DFC successfully decoded in FSK441 RW7A with 280ms ping / 2dB. OK1DFC used 10m dish and RW7A 3,7m dish + 300W PA, both stations used circular polarization and QRB was 1921km.

MS after FSK441

The revolution in DX operation was introduction of WSJT by K1JT in 2001. The new FSK441(A) mode with data speed of 8820 characters/minute made easy everyday 2m MS contacts during sporadic meteors.

Lately K1JT made more versions of FSK441, B and C. The difference between the versions is the build-in error correction in the B and C versions. The consequence of this error correction is however that the actual data speed decreases substantially. This data speed (8820 characters/minute for FSK441A) drops by approximately 57 percent (to 3780 characters/minute) when using FSK441C. This is the main reason why FSK441A is the default mode in IARU Region-1. So for 70cm MS is much better classical FSK441A however if you suffer during 70cm test from decoding problems, this would help you.

After few years of successful operation with FSK441 Joe, K1JT developed new experimental mode JTMS, which was found very usable for decoding of shorter underdense pings on 432 MHz. If you want play with JTMS, it's part of WSJT 10, you just have to create empty file in the WSJT folder called "experimental" and JTMS will appear in the WSJT main menu.

Later during 2017 K1JT introduced new software called WSJT-X, which removed FSK441 and newly included new MSK144 mode. This mode became very popular on 2m because it was redesigned to have 100% decoding without the FSK441's "trash" (similar to old PSK2 from DJ5HG) and more it removed classical meteor-scatter report system which was used since 50's (only strength in dB isn't good ideafor MS because it's only the matter of AGC settings in TRX) so it became very similar to FT8 mode and so it was possible to even use it via WSJT-Z with full automatic-mode The biggest problem of MSK144 if we compare it to FSK441 is the decoding problem with shorter pings below 100ms. Such a reflections are very common on 70cm and more MSK144 has problem with Doppler shift of head-echoes (special type of MS reflections on UHF) so FSK441 or JTMS is still preferred and recommended. The big part of VHF/UHF DXers now prefer new SW called MSHV from LZ2HV , which contains MSK as well as FSK.

From my point of view the MSK144 is step backwards for 70cm MS because especially Russian stations stopped to use ON4KST and such a chats rooms with English traffic so now it's much more difficult to find sked-partner for 70cm, where random contact is almost impossible. At the meantime Russians conducted new special VHF/UHF IRC Chat, which can be access without any registration at IRC.VHFDX.RU with chat-room #VHFDX, you just have to use some IRC client. If you want to write there in Russian language, you my try Deepl translator which is far better than Translator from Google.


Typical underdense ping on UHF (about 100ms long)

Reflections duration on 70cm is much lower then on 2m. During bigger showers such a Perseids isn't problem to observe on 2m bursts longer then 30seconds. In the same time the same meteor can produce on 70cm overdense reflection about 500ms long. So good time to try 70cm MS is just during major MS Showers - Quadrantids, η-Aquarids, Perseids and Geminids. Good information source is web of International Meteor Organization at and Meteor Sky View Virgo when you can see the actual position of radiant and where to beam your array (best condx are always 90deg off to radiant's position in Azimuth). It's said that the best MS shower for 70cm MS are Leonids because of it's high meteorid speed when it's entering the atmosphere however we must wait for next major maximum around year 2031. In general the best MS shower for 70cm needs 2 components - as big as possible entry-speed in atmosphere for the maximal ionizations and one the other hand good density of dust particles. As well at some cases ionizations depends on material as for an example Geminids particles are full of snow while for an example Perseids are iron-based particles which are able to survive longer life. By the way the MS reflectivity is also depended on average height of the atmosphere. During Solar maximum Earth's atmosphere is going to expand a bit which for an example is shorting life of some satellites at Earth's low orbit however also making possible of MS contacts on longer distance.

So how big is difference between the entry-speed of meteoroids in atmosphere? It can be seen at:

Shower ZHR V [km/s]
Leonids 15 var! 71
eta-Aquarids 50 var! 66
Orionids 20 66
Perseids 60 59
Lyrids 18 49
Quadrantids 110 41
delta-Aquarids 25 41
Geminids 150 35
Draconids 10 20

As you can see, Draconids are really slow and it was really confirmed that Draconids aren't so good on 144 MHz while on 50 MHz it's quite a popular MS shower. In fact Draconids was the first ever MS shower observed by the hamradio operators (on 6m) in 1946. From the perspective of December 2022, the Quadrantids and Perseids are getting weaker year by year and the best MS shower despite of it's low speed seems to be Geminids.  It's also should be noted that slower MS showers are able to making MS QSOs on shorter distance as the meteoroid starts to burning in lower altitude.


And how it's with radiant height above horizon? Due to theory the strongest reflections are observed when the radiant is at 0 degrees of elevation however in such a position most of meteors thank to bad entry angle are reflected from the atmosphere back into space so in low height reflections are strong but very rare. On the other hand when the radiant is in close to zenith, the numbers of meteors is high however reflections are minimal. Therefore the best position of radiant for the best compromise between strong reflections and their amount lay in height of about 45.

Usual reflections (or better said scatter) on 70cm are observed to be about 100ms long, the are called underdense (forget to try classic SSB as it's common on 144.0 MHz) and about 2-4 dB above noise level however few seconds bursts with strength over 20dB were observed. Some more info about theory of Meteor Scatter can be found at IMO's web:

So far the longest known reflection on 70cm was observed by GW4DGU from SM3AKW in year 2004 and it was 35seconds (they used HSCW). During my MS contact with UT5JCW he heard my signal about 20dB! over noise however this is extremely rare. With 1 or 2dB usual strong bursts over noise you need as good RX as possible and it's similar to 70cm EME. Personally I found MS on 70cm a bit more difficult then 70cm EME due to smaller activity and just few good windows per year. However don't be scared by this as 70cm MS is possible with your usual tropo equipment as well. 2m the ration of reflections is very low and during very good MS conditions aren't unusual contacts over 1h long. So the most important for such a 70cm MS test is very high level of patience. So far IARU R1 VHF manager handbook doesn't define the maximum length of skeds however common agreement is not going over 4 hours.

The one of the biggest problem of 70cm MS is the correct frequency (and that newcomers don't read the operation examples). FSK441 allow just the bandwidth of 400Hz and lot of usual 2m MS stations don't care about they precise frequency at all. However if you have on 2m transceiver with frequency indication 200Hz off, on 70cm it's 3 times more = 600Hz. Then, as I observed, isn't unusual situation that after 30minutes of your sked you get the first reflection, but you are not able to decode it as it is for example 800Hz off and you become little bit unhappy :-). In any case is good idea to use headsets during sked for ears monitoring, because then you can simply copy signals which are hundreds of Hz off. Another problem is due to very weak reflections clear band without QRM so sked you should set for late evening or early morning.

For further info regarding the MS traffic and when the QSO is finished I recommend to read excellent presentation from Bo, OZ2M ->

Software + hardware

Don't forget to switch off your Noise Blanker in your transceiver and if you can't turn-off AGC, turn on fast AGC. If you use external LNA, don't forget to turn off your internal preamplifier of your radio. Two preamplifiers in your RX chain can dramatically make your RX more sensitive to local QRM. It's a big problem in Czech Republic due to local CDMA provider "Ufon" with OFDM modulation just few hundreds KHz below 70cm DX segment. Another story is to use N-connectors with minimal attenuation because bad and cheap Chinese connector can kill your RX for more then 1dB (or more if you use more connectors of course. Pse don't even think about of PL-259 called "UHF" connectors! More at DJ3JJ's web).

I have been running MS in FSK441 for over last 5 years and I found that is very helpful to use for 70cm MS WSJT versions 4.34 and 5.96 together. In old 4 version you can set up QRN level and minimal ping duration so you have bigger chance to decode extreme weak pings then in new WSJT 6/7. From family of WSJT 4 versions I found the best version 4.34 as it's very fast with low requirements on your PC hardware. On the other hand WSJT6 allow set your Tol to 600Hz and it consist of SpecJT "waterfall". I use for TXing WSJT 4.34 and WSJT 6 is set up by "F3". Receiving of two different versions on one PC and one sound card helped me to decode plenty of normal undecodable reflections. These older versions you can download here.

For 70cm MS you should set your S = 1 (or better S = 0), QRN = 2 and maximal Tol 600 or 400Hz. If you find your partners signal and he hasn't heard you yet, tune on his signal for minimal DF. If he heard you and send you report, tune on his signal only with RIT. When you tune for minimal DF, stay on this frequency and set up your TOL = 100 or 50Hz and set your S = 0 + QRN = 0. In any case if you have problem to decode the signal correctly click on the signal in SpecJT or Spectrogram window. Also is good idea to do it, if WSJT decode automatically the bursts for example "OK1TEH RA3XX R2". If you click on the edge of burst in window you can decode the bursts in different way and WSJT then decode "A3XX R26 R26 OK".

The segment for 70cm skeds it's similar to 2m 432.370 MHz however you can run sked +- several kHz of course.

Usual 70cm MS station consist of about 300W and long yagi (33el), but with bigger patience you can work 70cm MS contact only with about 50 - 80W and 10el yagi (with BIG GUN station of course). For quite good chance is good to use 4xLY and 1kW but pay attention that for MS you need wide array beam characteristic. The ideal antenna for MS are vertically stacked yagies so the main lobe is wide in horizontal and reduced in vertical plane. For shorter test on 1000km+- distance is good to use elevation or shorter yagi. The extreme was my 70cm MS contact with IK0BZY. Enrico used just 9el yagi! and 400W. I don't think that it was aircrafts scatter as distance was 924km. Aircrafts signals on 70cm are against MS much longer and stronger (can be 59+20dB). By the way don't forget to pay attention to your transceiver frequency stability (QRH), one of solution which I use in my FT847 for EME you can see on next picture. The main problem of frequency stability of FT847 on 70cm is cold air from fan, which start blowing cold air after switched PTT on Xtal, which is getting cold. Solution is simple, just put on the Xtal some kind of polystyrene foam as it's shown on the following picture.

More photos of OK1TEH's 70cm equipment (PA, LNA etc) are at At last don't forget, that WSJT is for your PA like a FM so it must be capable to survive very long time with full power out (often much longer then for EME QSO on)! If you would like to read some information about my solid state PA, check out

Doppler shift

If you try contact on the big distance like 1500km you don't need to care about the Doppler shift at all. For example: During this Perseids I got very strong bursts from RK3WWF and DF of 99% of all bursts wasn't shifted more then 200Hz off. For shorter contacts (about 900km) it's a bit different and I'm thinking that some help could be use of SDR and some kind of MAP65 JT65 software for FSK441. It can be that some bursts (mainly head-echo type) can be shifted out of usual 2,4kHz SSB filter and you don't hear them. I'm thinking this because of comparison of 2m and 70cm during big shower: For example if you receive 30s bursts from Sweden in rate of 20x per 1 hour why do you RX only 2 reflections on 70cm? Where are that remained 18 bursts? Few KHz off? Or is it just the problem of different antennas? Unfortunately I don't have at home SDR radio so I can't check it. Please remember that this is just my "stupid" theory but if you have an idea about it I'm very interested in your opinion.

Note from December 2012: after extensive testing with FunCube SDR with 70cm MS on very short distances (TNX OK2POI) no reflections out of SSB filter were observed except one strange reflection however it was identified to be reflection not from an meteor but from satellite!


When to set skeds for 70cm during big MS showers?

Geminids in JO70FD          Quadrantids in JO70FD             Perseids in JO70FD

00:00 138 228 48 68 EL EA2   00:00 42 132 312 25 EL YO, G   00:00 39 129 309 35 EL  
01:00 174 264 84 72 EL F6, UA6   01:00 50 140 320 31 EL LZ, GM   01:00 50 140 320 49 EL LZ, GM
02:00 214 304 124 70 EL US7, G   02:00 57 147 327 40 EL LZ, GM   02:00 54 144 324 57 EL  
03:00 241 331 151 63 EL LZ, OY   03:00 65 155 335 48 EL LZ, OY   03:00 55 145 325 65 EL  
04:00 258 348 168 54 EL IK7   04:00 71 161 341 57 EL Z35, LA   04:00 53 143 323 73 EL  
05:00 270 0 180 44 EL SM4   05:00 78 168 348 66 EL IK7   05:00 35 125 305 80 EL  
06:00 281 11 191 35 EL SM5, IK0   06:00 86 176 356 76 EL IK7, SM3   06:00 345 75 255 82 EL  
07:00 291 21 201 25 EL OH, EA6   07:00 101 191 11 85 EL     07:00 312 42 222 76 EL  
08:00 301 31 211 17 EL OH, YL, ES   08:00 260 350 170 85 EL     08:00 305 35 215 69 EL  
09:00 311 41 221 9 EL UA3   09:00 274 4 184 75 EL     09:00 304 34 214 61 EL  
10:00 322 52 232 2 EL UA3   10:00 282 12 192 66 EL     10:00 307 37 217 53 EL  
11:00 333 63 243 -3     11:00 289 19 199 56 EL OH, EA6   11:00 312 42 222 46 EL  
12:00 346 76 256 -6     12:00 296 26 206 47 EL OH, YL, ES   12:00 317 47 227 39 EL  
13:00 359 89 269 -7     13:00 303 33 213 39 EL UA1, YL, EA6   13:00 323 53 233 33 EL  
14:00 11 101 281 -6     14:00 310 40 220 31 EL UA3, EA6   14:00 330 60 240 27 EL  
15:00 23 113 293 -3     15:00 319 49 229 24 EL UA3, EA5   15:00 338 68 248 23 EL  
16:00 36 126 306 1 EL YO, G   16:00 327 57 237 18 EL UA3, EA5   16:00 346 76 256 20 EL  
17:00 46 136 316 8 EL LZ, GM   17:00 336 66 246 14 EL UA4, EA2   17:00 354 84 264 19 EL  
18:00 57 147 327 15 EL SV2   18:00 345 75 255 11 EL UA4, F   18:00 2 92 272 18 EL  
19:00 67 157 337 24 EL LZ, OY   19:00 356 86 266 9 EL UA4, F   19:00 11 101 281 19 EL  
20:00 77 167 347 33 EL Z35, LA   20:00 6 96 276 9 EL UA6, G, F   20:00 19 109 289 22 EL  
21:00 87 177 357 42 EL IK7, SM3   21:00 15 105 285 11 EL UA6, G   21:00 26 116 296 25 EL  
22:00 100 190 10 52 EL IK0, OH0   22:00 25 115 295 15 EL UA6, G   22:00 34 124 304 30 EL  
23:00 115 205 25 61 EL UA3, EA6   23:00 34 124 304 19 EL YO, G   23:00 40 130 310 36 EL  
00:00 138 228 48 68 EL EA2   00:00 42 132 312 25 EL YO, GM   00:00 45 135 315 42 EL  

At the tables you can see when is radiant below horizon or too high and when is +- good time for skeds. For an example optimal time during Geminids for QSO into IK0BZY is around 06utc or around 21utc, because radiant is in 45deg of elevation and radiant aboit 90deg off to QSO path.

Alternative radiant's position

Everybody knows Virgo with it's nice AZ map. However sometimes there is problem to get running Java scripts. Good and simple solution is to use track radiant by simple using of good old WSJT and it's Astronomical data.

Note that RA is given in decimal hours. Some data needs to be converted into that format. The radiant favors communication paths that are perpendicular (+/- 90 off) to its position. On the example you can see that radiant is just 7,96deg above horizon and it's AZ is 47,05deg so for good MS QSO you should look 90deg off so for AZ of about 137 deg or 317deg.

RA/DEC for some major showers:

Quadrantids RA= 15.3 hrs DEC= 49

Perseids      RA=   3.1 hrs DEC= 58

Leonids       RA= 10.2 hrs DEC= 22

Geminids     RA=  7.5 hrs DEC= 32.6

(source: )

Active stations

During each big shower is common to work 1 or 2 full contacts in the whole Europe and number of active station is very very low. Some of these stations you can find in the following table. All of them should be QRV in FSK441:

Call      Grid   Antenna  Power    e-mail (ATT = @)
DK3WG     JO72GI  16x27el  1000W   dk3wg<ATT>
G4RGK     IO91ON   8x27el   400W   g4rgk<ATT>  
OH3AWW    KP11TL     1xLY     ?W   oh3aww<ATT> 
OK1DFC    JN79GW  8m DISH  1500W   ok1dfc<ATT>   
OK1TEH    JO70FD   1x23el   500W   ok1teh<ATT>   
PE1ITR    JO21QK   2x28el  1500W   rob<ATT>   
RK3WWF    KO72QU   4x26el  1000W   rk3wwf<ATT>
SM2CEW    KP15CR  8m dish    QRO   sm2cew<ATT> 
SM2A      K004NP   4x32el   800W   sthe<ATT>  
SM3BEI    JP81NG   4xLY      QRO   sm3bei<ATT>
SM3JQU    JP82QM   4x32el   750W   perolof.sjlander<ATT>
SM6FHZ    JO57XH   4xLY      QRO 

and maybe DK3BU, DJ8MS, PA2M, S50C + S53RM, ..

The list looks like big one but during shower are QRV just about 5 stations however future looks promising with increasing activity in Russia, Balkan and East countries.

best 73 & cu on UHF
Matej, OK1TEH -



Opinion of 70cm MS operators & article readers

Od: Christopher Bartram <>
Předmět: 432 MS
Datum: 9.1. 2009, 12:40

Hello Matej

Excuse my quick short reply, but I have heavy QRL at the moment!!

Here are some clips from a couple of emails I sent to Joe, DL8HCZ, after his/your note in DUBUS.
>I certainly wasn't one of the first to try 432MHz MS in the UK, although as G4DGU (IO91) I think I was one of the first to get close to making a QSO. Johnny Stace, G3CCH, now SK, was trying at least ten years before Karl, SM3AKW, and I ran our skeds. From memory, Johnny heard a few bursts, but speaking to him about it when I was planning to try, he was very doubtful that a QSO would ever be made! I think Peter Blair, G3LTF and others also ran tests a decade or more before I started to try.

>My skeds with Karl in the 1978 Perseids were almost successful. My log reads:

>12/8 0900 1410 SM3AKW 432.117MHz 36 26 4b 3p longest burst 10s NC
>13/8 0600 1300 SM3AKW 432.117MHz 26 26 4b NC

>My rig at that time was a single G3JVL loop yagi (6m boom), I was running 1kW output under a special licence, and the preamp used a prototype Microwave Associates bipolar transistor - this was just before the era of GaAsFETS! The noise figure measured using physically hot and cold loads was about 1.1dB.

>Unfortunately, at least for 432MS, the next three years Perseids were taken up with transatlantic tests on 144. That's another story which needs telling! If anyone is interested, there are reports in the RSGB magazine, Radcom, at that time.

>QRL and unsuitable QTHs kept me away from 432MS until I tried again in 2004 as GW4DGU.

>Running a 550W SSPA with 400W at the antenna, a pair of home-designed vertically stacked 15.7WL boom long yagis with about 20dBi gain, and a 0.5dB HEMT preamp, I worked from IO71SV:

>12/8 OK1DFC  JN79gw 1346km WSJT
>12/8 SM3AKW JP92ao 1786km HSCW
>12/8 S53RM    JN78hd  1525km WSJT
>12/12 DF5JJ    JO43xo   987km WSJT

>The QSO with DF5JJ was definitely by MS as the bursts had the characteristic (underdense burst) doppler shifts of 432MS.

>I have wanted to try in subsequent years, but both QRL and equipment problems have conspired against me!
>Who nowadays would be prepared to run a six hour sked?!

>> I was quite impressed by the length of some of the observed bursts on 432. 5sec, 10sec, 15 sec even up to 20 sec was reported in last years Perseids. Sometimes thought if such a 20sec burst would be good for a ping on 1296 MHz.....

>After working SM3AKW in 2004, I thought of suggesting a sked, but my 1296MHz equipment wasn't good enough! I may decide to get on 1.3GHz properly once my
24GHz EME system is working :-)

>I looked again through my old log books yesterday evening, and I see that I
>had another incomplete QSO with SM3AKW in the Quadrantids of 1979. I was also
>QRV from IO70 in the 1982 Perseids, and again, although I heard bursts from a
>few stations, I made no QSO...


I wrote-up my results with Karl in Radcom at the time (late 1979) and I think I also reported them in DUBUS, but I don't have copies. I seem to remember giving a talk at the RSGB VHF convention in 1979.

The physics of 432MHz MS isn't fundamentally different from that at 144. The major difference is that electron density in trails is usually not great enough to support 'overdense' reflection, although the long bursts observed demonstrate that there such conditions do exist, particularly in major showers.

It seems to me that most bursts are effectively 'underdense' refraction. Although underdense bursts do have doppler shift within the burst, caused by movement of the ionisation, I don't see a mechanism for usable bursts being doppler shifted by several kHz, except maybe in a transient manner. I'd be happy to be proved wrong!

A QSO on 1296 might just be possible in a major shower - eg Leonids. It might be difficult to get people to try, though!

It would be worth you reading the professional literature on meteor burst communications. I'm not up to date with that, but a quick google brought-up a reasonably recent book by someone called Schilling.

With modern technology 432MS is much more practicable than it used to be. Thanks for having the enthusiasm to start the ball rolling once again.

Vy 73


Od: Klaus Dreckshage <>
Předmět: AW: MS on 70/23cm
Datum: 9.1. 2009, 17:49

Hello Matej,

nice to here anything from here vie Mail. In the last 2-Yeas I made no QSOs via MS on 70cm, no Time and iam now active on short Wave.

On my QSOs I use only Gemenids and Perseids, they have more Rocks as normaly. The longest Reflektion I have, was 12sec in the QSO with SM2CEW. The max Doppler was 250Hz I can decode. Sometimes I have some more Doppler, but can't say it is from this Station I try to work. Most of time I try to work with 2 or 3 Station on same time.

On 23cm I never try a QSO. Also I don't know there is a article in DUBUS or other for MS on 70cm. There are not many Station's QRV on 70cm MS and have much time for this. Mostly they will work more on 2m sure QSO. I hope for more time in future for 70cm MS. If I have more Information, I give it to you.

Vy 73 de DL3YEE

Od: Paul <>
Kopie: ,
Předmět: Re: MS on 70/23cm
Datum: 10.1. 2009, 00:45

ok1teh Petrzilka wrote:
> Hello all
> I'd like to ask you about your experience with 70cm MS

a small number of successful ms qso's using both jt and hscw.

My take on this is that hscw is more successful as decodes by ear can cope with 'df' and Doppler / changes in level better than the automated Jt modes *but* its hard to sustain concentration for a couple of hours - conversely where JT modes are concerned you can be a little more relaxed when operating.

I agree Doppler seems more noticeable on short distance refl. Duration of refl, I cant quantify the difference between 50MHz,144MHz and 432MHz in numbers as I haven't enough data - but random reflections from 390MHz tetra carriers in PA are VERY few and far between - and these are very accurate signal sources (level/frequency) to use for such experiments from here.

My last test with DL8MS was a 2 hour qso with just about three or four pings to complete, interesting as he was running quite low power and I was running qro and a fairly big antenna, the duration and strength of pings at both ends was very similar, qro didn't seem to help much.

73 de Paul GW8IZR IO73TI

Od: Sine MERMAL - S53RM <>
Předmět: Re: MS on 70/23cm
Datum: 12.1. 2009, 11:21

Hello Matej.

I am very sorry and disappointed because I carefully click on SAVE every time when burst came on 70cm and on 2m, but nothing was saved on 2m PC. I discovered this on the evening in Saturday. My father has WSJT 5.9.6 on his laptop. Is there a problem with SAVE function in that version? Nothing was saved at all. So I can only comment on what I have noticed listening and looking simultaneously on both PC.

My father used 60W power on 2 m, so QSO with RA3WDK (same direction as RW3WR and 100km further) take more time to complete and I could compare 3 reflections. As expected, I noticed no reflections on 70cm outside time window when 2m reflections appears. Only strong enough reflections on 2m (in my case at least 15dB over noise) produce reflection on 70cm. We received, for example, around 15s burst on 2m and at its peak we receive 300ms burst on 70cm. There were many bursts on 2m that produced nothing on 70cm.

I did not noticed any significant Doppler shift during RW3WR sked (and other QSO's I made) on 70cm. Once on the right frequency, one can leave tuning knob as it is. So low number of reflections on 70cm is not caused because not hearing reflections that are outside receiver passband. We do not hear them because traces are not ionized enough to produce reflection also on 70cm. Wind speed at 100km altitude is up to 200km/h and ionized traces traveling at the same speed so Doppler shift is therefore below 100 Hz on 70cm. High Doppler shift during burst I noticed only at the beginning of the strong burst (on 2m and I also heard the same one time on 70cm), when ionized cloud is forming when dust particle entering earth atmosphere.

The longest burst I ever heard on 70cm was from RW3WR and it was 3s long. It was -500Hz off my frequency. This is practically the same amount I find for him later on your home page. Because significantly shorter reflections and smaller number of reflections, exact frequency is the most critical for completing 70cm MS QSO. Saving the band with Perseus SDR (I had it with me) would help, but I did not use it this time (have only two computers with us). Now I am working with of MENINA 432/21 transverter (prototype) and stability of my frequency is good. Before, with SSB Electronic 432/28 MHz transverter, the situation was bad. Stability of oscillator in SSB transverter is not good and I drifted 400Hz during my first MS QSO with GW4DGU back in 2004 Geminids (with 100W and 8x26el) and I did not decoded few bursts because of that. JT65 was almost impossible.

I do not remember any article about 70cm MS in any radioamateur magazine. I think I read something in ARRL Handbook once, but I am not sure. TNX for your article in last DUBUS.

I did not tried 23cm MS. I suppose, based on degradation of signals on 70 compare to 2m MS, that 23cm is almost impossible. But I would be very happy if someone will prove this wrong...

Sorry that we did not complete EME QSO this Saturday. I had problems with my elevation rotor. I saw you on waterfall, but no decode. Hope we will complete soon also EME.

73 Sine - S53RM

Hello Matej.

Before I have information that wind at 100km altitude is around 200km/h. Now I find on internet, that maximum wind speed is up to 360km/h. That gives you 150Hz of Doppler shift on 70cm. But we have antennas pointed to each other and reflections take place between us, when working 70cm MS, so Doppler effect is around 0Hz. (Like on 10 GHz rain scater, if rain is directly between two stations - one can hear very little distortion on signal). If we follow backscattering than Doppler on 70cm could be 300Hz.

73 Sine - S53RM

Od: Carl G. Mohlin <cegem97<AT>>
Předmět: Re:MS-question 70/23
Datum: 16.3. 2009, 15:30

Hello Matej,
Sorry for the delay. Actually I did answer it when I got home , but on sending it my server refused to take it and the msg disappeared. Well, I will do another try now. First I did my MS activity several years ago and only on HSCW so with FSK and waterfall assistance it should be EASY. Also freq accuracy is today available to everybody which is timesaving.
As to your questions.

1. Highest doppler on 70: No idea, as this could be anything depending on speed and direction of meteor giving the reflexion. On EME the moon on rise/set produces roughly max +/-1˝ kc of doppler on 432 .

2. Decoding ability. No experience.

3. Test 23cm. I have tried (in vain) to get EME guys on 23cm MS . Particularly for the first Leonides highintensity prediction (What year 2000 ??) I even did announced a CQ calling freq. Plenty doppler to be expected. I think a 23cm MS qso would be a world first! Will need power.

4.Best shower. My 432 qso:s have been on Perseides and Gemnides showers. Still not many qso's

5. Best burst 432. From a qso with GW4DGU got a ~50 seconds burst which partly reached S9 !! Just fantastic!! At that time had 8x21 Tonnas.

6. Max HSCW. I have just used 3000 lpm but more than that is possible but your rx AGC may be a limiting factor.

Articles on UHF MS in the past are lost in my memory. Cannot recall of any ? As far as I know it has just been 144 and below.. I expected MY/UA9FAD world record on 432 to be broken long time ago with the FSK44 coming into use. Apparantly CT1HZE has done it just recently and with just few km's. I dont remember how exactly we calculated our record. Might be some km's longer. hi

Regarding 23cm I think It might be possible to make dobbel/tripple hop aeroplane-scatter. As a matter of fact I'm sure many recorded MS qso on VHF are mixed modes including polution and aeroplanes. Vast number of Aeroplanes above Eu and polution too. hi

Ivo and I made our first OK/SM on 144 back in 1960 as I remember it We tried also 432 but never made it untill he had to leave. We used to work first on 144 and the go to 3rd harmonic on 432 to establish a common freq. hi

Good luck Matej. I will be on in the Dubus EME contest, 144 the first weekend CW and 432 the next.


 Od: Rytis LY4U <>
Předmět: Meteor Scatter 70cm
Datum: 20.11. 2009, 02:41 - včera v 02:41

Matej, this is what I would like to add about experiences on 432 MHz Meteor Scatter. Thanks for collecting this material, much appreciated.

Most notabe was UA9FAD LO88DA - LY2WR KO24OQ on 12-Aug-1998 21:30UTC, which was the world record at the time, later beaten by SM3AKW-UA9FAD, but still 2-nd(?) known result at 1932km.

During 1980-s and 90-s also worked IW5AVM(now IZ5EME), UA4API, SM2CEW, OE3JPC, YO2IS (first YO-LY on 70), YU1EV (first YU-LY on 70), PA3DZL, DL4MEA, DL1KDA (this one deserves special comment: during Leonids 1998 we met on 144.0 and during single 5 minute (or more) burst we talked with Alex like on tropo and arranged sked for 70cm, which we completed in 30min on 432.100 in HSCW). There were also numerous NC attempts with some reflections received: I5WBE, PA2CHR, HA7P. Likely there were other QSOs/attempts which I could not recall, unfortunately part of the old LY2WR logs have been lost during QRT period early 2000s-2008. All QSOs made using HSCW 1000-2500 LPM.

One particular attempt, deserving comment, was with a french group around Joel F6FHP IN94TR in Perseids 1999 - that would have been world record for 2191km but only one reflecion was received in 2 hours - by LY2WR and about 10-15second long (!) and as I recall, very loud. Rig used for all these QSO was FT-736R, 1.5KW PA and single 8m long yagi, location at that time was KO24OQ Vilnius TV tower 165m a.g.l., ~260m a.s.l.

What the difference from 2m MS is mainly importance of defining the reflection point in the sky, because of practical considerations - on 70cm stations use antennas with higher gain as compared to 2m (which is necessary) but resulting in much narrower main lobe, thus illuminating much smaller available area in the sky. As it is known, for QSOs in distances lesser than 2000km the antenna offset and elevetion is necessary for maximum efficiency, up to 30-50 degrees off the great-circle direction to a QSO partner for say 1000km, so the geometry factor plays a big role here. QSO's of 1500km and less all were made using some azimuth/elevation offset which proved effective. I may be wrong, but up to today (2009) I was not able to find better tool for calculating reflection point between two known locators in said meteor shower, than OH5IY's program (,DOS but works on my XP).

As for doppler, it should exist, but all my tests were done in HSCW thus not an important factor, can't comment on this. Another practical
recommendation - 70cm MS reflections happen as little as maybe 20 times less often than on 2m in similar conditions, so a great patience is essential in trying, but great satisfaction is also guaranteed once the 70cm MS reflections come finally! Wishing a success to all likeminded,



Note of OK1TEH: In 1998 LY2WR wrote:

IARU Region 1 MS record on 432 MHz (according to MS conference on the Internet)

According to a report by Rytis LY2BIL/LY2WR, MS works well also on 70 cm, however with more power, shorter antenna and big patience. After five failed attempts in May and on 10 to 11 August 1998 LY2WR made a QSO with  UA9FAD in MS CW on distance of 1931 km (KO24OQ-LO88DA). More LY2WR worked during this shower on 432 MHz SM2CEW and UA4API. Rytis has been running MS tests on 432 MHz since 1997 and he wrote  some brief advice for those who would like to try it too:

1.  Antenna: 1 to 4 yagies  for distances up to 1 500 km, elevation is needed. For bigger distance you need bigger arrays.

2. Output power: it is possible to make a contact with 200 W during shower peek, but for serious work you need 1 kW.

3. Receiver: good LNA at the antenna is a necessity.

4. Speed: In general, reflections on 432 MHz are much shorter then on 144 MHz so it's better to use the speed LPM 2000-3000.

5. The precise frequency is very important

6. Do not use the narrow CW filter.

7. Very important is the precise organization of skeds  and set up of  azimuth and elevation. Very good utility is program by OH5IY.

8. In conclusion: for MS on 70 centimeters big patience is needed, but then you'll  have bigger  joy of  completed QSO.


From To Received Bytes Subject LY2BIL VHF @ EU 1998-12-03 10:26 2344 MS-432 in Leonids

To  : VHF@EU

Hi all
Was very busy before leonids and didn't manage any skeds so had to arrange "on-go": during a long burst on 2m we exchanged sked-details with Alex DL1KDA and in 35min he was in log with 26/49 reports, on my side 5bursts, no pings, max 25sec, s9++. 17-Nov, 02:50UT. The only other sked on Nov-17 was with I5WBE, I received 40s of 59+20 reflection up to switch-over time, I replied immediatelly with "599 bk" in slow cw, hearing nothing in some 5 seconds I transmitted "I5WBE DE LY2WR 599 BK" in slow CW again. No reply so I resumed 2000lpm - ms procedure. As I was told later by Enrico, both transmissions were received by him. I assume the reflection lasted at least  100 seconds. Unfortunately sked was NC. On night 17-18 Nov had few more skeds, they were NIL of course (ZHR of 30-40 is clearly not enough for 432, or you must be very lucky!). As proposed I was doing many CQs on 432.0 hoping for random 1-burst QSOs between 00-07UT on Nov 17, but was only once hrd by PA0RDY who was so surprised he didn't  call me.

This experience proved me that ms-432 one-burst contacts were ABSOLUTELLY POSSIBLE in ZHR around 200 (IMO), having Leonids trails, there had to be some 10-15 stations making noise same time on same freq only.  Why 10-15? Simply because 432 MHz antennas illuminate much narrower part of the possible reflection area than on 2m. Hope we could do better here in Europe in 1999 Leonids,  I would repeat what I wrote before this years':

1. This may be the only lifetime chance to work some areas like F,I,S5,LZ,SV from LY on 432 MHz meteor-scatter, same for all of you, without using any special ms-procedures or high speed CW, simply 59-roger59-roger73! Just be QRV on same freq (432.0?) same time (frequent 3-5 minutes long reflections on 2m is a good indicator, eh?). EME-ers welcome!

2. Remember, Leonids radiant is over horizon between aprox. 22-12 hours local time (hi hi).
Well, on 2m I worked some 100 ms qso in about 5 hours, which gave 2 new squares. I'd better spend 100% time on 432.0 but who else was there? (PA0RDY - Rob, I'm gratefull for accepting my invitation).
p.s. Hope to QSO in BCC MS Contest, we will be using LYxxx callsign, calling CQ-(letter) and won't make self- or semi-self-announcements in cluster!
73 Rytis LY2BIL/LY2WR


More at