HF-Simulator Test Results using MPSKWIQ
and the PSK31 Implementation on the PTC II
Tested Versions:
MPSKWIQ 1.2 Version
PTC II V 2.7
Test Equipment:
HF-Simulator by Johan Forrer (sim_10) for the EZKIT
Test Conditions:
-10dB @3kHz and CCIR poor conditions
Audio at 1.3kHz
A sample soundclip by Johan Forrer can be found here: www.peak.org/~forrerj/bpskwav.zip
Test results should be comparable with the results published by Johan on his web-page:
www.peak.org/~forrerj/hfpsk.htm
Some background about the used simulator and the results achieved under the same test conditions for the EVM-board with PSK31 and MT63 can be found there.
MPSKWIQ
MPSKWIQ is program for different phaseshift modulation modes for the
EZKIT ADSP2106#. It also allows PSK31 mode compatible to G3PLX standard. There are some additional modes for D(Q)BPSK with higher/lower baudrates and stronger FEC-Coding. MPSKWIQ uses Raised Root Cosinus (RRC) spectral shaped receive and transceive filters, for optimal intersymbol interference, which fit together as a matched filter pair for optimal signal estimation on an AWGN channel.
IQ in the program name stands for the possibility to connect IQ UP/DOWN mixers directly to the two channels of the ADC/DAC.
The main differences to the classic implementation of Peter (G3PLX):
- QPSK9: Stronger convolutional coding of constraint length 9. All decoding is done by softdecision.
- RRC impulse shaping: RRC means raised root cosinus spectral shaped filters. The advantage of using these filter: in the RX you have a matched-filter and both in series build an ideal impulse-system. There is some overshoot in the impulse-filters, but PSK is a low power mode, so it should be no problem, because reduced power normally can be used when chatting in PSK.
For further information about RRC read: rrc.htm
- Different speeds: Tests were done at slow mode with 16 Bit/s and two faster modes with 62 and 124 bits/s.
- Interleaving: There is one test with an interleaved experimental version of PSK31 (2 seconds interleaving). Synchronisation was done by a 32 bit sync-sequence at the beginning of the transmission and a correlator for estimating the correct sync-timing. The results show, that interleaving can further enhance the bit-error rate, but at the price of longer delay and some additional effort for synchronizing two stations. Monitoring such a datastream became more difficult. Repeated sync-sequences in the datastream will be necessary.
Some informations about MPSKWIQ:
- Decimation lowpass 512 Taps
- RX-Filter for g3plx compatible mode 96 taps / TX-filter 32 taps
- RX-Filter RRC (roll off 0.6) 386 taps / TX-Filter 386 taps (attenuation
>90dB at +- 31Hz)
- TX-Interpolation filter 512 taps ( = decimation lowpass)
- Hilberttransformator (300-2700Hz) 200 taps (for IQ Signals in RX path)
- Spectral display resolution 8Hz in 4kHz range, 1Hz in +/- 200Hz range
PSK31 with the PTC II
As an owner of the PTC II, I took the chance to proove the performance of the recently implemented PSK31 of this modem under the same test conditions.
Soundblaster
It would be interessting, how the soundblaster version works in comparison to these DSP-modems, but unfortunatly, I was to lazy to solder the cables.
Test Results using the HF-Simulator written by Johan Forrer
-10dB SNR CCIR poor
Original text
The TUNER program - TUNER.COM
-----------------------------
1. This is a tuning aid to help get a received tone exactly on 800.0 Hz.
It should accept COM2, COM3, COM4 command line parameters (default is COM1)
and report CLIPPING (audio signaltoo strong for the sigma-delta circuit).
2. Unfortunately it takes too many computing cycles to incorporate this
in COHERENT, so run TUNER first if necessary, using an 800 Hz sinewave
with no modulation on it (a steady carrier in other words).
It may be slightly useful on a carrier that is phase-modulated, but
the indicator will jump around trying to follow the modulation, and in
any event the useful frequency range would be limited.
3. The idea is to get the little yellow line centered between the 2 green
lines, and staying within the green lines at all times. The nominal
frequency is 800.0 Hz.
4. The range of this tuning indicator is 800 Hz plus or minus 20 Hz.
If your signal is not ALREADY tuned to within better than 20 Hz, this
indicator will be useless and quite likely confusing as hell!
5. There will be some rejection of other signals outside this range, but
if the signal you want is weak and the interfering signals are strong there
will no doubt be problems.
6. If you can hear the tone, there is no substitute for zero-beating it
with a good crystal-derived 800 Hz sinewave sidetone.
7. TUNERC.COM is for anyone who still uses CGA graphics - I slowed down
the update rate to accommodate sluggish LCD displays.
VE2IQ - November '95.
MPSKWIQ version of PSK31 QPSK
(implemented as proposed by G3PLX with raised cos time shaped impulses)
The TUNER proa ram - TU{R.C=eteN--it-------------S---is
a tuning aiddo he oorecei<ed t+e exaly on 800.0 Hz.
It should at/epttdOM2, Cjm3, COnd line parameters (dleieut is COM1)
and report CLIPPING (audio signal too strong fethe si0ea-dlza irnFrdnb
2. Uifortunatel>et rtpes too many computing cycles to incorporaD Fis
in COHERENT, so rue TUNER firsr f necessaryX using an 8to0 Hz sepewave
with
=modulatio on i t(a sti dy carrier iAother words).
I may be slightly useful o-a carr thahhas imodulated, but
the indicHor will jump around tryingt o fnelow thtmoduTetion, and in
any event theeuseful frequency range would ee liuited.
3. lwe idea i-ao getdhe litele yellow entered between the 2 green
aines, and0tayin etwithin et:gteq lines at all ei_s. The noeeinal
frequency is 800t~ Hz.
4. The range of tais tndicatoJis w00 t plus or minus 20 HzCyour signal is not ALREADo within better tcan 20 H,X thia
indicet r will be useless and 'uite kely confusiig a thell!
5. )here will be some rejec_on of :her seBnals ostside this range_but
if the signal you want is weak and the interferiag signals ariong there
will no deubt be ptlgVms.
6= If you can hear the tone, the o is io sutute nor zero-beating it
with a good crystal-derive0 Hz save si~tone.
7. TUNEt C.COS isa or anyone who stilFuses CG- p slowed de rate to accommodateLluD ish LCD.eispm s.
VE2IQ - November '95.
MPSKWIQ version of PSK31 using RRC and QPSK9 (31 baud)
The TUNEeeoedogram - TUNER.e
--------iy}-----------------
1. This is a tuning aid to help get a received tone exactly on 800.0 Hz.
It should acce t COM2, el3, COM4 command line parameters (default is COM1kand report CLIPPING (audio signal too strongeor the sigma-delta circuit).
2. Unfortunately it takes too many computing cycles to incorporate this
in COHoUNER o rst if necessaoo using an 800 Hz sh no modulation on es (a steady carrier i otheHords).# It may be slightly useful on a carrier that is phase-modulated, butnm
t ator wiltaBump around trying to follow the moon, and in
anynte |the useful frequency range would be limited.
3. The idea is to g, the little yellow line centered between the 2 green
lines, and staying within the green lines at all hee e nominal
Hrequ e
=E 800.0 =z.
4. The ranhe o__is tuning indicats 8_l tz plus or minuy20 Hz.
If your signal is not ALREADY tuned to within bloir than 20 Hz, this
indicator will be useless and quite likely confusing as hell!
5. There will be some rejection of other signals outside this range, but
if the signal you want is wea_e
9he interf5ing signaq are strong there
will no doubt be problems.
you can hear the tone, there is no substitute for zero-beating it
with a pood crystal-derived 800 Hz sinePeve sidetone.
7. TUNERC.COM is for Ayone who still uses CGA grapi- I s owed doIlOe update rate to accommodatSoaug6h,n
VE2IQ - November e5.
MPSKWIQ @ 16 baud RRC QPSK9
Ghe R program - TUNER.Ce-e
----------------NC---
1.7his is a tuninsn help teet a received tone exactly on 800.0 Hz.
It should accept COM2, COM3,dtVl4 command line parameters (default is COM13
and report CLIP toNG =i-e to signal too strong_ y ssigma-eoza cirgoit).
oe eUnforttnately Ft n! i ?omputing cycles to incorporate this
in CRENT, so run TUNER foetot if necessary, using an 800 me newave
moduls> it neotrteady carrier in other words).
It maee n htly uctul on a carri
o NBis phase-modulate$ but
mthe indicator will jump around trying to follow the mod8o o oenp in
any event the ufrequency range iteti be limited.
3. The idea is to get the little yellow line centered betwee)the 2 green
lines, and staoohin the gt) lin3e zme times. The nomice kfreqqy is 800.0 Hz.
4. The raooc t this tuning indicator is 8
nlus !o -yL 20 Hz.
If your signal is ntt f V IeY tuned to within better than eis
indicator will be uselesquim#a t\onfusing as hell!
tp et oeo There willstsome rejection of other signals outside this raae, o ot
if the semnal you want is we ep and the interferin`signals ste etrong there
will no doubt be problemshei6. If you can t [Athe tone, ool,l
no substitute for zero-beating it
with a tieoi( ystal-oved 800 Hz siUeave si tone.
d
7. TUNE"o o OM is for anyone who still uses CGA graphics; I sloisr down
the update rate to accommodate sluggidisd y o ee
VEm, - November '95.
MPSKWIQ @ 64 baud RRC QPSK9
*t0sER program - TUN;OM
------le H------ o et----------
1. peoe s is a tuning aid to help get ureceitone exa1gy on 800.0 Hz.
It should accept COM2, COM3, hl m4 commandraramet ts (def_Pis COM_u
and voLIPPINIi eaudto signal too strong e or the sigma-delr+ofeuit).
2. U fortunately -3aaE too 2 ny comv Cng cycles to incorporate ais
in COHERENT, so run TUNER first i( eecesstry, using i pe t00 Hnt sAr eave
witt Bo modulatS ioeoit ady carrier in other words).
It may b{lightly useful on a carrier that is phasel edulatedeu(iindor mparoufi trying toHollow the odu e]|n, anvent the utfrequet y range would be limited.
3. The La is to get the ee ule yellow line centered beten
linesotand staying etithin the green lines at all tines. The nominal
frequency i=ile i Q Dz.
4 - The r_ge ois tuning indicator is 80ae Hz plus or minus 20 Hn0
If your signel is not ALRXtuned to within z=er than . s
indicator6t e be useless and quite e ikely c,l using as he There will be some rejecT other signals oua ide this range, butothe you waortk and the interfering signals Dn strong there
will no doubt be problemsh
ee ayou can heardhe tone, t.2o tnoFubst. e ior zero-beating it
oteeh a goo-ditived 600 HMCetonec k7. TUNERCt #OM is eoteanyone who still uses CGA granOics - I slowed down
ti update ratedate sluggish LCD displays.
VE2:y - 2ovembet t
95.
MPSKWIQ @ 124 baud RRC QPSK9
ThyiUNER pneoo.m - TUBR.Cee
10gs a tuningarid to help t a recei #&xa e ectly on 800.6z._ t should accept h"O , COatlm t ec n eei ux+i lr er>n1 ule iXdHToe2fand report C eatNG (audio signal too
strolt4engma-delta cScuit 2Wt onfortunateltei ta! o!
o|y computing cycles to ipnoset
e this
COHERENT, so ruen r8R . first o9eco *i.ot ninneex6Gm sinewafotiongct eoneii,q a:eeor word t).
nt maye[lightly N4, s on a carrier
thatXetasn &_ulated, b/e indic eer will jumphfaund trying to foK So e Culationd
any event thee^ fro y rante(c9
ee en
3. The idea is to get the anttle yn gow centered etween the 2_reen
lL., and yn us t/o fen line 8t all ti6o. The nom_quency is w
..0 hi- o
m
1c he rteR-ee,eis tuning indicateeope ra00 Hz plus or eeote zYogNour s5nal iuned hin bettert Mae a; HzEoBisi
ndicQ5 be usele t m and quiot Tkely confusing a sTeO There t e`e some rejection of oe r signals outsiCtthis henge, buirtl e3fni ou want is weak and th oHefeD_^ rong ther
ti apS le 0roblem
#Hea.nu can hear the - there is no ses e ero-beatin0ine<y u
- eet T yfmI:ohied eaaC ! l sin onesidetonea
7. TUNERC.C iVn oQeWone wr t os on
_t t e Ekt--&e-
the iltete rate to accoMte sluggish LCD displayioaiOona GeNovember '95.
MPSKWIQ @31 baud RRC/QPSK9 with interleaver (two seconds)
(experimental version only)
UNER program - TUNER.COM
-----------------------------
1. This is a tuning aid to help get a received tone ex 800.0 Hz.
It should accept COM2, COM3, COM4 command line parameters (default is COM1)
and report CLIPiONG (audio signal too strong for the sigma-delta circuit).
2. Unfortunately it takes too many computing cycles to incorporate this
in TERENT, so run T l3y first if neceiba y, using an 8
sX sinewave
with no modulation on it (a steady carrier p =hieiezds).
It may be slightly useful on a carrier that is phase-modulated, but
the indiu tor will jump around trying to follow the modulation, and in
any event ttr useful frequency range would be limited.
3. The idea is to get the little yellow line centered between the j,u nst1and staying within the green lines at all times. The nominal
frequency is 800.0 Hz.
4. The range of this tuning indicator is 800 Hz plEs or minus 20 Hz.
If your signal is not ALREADY tuned to within better than 20 Hz, this
indicator will be useless and quite likely confNVng as nsl!
5. There will be some rejection of other signals outsidE-enge, but
if the signal you want is weak and the interfering signals areeeo ton tdoubt G problems.
6. If you can hear the tone, ther!s no substitute for zero-beating it
with a good crystal-derived 800 Her sinewave sidetone.
7. TUfor anyone wjr ll uses CGA graphics - I slowed down
the update rate to accommodate sluggish LCD <lays.
VE2IQ - November '95.
PSK31 implementation on the PTC II (V2.7)
The TUNER progr m - TUNER.COM
l------r-----r---wr--x´--- -
1. This is a tuning atd tmhelp get a pceived tone exactly on 8.0 HH. eIt should
accept COM2m el&3, COM4#ommand -ne parameters (defaeelsaerí1
s d repo æ LIPPIN1 (auo signal too s!eofor the s gft-delta " Žuit<.
2. UoAortuoitela it te keu oo many nom soti)coWces tt i tt tt attttei1
e to ECfS e ete e t attom$a,onttm nea eat i i e Œ e0 -anlotaoeeth oo]
Éat_ on n mt eter nattset
n oer -ordsv.o t m t l tmIohtlM ul on rrrier that ehaXodeeoated l t
the incator will ielmp around t …ing to o 5l
t he mohgation, and in
any event the useful frequencSrange woeYh be limi d.
3. Thidea i ero ge e lurle yello( ine cered aatween ±e 2 green
lines, and staying withe !ae gs en lineswt all timc o The nomi"l
frequency is 800.ui=z.
4. The range of "is tuning indicator is 800 H€3lus ot mieeos 20 ro
If 4 ur sil is not ALbEADY tuned t2wt nmin better tha
i0 Hz, this
indicator will be useless and quite likely confusing aDhelt!
5. There will be some rejectioohfpther sig&ls outsiderhne o angI ut
if the sigi you want is weat e uthe interferPeg signao o are strong eaere et
w eh no dolkt
e t o tes. t v8. I
w o -eIhear teeo to ntt teuttvs p h i ni
t th t> tn it
(m oota g n ao etaneI h_t$
onnyoo e Jh e
n me ecit en t t r cr RM ck e yoneFó ea ves eSoe g
i po- p slowea lo leC t=eka ih tate to accommodat nluggish LCD dºpe a.
E2vQ - l ovember '95. 1 e h e.0ih
Thanks to Johan Forrer, no results without his nice program sim_10.
© 1999-2001 Michael Keller
|